=============================================================== Tag name: clm4_5_18_r275 Originator(s): erik (Erik Kluzek,UCAR/TSS,303-497-1326) Date: Mon Feb 5 15:02:53 MST 2018 One-line Summary: Update initial condition files, update fates, new CMIP6 pop-dens, more lnd_tuning_mod options, updated ndep for fully coupled Purpose of changes ------------------ Update the initial condition files for clm4_0, clm4_5, clm5_0 at f09_g16 for both CRUNCEpv7 forcing and GSWP3v1 forcing. Update FATES with science change to science-1.4.1_api-3.0.0. New CMIP6 population density file that has years beyond 1850 (the old one just had 1850, so was incorrect for historical or present-day). clm4_0 compsets now use CRUNCEPv7 forcing rather than the CRUNCEPv4 forcing. LND_TUNING_MODE is now used for initial conditions for different forcing. For fully coupled 1850 cases the ndep file is updated to a preliminary CMIP6 version. Fix the tools testing and build-namelist testing to work in the git checkout. Bugs fixed or introduced ------------------------ Issues fixed (include CTSM Issue #: #247 -- Clm40 1850 and HIST compsets are incorrect #246 -- New CMIP6 population density file is incorrect as it's constant in time #171 -- CLM is getting the mask from the clm namelist_defaults rather than from top level scripts Known bugs introduced in this tag (include github issue ID): #253 -- Bad new IC file for clm4_0 #254 -- Failing test ERS_D_Ld3.f10_f10_musgs.I1850Clm50BgcCrop.cheyenne_gnu.clm-default (sporadic glitch?) #255 -- PTCLMmkdata doesn't work in the CLM git checkout Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): LND_TUNING_MODE is set by default according to the compset the case is created with. If you change the CLM physics or DATM forcing, you'll want to update the LND_TUNING_MODE. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): LND_TUNING_MODE options changed to: clm4_5_CRUv7,clm4_5_GSWP3v1,clm4_5_cam6.0,clm5_0_cam6.0,clm5_0_CRUv7,clm5_0_GSWP3v1 Several new compset aliases added: CLM50: I1850Clm50Sp, I1850Clm50SpCru, I1850Clm50BgcCropCru CLM45: I1850Clm45BgcCruGs, I1850Clm45BgcGs CLM40 I1850Clm40CnGswGs (fixed I2000Clm40SpCruGs, I1850Clm40SpCruGs, and IHistClm40SpCruGs) Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files): New IC files for f09_g16 CMIP6 ndep file for fully coupled, pop-density beyond 1850 Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Setting of ndep_taxmode/ndep_varlist should be simplified in CLMBuildNamelist.pm Change name of setup_logic_lnd_tuning to setup_cmdl_lnd_tuning in CLMBuildNamelist.pm Changes to tests or testing: Change a few tests and add a few Add I1850Clm50SpCru, I1850Clm50BgcCropCru, I1850Clm40CnGswGs and I1850Clm50Sp tests at f09_g16 Change I1850Clm40SpCruGs f09_f09_mg16 to f09_g16 Change I2000Clm50BgcCruGs f09_g17 test to I1850Clm50BgcCruGs f09_g16 Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (134 comparison tests are different than clm4_5_18_r275) unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - OK PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - FAIL regular tests (aux_clm): cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_18_r275 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers: - what code configurations: clm4_0 for f09_g16 clm4_0 updated to CRUv7 Forcing all clm4_5 and clm5_0 configurations, FATES-update - what platforms/compilers: All - nature of change: initialization, BGC for HIST/2000 compsets Detailed list of changes ------------------------ List any externals directories updated (cime, rtm, mosart, cism, fates, etc.): FATES fates to fates_s1.4.1_a3.0.0_rev3 Pull Requests that document changes (include PR ids): (https://github.com/ESCOMP/ctsm/pull) #252 -- New IC files #251 -- README update #248 -- Documentation of SMB in namelist_definition #228 -- Get tools and testing working =============================================================== =============================================================== Tag name: clm4_5_18_r274 Originator(s): sacks (Bill Sacks) Date: Tue Jan 30 05:55:53 MST 2018 One-line Summary: Fix auto-detection of CIME_MODEL in a standalone checkout Purpose of changes ------------------ The auto-detection of whether CIME_MODEL is acme or cesm is broken in standalone checkouts of clm4_5_18_r273. (This auto-detection relied on whether there was an SVN_EXTERNAL_DIRECTORIES file present at the top level.) This tag points to a new cime version that fixes this issue. Bugs fixed or introduced ------------------------ Issues fixed (include CTSM Issue #): - #238 - clm4_5_18_r273 requires setting CIME_MODEL Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Erik Kluzek Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - not run unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - not run regular tests (aux_clm): cheyenne_intel ---- pass cheyenne_gnu ------ pass hobart_nag -------- pass hobart_pgi -------- pass hobart_intel ------ pass Note: I ran the cheyenne tests with CIME_MODEL=cesm defined in my environment, and the hobart tests without this setting. CLM tag used for the baseline comparisons: clm4_5_18_r273 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any externals directories updated (cime, rtm, mosart, cism, fates, etc.): - cime: clm4518/n03/cime5.4.0-alpha.03 -> clm4518/n04/cime5.4.0-alpha.03 Fix to auto-detect that CIME_MODEL=cesm based on presence of manage_externals rather than SVN_EXTERNAL_DIRECTORIES List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Fix the documentation of some expected fails M cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_18_r273 Originator(s): sacks (Bill Sacks) Date: Fri Jan 26 15:26:06 MST 2018 One-line Summary: Support a standalone checkout from git Purpose of changes ------------------ This PR allows a standalone git-based checkout to work. Starting with this tag, we'll be using git rather than svn for CLM/CTSM development. The biggest changes here are pulling in manage_externals, which serves the role of svn externals. After cloning the repository, you can get all of the necessary externals by running: ./manage_externals/checkout_externals See the file README_EXTERNALS.rst at the top level of the repository for more details on using this tool. Other than that, this PR mostly involves changes to allow you to run from the new directory structure of a standalone checkout: Now all of the CLM/CTSM directories appear at the top level of the checkout, rather than nested under components/clm. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include github issue ID): - https://github.com/ESCOMP/ctsm/issues/220 For new standalone organization, need to generalize paths for tools and testing Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - The svn trunk is now frozen. All future development will occur in the master branch of https://github.com/ESCOMP/ctsm - A standalone git checkout will have all CLM directories (src, bld, etc.) at the top level, rather than under components/clm. - To get all externals after cloning the git repository, run ./manage_externals/checkout_externals See the file README_EXTERNALS.rst at the top level of the repository for more details on using this tool. - More details on working with the git repository will be coming soon Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - build-namelist unit tests currently do not run (#220) - tools tests, and the tools builds themselves, are expected not to work (#220) Changes to tests or testing: none Code reviewed by: Ben Andre, Erik Kluzek Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - FAIL (#220) unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - NOT RUN; EXPECTED TO FAIL (#220) PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - NOT RUN; MAY FAIL DUE TO #220 regular tests (aux_clm): cheyenne_intel ---- pass cheyenne_gnu ------ pass hobart_nag -------- pass hobart_pgi -------- pass hobart_intel ------ pass CLM tag used for the baseline comparisons: clm4_5_18_r272 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any externals directories updated (cime, rtm, mosart, cism, fates, etc.): - cime: clm4518/n02/cime5.4.0-alpha.03 -> clm4518/n03/cime5.4.0-alpha.03 Has some changes needed for new standalone directory structure to work (these changes are now on cime master, though see #225 before updating to cime master) - cism: cism2_1_40 -> cism2_1_46 Points to cism in git cism2_1_41, 42, 43 and 44 contained very minor changes cism2_1_45 and 46 are the git migration List all files eliminated: ========= Handle externals via manage_externals rather than svn externals D README_EXTERNALS D SVN_EXTERNAL_DIRECTORIES D SVN_EXTERNAL_DIRECTORIES.standalone List all files added and what they do: ========= Added this directory and everything below it. Pulled in via git subtree from https://github.com/NCAR/manage_externals/. This is the new tool for pulling in and updating externals (replacing svn externals). A manage_externals/ ========= Handle externals via manage_externals rather than svn externals A Externals.cfg A Externals_CLM.cfg A README_EXTERNALS.rst ========= Handle new directory structure of standalone checkouts A .config_files.xml ========= Add some .gitignore files A src/dyn_subgrid/.gitignore A src/unit_test_shr/.gitignore A src/unit_test_stubs/main/.gitignore A src/unit_test_stubs/utils/.gitignore A src/utils/.gitignore A cime_config/.gitignore List all existing files that have been modified, and describe the changes: ========= Handle new directory structure of standalone checkouts M bld/CLMBuildNamelist.pm M bld/configure M cime_config/buildcpp M cime_config/buildlib M cime_config/buildnml M src/CMakeLists.txt ========= Documentation updates for move to git M .CLMTrunkChecklist M .gitignore M LICENSE M README.rst M src/README.unit_testing ========= Separately document expected fails for aux_clm vs fates test lists M cime_config/testdefs/ExpectedTestFails.xml ========= ChangeLog now only stored in doc directory, not in top-level directory R095 .ChangeLog_template doc/.ChangeLog_template R070 UpDateChangeLog.pl doc/UpdateChangelog.pl =============================================================== =============================================================== Tag name: clm4_5_18_r272 Originator(s): erik (Erik Kluzek) Date: Thu Jan 25 01:17:01 MST 2018 One-line Summary: Bring in latest FATES release version to CLM trunk: fates_s1.4.1_a3.0.0_rev2 Purpose of changes ------------------ Bring in the latest version of FATES from NGEET/fates-release and the associated interface changes in CLM from NGEET/fates-clm. The new version is 1.4.1 for science and version 3.0 of the API updated from 1.3.0 for science and 1.0.0 for the API. Bugs fixed or introduced ------------------------ Issues fixed (include CCTSM Issue #): #219 #219 -- Options listed incorrectly in testlist_clm.xml Known bugs introduced in this tag (include github issue ID): (or found) #216 -- PE layout for hcru_hcru needs to be increased so that it can run #213 -- CPLHIST case is matching the wrong CLM_NML_USE_CASE #212 -- Can't use a non-fates restart file to interpolate to a simulation with fates #211 -- Can't handle interpolating from a fates restart file because of the cohort dimension Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Simple FATES initial-condition files were added for present day at f45 and 1x1_brazil. FATES IC can't be interpolated from so you either have to have an exact match or do a cold start. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New FATES namelist items: use_fates_planthydro, use_fates_ed_st3, use_fates_ed_prescribed_phys, use_fates_inventory_init, fates_inventory_ctrl_filename,use_fates_logging Changes made to namelist defaults (e.g., changed parameter values): when FATES on, non-fates finidat files won't be matched Changes to the datasets (e.g., parameter, surface or initial files): New fates param file and finidat files for f45 and 1x1_brazil Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Changes to tests or testing: ed test list removed and updated fates list added in with new version=2.0 format, fates test list now sets machine/compiler rather than using generic "fates" for both. Code reviewed by: self, rgknox, ckoven Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (33 fates namelist tests are different) unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_18_r271 Answer changes -------------- Changes answers relative to baseline: Only simulations with fates on Summarize any changes to answers, i.e., - what code configurations: Only "-bgc fates" - what platforms/compilers: All - nature of change: new climate, fates science updated from 1.3.0 to 1.4.1 Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): rtm, mosart, fates rtm to rtm1_0_65 mosart to mosart1_0_30 fates to fates_s1.4.1_a3.0.0_rev2 List all files eliminated: fates test directory changed to one capitolized to be similar to new tests added D components/clm/cime_config/testdefs/testmods_dirs/clm/fates/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/fates/shell_commands List all files added and what they do: New tests for exercising more Fates options, all are based on the Fates test-mod and add a few changes on top of it A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesShortList/user_nl_clm ------- Short list of hist_fincl1 A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesShortList/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesST3/user_nl_clm ------------- use_fates_ed_st3=T A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesST3/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesAllVars/user_nl_clm --------- Long list of FATES variables A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesAllVars/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesAllVars/shell_commands ----- Cold-start A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesLogging/user_nl_clm -------- use_fates_logging=T A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesLogging/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesColdDef/user_nl_clm -------- Longer list of hist_fincl1 A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesColdDef/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesColdDef/shell_commands ----- Cold-start A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesPPhys/user_nl_clm ---------- use_fates_ed_prescribed_phys=T A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesPPhys/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/Fates/user_nl_clm -------------- Moved from fates directory, longer hist_fincl1 list A components/clm/cime_config/testdefs/testmods_dirs/clm/Fates/shell_commands ----------- Moved from fates directory, no longer a cold-start A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesNoFire/user_nl_clm -------- use_fates_spitfire=F A components/clm/cime_config/testdefs/testmods_dirs/clm/FatesNoFire/include_user_mods List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm - Handle new FATES namelist items, send use_fates down for setting of finidat, if use_fates_inventory_init=T, make sure fates_inventory_ctrl_filename is set M components/clm/bld/unit_testers/build-namelist_test.pl - Change ED to FATES, test new FATES features M components/clm/bld/namelist_files/namelist_defaults_overall.xml -- clm_start_type is arb_ic when use_fates=T M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Update fates var descriptsion, and add new namelist items: use_fates_logging, use_fates_planthydro, use_fates_ed_st3, use_fates_ed_prescribed_phys, use_fates_inventory_init, fates_inventory_ctrl_filename, remove use of ED for FATES M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Update fates_paramfile, use_init_interp=T only if use_fates=F, add use_fates=F to init_interp_attributes, add use_fates setting for finidat files, and add new f45 and 1x1_brazil FATES finidat files, by default new use_fates_* options are False. M components/clm/cime_config/config_compsets.xml ------------- Add a stub-GLC FATES compset: I2000Clm50FatesCruGs M components/clm/cime_config/testdefs/ExpectedTestFails.xml -- Many new expected fails for #221 and NGEET/fates#315 M components/clm/cime_config/testdefs/testlist_clm.xml ------- Remove "ed" test-category, new "fates" test category Move to after rather than nested inside (issue #219). M components/clm/src/utils/clmfates_paraminterfaceMod.F90 - Add write about dimension shape on abort M components/clm/src/utils/clmfates_interfaceMod.F90 ------ New interface control variables, add call to FatesReportParameters, use_fates_plant_hydro=>use_fates_planthydro, fix restart_map, zero_site=>init_site_vars use numpft_ed M components/clm/src/main/decompInitMod.F90 --------- Set gi rather than ci for cohorts M components/clm/src/main/subgridMod.F90 ------------ Add call subgrid_get_info_cohort, change subgrid_get_info_natveg call M components/clm/src/main/histFileMod.F90 ----------- Change some of the parameter names, EDTypesMod interfaces get moved to FatesInterfaceMod, nlevsclass_ed=>nlevsclass, maxpft=>numpft_ed, M components/clm/src/main/controlMod.F90 ------------ Add new fates flags to namelist and broadcast/print them M components/clm/src/main/clm_varctl.F90 ------------ New FATES flags, change ED=>FATES and some formatting M components/clm/src/main/initGridCellsMod.F90 ------ ncohorts=>npatches M components/clm/src/main/initVerticalMod.F90 ------- Remove call to ed_hist_scpfmaps M components/clm/src/main/readParamsMod.F90 --------- Remove call to FatesReadParameters M components/clm/src/biogeophys/CanopyFluxesMod.F90 - Add blank line M components/clm/src/biogeophys/OzoneMod.F90 -------- Comment out setting of o3* variables to spval M components/clm/src/CMakeLists.txt -- Add fates/main to unit-test directories =============================================================== =============================================================== Tag name: clm4_5_18_r271 Originator(s): erik (Erik Kluzek) Date: Sat Jan 20 20:27:00 MST 2018 One-line Summary: Update testlist to v2 and remove yellowstone Purpose of changes ------------------ Redo the testlist in version 2 format and remove yellowstone from it. Also update mosart and rtm with updated test lists. The new testlist format also allows wallclock times to be set (and they are) so that they can run in the queue faster. Change some files that referenced bugzilla ID's to github issue #'s. Update tools testing for gen_domain to use version under cime/tools rather than the copy under clm. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include github issue ID): #216 -- PE layout for hcru_hcru #213 -- CPLHIST case is matching the wrong CLM_NML_USE_CASE #212 -- Can't use a non-fates restart file to interpolate to a simulation with fates #211 -- Can't handle interpolating from a fates restart file because of the cohort dimension Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): MEMLEAK, TPUTCOMP and MEMCOMP tests now seem to be robust and can be trusted. Changes to tests or testing: Test list completely redone Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - OK PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - OK regular tests (aux_clm): cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- PASS hobart_intel ------ PASS CLM tag used for the baseline comparisons: clm4_5_18_r270 Answer changes -------------- Changes answers relative to baseline: no bit-for-bit, no source code changed Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, mosart, rtm cime to clm4518/n02/cime5.4.0-alpha.03 which allows CLM grids in testlist mosart to mosart1_0_29 with redone test list in v2 format rtm to rtm1_0_64 with redone test list in v2 format List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M parse_cime.cs.status -- MEMLEAK, MEMCOMP, and TPUTCOMP don't need to be ignored anymore M .ChangeLog_template -- Remove bugzilla in place for github issue ids Update to test gen_domain under cime/tools rather than the copy under components/clm/tools/gen_domain The copy won't be able to be done in the new github world. M components/clm/test/tools/TBLCFGtools.sh ----- Set name of test directory by basename (so relative path can be passed in) M components/clm/test/tools/input_tests_master - gen_domain path is relative to tools directory but under cime now M components/clm/test/tools/test_driver.sh ----- Remove yellowstone set "--mpilib mpi-serial" for TOOLS CFG string M components/clm/test/tools/TSMCFGtools.sh ----- Set name of test directory by basename M components/clm/test/tools/TCBCFGtools.sh ----- Set name of test directory by basename M components/clm/bld/unit_testers/build-namelist_test.pl - Fix number of tests M components/clm/cime_config/testdefs/testlist_clm.xml ------- Completely redo in version 2 format M components/clm/cime_config/testdefs/ExpectedTestFails.xml -- Update to github issues rather than bugzilla =============================================================== =============================================================== Tag name: clm4_5_18_r270 Originator(s): sacks (Bill Sacks) Date: Wed Dec 20 12:04:25 MST 2017 One-line Summary: Always use multiple elevation classes for glacier, even with stub glc Purpose of changes ------------------ Up until now, we have been using the glacier_mec (multiple elevation class) scheme when running with CISM, but the old non-elevation class scheme when running with a stub glacier model (SGLC). This is problematic because it means that simply switching from CISM to SGLC fundamentally changes CLM's glacier physics. (Note: with the current configuration, the multiple elevation class scheme is enabled over Greenland and Antarctica but not elsewhere.) There are a few use cases where this is important: - Single-point and regional runs over glacier regions: In this case, you cannot use CISM. Members of the LIWG want to do runs like this so they can do single-point testing with the multiple elevation class scheme. - Currently, CISM does not support a Gregorian calendar (i.e., with leap years). So cases with a Gregorian calendar need to use SGLC. This includes data assimilation and CAM specified dynamics runs. We want the physics of these runs to match the standard CLM physics as closely as is feasible. - There may be other cases where a user doesn't want to include CISM because of the extra complexity this involves, such as when setting up a new grid (which would require additional mapping files for CISM). Again, we want the physics of these runs to match the standard CLM physics as closely as is feasible. Note that runs with SGLC will still differ somewhat from runs with CISM, even when CISM is running in the typical NOEVOLVE mode. This is because there will be differences in % glacier and topographic heights over Greenland: in a run with CISM, this information comes from CISM at the start of the run so that CLM and CISM agree in this respect; in a run with SGLC this information comes from CLM's surface dataset. However: Running with SGLC now gives the same answers as running with CISM with GLC_TWO_WAY_COUPLING set to FALSE. (These both differ from the standard configuration that uses CISM with GLC_TWO_WAY_COUPLING TRUE just over Greenland.) Mostly-unrelated change: Remove tropicAtl grid Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): - https://github.com/ESCOMP/ctsm/issues/158 now causes test failures for ERS_Lm20_Mmpi-serial.1x1_smallvilleIA.I2000Clm50BgcCropGs.yellowstone_pgi.clm-monthly, since we're now using a new and buggy surface dataset for that case. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - If you use an finidat file from an old run that used SGLC, you'll need to run init_interp and will need to set init_interp_fill_missing_with_natveg = .true. in user_nl_clm. Note that this will initialize glacier_mec landunits with information from vegetated landunits, which is not a very good initialization, so you'll want to spin up for a while to spin up the new glacier_mec landunits. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): no Changes made to namelist defaults (e.g., changed parameter values): no Changes to the datasets (e.g., parameter, surface or initial files): - New surface datasets for single-point and regional runs: These now have the data necessary to run with the multiple elevation class glacier scheme. None of our out-of-the-box single-point or regional datasets have any glacier cover, so this doesn't lead to substantive changes. Note, though, that the smallville datasets hadn't been updated in a while, so there are bigger changes to the smallville datasets. The new smallville datasets are buggy (due to https://github.com/ESCOMP/ctsm/issues/158), but the old ones wouldn't have worked with the new code due to missing glacier_mec information. Note that I have NOT changed the landuse.timeseries files, because those don't change in any substantive way with the setting of -glc_nec in mksurfdata_map. - Runs with SGLC now point to different finidat files by default: they now point to finidat files that were generated with glacier_mec, which may differ substantially from the files that were generated without glacier_mec (since I think these came from completely different spinups). Substantial timing or memory changes: Expect some performance degradation in SGLC runs, since now we're using multiple elevation classes over Greenland and Antarctica, though I haven't investigated it carefully. Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Delete tropicAtl_subset tests: these tests didn't provide much value, and the code they cover is also covered by unit tests Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - ok prove test_build_namelist fails as on trunk main build-namelist tests pass, clm4_5 and clm5_0 lnd_in files differ from trunk as expected unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - ok Tests pass other than these known issues: smfg4 TSMscript_tools.sh PTCLM PTCLMmkdata PTCLM_USUMB_Global_clm4_5^buildtools See https://github.com/ESCOMP/ctsm/issues/188 blf84 TBLscript_tools.sh PTCLM PTCLMmkdata PTCLM_USUMB_clm4_5^buildtools blfc4 TBLscript_tools.sh PTCLM PTCLMmkdata PTCLM_USUMB_Cycle_clm4_5^buildtools See https://github.com/ESCOMP/ctsm/issues/187 PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - ok 1 failure, as on the trunk compare tests: 40 PASS, 1 Fail; on the trunk, all 41 Failed. I'm not sure why my branch improves things in this respect. regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi --- ok yellowstone_gnu --- ok cheyenne_intel ---- ok cheyenne_gnu ------ ok hobart_nag -------- ok hobart_pgi -------- ok hobart_intel ------ ok Tests pass. Baseline comparisons fail for compsets with SGLC (both global tests and single-point and regional tests). This is expected and is described in detail below. CLM tag used for the baseline comparisons: clm4_5_18_r269 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All compsets with SGLC (as opposed to CISM) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Potentially new climate, though not investigated carefully Answers change in runs with SGLC due to the use of different out-of-the-box finidat files: they now point to finidat files that were generated with glacier_mec, which may differ substantially from the files that were generated without glacier_mec (since I think these came from completely different spinups). For global runs, even without finidat differences, answers change over Greenland and Antarctica, since we're now using the multiple-elevation-class scheme there. Answers are the same outside of Greenland and Antarctica, where we still use a single glacier elevation class. (Based on spot-checking FSH in a 3-year cold start run.) (For single-point and regional runs, differences are mainly just due to changes in finidat. One exception may be smallville tests, since the smallville surface datasets have changed more extensively; I did not investigate the smallville answer changes carefully.) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime: cime5.4.0-alpha.03 -> billsacks/always_glcmec_n01 The changes here are: - always use GLC_NEC=10, except for CLM40 - delete references to tropicAtl List all files eliminated: ========= These glacierMEC use cases weren't accomplishing anything: compared with the similar non-glacierMEC use cases, they just set the glc_nec namelist variable, but this variable didn't even exist! So I have switched cases to use the corresponding non-glacierMEC use cases D components/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml D components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml D components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml D components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml D components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml D components/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml D components/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml ========= This use case appeared to be unused D components/clm/bld/namelist_files/use_cases/glacierMEC_pd.xml ========= These tropicAtl_subset tests were hard to maintain and didn't provide much value: the code they covered is also covered via unit tests D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/README D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/README D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/README D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/README D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Always use -glc_nec 10, delete tropicAtl, and other minor fixes M components/clm/tools/mksurfdata_map/Makefile.data ========= Require glc_nec >= 1 when making surface datasets M components/clm/tools/mksurfdata_map/src/mkfileMod.F90 M components/clm/tools/mksurfdata_map/src/mkglcmecMod.F90 M components/clm/tools/mksurfdata_map/src/mksurfdat.F90 M components/clm/tools/mksurfdata_map/mksurfdata.pl ========= Require glc_nec >= 1 for clm45/clm50 M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/unit_testers/build-namelist_test.pl M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/cime_config/buildnml ========= Remove istice; always use glacier_mec type M components/clm/src/biogeochem/DryDepVelocity.F90 M components/clm/src/main/controlMod.F90 M components/clm/src/main/LandunitType.F90 M components/clm/src/main/TopoMod.F90 M components/clm/src/main/filterMod.F90 M components/clm/src/main/lnd2glcMod.F90 M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/subgridAveMod.F90 M components/clm/src/main/initGridCellsMod.F90 M components/clm/src/main/initVerticalMod.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/surfrdMod.F90 M components/clm/src/main/clm_instMod.F90 M components/clm/src/main/landunit_varcon.F90 M components/clm/src/main/subgridWeightsMod.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/subgridMod.F90 M components/clm/src/main/ColumnType.F90 M components/clm/src/main/histFileMod.F90 M components/clm/src/main/restFileMod.F90 M components/clm/src/biogeophys/WaterStateType.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/TemperatureType.F90 M components/clm/src/biogeophys/SurfaceAlbedoMod.F90 M components/clm/src/biogeophys/PhotosynthesisMod.F90 M components/clm/src/biogeophys/CanopyTemperatureMod.F90 M components/clm/src/biogeophys/SoilHydrologyInitTimeConstMod.F90 M components/clm/src/biogeophys/SurfaceResistanceMod.F90 M components/clm/src/biogeophys/TotalWaterAndHeatMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/GlacierSurfaceMassBalanceMod.F90 M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/dyn_subgrid/dynInitColumnsMod.F90 M components/clm/src/dyn_subgrid/dynLandunitAreaMod.F90 M components/clm/src/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights.pf M components/clm/src/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/cpl/clm_cpl_indices.F90 M components/clm/src/cpl/lnd_import_export.F90 M components/clm/src/cpl/lnd_comp_mct.F90 ========= In addition to supporting the main changes of this tag, also do a fairly major refactor: set fields internally rather than letting lnd_import set the fields M components/clm/src/main/glc2lndMod.F90 ========= Fixes to unit tests for changes in glc2lnd M components/clm/src/main/test/topo_test/test_topo.pf ========= Remove tropicAtl tests and references to tropicAtl grid (see above for more notes on this) M components/clm/cime_config/testdefs/testlist_clm.xml M components/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M components/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M components/clm/bld/namelist_files/namelist_definition_clm4_0.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml M components/clm/bld/namelist_files/namelist_defaults_overall.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_0.xml M components/clm/bld/namelist_files/checkmapfiles.ncl M components/clm/doc/UsersGuide/adding_files.xml ========= No longer use glacierMEC-specific use cases (see above for more notes on this) M components/clm/cime_config/config_component.xml ========= Fix creation date of files pointed to by tests M components/clm/test/tools/nl_files/PTCLM_USUMB_clm4_5 M components/clm/test/tools/nl_files/PTCLM_USUMB_Cycle_clm4_5 M components/clm/test/tools/nl_files/PTCLM_USUMB_Global_clm4_5 ========= No longer need init_interp_fill_missing_with_natveg (I think this could have been removed a while ago) M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KSinkMOut/user_nl_clm ========= New expected failure: ERS_Lm20_Mmpi-serial.1x1_smallvilleIA.I2000Clm50BgcCropGs.yellowstone_pgi.clm-monthly M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_18_r269 Originator(s): erik (Erik Kluzek,UCAR/TSS,303-497-1326) Date: Sat Dec 16 19:49:41 MST 2017 One-line Summary: Move externals to version in github Purpose of changes ------------------ Move the CLM externals to the version in github for MOSART, RTM and PTCLM. For PTCLM this meant updating to a version that didn't take advantage of subversion keywords. I also updated to a version that can properly compare to the files it keeps even with differences of user, machine, path, or date. Bugs fixed or introduced None ------------------------ Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self,andre Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - OK regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_18_r268 Answer changes -------------- Changes answers relative to baseline: no bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): rtm, mosart, PTCLM mosart1_0_28 to github version rtm1_0_63 to github version PTCLM to PTCLM2_171216c and github version List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: None =============================================================== =============================================================== Tag name: clm4_5_18_r268 Originator(s): erik (Erik Kluzek) Date: Mon Dec 11 16:47:24 MST 2017 One-line Summary: Fix calculation of stomatal resistence used in dry-deposition, increase threshold of totvegc for soil decomposition to 0.1 Purpose of changes ------------------ Changes from Louisa Emmons to fix the calculation of stomatal resistence in the dry-deposition code so it's the inverse of the previous calculation (and the arbitrary multiplicitive factor to reduce it was removed). Add new optional history fields for dry-deposition. Increase the threshold of total vegetation carbon for when decomposition pools are reset from 1.e-8 to 0.1. This change comes from Dave Lawrence and Keith Oleson. The code was also modified to use better names for this rather than ccrit and ncrit which came in clm4_5_17_r264. Add some new checking for these settings at initialization. Add some changes to the default CMIP6 output from Keith Oleson. Add in f05 fsurdat and one landuse.timeseries datasets. Add check in SSP test that makes sure the compset used includes MOSART as expected. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2545, 2546 2526 -- Fix calculation of stomatal resistance for dry deposition 2545 -- SSP test assumes that MOSART is being used, and dies inexplicably if not Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Add new optional history fields when drydeposition is on: DRYDEPV_*, RS_DRYDEP_O3 Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files) Add f05 fsurdat and one landuse.timeseries file: Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): RS_DRYDEP_O3 history field is hardcoded for Ozone. In most cases the same value would apply to other species, but it is hardcoded to give values for Ozone. Changes to tests or testing: Add ciso SSP test Code reviewed by: self,tilmes,emmons,fvitt Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_18_r267 Answer changes -------------- Changes answers relative to baseline: Yes, but only for drydep and/or exit_spinup Summarize any changes to answers, i.e., - what code configurations: with drydep on, or in exit_spinup mode for either clm4_5 or clm5_0 - what platforms/compilers: All - nature of change: modest Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: Add new ciso_rtmColdSSP testmod -- includes ciso and rtmColdSSP testmods A /clm/cime_config/testdefs/testmods_dirs/clm/ciso_rtmColdSSP/include_user_mods List all existing files that have been modified, and describe the changes: M components/clm/bld/unit_testers/build-namelist_test.pl -- Increase test numbers as now testing f05 datasets M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ---- Add f05 fsurdat files and one landuse.timeseries file M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- Add f05 to list of supported resolutions M components/clm/cime_config/testdefs/testlist_clm.xml --------------- Add SSP ciso test M components/clm/cime_config/usermods_dirs/cmip6_output/user_nl_clm -- Added FCTR, FCEV, FGEV, FIRE, FSR, and HTOP; and removed QVEGT; all from the h1 file (fincl2) (from Keith Oleson) M /clm/cime_config/SystemTests/ssp.py ---------------------- Update comments, add expect that compset uses mosart M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 --- Change ncrit to totvegcthresh and SetNCrit to SetTotVgCThresh add abort if totvegcthresh is unset, and check that it's greater than zero M components/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 - Set totvegcthresh to 0.1 and pass down via setTotVgCThresh methods M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 ----- Change ccrit to totvegcthresh and SetNCrit to SetTotVgCThresh add abort if totvegcthresh is unset, and check that it's greater than zero M components/clm/src/biogeochem/DryDepVelocity.F90 ------------------------ Modifications from Louisa Emmons, Corrected the equation calculating stomatal resistance from rssun and rssha, add new array rs_drydep_patch, add new history output possibilities: DRYDEPV_*, RS_DRYDEP_O3, use spval instead of hardcoded magic number 1.e36 =============================================================== =============================================================== Tag name: clm4_5_18_r267 Originator(s): sacks (Bill Sacks) Date: Thu Dec 7 19:26:48 MST 2017 One-line Summary: Fixes to accumulator fields Purpose of changes ------------------ (1) Only accumulate accumulators over active (typically, non-zero-weight) patches/columns. This changes answers for transient runs. Previously, accumulator fields were being accumulated even when a patch/column was inactive. In many cases this was relatively harmless, but in some cases it was accumulating garbage like spval. Then, once the point became active, it could take a while for the garbage accumulated values to be worked out. This changes answers for transient runs. I have examined all accumulator fields used in a typical BgcCrop run (this does NOT include fields that are only used for CNDV or FATES). Most changes only last for the first month or so that a point is active, but there are larger changes for: - PREC10 and PREC60: The changes in these are probably the most extreme. In the old code, these started as spval (rather than a blank slate) over newly-active patches. PREC10 is garbage for the first few years after a point becomes active. I didn't look at PREC60, but its logic is the same as PREC10, so I would expect it to be garbage for about the first 20 years after a point becomes active. PREC10 is used in the fire code and for stress deciduous phenology. PREC60 is used in the fire code. - AnnET: In the old code, this started as 0 (rather than a blank slate) over newly-active points. This affects answers for the first year or two. This is used in N fixation code. - FSUN24 and FSUN240: In the old code, these started out as spval; it takes a few years for the FSUN240 accumulator to come back down to reasonable values (even though it's a 10-day accumulator, its "exponential running average" behavior means that it takes much longer for memory of a huge value like spval to disappear). These are just used in VOCEmissionsMod. I have not investigated how much these changes in accumulators affect the science of the model. The biggest effects will probably be in the fire code in transient runs, but I haven't looked at how big of an effect these errors in PREC60 have. There are also potentially very large changes in CNDV, due to change in behavior of some patch-level accumulators which are needed for PFT establishment (and thus should be available over inactive patches, but no longer are valid for inactive patches). However, these fields were already buggy (see bug 2537). (2) For startup/hybrid runs with non-blank finidat (including runs with use_init_interp): Continue accumulating continuously from where things were in the finidat file. Previously, many of the accumulators were being reset in startup/hybrid runs. This appeared to be an accident of the implementation details, whose fix was easy as a side-effect of the fix for (1). This shouldn't have a huge effect: most accumulators are order 10 days, with the longest being 365 days, so differences should work themselves out within the first year or so. But this could be important if you expect a spinup run to transition continuously into a transient run (for example). This changes answers for any startup or hybrid case with non-blank finidat, where the restart file is from this tag or later. It does NOT change answers for old restart files. (These old files are missing the new _NSTEPS fields associated with each accumulator field, and so nsteps values are set to 0 in a startup/hybrid run, similarly to the old behavior.) Also, a somewhat-unrelated change: Abort if a field is missing from the restart file in a branch run (just as we do for a continue run); tell the user to do a startup or hybrid run in this case. My reason for folding this change into this tag was so that I don't need to worry about backwards compatibility of branch runs with old restart files. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2526: Some CLM accumulated fields (via accumulMod) are incorrect with dynamic landunits Known bugs introduced in this tag (include bugzilla ID): - The following pre-existing bugs may be somewhat worse now: - 2537: Climate accumulator fields incorrect for CNDV - 2539: CNDV logic for agdd20 and tmomin20 needs to be revised to account for when a given patch/column becomes active Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Can no longer do a branch run if some fields are missing from the restart file, either due to code changes or configuration changes (need to do hybrid or startup). - Cannot do a branch from restart files generated with prior tags, because the _NSTEPS restart fields associated with accumulation fields are missing. - Startup or hybrid runs using restart files that were generated from prior tags will not take advantage of fix (2) above. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - not run unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - not run regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi --- ok yellowstone_gnu --- ok cheyenne_intel ---- ok cheyenne_gnu ------ ok hobart_nag -------- ok hobart_pgi -------- ok hobart_intel ------ ok ok means tests pass, answer changes as expected (see below) CLM tag used for the baseline comparisons: clm4_5_17_r266 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: - Cases with transient vegetation (Hist, Dv) - Startup / hybrid cases using restart files generated from this code base. In the test suite, this just shows up as answer changes in ERI and SSP tests, since other tests use old initial conditions files. - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff. Not yet investigated as to whether this is the same or different climate. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Cover accumulMod with unit tests A components/clm/src/main/test/accumul_test/test_accumul.pf A components/clm/src/main/test/accumul_test/CMakeLists.txt A components/clm/src/main/test/accumul_test List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/src/main/accumulMod.F90 M components/clm/src/biogeophys/TemperatureType.F90 ========= Change PREC365 to col-level: with these changes, when it was at patch-level, it would be invalid over not-yet-active patches - yet it's needed there for the sake of PFT establishment in CNDV. Also initialize pbot to spval now that bug 2526 is resolved. M components/clm/src/main/atm2lndType.F90 M components/clm/src/biogeochem/CNDVEstablishmentMod.F90 ========= Need 'active' flag in gridcell_type (always true), so that gridcell_type has the same interface as other subgrid types in this respect M components/clm/src/main/GridcellType.F90 ========= Abort if a field is missing from the restart file in a branch run (see above for details) M components/clm/src/utils/restUtilMod.F90.in ========= Cover accumulMod with unit tests M components/clm/src/main/test/CMakeLists.txt ========= Changes needed so that code under test can call get_proc_bounds M components/clm/src/unit_test_shr/unittestSubgridMod.F90 =============================================================== =============================================================== Tag name: clm4_5_17_r266 Originator(s): sacks (Bill Sacks) Date: Mon Nov 27 19:43:39 MST 2017 One-line Summary: Avoid negative snow densities Purpose of changes ------------------ For very low temperatures, fresh snow densities were negative. This was only happening at extremely low temperatures (below -100 deg C), but caused snow balance errors when it occurred. In exploring the LoTmpDnsSlater2017 parameterization in NewSnowBulkDensity, I came away with the following thoughts: It seems like there should be a limit on the temperature-dependent bifall value. I extracted this function into python and played with it for a few minutes: else if (new_snow_density == LoTmpDnsSlater2017) then ! Andrew Slater: A temp of about -15C gives the nicest ! "blower" powder, but as you get colder the flake size decreases so ! density goes up. e.g. the smaller snow crystals from the Arctic and Antarctic ! winters bifall(c) = -(50._r8/15._r8 + 0.0333_r8*15_r8)*(forc_t(c)-tfrz) - 0.0333_r8*(forc_t(c)-tfrz)**2 end if It looks like the intention is to have snow densities that increase with decreasing temperature (below -15 C), but the effect of this parabolic function is to have densities start coming back down as temperature decreases too low (below about -58 C). My sense is that we should do one of the following in the above block of code: (1) Limit bifall(c) to be no less than 50... my read of this code is that temperature-dependent bifall is never meant to go below 50. (2) For this cold-temperature-based bifall, use an "effective" forc_t that is never colder than about -58 C – or the point at which this function begins to decrease with further-decreasing temperatures. (We could determine the exact temperature at which this function turns over... -58 C is just an approximation.) I'm not positive that this is what's intended, but my read of the comment suggests that densities were never intended to start decreasing with further-cooling temperatures. See comments and figures in http://bugs.cgd.ucar.edu/show_bug.cgi?id=2536 for more details. Leo van Kampenhout and Jan Lenaerts support limitation option 2, which is my preference as well. Thus, this tag implements that limitation. This change changes answers for (downscaled) atmospheric forcing temperatures below -57.55 deg C. These very low temperatures are rare, but they do sometimes occur, both in coupled runs and in offline runs. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2536: Snow balance error Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Leo van Kampenhout Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - not run unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - not run regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi --- ok yellowstone_gnu --- ok cheyenne_intel ---- ok cheyenne_gnu ------ ok hobart_nag -------- ok hobart_pgi -------- ok hobart_intel ------ ok ok means tests pass, answers change as expected CLM tag used for the baseline comparisons: clm4_5_17_r265 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff/same climate Many (but not all) tests show answer changes. Answers should only change for (downscaled) forcing temperatures below -57.55 deg C. I spot-checked one test with answer changes (SMS_D_Ld3.f10_f10_musgs.I1850Clm50BgcCrop.cheyenne_intel.clm-default). I checked H2OSNO and FSH in this test, in the last day's h0 file; these fields differ (relative to the baseline) in only one gridcell. TBOT was about -55 deg C in the daily average of the last day, so it seems reasonable that this forcing temperature was below -57.55 deg C at some point in the 3-day simulation. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_newSnowBulkDensity.pf =============================================================== =============================================================== Tag name: clm4_5_17_r265 Originator(s): erik (Erik Kluzek) Date: Sat Nov 18 12:40:20 MST 2017 One-line Summary: Update the clm50 parameter file for better behavior with prognostic crop, and fix a bug in clm50 urban model Purpose of changes ------------------ Update a few parameters to tweak prognostic crop harvest yields, adjusting FUN costs, and changing C4 crops and wheat by Keith Oleson, Dave Lawrence, and Danica Lombardozzi. Danica had this to say: Overall, the wheat parameters work well (give us correct Fertilized - Unfertilized behavior and reasonable crop yields), but the C4 crop parameters are less satisfying. The 2A parameter changes still have trouble with the fertilized - unfertilized behavior in some regions and lower yields, but seems to be the best combined set of behaviors since there seems to be a direct tradeoff between the fertilized behavior and crop yields Also a fix to urban initialization of rootfr for pervious road. It had used a magic number to assuming nlevsoi was 10. Now we divide by the actual nlevsoi. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2434 2434 -- urban assumed to have 10 soil layers Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): clm50 params file Changes to the datasets (e.g., parameter, surface or initial files): clm50 parameter file Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self,oleson Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_17_r264 Answer changes -------------- Changes answers relative to baseline: YES, for CLM50 Summarize any changes to answers, i.e., - what code configurations: all clm50 (except non-urban and non-crop) - what platforms/compilers: all - nature of change: crop behavior changes, urban working better Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- new params file M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 ----- Divide by nlevsoi rather than assume nlevsoi is 10 with a magic number. =============================================================== =============================================================== Tag name: clm4_5_17_r264 Originator(s): erik (Erik Kluzek) Date: Thu Nov 16 13:35:59 MST 2017 One-line Summary: Changes from Dave Lawrenece on resetting soil carbon for spinup Purpose of changes ------------------ Changes from Dave Lawrence so that total vegetation carbon (totvegc) is used for resetting soil carbon stocks during the AD spinup exit (to determine if you should reset both carbon and nitrogen). Also read in totvegc for C13, and C14. These are then passed down into soilbiogechemistry nitrogenstate and carbonstate restart to determine if the decomposition pools should be reset. The resetting is only done over non-crop vegetation (ignoring both prognostic and generic crop types) and is only done on the exit_spinup phase going out of accellerated decomposition mode. Fix a couple small issues that in principle could change answers, but we didn't see answer changes in testing. There was an error in how rootfr was initialized if it wasn't on the restart file. Made sure the crop patch level restart variables are read after the enter_spinup code. Bugs fixed or introduced ------------------------ 2406 -- Crop patch-level restart variables read after enter_spinup code 2426 -- Error in rootfr for SoilStateType restart 2447 -- dayl_factor not just exactly zero, but small Bugs fixed (include bugzilla ID): Known bugs introduced in this tag (include bugzilla ID): Following found 2534 -- rootfr_road_perv calculation now incorrect for CLM50 2533 -- c13/c14 isotope time-series is awkward for 1850 control (convert to streams file) 2531 -- FATES doesn't call SoilBiogeochemCompetitionInit Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): SetNCrit in SoilBiogeochemNitrogenStateType.F90, should be SetCCrit, because totvegc is used rather than totvegn. SoilBiogeochemPrecisionControlInit was added as a new method that could be extended to set ncrit and ccrit from namelist, or to set them from the settings for CNPrecisionControl that are read in by namelist Changes to tests or testing: None Code reviewed by: self,dlawren Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular (without testing on cheyenne when tag was made) [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- Ran on yellowstone OK cheyenne_gnu ------ Ran on yellowstone OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_17_r263 Answer changes -------------- Changes answers relative to baseline: Yes, but only for spinup! Summarize any changes to answers, i.e., - what code configurations: Only during spinup (at exit_spinup stage) - what platforms/compilers: All - nature of change: minor for spinup Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNVegCarbonStateType.F90 --- Add totvegc_col (and cor c13/c14) to restart M components/clm/src/biogeochem/CNVegetationFacade.F90 ----- Add get_totvegc_col to get the column level totvegc M components/clm/src/main/clm_initializeMod.F90 ------------ Remove use of CNDriverInit, not needed M components/clm/src/main/clm_instMod.F90 ------------------ Add call to SoilBiogeochemCompetitionInit do vegetation restart before soil bgc restart, pass totvegc down to soilbiogeochem_carbonstate restart (and nitrogenstate) M components/clm/src/biogeophys/PhotosynthesisMod.F90 ------ Fix bug 2447 checking for small dayl_factor and not just exactly zero M components/clm/src/biogeophys/SoilStateType.F90 ---------- Fix bug 2446 initialize rootfr only if rootfr not read in M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 ---- Add SetNCrit to set new module variable ncrit pass totvegc_col into restart, if it is lower than ncrit (and not crop) reset decomposition pools to zero M components/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 -- Add SoilBiogeochemPrecisionControlInit, to set ncrit and ccrit now module variables and pass them down to carbonstate and nitrogenstate M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 ------ Add SetCCrit to set new module variable ccrit pass totvegc_col into restart, if it is lower than ccrit (and not crop) reset decomposition pools to zero save restart_file_spinup_state and read it in with C12 =============================================================== =============================================================== Tag name: clm4_5_17_r263 Originator(s): erik (Erik Kluzek) Date: Wed Nov 8 15:36:41 MST 2017 One-line Summary: Drydep vel. for SO2 doubled, influence of organic mater in soil calmed, max daylength not hardwired to present day Purpose of changes ------------------ Keep the hydraulic conductivity for organic soil higher than it is for mineral soil. This was found to be an issue and fixed by Sean Swenson, Dave Lawrence, and Keith Oleson. Double dry deposition velocity for SO2 (Jean-Francois Lamaruque) for both clm4_5 and clm5_0. Fix bug for value of NACTIVE at first time-step. Partial fix so that max daylengh isn't hardwired to present day at least at initialization (see caveat below). Move the sample files for the mkprocdata_map tool (for mapping between unstructured grids and regular lat lon grids) out of CLM and into $CSMDATA. Update the c13/c14 timeseries files to the 2.0 version as per Keith Lindsay. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2530 -- organic soil influence varies wildly 2304 -- NACTIVE at first time step 1843 -- Max daylength now uses orbital parameters at initialization rather than hardwired to present day (partial) Known bugs introduced in this tag: 2529 -- SSP case that runs on cheyenne fails on hobart Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): c13/c14 time-series files Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Max daylength will still be wrong in cases where variable_years is used, where the orbit changes after initializaiton. It'll start out correct and begin to diverge each year. Since, this feature is untested it's not worth going through the additional changes to correct it during the run. At the time the tag was made, cheyenne was down, so there are six tests that weren't confirmed to pass: ERS_D_Ld3.f19_f19_mg16.IHistClm40SpCruGs.cheyenne_intel.clm-40default ERS_Ly5_P60x1.f10_f10_musgs.IHistClm50BgcCrop.cheyenne_intel.clm-cropMonthOutput SMS_D_Ld5.f10_f10_musgs.I2000Clm50Fates.cheyenne_intel.clm-fates SMS_D_Ly2.1x1_numaIA.IHistClm50BgcCropGs.cheyenne_intel.clm-ciso_bombspike1963 SMS_Ld1.f19_f19_mg16.IHistClm40SpCruGs.cheyenne_intel.clm-40default SMS_Ld1.f19_g17.I2000Clm50Vic.cheyenne_intel.clm-default Changes to tests or testing: tools testing modified to point to files in CSMDATA Code reviewed by: self,lamar,oleson,sacks,dlawren Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular, tools [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_16_r262 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: clm4_5 and clm5_0 - what platforms/compilers: all - nature of change: changes answers with bug fixes in a significant way If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: oleson/csm/clm50_r243_1deg_GSWP3V2_cropopt_nsc_hksat_hist Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: Remove sample files, move to CSMDATA, remove sample namelist D components/clm/tools/mkprocdata_map/clm4054_f19g16_I2000.clm2.h0.2000-01_c170430.nc D components/clm/tools/mkprocdata_map/map_ne30np4_nomask_to_fv1.9x2.5_nomask_aave_da_c121107.nc D components/clm/tools/mkprocdata_map/clm4054_ne30g16_I2000.clm2.h0.2000-01_c170430.nc D components/clm/tools/mkprocdata_map/mkprocdata_map_in List all files added and what they do: A components/clm/tools/mkprocdata_map/README.filedescriptions -- Describe the files in this directory List all existing files that have been modified, and describe the changes: M components/clm/test/tools/nl_files/mkprocdata_ne30_to_f19_I2000 - Point to files in CSMDATA rather than mkprocdata_map dir. M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- Point to new v2.0 of the c13/c14 timeseries files M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 -- Fix bug 2530 for organic mater influence in soil M components/clm/src/biogeochem/DryDepVelocity.F90 ------------- Double SO2 dry deposition rate M components/clm/src/biogeochem/CNFUNMod.F90 ------------------- Bug, 2304 for NACTIVE on history for first time-step M components/clm/src/biogeochem/ch4Mod.F90 --------------------- Bug, 2528, floating overflow for methane (doesn't change answers) M components/clm/src/main/clm_initializeMod.F90 ---------------- Bug, 1843, max daylength no longer hardwired to present day at initialization =============================================================== =============================================================== Tag name: clm4_5_16_r262 Originator(s): sacks (Bill Sacks) Date: Fri Oct 27 10:17:54 MDT 2017 One-line Summary: Rename atm2lnd history fields for downscaled fields, properly turn on vic for clm45, and other minor fixes Purpose of changes ------------------ (1) Rename atm2lnd history fields for downscaled fields. The downscaled version is now the default. Non-downscaled history fields are now noted explicitly with a "_FROM_ATM" suffix. A detailed list of changes is given below. (2) Properly turn on VIC for CLM45 compsets (fixes bug 2465) (3) Use correct PE layouts for CISM1 (4) Remove CWDC_HR diagnostic field, which was always 0 (fixes bug 2502) (5) Turn on U10_DUST diagnostic field by default (fixes bug 2246) (6) Fix documentation of some variables Detailed list of changes related to (1): Rain and snow: Old: - RAIN: "atmospheric rain" - NOT downscaled - Rainf: "atmospheric rain" - NOT downscaled - SNOW: "atmospheric snow" - NOT downscaled - RAIN_REPARTITIONED: "atmospheric rain, after rain/snow repartitioning based on temperature" - DOWNSCALED - SNOW_REPARTITIONED: "atmospheric snow, after rain/snow repartitioning based on temperature" - DOWNSCALED New: - RAIN: "atmospheric rain, after rain/snow repartitioning based on temperature" - DOWNSCALED - Rainf: "atmospheric rain, after rain/snow repartitioning based on temperature" - DOWNSCALED - SNOW: "atmospheric snow, after rain/snow repartitioning based on temperature" - DOWNSCALED - RAIN_FROM_ATM: "atmospheric rain received from atmosphere (pre-repartitioning)" - NOT downscaled - SNOW_FROM_ATM: "atmospheric snow received from atmosphere (pre-repartitioning)" - NOT downscaled Air temperature: Old: - TBOT: "atmospheric air temperature" - NOT downscaled - Tair: "atmospheric air temperature" - NOT downscaled - Tair_downscaled: "atmospheric air temperature downscaled to columns" - DOWNSCALED New: - TBOT: "atmospheric air temperature (downscaled to columns in glacier regions)" - DOWNSCALED - Tair: "atmospheric air temperature (downscaled to columns in glacier regions)" - DOWNSCALED - Tair_from_atm: "atmospheric air temperature received from atmosphere (pre-downscaling)" - NOT downscaled Atmospheric pressure: Old: - PBOT: "atmospheric pressure" - NOT downscaled - Psurf: "surface pressure" - NOT downscaled New: - PBOT: "atmospheric pressure at surface (downscaled to columns in glacier regions)" - DOWNSCALED - Psurf: "atmospheric pressure at surface (downscaled to columns in glacier regions)" - DOWNSCALED Atmospheric longwave radiation: Old: - FLDS: "atmospheric longwave radiation" - NOT downscaled - LWdown: "atmospheric longwave radiation" - NOT downscaled New: - FLDS: "atmospheric longwave radiation (downscaled to columns in glacier regions)" - DOWNSCALED - LWdown: "atmospheric longwave radiation (downscaled to columns in glacier regions)" - DOWNSCALED Potential temperature: Old: - THBOT: "atmospheric air potential temperature" - NOT downscaled New - THBOT: "atmospheric air potential temperature (downscaled to columns in glacier regions)" - DOWNSCALED Specific humidity: Old: - QBOT: "atmospheric specific humidity" - NOT downscaled - Qair: "atmospheric specific humidity" - NOT downscaled New: - QBOT: "atmospheric specific humidity (downscaled to columns in glacier regions)" - DOWNSCALED - Qair: "atmospheric specific humidity (downscaled to columns in glacier regions)" - DOWNSCALED Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2192: wrong units in comment for variables qflx_ev_snow, qflx_ev_soil, qflx_ev_h2osfc - 2246: Turn on U10_DUST history field variable by default - 2363: inline documentation is wrong in SoilTemperatureMod.F90 - 2401: Need to put in an "override" for PE layouts for CISM1 - 2465: ICLM45VIC compset doesn't actually turn VIC on - 2502: CWDC_HR is identically zero - 2503: Wrong units for FAREA_BURNED Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Changed meaning of some atm2lnd history fields: see above Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self; Sean Swenson did a conceptual review of the changes in atm2lnd diagnostic fields Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - not run unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - not run regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi --- ok yellowstone_gnu --- ok cheyenne_intel ---- ok cheyenne_gnu ------ ok hobart_nag -------- ok hobart_pgi -------- ok hobart_intel ------ ok ok means tests pass, answers change for some atm2lnd diagnostic fields as noted below CLM tag used for the baseline comparisons: clm4_5_16_r261 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): - For most configurations: Changes diagnostic fields only (see below for list) - Changes CLM45 compsets with VIC: actually turns on VIC (there are no CLM45 tests with VIC in the test suite) Changes the following fields: FLDS, PBOT, QBOT, RAIN, SNOW, TBOT, THBOT: These now give downscaled versions of these fields. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Rename atm2lnd history fields for downscaled fields M components/clm/src/main/atm2lndType.F90 ========= Properly turn on VIC for CLM45 compsets M components/clm/cime_config/config_component.xml ========= Use correct PE layouts for cism1 M components/clm/cime_config/config_pes.xml ========= Remove CWDC_HR diagnostic field M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 ========= Turn on U10_DUST by default M components/clm/src/biogeophys/FrictionVelocityMod.F90 ========= Fix various documentation M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/src/biogeophys/BareGroundFluxesMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/SoilFluxesMod.F90 M components/clm/src/biogeophys/CanopyFluxesMod.F90 M components/clm/src/biogeophys/SoilHydrologyMod.F90 M components/clm/src/biogeochem/CNVegStateType.F90 =============================================================== =============================================================== Tag name: clm4_5_16_r261 Originator(s): sacks (Bill Sacks) Date: Wed Oct 25 07:37:24 MDT 2017 One-line Summary: Add option to reset snow over glacier columns Purpose of changes ------------------ We have had an option to reset snow over non-glacier columns, using the reset_snow namelist variable. This is useful when transitioning from a spinup with one set of atmospheric forcings to a run with substantially different atmospheric forcings. We originally did not provide an option to reset snow over glacier columns, because we couldn't find a satisfactory way to determine which glacier columns should and should not be reset. However, members of the LIWG have found that it's hard to analyze some test runs without this option. This tag, therefore, provides an option to reset snow over glacier columns, using a new namelist flag reset_snow_glc. You can optionally provide an elevation threshold for this resetting: By setting reset_snow_glc_ela to some elevation (in meters), only glacier columns below this elevation get their snow packs reset. Note that you run the risk of baking in answers with this (how much of the resulting spatial patterns are due to your choice of where to reset snow?), but this can still be useful in some cases. In general, this is meant for testing, and not for "real" scientific production runs. Original changes were from Leo van Kampenhout, with some refactoring by Bill Sacks. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): - 2525: test_build_namelist.pl fails in clm4_5_16_r260 (this bug existed prior to this tag, but was first discovered here) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variables: - reset_snow_glc: logical; if set to .true., resets the snow pack over glacier columns - reset_snow_glc_ela: elevation threshold for resetting snow pack over glacier columns Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - reseedresetsnow testmod turns on snow resetting over glacier points, and so changes answers. It uses the reset_snow_glc_ela threshold that Leo has been using. Code reviewed by: self, Leo van Kampenhout Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - ok (run on yellowstone) Tests in build-namelist_test.pl pass but change answers as expected due to new namelist options. NOTE: prove test_build_namelist.pl fails (see bug 2525) unit-tests (components/clm/src): cheyenne - pass tools-tests (components/clm/test/tools): cheyenne - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - not run regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi --- ok yellowstone_gnu --- ok cheyenne_intel ---- ok cheyenne_gnu ------ ok hobart_nag -------- ok hobart_pgi -------- ok hobart_intel ------ ok ok means tests pass, NLCOMP failures as expected. Also, expected baseline failure: ERS_D.f19_g17.I1850Clm50BgcCrop.cheyenne_intel.clm-reseedresetsnow However, there were also baseline failures in these tests, due to problems with r260: (1) SMS_Ld5.f19_g17.IHistClm50Bgc.cheyenne_intel.clm-decStart SMS_Ld5.f19_g17.IHistClm50Bgc.yellowstone_pgi.clm-decStart SMS_Ld5.f19_g17.IHistClm50Bgc.yellowstone_intel.clm-decStart These are due to problems with the baseline, in terms of which cism history files are present. Baseline comparisons failed in the same way when using out-of-the-box r260. The current code seems correct in this respect, so this is just a problem with the baseline directories. (2) SMS_D_Ly2.1x1_brazil.IHistClm50BgcQianGs.cheyenne_intel.clm-ciso_bombspike1963 Differences in C14 history fields. Baseline comparisons failed in the same way when using out-of-the-box r260. I can't tell if the baselines are correct or the trunk code is correct. I have asked Erik to look into this. CLM tag used for the baseline comparisons: clm4_5_16_r260 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Only tests with reseedresetsnow testmod - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate for tests with this testmod If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Unit tests of new changes M components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_SnowCappingExcess.pf ========= Pass down new arguments M components/clm/src/main/clm_driver.F90 M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 ========= New namelist variables M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Changes to reseedresetsnow testmod to turn on the new option M components/clm/cime_config/testdefs/testmods_dirs/clm/reseedresetsnow/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_16_r260 Originator(s): erik (Erik Kluzek) Date: Tue Oct 24 15:59:41 MDT 2017 One-line Summary: Update paramater file for CLM50 as well as fates, fix a few issues Purpose of changes ------------------ Update the parameter file for CLM5.0, changing: ekc_active,ekn_active,kn_nonmyc,psi50, and stem_leaf. Dave Lawrence, Keith Oleson, Will Wieder, and Rosie Fischer were involved in changing these parameters to increase survivability of BDT Tropical and NDT Boreal and to slightly decrease the over-productivity of BDS boreal. The version of FATES used was updated to one with the same API, but updated science (science_1.3.0 api_1.0.0). There were a couple issues fixed as well. PTCLM was updated to fix a few issues and also allow us to test running with an updated PTCLM user-mods-directory. So the US-UMB testing was updated to point to a freshly created directory. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2524 -- Run with decStart tests 2523 -- Run with cnonly 2522 -- PTCLMmkdata doesn't handle crop on or off well in latest CLM45/CLM50 2521 -- Add warning for when PTCLMmkdata sets MPILIB to mpi-serial by default Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): -crop option becomes a toggle for CLM_BLDNML_OPTS (so -no-crop is valid as well) Same setup for mksurfdata.pl was introduced as well (and for PTCLMmkdata) Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): New parameter file for CLM5.0 Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: USUMB tests updated, tests for PTCLM updated Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular, tools [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (118 compare tests fail) unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - OK (one expected fail) regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_16_r259 Answer changes -------------- Changes answers relative to baseline: Yes for CLM50 and/or with fates Summarize any changes to answers, i.e., - what code configurations: all configurations with CLM50 or CLM45 with Fates - what platforms/compilers: All - nature of change: Improvements in survivability of certain trees Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): PTCLM PTCLM to PTCLM2_171024b List all files eliminated: None List all files added and what they do: Missing file from previous tag A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart/user_nl_cism List all existing files that have been modified, and describe the changes: M components/clm/test/tools/TSMscript_tools.sh --------------------- Update PTCLM version of mapping data being pointed to M components/clm/test/tools/nl_files/mksrfdt_10x15_1850 ------------ Change -no_crop option to -no-crop M components/clm/test/tools/nl_files/mksrfdt_1x1_vancouverCAN_2000 - Change -no_crop option to -no-crop M components/clm/tools/mksurfdata_map/mksurfdata.pl ---------------- Change -no_crop option to -no-crop toggle M components/clm/bld/CLMBuildNamelist.pm --------------------------- Change -crop option to a toggle (so -no-crop is valid) Fix bug 2523 by setting bgc_spinup M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Update parameter file for CLM50 M components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/user_nl_clm ---- Point to c171024 version of PTCLM data M components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/shell_commands - Point to c171024 version of PTCLM data =============================================================== =============================================================== Tag name: clm4_5_16_r259 Originator(s): erik (Erik Kluzek) Date: Tue Oct 17 10:39:36 MDT 2017 One-line Summary: Update to latest cime from cesm2_0_beta07 and config_components version 3 Purpose of changes ------------------ Update CLM to use version 3 for config_components.xml which has additional error checking of compset names. Also update it for mosart, cism and rtm. And cime is updated to the version used in cesm2_0_beta07. There are a few new features brought in with the latest cime that are described below. CLM now requires one and only one modifier for what type it is in the compset name (i.e. only one "%" as in CLM45%BGC or CLM50%SP). Will now abort when CLM build-namelist encounters a warning (after printing information on all warnings). To ignore the warnings and continue add the new "-ignore_warnings" to CLM_BLDNML_OPTS. "-irrig" is now an option to CLM_BLDNML_OPTS only for clm4_0, set "irrigate" namelist item in your user_nl_clm file for clm4_5 or clm5_0. By default build-namelist sets irrigate to true for transient or present-day configurations if crop is on and clm5_0 (unless DV on). Get PTCLMmkdata fully converted over to cheyenne, and create a new set of files for testing it's ability to create cases. Checking for .true. or .false. in CLM build-namelist is now more robust and allows for "F" or "T" as newer FORTRAN versions do. Fix several bugs as described below. PTCLM and CLM tools were both verified to work and pass testing. The PE layout for f10 on hobart was changed to a single-node so that it would compare exactly to previous versions. Fix a bug for NCK tests in RTM and MOSART. Get mksurfdata_map working correctly for f05. Error message when you set finidat=' ' is more clear on what to do. You can also now do a cold-start for a hybrid case prints a warning), but not allowed on a branch case. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2519 -- Trouble setting use_aereoxid_prog=.false./ 2515 -- Setting finidat=' ' gives cryptic error message 2510 -- Mapping files from the 1km grid can only be created on geyser 2509 -- Bad mapping files for f05 resolution 2504 -- hybrid/branch cases don't allow CLM_FORCE_COLDSTART to be on 2494 -- MOSART test NLCOMP fails due to duplicate mosart_inparam in user_nl_mosart 2477 -- Multiinstance mosart with REFCASE does not initialize Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): This includes an update to cime, so some behavior is a little different than before (see section below) Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): CLM now allows one and only one modifier for the type of configuration in the compset name (i.e. CLM50%BGC or CLM45%SP (only one "%" modifier is allowed) ). "-irrig" is now only allowed as an option to CLM_BLDNML_OPTS for clm4_0, for clm4_5/clm5_0 the handling of irrigate is handled inside CLM build-namelist. Remove some no longer needed forcing definitions for compsets (the year range at the beginning of the compset name). Now abort on build-namelist warnings (but adding "-ignore_warnings" option to CLM_BLDNML_OPTS allow it to run without it). Remove compset forcings that aren't being used: 5505, C2R6, C2R8, C2R4, C1, GEOS (start of long compset names) Changes to CIME user interface: - Eleminate --user-compset. Clean up caseroot directory upon failed creation. - Check coupler fields for nan. - New xml variable, readmode, - xmlquery --group has been changed go xmlquery --get-group - New case.qstatus tool. - Autodeletion of broken case dirs. - Add --mail-user and --mail-type to case.submit. - Remove --user-compset from create_newcase, allow compset longname in testname argument to create_test. - New argument to create_clone (--user-mods-dir to also point to user-mods directory that will be used) Changes made to namelist defaults (e.g., changed parameter values): Correct SCRIP grid file for f05 and add mapping files for them Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: clm-decStarts cism history is a bit different (shown as changing answers) Allow longname in tests - Add --pesfile option to create_test. - New --timeout flag for wait_for_tests - CLM now needs to define its own LII and SSP tests (they already were put in place, now it's actually using them) Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular, tools [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (216 compare tests fail comparing to clm4_5_16_r258 tag) unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): cheyenne - OK regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_16_r258 Answer changes -------------- Changes answers relative to baseline: No (bit for bit) clm/decStart tests were changed though Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, cism, rtm, mosart cime to cime5.4.0-alpha.03 cism to cism2_1_40 rtm to rtm1_0_63 mosart to mosart1_0_28 List all files eliminated: Remove -irrig option for clm4_5/clm5_0 tests D components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/shell_commands List all files added and what they do: A components/clm/bld/namelist_files/LogMessages.pm - Log message, warning, exit and abort handling for perl List all existing files that have been modified, and describe the changes: Update to new PTCLM directory for testing PTCLM cases M components/clm/test/tools/nl_files/PTCLM_USUMB_Cycle_clm4_5 -- Change map_gdate M components/clm/test/tools/nl_files/PTCLM_USUMB_Global_clm4_5 - Change map_gdate M components/clm/test/tools/nl_files/PTCLM_USUMB_clm4_5 --------- Change map_gdate M components/clm/test/tools/TSMscript_tools.sh -- Change user_mods directory used for PTCLM case M components/clm/tools/mkmapdata/regridbatch.sh -- Get working on cheyenne M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Working f05 SCRIP grid file M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ----- Remove irrig M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ------- Change irrig to irrigate add mapping files for f05 Set irrigate for use-cases, .true. for use_crop=T, use_cndv=F and not 1850 M components/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M components/clm/bld/namelist_files/use_cases/stdurbpt_pd.xml M components/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_transient.xml M components/clm/bld/namelist_files/use_cases/20thC_transient.xml M components/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/glacierMEC_pd.xml M components/clm/bld/namelist_files/use_cases/2000_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M components/clm/bld/CLMBuildNamelist.pm -- moving warnings, logging, exit, and fatal_error to LogMessages Add -ignore_warnings, -irrig is now only for clm4_0, pass $opts to add_default, not just {test}, use value_is_true for all logical evalutions (extend it to all T or F), warnings will all be written out, and abort at the end (unless -ignore_warnings is used) don't die if clm_usr_name is set and didn't find user datasets, add final_exit on completion M components/clm/bld/unit_testers/build-namelist_test.pl - Add tests for -ignore_warnings option handle -irrig for clm4_5 M components/clm/cime_config/config_compsets.xml -- Add I2000Clm50BgcCruGs compset M components/clm/cime_config/config_pes.xml ------- Set PE count for f10 on hobart to single node M components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/user_nl_clm --------------- turn irrigate on M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm - turn irrigate on M components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/user_nl_clm ------- turn irrigate on M components/clm/cime_config/testdefs/testlist_clm.xml - 1x1_camdenNJ test with stub-GLC: I2000Clm50BgcCruGs M components/clm/cime_config/config_component.xml --- Update to config_component version 3. CLM must have 1 and exactly one modifier, move setting of -irrig for 2000 and transient compsets to inside of CLM build-namelist remove some no longer needed forcing definitions for compsets M components/clm/cime_config/buildnml - Print warning if branch and coldstart, allow colstart with hybrid Add warning if doing a coldstart M components/clm/src/CMakeLists.txt - Add additional subdirectory for unit testing (from Bill Sacks) M components/clm/src/main/controlMod.F90 - Allow use_crop=F and irrigate M parse_cime.cs.status - Seperate out differences when baselines don't exist =============================================================== =============================================================== Tag name: clm4_5_16_r258 Originator(s): erik (Erik Kluzek) Date: Fri Oct 6 22:25:29 MDT 2017 One-line Summary: Revert change to fire, set n_melt_glcmec=10, update mosart to version that doesn't accumulate water in short rivers Purpose of changes ------------------ Revert the changes to fire that came in clm4_5_16_r257. Add some notes that clarify that the code as is, is correct. Set n_melt_glcmec=10.0 for clm5.0 (This should have happened in the fresh snow grain size change in clm4_5_16_r256). Update mosart to a version that doesn't accumulate water in short rivers. This is done by explicitly making sure river length isn't smaller than a minimum value. Bugs fixed or introduced ------------------------ 2516 -- Mistaken bug for fire mortality calculations for some Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): Changed n_melt_glcmec to 10.0d00 for clm5_0 (same as clm4_5, but double precision) Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Some of the fire mortality code is confusing, we added some notes to clarify that's it's correct Changes to tests or testing: None Code reviewed by: self, dlawren, oleson, swenson Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (231 compare tests fail) unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_16_r257 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 and/or with mosart - what platforms/compilers: all - nature of change: small Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): mosart mosart to rlen_min_n01_mosart1_0_26 (will bring trunk version of mosart in an upcoming tag) List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - set M components/clm/src/biogeochem/CNFireBaseMod.F90 --- Revert fire changes (bug 2516) add comments use fm_droot as it's used to represent fraction of transport from livestem/livecroot to litter or to represent the fraction of plant-tissue mortality for deadstem/deadcroot M components/clm/src/biogeochem/CNFireLi2014Mod.F90 - Revert fire changes =============================================================== =============================================================== Tag name: clm4_5_16_r257 Originator(s): erik (Erik Kluzek) Date: Fri Sep 29 10:17:36 MDT 2017 One-line Summary: Fix two bugs found by Hui Tang, one for Carbon isotopes and one in the fire model Purpose of changes ------------------ This is a fix for two bugs found by Hui Tang. One is for Carbon isotopes where a fix that came in for clm4_5_11_r188, used leafc for too many of the isotope flux calculation calls. This changes it so the correct terms are used for the particular flux. The other fix is in the fire model where fire emissions were using the wrong fire related mortality term for dead stem and should have been using a different type. Now it matches the flux type, as well as the term used in combustion completness for the particular fire emission. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2517 -- Bug in CNFireFluxes for some fire fluxes 2516 -- Leafc wrongly used in too many CIsoFluxCalc calls. Known bugs introduced in this tag (include bugzilla ID): (new one found) 2518 -- --- warnings from CLM build-namelist don't get sent out of preview_namelist Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self,oleson,dlawren,Hui Tang Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK CLM tag used for the baseline comparisons: clm4_5_16_r256 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 with CN and fire on, and also for carbon isotopes - what platforms/compilers: all - nature of change: very small impact If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: /oleson/clm50_r255_fireiso_1deg_GSWP3V1_1850 URL for LMWG diagnostics output used to validate new climate: http://webext.cgd.ucar.edu/I1850/clm50_r255_fireiso_1deg_GSWP3V1_1850/lnd/clm50_r255_fireiso_1deg_GSWP3V1_1850_21_50-clm50_r255_coniso_1deg_GSWP3V1_1850_21_50/setsIndex.html Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNFireBaseMod.F90 --- Some fire emissions were using fire related mortality for dead stem and should have been using a different type it should match the one used in the combustion completeness factor and the main type M components/clm/src/biogeochem/CNFireLi2014Mod.F90 - Same change as above M components/clm/src/biogeochem/CNCIsoFluxMod.F90 -- In the fix for bug 2116 that came in with clm4_5_11_r188, leafc was used in too many CNCIsoFluxCalc calls this corrects it to use the correct term =============================================================== =============================================================== Tag name: clm4_5_16_r256 Originator(s): erik (Erik Kluzek) Date: Wed Sep 20 15:42:17 MDT 2017 One-line Summary: Fresh snow grain radius is temperature dependent Purpose of changes ------------------ Changes from Leo VanKampenhout to have fresh snow grain size a function of air temperature for CLM50. Behavior is the same as before for CLM45 with fresh snow grain size a fixed parameter. For CLM50 it's that same value (54.526) less than -30C and linear with T up to 204.526 at 0C, and 204.526 for temperatures greater than 0C. Bugs fixed or introduced ------------------------ Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Added namelist item: fresh_snw_rds_max Changes made to namelist defaults (e.g., changed parameter values): fresh_snw_rds_max is snw_rds_min for clm4_5 fresh_snw_rds_max is 204.526 for clm5_0 Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self,L.vanKampenhout@uu.nl Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (343 compare tests fail since namelists change) unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu - OK cheyenne_intel - OK cheyenne_gnu - OK hobart_nag - OK hobart_pgi - PASS hobart_intel - PASS CLM tag used for the baseline comparisons: clm4_5_16_r255 Answer changes -------------- Changes answers relative to baseline: Yes, for clm5_0 If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm5_0 - what platforms/compilers: All - nature of change: more snow melt at expense of ice melt If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: oleson clm50_r251_SNOWGRAIN_2000 URL for LMWG diagnostics output used to validate new climate: http://webext.cgd.ucar.edu/B1850/b.e20.B1850.f09_g17.pi_control.all.196/lnd/b.e20.B1850.f09_g17.pi_control.all.196.8_27-b.e20.B1850.f09_g17.pi_control.all.195.8_27/set1/set1.html Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SnowSnicarMod.F90 -- Pass atm2lnd_inst in, make sure snw_rds isn't less than the min parameter create a function for setting FreshSnowRadius, dependent on atm2lnd_inst (for air temperature), the function sets to min parameter if fresh_snw_rds_max is same as min (clm45), or to min at less than -30C linear between fresh_snw_rds_max and min parameter up to 0C, and then at fresh_snw_rds_max value for clm50. M components/clm/src/biogeophys/AerosolMod.F90 ----- Add InitReadNML and fresh_snw_rds_max M components/clm/src/main/clm_driver.F90 ----------- Pass atm2lnd_inst down to SnowAgeGrain M components/clm/src/main/clm_instMod.F90 ---------- Pass NLFilename to aerosol_inst%Init M components/clm/bld/CLMBuildNamelist.pm - Set fresh_snw_rds_max, and add aerosol namelist M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Set fresh_snw_rds_max for clm45 and clm50 M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Define fresh_snw_rds_max =============================================================== =============================================================== Tag name: clm4_5_16_r255 Originator(s): erik (Erik Kluzek) Date: Mon Sep 18 21:21:29 MDT 2017 One-line Summary: Changes to crop and enabling Carbon isotopes for crop and some miscellaneous small changes Purpose of changes ------------------ Changes for crop on as well as enabling carbon isotopes when crop is on. Fix some small answer changing issues. Also bring in new timeseries datasets for C14 as well as now C13. The C14 dataset has hardwired latitude sectors it uses (so previous file can NOT be used). Properly handle and update C13/C14 variables when crop is on. Handle CIsoFluxCalc calls for a list of crop specific fields. Add cropseedc_deficit, to storvegc as well as totveg. Send livestemc to litter for prognostic crop. Ignore downreg for carbon isotopes in Fractionation subroutine. For dynamic roots, use rooting values on parameter dataset which is set to values from PFT=1. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2508 -- Incorrect landuse.timeseries datasets for f09 and hcru for rcp=8.5 in r254 2489 -- Set c/nnegcrit for clm45/clm50 (10X higher for clm45 than clm50) 2412 -- Initialize rootfr to PFT==1 value for dynamic roots -- update params files 2396 -- make subtraction the default behavior for clm4_5. 2316 -- update carbon isotopes same as C12: CNVegCarbonStateType.F90 2211 -- Fix time used for Ciso time series Known bugs introduced in this tag (include bugzilla ID): 2511 -- Overflow with Carbon isotopes Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Change to restarts when c-isotopes on Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist items: use_c13_timeseries, atm_c13_filename New history vars: C13_GRAINC, C13_CROPSEEDC_DEFICIT, C14_GRAINC, C14_CROPSEEDC_DEFICIT Changes made to namelist defaults (e.g., changed parameter values): Set cnegcrit/nnegcrit for CLM45/CLM50 Changes to the datasets (e.g., parameter, surface or initial files): New C13 and C14 timeseries files New params files, but only answer changing part is for dynamic roots (setup bare-soil on file to match what's used for PFT=1) Bring in correct datasets for rcp8.5 for f09 and hcru 1x1_numaIA 1850 and 1850-2000 transient files Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): The C13/C14 timeseries should be implemented with streams for greater flexibility, robustness, and code reuse. The new C14 timeseries file is hardcoded for three specific latitude sectors, and thus can't use the previous file, nor a file of a different configuration. Downreg is passed down from get_downreg_patch in clm_driver, to CanopyFluxes to Fractionation, but no longer used (it should be removed). Note the following from Dave: While debugging crop isotope code, found that cpool_patch and frootc_patch could occasionally be very small but nonzero numbers after crop harvest, which persists through to next planting and for reasons that could not 100% isolate, caused C12/C13 ratios to occasionally go out of bounds. Zeroing out these small pools and putting them into the flux to the atmosphere solved many of the crop isotope problems Changes to tests or testing: Added some tests, two of the new tests fail due to bug 2511 Code reviewed by: self, koven, dlawren, dll, oleson Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS (234 comparison tests fail) unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu - OK cheyenne_intel - OK cheyenne_gnu - OK hobart_nag - OK hobart_pgi - OK hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_16_r254 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: Crop, Carbon-isotopes - what platforms/compilers: All - nature of change: small change to crop with missing updates, crop with carbon-isotopes are now enabled Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: D components/clm/src/biogeochem/C14BompbSpikeMod.F90 --- Change name to more general List all files added and what they do: A components/clm/src/biogeochem/CNCIsoAtmTimeSeriesReadMod.F90 -- Handle time series for both C13 and C14 List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNPhenologyMod.F90 -------- Remove subtract_cropseed always do when crop on, pass c13/c14 instances down to CNPhenology and CropPhenology. After setting C12-leafc_xfer/totvegc set the c13/c14 leafc_xfer/totvegc as well M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 ---- Change location of xsmrpool update. Dave added this note: While debugging crop isotope code, found that cpool_patch and frootc_patch could occasionally be very small but nonzero numbers after crop harvest, which persists through to next planting and for reasons that could not 100% isolate, caused C12/C13 ratios to occasionally go out of bounds. Zeroing out these small pools and putting them into the flux to the atmosphere solved many of the crop isotope problems M components/clm/src/biogeochem/CNCIsoFluxMod.F90 --------- Add CIsoFluxCalc calls for: soilc_change, cpool_to_resp, grainc_xfer_to_|grainc|food|seed, crop_seedc_to_leaf, grain_|curmr|xsmr, cpool_|grain|grain_storage|_gr, cpool_to_|grainc|grainc_storage, transfer_grain_gr, grainc_storage_to_xfer, livestemc_to_litter, and if use_grainproduct for ciso-grainc_to_cropprodc/grain_mr pass iso totvegc_path for hrv_xsmrpool_to_atm CIsoFluxCalc, section adding livestemc to litter for crop add additional check for ciso_state(i) /= 0 for divide by check in CIsoFluxCalc M components/clm/src/biogeochem/CNPrecisionControlMod.F90 - Handle c13/c14 TruncateCandNStates for grainc/storage/xfer, cropseedc_deficit_patch and xsmrpool M components/clm/src/biogeochem/CNVegCarbonStateType.F90 -- Add: C13_GRAINC, C13_CROPSEEDC_DEFICIT, C14_GRAINC, C14_CROPSEEDC_DEFICIT restartvar calls explicit for c12, c13, c14: grainc, grainc_storage, grainc_xfer (with c13 and c14 for ciso) add cropseedc_deficit, to storvegc as well as totvegc, fix bug 2316 make cisotopes the same as C12 M components/clm/src/biogeochem/CNDriverMod.F90 ----------- Pass more c13/c14 instances down to CNPhenology calls M components/clm/src/main/controlMod.F90 ------------------ Add namelist items: use_c13_timeseries, atm_c13_filename remove abort when crop and use_c13/use_c14 is on M components/clm/src/main/clm_initializeMod.F90 ----------- Add call to C13_init_TimeSeries M components/clm/src/biogeophys/PhotosynthesisMod.F90 ----- Handle C13 timeseries (call C13TimeSeries), put C14 bombspike data in hard-wired latitude sectors, ignore downreg for carbon isotopes in Fractionation subroutine M components/clm/src/biogeophys/RootBiophysMod.F90 -------- Use parameters for roots from params file for bare-soil rather than an explicit if statement for it M components/clm/bld/CLMBuildNamelist.pm ------------------- Add cnegcrit/nnegcrit, allow crop and C-isotopes to be on at the same time, add c13 time-series M components/clm/bld/unit_testers/build-namelist_test.pl --- Change testing for time-series M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ----- New datasets for atm_c14_filename, and atm_c13_filename, new paramdata files, set cnegcrit/nnegcrit, flanduse_timeseries for f09 and hcru @ rcp=8.5 M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml --- Add: use_c13_timeseries, atm_c13_filename update similar documenation on c14 bombspike M components/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml -- Same as changes above M components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_bombspike1963/user_nl_clm - Set use_c13_timeseries=T M components/clm/cime_config/testdefs/testlist_clm.xml - Change two ciso transient tests to include crop, add transient crop ciso test M components/clm/cime_config/testdefs/ExpectedTestFails.xml -- Add expected fails for new ciso overflow bug 2511 =============================================================== =============================================================== Tag name: clm4_5_16_r254 Originator(s): erik (Erik Kluzek) Date: Fri Sep 1 00:09:25 MDT 2017 One-line Summary: Update surface datasets, and point to new CMIP6 population density dataset Purpose of changes ------------------ Update all surface datasets to new versions that have updated soil albedo, and use latest PFTCFT files. Also get mksurdata_map fully working on cheyenne for batch mode. Point to latest CMIP6 population density dataset for historical case, this is with HYDE3.2 data. New fsurdat files for 16pft are with irrigation on. If ask for irrig=F and 16pft won't be able to find any files. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2479 --- Irrigation is occurring in non-prognostic-crop runs Known bugs introduced in this tag (include bugzilla ID): (found ...) 2506 --- ntree used before it's set in pftconMod.F90 2017-08-25 2505 --- Set stream_year_first_urbantv/last for CLM4.5 just for consistency 2017-08-25 2504 --- hybrid/branch cases don't allow CLM_FORCE_COLDSTART to be on 2017-08-18 2503 --- Wrong units for FAREA_BURNED 2017-08-16 2502 --- CWDC_HR is identically zero 2017-08-16 2500 --- ne30 case fails on hobart_nag due to floating overflow Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): -irrig .true. is needed without -crop fsurdat files without crop are all just with irrigation on. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): Datasets Changes to the datasets (e.g., parameter, surface or initial files): new fsurdat, landuse.timeseries, popdens Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: none Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS tools-tests (components/clm/test/tools): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu - OK cheyenne_intel - OK cheyenne_gnu - OK hobart_nag - OK hobart_pgi - OK hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_16_r253 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): similar climate CLM4.5/CLM5.0 surface datasets change with latest soil albedo and small changes form raw PFTCFT files. also an updated population-density dataset used for historical Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/tools/mksurfdata_map/mksurfdata.pl -- Add rundir option, if 16pfts add Irrig to filename M components/clm/tools/mksurfdata_map/Makefile.data -- Get working on cheyenne or yellowstone M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ------- Default resolution and sim_year, set irrig to .true. when crop off, update fsurdat, flanduse_timeseries, stream_fldfilename_popdens M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Update soil-color and PFTCFT files M components/clm/bld/CLMBuildNamelist.pm --- check irrig for fsurdat files, fsurdat with 16pft is with irrig=T make sure set only set irrigate when use_crop on M components/clm/bld/unit_testers/build-namelist_test.pl -- More tests w clm5_0, irrigate=T set without crop M components/clm/bld/unit_testers/NMLTest/CompFiles.pm ---- Add more documentation M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm - Update finidat with updated fsurdat file M components/clm/src/biogeophys/PhotosynthesisMod.F90 -- FPSN_W? variables are default off, and instant averaged =============================================================== =============================================================== Tag name: clm4_5_16_r253 Originator(s): erik (Erik Kluzek) Date: Fri Aug 4 20:23:25 MDT 2017 One-line Summary: Check on reasonable import/export from CLM, check ndep units from input file Purpose of changes ------------------ Do some reasonable physical checks on the import/export state to/from CLM. Checking for NaN's and that solar should be greater or equal to zero, LW down, specific humidity greater than zero. Check the units on the input nitrogen deposition file and if per year divide by seconds per year, leave alone if per seconds, and die if units aren't recognized. Increase the tolerence for warnings for Nitrogen, Carbon, and soil energy balance. Add missing MPI broadcasts for two ndep namelist variables. Fix an LII test that was failing. Change the PE layout for f4x5 grids for yellowstone and cheyenne. Add -die_on_duplicate option to parse_cime.cs.status and have summary option add notes about expected fails. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): [If none, remove this line] 2498 -- missing MPI bcast 2499 -- Update finidat file so LII case will work 2484 -- Check for units of ndep only divide by seconds in a year if units are per year Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] NGEET/fates-clm/#16 -- FATES changes answers at a level of at least roundoff when the PE layout changes NGEET/fates-clm/#15 -- CLM-FATES can't restart when PE layout is change ESMCI/cime/#1793 ----- Run time error for mpi-serial case on cheyenne_intel when created with aux_clm create_testd Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): units for ndep input files can either be g(N)/m2/yr or g(N)/m2/s, will now abort for anything else Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: Added test with CMIP6 ndep file Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne - PASS unit-tests (components/clm/src): cheyenne - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu - OK cheyenne_intel - OK cheyenne_gnu - OK hobart_nag - OK hobart_pgi - OK hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_16_r253 Answer changes -------------- Changes answers relative to baseline: No bit-for-bit (although FATES tests change because of the PE layout change) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: Add test mods for case with CMIP6 ndep forcing that has different units/variable name A + components/clm/cime_config/testdefs/testmods_dirs/clm/clm50CMIP6frc List all existing files that have been modified, and describe the changes: M components/clm/cime_config/config_pes.xml ------------- Update f45 PE layout for yellowstone/cheyenne M components/clm/cime_config/testdefs/testlist_clm.xml -- Add clm50CMIP6frc test M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm - Update finidat file so LII test will pass M components/clm/src/biogeochem/CNBalanceCheckMod.F90 - Lower amount of warnings written out for C and N by order of mag. M components/clm/src/biogeochem/CropType.F90 ---------- Add if-masterproc to write statement M components/clm/src/biogeophys/BalanceCheckMod.F90 --- Lower amount of warnings written out for soil energy balance M components/clm/src/cpl/lnd_import_export.F90 -------- Add checking for physically bad import or export data ensuring no NaN's are exchanged, solar >= 0, LW-dn > 0, q > 0 M components/clm/src/main/ncdio_pio.F90.in ------------ Remove write to log file from all tasks M components/clm/src/main/ndepStreamMod.F90 ----------- Do mpi_bast for ndep_varlist/ndep_taxmode, check units of input ndep variable and divide by seconds in a year if g(N)/m2/yr, but leave alone for g(N)/m2/s otherwise die M parse_cime.cs.status - Add -die_on_duplicate option, and have summarize option make notes about expected fails =============================================================== =============================================================== Tag name: clm4_5_16_r252 Originator(s): erik (Erik Kluzek) Date: Mon Jul 24 16:57:42 MDT 2017 One-line Summary: Update parameter file for some crop albedo issues, and fix a bug in urban albedo for nightime Purpose of changes ------------------ Bring in a new CLM50 parameter file that updates a few fields for prognostic crop with affect on crop albedo (rholnir,taulnir,taulvis,xl). Bring in changes to fix urban albedo for nighttime from Keith Oleson. Rather than set albedo's for the urban canyon all to 1.0, set them to the viewfactors, as that will result in combined albedo of one, whereas the previous way ended up with albedo's greater than one. Here's a description from Keith about the change in urban albedo. " For diffuse radiation, which is independent of solar zenith angle, the road albedos appear to be exactly equal to the road-sky view factor. And the wall albedos are equal to the wall-sky view factor. So the albedos are the albedo of the surface (1.0) times the appropriate view factor. This makes sense. All incident radiation entering the canyon should be reflected to the sky and the contribution of each surface is according to the view factor of each surface to the sky as determined by the height to width ratio. The view factors of the two walls and the road to the sky sums to 1.0. So setting the diffuse albedos to these values when the sun is below the horizon results in a scaled diffuse albedo of 1.0. For direct radiation, it's more complicated because the solar zenith angle is involved when running the radiation model. However, it seems reasonable to assume the same albedos for direct as for diffuse." Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2492 -- Urban scalings can lead to albedo values greater than 1 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files):New params file for CLM50 Updates rholnir,taulnir,taulvis,xl for prognostic crop fields Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self,oleson Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: cheyenne ---- PASS (60 clm50 tests are different as expected) unit-tests (components/clm/src): cheyenne ---- PASS regular tests (aux_clm): (31 Clm45 tests are different than baseline in cpl hist files for: l2x_Sl_avsdr ("dr" and "df" and "av" and "an") as expected) yellowstone_intel - OK yellowstone_pgi --- OK yellowstone_gnu --- OK cheyenne_intel ---- OK cheyenne_gnu ------ OK hobart_nag -------- OK hobart_pgi -------- OK hobart_intel ------ OK Tests: ERS_Ld5_Mmpi-serial.1x1_vancouverCAN.I1PtClm45SpGs.cheyenne_intel.clm-default, ERS_Ld5_Mmpi-serial.1x1_vancouverCAN.I1PtClm45SpGs.yellowstone_intel.clm-default show: FSR, FSRND, and FSRVD different CLM tag used for the baseline comparisons: clm4_5_16_r251 Answer changes -------------- Changes answers relative to baseline: Yes for CLM50, and for CLM45 when coupled to CAM Summarize any changes to answers, i.e., - what code configurations: CLM50 (CLM45 when coupled to CAM and albedos passed to coupler for any case) - what platforms/compilers: all - nature of change: similar climate If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: oleson/b.e20.B1850.f09_g17.pi_control.all.186.alb URL for LMWG diagnostics output used to validate new climate: http://webext.cgd.ucar.edu/B1850/b.e20.B1850.f09_g17.pi_control.all.186.alb/lnd/b.e20.B1850.f09_g17.pi_control.all.186.alb.17_36-b.e20.B1850.f09_g17.pi_control.all.186.17_36/setsIndex.html Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - Updated params file for CLM50 M components/clm/src/biogeophys/UrbanAlbedoMod.F90 --------------- Fix bug 2492. Set albedo's of urban surfaces to view factors, by default and for solar zenith angles below horizon. =============================================================== =============================================================== Tag name: clm4_5_16_r251 Originator(s): erik (Erik Kluzek) Date: Fri Jul 14 01:38:38 MDT 2017 One-line Summary: Update mksurfdata_map for soil depth/color, add new mapping files Purpose of changes ------------------ Bring in new soil color (from Peter Lawrence) dataset for mksurfdata_map. Also bring in a new soil depth dataset that uses a mask for when soil depth > 0. Bring in a version of PTCLM that's working with this cime. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): New mapping files for soil depth for mksurfdata_map New soil color data for mksurfdata_map New soil depth dataset on the map defined by when soil depth > 0 Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: None Changes to tests or testing: None Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: tools [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - PASS cheyenne - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK CLM tag used for the baseline comparisons: none Answer changes -------------- Changes answers relative to baseline: No, bit-for-bit (CLM code wasn't changed) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): PTCLM PTCLM to PTCLM2_170706 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/tools/mkmapdata/regridbatch.sh -- go back to using 8 processors M components/clm/tools/mkmapdata/mkmapdata.sh ---- add mapping for 5x5min_ORNL-Soil, change check for $MPIEXEC, so ignores arguments M components/clm/tools/mksurfdata_map/mksurfdata.pl -- Remove extra "_" in "__CMIP6_" M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- New 5x5min-ORNL-Soil mapping files M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - SCRIP grid file for 5x5min-ORNL-Soil, change soildepth mask to ORNL-Soil update soil color and soil depth datasets M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add ORNL-Soil mask =============================================================== =============================================================== Tag name: clm4_5_16_r250 Originator(s): erik (Erik Kluzek) Date: Thu Jul 13 13:54:13 MDT 2017 One-line Summary: Update finundation dataset, new fsurdat files with updated soil color and soil depth, update mosart areas, fix cheyenne_gnu Purpose of changes ------------------ Update all fsurdat datasets with updated soil color and soil depth. Soil color as modified by Peter Lawrence. Soil depth was modified to use a mask for soil depth>0 which makes the minimum soil depth 3cm (model still has a minimum of 40cm). Also updated area's for the mosart routing file (the previous areas were off by up to about 1%). the update to the area's will influence fluxes. Also fix a problem so that cheyenne_gnu will work (remove pure attribute from two functions). Also bring in the updated finundation inversion from Sean Swenson. This will change the methane output significantly for CLM50BGC cases. The previous inversion file was incorrect and shouldn't be used! The new file is also at a higher resolution 1-degree rather than 2-degree. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2493 (gnu6.3.0 compiler doesn't like "pure" functions that use outside modules, especially public ones) Known bugs introduced in this tag (include bugzilla ID): 2491 (Error in Banddiagonal in latest CESM simulation) (found in Cecile's coupled simulation) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): Just datasets see below Changes to the datasets (e.g., parameter, surface or initial files): Update all fsurdat and the stream_fldfilename_ch4finundated file fsurdat files have soil color and soil depth updated. The finundated file is now at f09 resolution rather than f19, and the previous file was incorrect. Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu - OK cheyenne_intel - OK cheyenne_gnu - OK hobart_nag - OK hobart_pgi - OK hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_16_r249 Answer changes -------------- Changes answers relative to baseline: Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 or with MOSART - what platforms/compilers: all - nature of change: similar climate, methane outputs are substantially changed Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): mosart mosart update to mosart1_0_26 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SnowHydrologyMod.F90 -- Remove pure attribute from OverburdenCompactionVionnet2012 M components/clm/src/biogeophys/IrrigationMod.F90 ----- Remove purse attribute from PointNeedsCheckForIrrig M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- Update all fsurdat files and stream_fldfilename_ch4finundated =============================================================== =============================================================== Tag name: clm4_5_16_r249 Originator(s): sacks (Bill Sacks) Date: Thu Jul 6 14:32:40 MDT 2017 One-line Summary: All new compsets, reworked test lists, and related fixes Purpose of changes ------------------ The changes here were a joint effort between Dave Lawrence, Ben Andre, Erik Kluzek, Mariana Vertenstein and Bill Sacks (1) All new compset aliases, with the following changes: (a) Most compset aliases are now for CLM5, with fewer CLM45 compsets and even fewer CLM40 compsets. (b) Compset aliases now use capitalization of only the first letter of each piece, to avoid the alphabet soup problem. For example, IHistClm50BgcCrop (c) Some other tweaks have been made to compset aliases, such as always listing the time period and Sp (satellite phenology) explicitly (d) Compsets use GSWP3 forcing unless the alias specifies otherwise (e) Compsets use CISM2%NOEVOLVE - and so use glc_mec in CLM - unless the alias has Gs (for GLC - stub) (f) An evolving ice sheet (CISM2%EVOLVE) is now denoted by having a final 'G' in the compset alias (2) Completely reworked test lists: (a) Use new aliases (b) Changed most tests to CLM5 (c) Merged the aux_clm45 and aux_clm40 test categories into aux_clm (d) Removed the 'null' test category, which was used for SBN tests (e) Some minor additional tweaks to the aux_clm test list to ensure appropriate coverage and remove some redundant tests (f) For a few testmods directories, removed some settings that had nothing to do with those testmods (use_dynroot, use_hydrstress) (3) Fixed VIC: (a) Previously, VIC compsets were not actually turning VIC on; now they are (b) Fixed namelist defaults for VIC with CLM50 (4) Updated RTM and MOSART externals: changed their test lists and fixed a MOSART bug (5) Updated CISM external; this changes answers for CISM when ice evolution is turned off, due to changing the lnd -> glc mapping in this case (6) Point to a CIME branch that fixes some bugs Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2473: Change order of tests run with the LII test - 2474: CLM50 CN compsets seem to be set up wrong - 2480: vichydro option doesn't work with CLM50 as it needs bedrock off and lower_boundary_condition=3 (with aquifer layer) CIME Issues fixed (include issue #): I think some issues are fixed here, but I'm not sure which ones Known bugs introduced in this tag (include bugzilla ID): - 2485: CLM's 'ed' testlist still has old-style compsets - 2486: Failure in camdenNJ test in prebeta test list - 2488: Floating overflow in SoilBiogeochemNitrifDenitrifMod.F90 (newly-discovered, but presumably existed in previous tags) - 2489: Clm45BgcCrop case dies due to negative leafn (newly-discovered, but presumably existed in previous tags) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - All new compset aliases Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - All new compset aliases Changes made to namelist defaults (e.g., changed parameter values): - Changed some defaults when running with VIC Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - All new aliases means baseline comparisons all BFAIL Changes to tests or testing: - Massive overhaul - see notes elsewhere in this ChangeLog entry Code reviewed by: This was a joint effort by Dave Lawrence, Ben Andre, Erik Kluzek, Mariana Vertenstein and Bill Sacks. Various pieces were reviewed by different people in this list. Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok pass except for 'compare file.*DNE', presumably due to new tests added unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu - ok cheyenne_intel - ok hobart_nag - ok hobart_pgi - ok hobart_intel - ok Most tests pass; baseline comparisons BFAIL. Bugs opened (and Expected Fails list updated) for remaining failures. Ran most testing on r85975; just ran relevant tests on latest revision with changed compset (changed IHistClm50BgcGs to IHistClm50BgcQianGs) CLM tag used for the baseline comparisons: N/A Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All tests BFAIL due to new compset aliases. Because of this, I'm not completely sure of the answer changes, but I think they are: - Tests with CISM2%NOEVOLVE change answers in fields on the glc grid due to new CISM external - CLM50 tests with Vic change answers because now they're actually using VIC - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate for tests that change answers If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime - rtm - mosart - cism List all files eliminated: ========= No longer need these testmods D components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC D components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/README D components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC D components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/user_nl_datm D components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5 D components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12 List all files added and what they do: ========= Replaces af_bias_v5 A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v7/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v7/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v7/user_nl_datm A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v7 List all existing files that have been modified, and describe the changes: ========= Reworked set of compsets and aliases M components/clm/cime_config/config_compsets.xml ========= Reworked test list as described above M components/clm/cime_config/testdefs/testlist_clm.xml ========= Stop setting glc_snow_persistence_max_days in glcMEC testmods. It made sense to set this for testing when most of our tests used CLM45, because the default snow persistence was 20 years. But for CLM50 the default snow persistence is 0, so it no longer makes sense to set this value specially for testing. M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_long/user_nl_clm ========= Apply CN and VIC for CLM50 M components/clm/cime_config/config_component.xml ========= Use compatible settings for VIC M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/unit_testers/build-namelist_test.pl ========= Fix for new glacier region M components/clm/cime_config/usermods_dirs/cmip6_glaciers/user_nl_clm ========= Make SSP work with mosart instead of rtm M components/clm/cime_config/SystemTests/ssp.py ========= Change order of LII test to support re-creating initial conditions files (rather than requiring a separate SMS_Ln1 test) M components/clm/cime_config/SystemTests/lii.py M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/README ========= New, compatible initial conditions M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm ========= No longer set use_dynroot = .true., use_hydrstress = .false. M components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/user_nl_clm ========= No longer set use_hydrstress = .false. M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm ========= Update ExpectedTestFails list for new tests; new failures are: SMS_Ld5_D_P24x1.f10_f10_musgs.IHistClm50Bgc.hobart_nag.clm-decStart ERP_D_P24x1.f10_f10_musgs.IHistClm50Bgc.hobart_nag.clm-decStart ERP_D.f10_f10_musgs.IHistClm50Bgc.yellowstone_gnu.clm-decStart ERS_Ly5_P72x1.f10_f10_musgs.IHistClm45BgcCrop.cheyenne_intel.clm-cropMonthOutput M components/clm/cime_config/testdefs/ExpectedTestFails.xml ========= Update documentation of system testing M .ChangeLog_template =============================================================== =============================================================== Tag name: clm4_5_16_r248 Originator(s): sacks (Bill Sacks) Date: Wed Jun 28 15:50:13 MDT 2017 One-line Summary: Melt most ice runoff Purpose of changes ------------------ Up until now, excess snow in CLM (accumulated beyond the snow capping limit) has been sent to the ocean as an ice runoff flux. However, this caused problems in some regions, in that it cooled an already-cold ocean, leading to excessive sea ice growth. Keith Lindsay, Marika Holland and others suggested changing CLM so that it sends this flux as liquid rather than ice runoff - having the atmosphere melt the ice rather than having the ocean do so. These changes are implemented here. This has now been tested in many fully-coupled simulations, and it indeed solves the problem of excessive sea ice growth, without having any noticeable detrimental effects on other aspects of the simulation. Specifically, this affects ice runoff from (a) glacier_mec columns outside of Greenland and Antarctica, and (b) all other column types (vegetated, etc.). We still send ice runoff from Greenland and Antarctica to simulate the ice calving flux that occurs in those regions. In addition, we still send an ice runoff flux generated from dynamic landunit adjustments. This new behavior is used by default in CLM50, but not in CLM45. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variable: melt_non_icesheet_ice_runoff - New namelist variable: glacier_region_ice_runoff_behavior Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok tests pass, namelists change as expected unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - ok hobart_nag --- ok hobart_pgi --- ok hobart_intel - ok ok means tests pass, answers change as expected Ran most testing on melt_ice_runoff_n05_clm4_5_16_r247; ran hobart on latest (which included a commit to split long lines) CLM tag used for the baseline comparisons: clm4_5_16_r247 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM50 cases - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate This changes runoff terms (liquid and ice) and the sensible heat flux sent to the atmosphere. In addition, there are limited changes in some methane-related terms; this is presumably due to the dependence of finundated on TWS, and the fact that TWS changes due to the changes in these runoff terms. There are also diffs in x2r_Flrl_irrig for this test (and the similar yellowstone test): ERS_D_Ld10.T31_g37_gl4.IGHISTCLM50.cheyenne_intel.clm-glcMEC_decrease. It makes sense that there would be x2r irrig diffs due to diffs in volr, but I was surprised that this test had any irrigation. It turns out this is due to bug #2479. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: b.e20.B1850.f09_g17.pi_control.all.149 URL for LMWG diagnostics output used to validate new climate: ??? Note that the climate tests were done on an earlier, kludgey implementation. I have verified that the latest implementation is bit-for-bit with that earlier implementation via a 5-year run of compset 1850_DATM%GSWP3_CLM50%BGC-CROP_SICE_SOCN_MOSART_CISM2%NOEVOLVE_SWAV at resolution f09_g16, producing annual-average cpl history files. (For the original, used https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/melt_ice_runoff_orig_tags/melt_ice_runoff_orig_n01_clm4_5_16_r246.) There were differences in a few CLM diagnostic fields, as expected (because I have tweaked some diagnostic fields in the latest version), but the cpl avghist files were bit-for-bit. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes as described above M components/clm/src/main/lnd2atmMod.F90 M components/clm/src/main/lnd2atmType.F90 ========= Changed long name in some runoff terms: removed statements that they do not include QSNWCPICE, because now they *do* include the *melted* portion of QSNWCPICE M components/clm/src/biogeophys/WaterfluxType.F90 ========= New glacier region-specific behavior M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/main/test/glcBehavior_test/test_glcBehavior.pf M components/clm/src/main/clm_driver.F90 ========= New namelist flags controlling the behavior M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm M components/clm/src/main/clm_instMod.F90 =============================================================== =============================================================== Tag name: clm4_5_16_r247 Originator(s): sacks (Bill Sacks) Date: Mon Jun 26 19:52:20 MDT 2017 One-line Summary: New GLACIER_REGION field with CISM domain split in two Purpose of changes ------------------ In an upcoming tag, we'll need a separate GLACIER_REGION for the purpose of determining where ice runoff should be melted: we do not want this melt to happen over Greenland itself, but we do want it to happen in peripheral parts of the CISM domain (and in particular, we want it to happen over the Canadian archipelago, which is within the CISM domain, and thus was within the "Greenland" GLACIER_REGION). (We still need a GLACIER_REGION that encompasses areas outside Greenland itself in case ice extent grows a bit beyond the current boundaries. So we truly need two separate GLACIER_REGIONs for (a) Greenland itself, and (b) peripheral areas of the CISM domain.) This tag updates the surface datasets and related fields to have this extra GLACIER_REGION. In addition to updating the GLACIER_REGION raw data file, this also introduces a new regridding method that is applied to GLACIER_REGION: this takes the maximum value from any source point (NOT the dominant value). This is important for the new GLACIER_REGION, where we want a destination (CLM) cell to be considered in region 2 (Greenland itself) if ANY overlapping source points are in region 2. This means that the ordering of regions in the GLACIER_REGION raw data file is important: preference will be given to higher-numbered regions. Note that I just changed surface datasets that had time stamp 170428 or later. This excluded some older single point datasets, for which the new glacier region doesn't matter anyway: - surfdata_0.125x0.125_simyr2000_c150114.nc - surfdata_0.125x0.125_mp24_simyr2000_c150114.nc - surfdata_1x1_smallvilleIA_78pfts_simyr2000_c160127.nc - surfdata_1x1_smallvilleIA_78pfts_simyr1850_c160127.nc - surfdata_1x1_tropicAtl_16pfts_simyr1850_c160127.nc (used in tropicAtl_subset testmod) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - There are some additional subtleties now for setting up a run with different glacier domains (e.g., a paleoclimate run) Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): - New surface datasets - just differ in GLACIER_REGION Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - There are some additional subtleties now for setting up a run with different glacier domains (e.g., a paleoclimate run) Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - ok PASSed except expected failures: - baseline comparisons failed for mksurfdata_map tests - PTCLM tests failed as expected Also ran mksurfdata_map unit tests: - new-style (cmake / pfunit-based): pass - old-style: same failures as on trunk PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass cheyenne_intel - pass hobart_nag --- pass hobart_pgi --- pass hobart_intel - pass Ran tests on glacier_region_canada_n03_clm4_5_16_r246 - After that, modified the default mksurfdata_map namelist used for testing: pointed to updated file CLM tag used for the baseline comparisons: clm4_5_16_r246 Answer changes -------------- Changes answers relative to baseline: NO - bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: List all files added and what they do: ========= Get mkindexmapMod under some pfunit-based unit tests - required stubbing out a bunch of netcdf stuff A components/clm/tools/mksurfdata_map/src/unit_test_stubs A components/clm/tools/mksurfdata_map/src/unit_test_stubs/mkncdio.F90 A components/clm/tools/mksurfdata_map/src/test/mkindexmap_test A components/clm/tools/mksurfdata_map/src/test/mkindexmap_test/test_mkindexmap.pf A components/clm/tools/mksurfdata_map/src/test/mkindexmap_test/CMakeLists.txt ========= Add get_max_indices and associated unit tests A components/clm/tools/mksurfdata_map/src/test/mkgridmap_test A components/clm/tools/mksurfdata_map/src/test/mkgridmap_test/CMakeLists.txt A components/clm/tools/mksurfdata_map/src/test/mkgridmap_test/test_mkgridmap.pf List all existing files that have been modified, and describe the changes: ========= New raw dataset for glacier region M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml M components/clm/tools/mksurfdata_map/mksurfdata_map.namelist ========= Get mksurfdata_map pfunit-based unit tests passing with new cime M components/clm/tools/mksurfdata_map/src/README.unit_testing M components/clm/tools/mksurfdata_map/src/CMakeLists.txt ========= Add get_max_indices and associated unit tests; use get_max_indices to regrid glacier region M components/clm/tools/mksurfdata_map/src/mkgridmapMod.F90 M components/clm/tools/mksurfdata_map/src/mkindexmapMod.F90 M components/clm/tools/mksurfdata_map/src/mkglacierregionMod.F90 M components/clm/tools/mksurfdata_map/src/test/CMakeLists.txt ========= Allow building mkindexmapMod for unit tests - required stubbing out a bunch of netcdf stuff M components/clm/tools/mksurfdata_map/src/mkncdio.F90 ========= Point to new surface datasets M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Extended namelist array for new region M components/clm/cime_config/SystemTests/lvg.py ========= Removed duplicate makefile rules M components/clm/tools/mksurfdata_map/Makefile.data =============================================================== =============================================================== Tag name: clm4_5_16_r246 Originator(s): sacks (Bill Sacks) Date: Wed Jun 14 12:40:31 MDT 2017 One-line Summary: Update to latest cime Purpose of changes ------------------ Main change is to update to the latest version of cime. One of the changes in this cime version is that component-specific system tests (e.g., LII, SSP) can now be defined in the component rather than in cime. So this tag defines LII and SSP in CLM. In addition, this adds an LVG test (Land Virtual Glaciers), which verifies that adding virtual glacier columns doesn't change answers. Another significant change in this cime version is that the lnd -> glc downscaling (i.e., remapping) method has been changed to be bilinear with a later conservation correction, rather than area-conservative. This leads to smooth fields on the CISM grid. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New drv_in namelist variable 'glc_renormalize_smb' - controls whether to apply global corrections to the downscaled lnd -> glc smb for conservation Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - Moves LII and SSP test definitions into CLM - Adds new test type, LVG - Adds an LVG test Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass cheyenne_intel - ok hobart_nag --- pass hobart_pgi --- ok hobart_intel - pass ok means tests pass, some answers change as noted below CLM tag used for the baseline comparisons: clm4_5_16_r245 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: see below - what platforms/compilers: see below - nature of change (roundoff; larger than roundoff/same climate; new climate): see below (1) Changes answers for cases with CISM, due to new downscaling method. These changes show up only for long tests with CISM, and are potentially climate-changing: FAIL SMS_Lm37.T31_g37_gl4.IG1850CLM50.cheyenne_intel.clm-glcMEC_long BASELINE clm4_5_16_r245 FAIL SMS_Lm37.T31_g37_gl4.IG1850CLM50.yellowstone_intel.clm-glcMEC_long BASELINE clm4_5_16_r245 (2) Changes answers for hobart-pgi tests, due to new version of the pgi compiler on hobart. Magnitude of the change not investigated. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime: cime5.3.0-alpha.21 -> cime5.3.0-alpha.27 List all files eliminated: none List all files added and what they do: ========= CLM-specific system tests are now defined here rather than in cime. Also, added an LVG test type: Land Virtual Glacier. A components/clm/cime_config/config_tests.xml A components/clm/cime_config/SystemTests A components/clm/cime_config/SystemTests/lii.py A components/clm/cime_config/SystemTests/ssp.py A components/clm/cime_config/SystemTests/lvg.py List all existing files that have been modified, and describe the changes: ========= Ran through -cleanxml, and added an LVG test M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_16_r245 Originator(s): sacks (Bill Sacks) Date: Wed Jun 14 12:13:54 MDT 2017 One-line Summary: Only adjust glc_mec topographic heights if glc_do_dynglacier is true Purpose of changes ------------------ Two changes in this tag: (1) Only adjust glc_mec topographic heights if glc_do_dynglacier is true (this namelist variable is controlled by the GLC_TWO_WAY_COUPLING xml variable). Previously, CLM updated its glacier topographic heights even if this variable was false (so we were not updating areas from CISM). Bill Lipscomb and Jeremy Fyke agreed a while ago that we should change this. None of us could remember why we initially set it up to always update topographic heights, but we felt that it's more intuitive to avoid updating topographic heights if we're not updating glacier areas. Note that setting GLC_TWO_WAY_COUPLING=FALSE is *not* a common thing to do, so this change does not affect most cases. (The typical way of doing one-way coupling with CISM is to use CISM2%NOEVOLVE, which turns off ice evolution but keeps GLC_TWO_WAY_COUPLING=TRUE in order to update CLM's glacier areas and topographic heights at initialization.) (2) Adds some user mods and testmods (currently unexercised) for cmip6 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: New test mods directories, currently not exercised (I'll add tests soon) Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - ok hobart_nag --- ok hobart_pgi --- ok hobart_intel - ok ok means tests pass, a few expected baseline comparison failures CLM tag used for the baseline comparisons: clm4_5_16_r244 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Cases with GLC_TWO_WAY_COUPLING=FALSE - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate This affects tests with glcMEC_changeFlags and glcMEC_spunup_1way testmods Also, NLCOMP failures for a few tests, I think explainable by the process Erik used to update some testmods to get LII tests passing: FAIL SMS_D_Ln1.f09_g17.ICLM45BGC.yellowstone_gnu.clm-defaultf09IC NLCOMP FAIL ERP.f19_g17_gl4.IG1850CLM45.cheyenne_intel.clm-glcMEC_spunup_1way NLCOMP FAIL SMS_D_Ln1.f19_g17_gl4.IG1850CLM45.cheyenne_intel.clm-glcMEC_spunup_1way NLCOMP If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: ========= Renamed D components/clm/cime_config/usermods_dirs/clm/clm5_OutputCrop/user_nl_clm D components/clm/cime_config/usermods_dirs/clm/clm5_OutputCrop/shell_commands D components/clm/cime_config/usermods_dirs/clm/clm5_OutputCrop D components/clm/cime_config/usermods_dirs/clm List all files added and what they do: ========= Top-level cmip6 usermods directory, which will include other usermods directories needed for the cmip6 runs A components/clm/cime_config/usermods_dirs/cmip6/include_user_mods A components/clm/cime_config/usermods_dirs/cmip6 ========= Renamed from clm5_OutputCrop; also added some glacier output fields needed for cmip6 runs A components/clm/cime_config/usermods_dirs/cmip6_output/user_nl_clm A components/clm/cime_config/usermods_dirs/cmip6_output/shell_commands A components/clm/cime_config/usermods_dirs/cmip6_output/README A components/clm/cime_config/usermods_dirs/cmip6_output ========= Turns on virtual glacier columns over Antarctica. This is desired by the LIWG for cmip6 runs, so that we can have the Antarctica output that would be needed to later run an ice sheet model over Antarctica. However, it comes with a 10% performance cost. A components/clm/cime_config/usermods_dirs/cmip6_glaciers/user_nl_clm A components/clm/cime_config/usermods_dirs/cmip6_glaciers ========= Test mod that allows testing the cmip6 usermods (currently not exercised with any tests) A components/clm/cime_config/testdefs/testmods_dirs/clm/cmip6/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/cmip6 ========= Test mod that allows running the new LVG (virtual glaciers) test (currently unexercised) A components/clm/cime_config/testdefs/testmods_dirs/clm/no_vector_output/README A components/clm/cime_config/testdefs/testmods_dirs/clm/no_vector_output/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/no_vector_output/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/no_vector_output List all existing files that have been modified, and describe the changes: ========= Only adjust topographic heights if glc_do_dynglacier is true M components/clm/src/main/TopoMod.F90 M components/clm/src/main/test/topo_test/test_topo.pf M components/clm/src/main/glc2lndMod.F90 ========= Add _musgs to f10_f10 grid M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_16_r244 Originator(s): erik (Erik Kluzek) Date: Fri Jun 9 14:26:07 MDT 2017 One-line Summary: Update cime and cism externals, changes answers for IG compsets, test g17 grids, fix a few issues, update mksurfdata_map Purpose of changes ------------------ Update the cime version which includes some important updates. Remove lt_archive, new script query_config to query compsets and grids etcetera. Force user to always use case.submit rather than allow them to submit to batch queue on their own. Some updates to handling of namelist items in user_nl_clm. All grids now contain a mask variable. Also fixes a couple glitches from the previous version, see bugs fixed below. Update cism which changes answers (for cases running with CISM) as 4km grid was changed. Fix some small issues with CLM. Fix interpolation issue with GSSUN and GSSHA, use shr_mpi needed for unit-testing, and fix an issue with a use of MCT routines for clm40. Several updates to mksurfdata_map, fixing problems, starting to get it working on cheyenne, and upgrading to latest CMIP6 PFT datasets from Peter Lawrence. -Update cime and cism. -Unit tests moved to cime/scripts/fortran_unit_testing/. -Include Jim Edward's configcompsetupdate changes here. -Convert mosart half-degree file from NetCDF-4 to NetCDF-3. -Change grids in test lists so that mask is added to grids that need it (i.e. f45_f45 becomes f45_f45_mg37). -Add note about supported compset IGM1850GSWCLM50BGCCROP for f09_g17, f19_g17. -Use grids back to g16 as g17 domain files aren't there. -Get LII tests working. -Rerun the list of tests that failed, see if works again because of a system issue. -use old g16 mask for CLM40 tests. -Add CMIP5/CMIP6 names to surfdata file creation. -Bring in mksrfunreplu branch. -New rawdata PFT datasets from Peter. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): [If none, remove this line] 2468 GSSUN/GSSHA interp issue 2440 change to orderedPoints for clm40 as well 2425 mksurfdata_map changes from Sean didn't get incorporated 2419 Irrigated areas are low in mksurfdata_map 2399 Problem with urban points with latest mksurfdata_map CIME Issues fixed (include issue #): [If none, remove this line] #1215 -- (in this version of cime) TestStatus sometimes needs to have write permission added #1155 -- manage_case doesn't work (renamed to query_config) #848 -- ./xmlquery --list-all doesn't work (refactored) Known bugs introduced in this tag (include bugzilla ID): 2471 PTCLM doesn't work 2470 CLM is getting the mask from the clm namelist_defaults rather than from top level scripts 2469 Cleanup restartvar_{DIMS}d_{TYPE} to ONLY be valid for a 1d interface 2467 Case on non-year boundary dies with error about ignoring ic_date for a crop case, even though for a branch or hybrid case 2465 ICLM45VIC compset doesn't actually turn VIC on 2464 allowlakeprod are outputing standard variables instead of the "lake" specific versions... Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): cime update brought in quite a few changes - Add skip-preview-namelist to case.submit and case_run. - Renamed manage_case to query_config. - manage_case -> query_config --compset -> --compsets --machine -> --machines added 'all' option for --compsets added 'current' and to --machines - Remove all references to lt_archive which will not be supported in CESM2.0. - Add Support for '+=' in Namelists. - Add --component Option to preview_namelist. - Add Support of true and false in Namelist. - Force user to always go through case.submit. - Adds support for namelist entries of the form foo(3) = 'a' - esmf_logging (and therefore drv_in) are now optional - New run time XML variable: ESMF_LOGFILE_KIND - New --answer flag to create_newcase -Update so that all relevent grids contain a mask variable. - Add preview_run to case.run Changes made to namelist defaults (e.g., changed parameter values): Added mapping files for 0.25x0.25_MODIS Changes to the datasets (e.g., parameter, surface or initial files): Just to datasets to create surface datasets Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - Use ESMF7 on yellowstone and cheyenne. - Big cs.status upgrade. - Add --input-dir argument to create_test Changes to tests or testing: Change most gx1v6 masks to gx1v7 Keep clm40 tests at gx1v6 Code reviewed by: self, jedwards (config), mvertens, sacks (unit-tests) Did you follow the steps in .CLMTrunkChecklist: Yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - OK PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - FAIL regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK cheyenne_intel - OK hobart_nag --- OK hobart_pgi --- PASS hobart_intel - PASS CLM tag used for the baseline comparisons: clm4_5_16_r243 Answer changes -------------- Changes answers relative to baseline: Yes for IG compsets with active glacier model Summarize any changes to answers, i.e., - what code configurations: Any with CISM2 - what platforms/compilers: All - nature of change: minor Changes answers for cases with CISM2 at 4km (the default resolution for CISM2 cases) - Even changes answers for standard one-way coupled cases (CISM2%NOEVOLVE) - Fields on the CISM grid will have DIMSIZEDIFF - Fields sent to CLM have roundoff-level differences - time variable changes for all cases with CISM Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, cism, mosart cime to cime5.3.0-alpha.21 cism to cism2_1_33 mosart to mosart1_0_24 List all files eliminated: D components/clm/tools/SVN_EXTERNAL_DIRECTORIES List all files added and what they do: M components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/README --- Add description of this testmods and how to update the finidat file for it A components/clm/test/tools/tests_pretag_cheyenne_nompi ---- Add test list for cheyenne List all existing files that have been modified, and describe the changes: M components/clm/test/tools/TCBCFGtools.sh - Source $INITMODULES M components/clm/test/tools/test_driver.sh - Add cheyenne, set INITMODULES M components/clm/test/tools/TSMCFGtools.sh - Handle CSMDATA M components/clm/test/tools/nl_files/gen_domain.ne30.runoptions - Use CSMDATA for path to mapping file M components/clm/test/tools/nl_files/gen_domain.T31.runoptions -- Use CSMDATA for path to mapping file M components/clm/tools/mkmapdata/regridbatch.sh ---------------- Set PBS for cheyenne, increase regrid_num_proc to 36 M components/clm/tools/mkmapdata/mkmapdata.sh ------------------ Increase REGRID_PROC to 36, show defaults in --help, change so -b means run with mpi, also create 0.25x0.25_MODIS maps, add cheyenne, rm hopper M components/clm/tools/mksurfdata_map/Makefile.data ------------ Add global-present-T42 M components/clm/tools/mksurfdata_map/src/mkharvestMod.F90 ----- Change harvest_fieldnames and string_undef M components/clm/tools/mksurfdata_map/src/mkgridmapMod.F90 ----- Add gridmap_areaave_scs interface which multiplys weight by ratio of source over destination weight (set to zero if destination weight is zero) M components/clm/tools/mksurfdata_map/src/mkpftMod.F90 --------- Use gridmap_areaave_scs interface rather than default one, passing down source weight of pctnatveg_i and desitination weight of pctnatveg_o (both mulitipled by 0.01) if zeroing out also set pctnatveg_o, pctcrop_o, pct_nat_pft_o, pct_cft_o to zero, and first entry of pct_nat_pft_o and pct_cft_o to 100 M components/clm/tools/mksurfdata_map/mksurfdata_map.namelist -- Update with new files M components/clm/tools/mksurfdata_map/mksurfdata.pl ------------ Add cmip_series to output filenames, use vegtyp to get harvest files, M components/clm/tools/mksurfdata_map/landuse_timeseries_hist_78pfts_simyr1850-2015.txt - Update with new datafiles. Note, each year is repeated once to point to same data (to be used for harvesting). This allows harvest files to be seperate. M components/clm/bld/unit_testers/build-namelist_test.pl --------------- Don't test 0.25x0.25 grid M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Add 0.25x0.25 MODIS mapping file, harvest, PFT, LAI, and soil-color are 0.25x0.25_MODIS, harvest files are: mksrf_fvegtyp, Update: LAI, soil-color, and PFT files, remove harvest files M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ----- Add 0.25x0.25 to hgrid and res, add 2015 to sim_year, and 1980-2015 to sim_year_range, M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ------- Add 0.x25x0.25_MODIS mapping files M components/clm/cime_config/config_compsets.xml ------- Changes from Jim Edwards to update to version 2.0 which will do xmllint checking on the file, add for f09, f19 IGM1850GSWCLM50BGCCROP compsets M components/clm/cime_config/testdefs/testlist_clm.xml - Add SMS _Ln1 tests to correspond to each LII test (to make updating finidat files for LII tests easier, update grids to include mask when needed, and update most g16 mask tests to g17 (except clm40 tests which need to remain at g16) M components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/user_nl_clm ------- Update finidat file, set use_init_interp to True and add notes on how to update it M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm - Update finidat file, set use_init_interp to True and add notes on how to update it M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/README ------ Add notes on how to update the finidat file M components/clm/SVN_EXTERNAL_DIRECTORIES - Point to PTCLM here M components/clm/src_clm40/main/GetGlobalValuesMod.F90 - Change mct_gsmap_OP to mct_gsmap_orderedPoints (bug 2440) M components/clm/src_clm40/main/ncdio_pio.F90.in ------- Change mct_gsmap_OP to mct_gsmap_orderedPoints (bug 2440) M components/clm/src/biogeochem/CNSharedParamsMod.F90 -- Use shr_mpi_bcast rather than straight MPI (important for new unit tests) M components/clm/src/README.unit_testing --------------- Updated notes on how to do it from Bill Sacks M components/clm/src/biogeophys/PhotosynthesisMod.F90 -- Set switchdim=T for GSSUN/GSSHA (bug 2468) M components/clm/src/CMakeLists.txt -------------------- Add include(CIME_initial_setup) =============================================================== =============================================================== Tag name: clm4_5_16_r243 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue May 23 16:27:42 MDT 2017 One-line Summary: History output cleanup from Dave Lawrence Purpose of changes ------------------ * Reduce default history output by setting numerous variables to default=inactive, limiting some 3d vars to active soil layers only. * Update default history output for CLM5. * Added diagnostic new output for DWT_WOOD_PRODUCTC_GAIN_PATCH, GS_MOL, lake_icefracsurf_col, h2osoi_liq_tot_col, h2osoi_ice_tot_col * Remove double accounting for CNSoy fixation when FUN is on. * Bug fixes to allow landunit-level and column-level averaging and vector output using the hist_type1d_pertape namelist option (LAND for land-unit, COL for column-level). Note that if a user asks for a variable at a subgrid level that is not available at that subgrid level, the model will crash ungracefully. An example would be asking for TOT_WOODPRODC at the land-unit level when this variable is only available at the grid cell level, or asking for H2OSOI at the pft-level when H2OSOI is a column-level variable. Code may be added in a future tag that will inform the user of the error. * Add usermods dir for recommended clm5 crop default output. To use the usermod when creating a new case: ./create_newcase --compset HIST_DATM%GSWP3_CLM50%BGC-CROP_SICE_SOCN_MOSART_CISM2_SWAV \ --user-mods-dir clm/clm5_OutputCrop \ ... machine, compiler, grid, case name, etc... Bugs fixed or introduced ------------------------ None, but see note below in testing about c14 bombspike test. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): The default history output has been reduced. You may need to explicitly request fields of interest from the namelist. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Dave Lawrence, Ben Andre Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - all tests pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - ok, see note below hobart_nag - ok hobart_pgi - ok hobart_intel - ok CLM tag used for the baseline comparisons: clm4_5_16_r242, no yellowstone intel baselines were generated for clm4_5_16_r242 Baselines comparisons for clm45/clm5 tests show: 1. changes in fields sizes (expected, restrict vertical fields to nlevsoi) 2. different fields on current and baseline files (expected, new default active/inactive) 3. no RMS differences, except the following tests: These tests failed baseline comparisons with RMS differences in testing due to the soy nitrogen double accounting fix. ERP_P60x2_Lm36.f10_f10.ICRUCLM50BGCCROP.cheyenne_intel.clm-clm50cropIrrigMonth_interp ERS_Ly3.f10_f10.I1850CRUCLM50BGCCROP.cheyenne_intel.clm-clm50KSinkMOut ERS_Ly5_P60x1.f10_f10.IHISTCLM50BGCCROP.cheyenne_intel.clm-cropMonthOutput ERP_P60x2_Lm36.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-clm50cropIrrigMonth_interp ERS_Ly3.f10_f10.I1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KSinkMOut ERS_Ly5_P60x1.f10_f10.IHISTCLM50BGCCROP.yellowstone_intel.clm-cropMonthOutput The ciso-c14 bombspike1963 test failed baseline test with RMS differences in C14_* fields on cheyenne_intel. This did not have differences when tested against r273 and yellowstone_intel. This seems to be intermintent, there was at least one test run with r242 cheyenne_intel where this passed without RMS differences. SMS_D_Ly2.1x1_brazil.IHISTCLM50BGC.cheyenne_intel.clm-ciso_bombspike1963 Answer changes -------------- Changes answers relative to baseline: some clm5 compsets If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: removal of double accounting for soy nitrogen fixation from clm5 crop runs with fun active. See tests above. - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): unknown, presumed larger than roundoff same climate. If bitwise differences were observed, how did you show they were no worse than roundoff? n/a If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: n/a URL for LMWG diagnostics output used to validate new climate: n/a Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: New usermods with default output for clm5 crop runs A clm/cime_config/usermods_dirs/clm/clm5_OutputCrop A clm/cime_config/usermods_dirs/clm/clm5_OutputCrop/user_nl_clm A clm/cime_config/usermods_dirs/clm/clm5_OutputCrop/shell_commands List all existing files that have been modified, and describe the changes: M clm/cime_config/config_component.xml - update component description from Erik Kluzek Set history fields active/inactive: M clm/src/biogeochem/CNVegCarbonStateType.F90 M clm/src/biogeochem/CNVegNitrogenStateType.F90 M clm/src/biogeochem/CNVegNitrogenFluxType.F90 M clm/src/biogeochem/CropType.F90 M clm/src/biogeophys/EnergyFluxType.F90 - don't output btran unless hydrstress active M clm/src/biogeophys/TemperatureType.F90 MM clm/src/biogeophys/HumanIndexMod.F90 M clm/src/biogeophys/SoilHydrologyType.F90 M clm/src/biogeophys/SolarAbsorbedType.F90 M clm/src/biogeophys/CanopyStateType.F90 M clm/src/biogeophys/SurfaceRadiationMod.F90 M clm/src/biogeophys/WaterfluxType.F90 - variable renaming M clm/src/biogeophys/SoilStateType.F90 New diag output, set default active/inactive fields: M clm/src/biogeochem/CNVegCarbonFluxType.F90 M clm/src/biogeochem/CNProductsMod.F90 M clm/src/biogeophys/PhotosynthesisMod.F90 M clm/src/biogeophys/LakeStateType.F90 M clm/src/biogeophys/LakeTemperatureMod.F90 M clm/src/biogeophys/HydrologyNoDrainageMod.F90 - vertically summed water state calc M clm/src/biogeophys/WaterStateType.F90 Default active/inactive history fields, restrict output to nlevsoi: M clm/src/biogeochem/ch4Mod.F90 M clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 M clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 M clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 M clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 FUN Soy Nitrogen Fixation double accounting: M clm/src/biogeochem/CNDriverMod.F90 Subgrid mapping: M clm/src/main/histFileMod.F90 M clm/src/main/subgridAveMod.F90 =============================================================== =============================================================== Tag name: clm4_5_16_r242 Originator(s): erik (Erik Kluzek) Date: Sun May 21 21:16:34 MDT 2017 One-line Summary: Use finundated streams for CLM50, and use ndep if it comes from cpl Purpose of changes ------------------ For methane, read in inversion for TWS to Prigent Satellite dataset from a streams file for CLM50. For CLM45 still read in ZWT inversion parameters from the surface dataset. You can read it from the streams file if you provide it to the namelist. Answers change for CLM50BGC compsets because of this change to methane. The change does affect results outside of methane after enough time. Add option to use nitrogen deposition sent from the coupler rather than from datasets (from mvertens). Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): bug 2462 -- TWS restart bug 2457 -- remove FIND_HIST bug 2454 -- species masterlist confusion bug 2443 -- ndep bug 2456 -- rm extra settings bug 2463 -- T42 SP dataset Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist item for finundated streams stream_fldfilename_ch4finundated. It can be used for both TWS_inversion which is now the CLM50 default, or the ZWT_inversion (CLM45 default), but by default CLM45 is reading ZWT inversion parameters from the surface dataset Changes made to namelist defaults (e.g., changed parameter values): ZWT_inversion for CLM45 and TWS_inversion for CLM50 for inundated fraction Changes to the datasets (e.g., parameter, surface or initial files): Add 2000 T42 fsurdata file Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: Add two new tests Add test for hourly radiation with lightning resolution at half-degree And test with reseeding and reset of snow Code reviewed by: self,mvertens Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (94 compare tests FAIL) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - Didn't run yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK cheyenne_intel - OK hobart_nag --- PASS hobart_pgi --- PASS hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_16_r241 Answer changes -------------- Changes answers relative to baseline: Yes, CLM50 Summarize any changes to answers, i.e., - what code configurations: CLM50 (with finundated streams for TWS_inversion) - what platforms/compilers: All - nature of change: similar climate, but methane changes (which does feedback into system) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: None List all files added and what they do: Add new tests A components/clm/cime_config/testdefs/testmods_dirs/clm/rad_hrly_light_res_half/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/rad_hrly_light_res_half/user_nl_datm A components/clm/cime_config/testdefs/testmods_dirs/clm/reseedresetsnow/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/reseedresetsnow/include_user_mods List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm -------------------- Remove fin_use_fsat, set finundation_method and get finundated streams file if TWS_inversion, add ch4finundated namelist group M components/clm/bld/unit_testers/build-namelist_test.pl ---- Add test for change to finundated if not methane M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Remove extra settings of use_init_interp, add T42 present day fsurdat file, remove fin_use_fsat, set finundation_method to TWS_inversion for CLM50 and ZWT_inversion for CLM45, set finundation_res=1.9x2.5 and set stream_fldfilename_ch4finundated file for it M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add: stream_fldfilename_ch4finundated, finundation_res remove fin_use_fsat, correct finundation_method documentation M components/clm/cime_config/config_compsets.xml ------- Add IGMCRUCLM50SP, IGMCRUCLM50BGCROP compsets M components/clm/cime_config/testdefs/testlist_clm.xml - Change prealpha/prebeta yellowstone_intel/gnu tests to cheyenne add new tests for reseedresetsnow and rad_hrly_light_res_half M components/clm/src/main/clm_instMod.F90 -------- Pass NLFilename down to ch4 Init M components/clm/src/main/clm_varctl.F90 --------- Add ndep_from_cpl variable M components/clm/src/main/clm_driver.F90 --------- If ndep_from_cpl is true don't do ndep_interp M components/clm/src/main/clm_initializeMod.F90 -- If ndep_from_cpl is true don't do ndep_init and ndep_interp M components/clm/src/main/histFileMod.F90 -------- Add error check if field name is too long M components/clm/src/biogeophys/WaterStateType.F90 - Add TWS to restart file M components/clm/src/cpl/clm_cpl_indices.F90 --- Set ndep_from_cpl based on existance of Faxa_nhx and Faxa_noy M components/clm/src/cpl/lnd_import_export.F90 - Set forc_ndep_grc if ndep_from_cpl is set, convert units M components/clm/src/biogeochem/SpeciesNonIsotopeType.F90 --- Add trim's to strings to make sure blanks aren't added M components/clm/src/biogeochem/ch4Mod.F90 ------------------ Pass NLFilename to Init, add ch4findstream data and Init to it Remove fws_a/b/hist, remove fin_use_fsat use finundation_mtd Call ch4findstream%CalcFinundated is finundated streams are being used (stream_fldfilename_ch4finundated is set), or if finundation_mtd_h2osfc, M components/clm/src/biogeochem/ch4FInundatedStreamType.F90 - Add UseStreams method, get streams working properly and M components/clm/src/biogeochem/SpeciesIsotopeType.F90 ------ Add trim for strings to make sure blanks don't happen M components/clm/src/biogeochem/ch4varcon.F90 --------------- Remove fin_use_fsat =============================================================== =============================================================== Tag name: clm4_5_16_r241 Originator(s): sacks (Bill Sacks) Date: Tue May 16 20:43:43 MDT 2017 One-line Summary: Fix ch4 subtle dependence on processor count Purpose of changes ------------------ When we moved to using interpolated initial conditions in most cases, long tests that change processor counts (ERP, PEM) began to fail (bug 2458). I tracked this problem down to annsum_npp, annsum_counter, and specifically bug 1597. I found that, in both the original finidat and finidat_interp_dest, we have: - For coltype = 1, counter can be 0, 3600 or 5400 - For coltype = 200-something (crop), counter is always 5400 - For other coltypes, counter = 0 (but these shouldn't matter) My best guess is that there is some interaction between (1) the history of an initial conditions file, and (2) dynamic landunits, which causes annsum_counter to differ by column. Specifically, with (1): I think that problems arise when an initial conditions file has (at some point in its history) arisen from a run that didn't run for a complete year; e.g., this can happen in the course of making an interp'ed initial conditions file - for which we typically run for one or a few time steps. This can lead to a misalignment of the resetting of annsum_counter relative to when dynamic landunits kick in, and thus the counter can differ for different columns. This tag fixes the problem by making each column dependent on its own count, rather than using the count from the first column on that MPI task. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 1597: awkward conditionals to determine end of year in CN AnnualUpdate - 2458: ERP clm5 restart failure in clm4_5_16_r237 Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] - https://github.com/NCAR/CLM/issues/7 : CN annsum_counter can trigger mid-year, and at different times for different columns Decision in talking with Erik and Keith was that we would not try to resolve this right now - it seems that it can cause a bit of confusion, but no scientific problems as far as we can tell. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Erik; conceptually reviewed by Keith Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - pending; I'll check them post-tag and open an issue if any fail hobart_nag --- ok hobart_pgi --- not run hobart_intel - ok ok means tests pass, some baselines fail as expected CLM tag used for the baseline comparisons: clm4_5_16_r240 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Many configurations, when run for at least a year - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; expected to be same climate, but not investigated carefully. Answer changes can appear whenever we have an initial conditions file where different columns have different values of the annsum_counter restart field. This is the case for present-day restart files, as well as at least one year-1850 restart file (clmi.I1850CRUCLM45BGC.0241-01-01.1.9x2.5_g1v6_simyr1850_c170308.nc). The impact of this problem is expected to be small, at least if the differences in annsum_counter values are small (which they are for the two initial conditions files I have checked, and I expect this is the case for others, too): In this case, the impact is that some annually-accumulated fields reach their end-of-year mark a few time steps too early or late. However, I have not investigated the impact of this bug fix carefully. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/biogeochem/CNAnnualUpdateMod.F90 ========= New function to create a sub-filter based on an existing filter and a column-level logical array M components/clm/src/main/filterColMod.F90 M components/clm/src/main/test/filter_test/test_filter_col.pf ========= Remove no-longer-failing tests M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_16_r240 Originator(s): sacks (Bill Sacks) Date: Mon May 15 14:48:22 MDT 2017 One-line Summary: Make glc_dyn_runoff_routing real-valued Purpose of changes ------------------ The glc_dyn_runoff_routing variable tracks which CLM grid cells overlap fully-coupled glc (CISM) grid cells, and which do not: In those that do not, CLM needs to send the snow capping flux directly to runoff. Previously, this value was a logical. However, we (Bill Lipscomb and Bill Sacks) realized that this really needs to be real-valued - having a value between 0 and 1 where a CLM cell partially overlaps the CISM domain. This is needed for complete conservation when we are running with a fully-coupled (two-way-coupled) glc component. This changes answers for cases with two-way-coupled CISM, but does NOT change answers for the typical one-way-coupled (%NOEVOLVE) case. Main changes are from Bill Lipscomb. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - There is some duplicate logic between CLM and the coupler in determining the fraction of the CLM grid cell that overlaps with the active glc domain. Ideally, we'd move all of the relevant logic into the coupler and avoid needing it in CLM. See the TODO note in glc2lndMod.F90 Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - ok hobart_nag --- ok hobart_pgi --- not run hobart_intel - ok ok means tests pass, baselines fail as expected CLM tag used for the baseline comparisons: clm4_5_16_r239 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: cases with CISM two-way coupled (i.e., without %NOEVOLVE) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff, but only changes answers around the margin of Greenland, so expected to be same climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/src/main/test/topo_test/test_topo.pf M components/clm/src/main/glc2lndMod.F90 M components/clm/src/biogeophys/GlacierSurfaceMassBalanceMod.F90 ========= Add a test to the xfail list which fails sporadically due to bug 2461 M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_16_r239 Originator(s): sacks (Bill Sacks) Date: Sat May 13 20:35:52 MDT 2017 One-line Summary: Avoid division by zero in snow WindDriftCompaction Purpose of changes ------------------ Tweak the equations in WindDriftCompaction to avoid division by 0, to fix bug 2460. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2460: Floating point trap in FWAMIP compset Known bugs introduced in this tag (include bugzilla ID): - 2461: Multi-instance tests failing sporadically This has actually been an issue for the last few tags, likely related to bug 2289 - default finidat_interp_dest file name should have instance number Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Dave Lawrence Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok cheyenne_intel - ok Did not do baseline comparisons for cheyenne tests hobart_nag --- ok hobart_pgi --- not run hobart_intel - ok ok means tests pass, baseline comparisons fail for clm50 tests as expected CLM tag used for the baseline comparisons: clm4_5_16_r238 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? via analysis of the change: the new equation is algebraically identical to the old; essentially we now have X*Y/Z rather than X/(Z/Y) If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main change, as described above M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Note sporadic multi-instance failure; remove a test from the xfail list that is now passing M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_16_r238 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Sat May 13 11:45:15 MDT 2017 One-line Summary: ED is now an external library 'fates' Purpose of changes ------------------ The ongoing evolution of the 'ED' library is being done by NGEE-Tropics as 'FATES'. This change removes the 'ED' directory and replace it with link an external release of fates. The primary changes are to finish creation of an interface layer to separate fates from CLM. There are also numerous science improvements in the linked fates library. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2459 - patch from Erik Kluzek Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Changes made to namelist defaults (e.g., changed parameter values): * ED has been completely rebranded as fates. This includes: * New/renamed compsets - ICLM45ED is now ICLM45FATES, ICLM50FATES. * Renamed name list variables: * use_ed --> use_fates * use_ed_spitfire --> use_fates_spitfire * New namelist variable: fates_paramfile, this can point to either the clm parameter file, or a standalone fates parameter file depending on user preference. All fates dimensions and variables are namespaced by 'fates_'. * Fates can not restart and interpolate from a CLM-{SP,CN,BGC} initial conditions file. It must be run in cold start or restart from a CLM-fates restart file. Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Changes to tests or testing: * Add tests to aux_clm45 using the fates compsets and renamed the ed test mods to fates and fatesFire. * Added the cheyenne_intel test list to aux_clm45 * all namelist tests are expected to fail because of use_ed -> use_fates rename. Code reviewed by: Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - all tests pass tools-tests (components/clm/test/tools): yellowstone - n/a not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass cheyenne_intel - [ experimental, see xfails list ] pass w/ xfails hobart_nag - pass w/ xfails hobart_pgi - not run per CLM-CMT discussion, see bugz 2447, 2455 hobart_intel - pass CLM tag used for the baseline comparisons: clm4_5_16_r237 Answer changes -------------- Changes answers relative to baseline: answer changing for ED/fates compsets, but bit for bit for all other clm45 and clm50 compsets Verified not answer changing for CLM50 compsets via test suite, and 20 year simulation by Dave Lawrence and Keith Olseson. If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ED/fates compsets only - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): unknown If bitwise differences were observed, how did you show they were no worse than roundoff? n/a If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: n/a URL for LMWG diagnostics output used to validate new climate: n/a Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): A components/clm/SVN_EXTERNAL_DIRECTORIES - new SVN_EXTERNALS file for fates A components/clm/src/fates List all files eliminated: D components/clm/src/ED - removed and replaced by fates external List all files reamed: rename testmod edTest --> fates default now with fire off A testmods_dirs/clm/fates A testmods_dirs/clm/fates/user_nl_clm A testmods_dirs/clm/fates/shell_commands D testmods_dirs/clm/edTest D testmods_dirs/clm/edTest/user_nl_clm D testmods_dirs/clm/edTest/shell_commands rename testmod edNoFire --> fatesFire A testmods_dirs/clm/fatesFire A testmods_dirs/clm/fatesFire/user_nl_clm A testmods_dirs/clm/fatesFire/include_user_mods A testmods_dirs/clm/fatesFire/shell_commands D testmods_dirs/clm/edNoFire D testmods_dirs/clm/edNoFire/user_nl_clm D testmods_dirs/clm/edNoFire/include_user_mods D testmods_dirs/clm/edNoFire/shell_commands List all files added and what they do: List all existing files that have been modified, and describe the changes: M .ChangeLog_template - include cheyenne_intel M parse_cime.cs.status - patch from Erik Kluzek bugz 2459 Point build to new src/fates directory instead of src/ED M clm/cime_config/buildlib M clm/bld/configure Update namelist with new fates paramfile, rename use_ed to use_fates in namelist logic, rename functions to be fates instead of ed. Update tests. M clm/bld/CLMBuildNamelist.pm M clm/bld/unit_testers/build-namelist_test.pl M clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M clm/bld/test_build_namelist/t/test_do_transient_pfts.pm A clm/bld/test_build_namelist/t/test_fates_mode.pm D clm/bld/test_build_namelist/t/test_ed_mode.pm M clm/bld/test_build_namelist/t/test_do_harvest.pm M clm/bld/test_build_namelist/t/test_do_transient_crops.pm M clm/bld/namelist_files/namelist_definition_clm4_5.xml M clm/bld/namelist_files/namelist_defaults_clm4_5.xml Update test for fates, cheyenne_intel M clm/cime_config/testdefs/ExpectedTestFails.xml M clm/cime_config/testdefs/testlist_clm.xml Stub fates parameter interface for unit tests: M components/clm/src/CMakeLists.txt A + components/clm/src/unit_test_stubs/utils/clmfates_paraminterfaceMod_stub.F90 M components/clm/src/unit_test_stubs/utils/CMakeLists.txt Rename ED compset FATES M clm/cime_config/config_compsets.xml M clm/cime_config/config_component.xml Fates interface A clm/src/utils/clmfates_paraminterfaceMod.F90 M clm/src/utils/clmfates_interfaceMod.F90 Rename use_ed to use_fates and other rebranding M clm/src/biogeochem/CNCStateUpdate1Mod.F90 M clm/src/biogeochem/CNMRespMod.F90 M clm/src/biogeochem/CNVegetationFacade.F90 M clm/src/biogeochem/CNDriverMod.F90 M clm/src/main/clm_varpar.F90 M clm/src/main/subgridRestMod.F90 M clm/src/main/atm2lndType.F90 M clm/src/main/restFileMod.F90 M clm/src/main/initSubgridMod.F90 M clm/src/main/initVerticalMod.F90 M clm/src/main/surfrdMod.F90 M clm/src/biogeophys/PhotosynthesisMod.F90 M clm/src/biogeophys/HydrologyDrainageMod.F90 M clm/src/biogeophys/SurfaceAlbedoMod.F90 M clm/src/biogeophys/SurfaceRadiationMod.F90 M clm/src/biogeophys/SoilMoistStressMod.F90 M clm/src/biogeophys/CanopyStateType.F90 M clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M clm/src/dyn_subgrid/dynEDMod.F90 M clm/src/dyn_subgrid/dynSubgridControlMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeCNMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemPotentialMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 M clm/src/soilbiogeochem/SoilBiogeochemDecompMod.F90 M clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 Update logic for fates compatibility and to use clm_fates interface M clm/src/biogeochem/EDBGCDynMod.F90 M clm/src/main/subgridWeightsMod.F90 M clm/src/main/decompInitMod.F90 M clm/src/main/clm_initializeMod.F90 M clm/src/main/subgridMod.F90 M clm/src/main/histFileMod.F90 M clm/src/main/controlMod.F90 M clm/src/main/clm_varctl.F90 M clm/src/main/clm_driver.F90 M clm/src/main/pftconMod.F90 M clm/src/main/initGridCellsMod.F90 M clm/src/main/decompMod.F90 M clm/src/main/clm_instMod.F90 M clm/src/main/readParamsMod.F90 M clm/src/biogeophys/CanopyTemperatureMod.F90 M clm/src/biogeophys/HydrologyNoDrainageMod.F90 M clm/src/biogeophys/BalanceCheckMod.F90 M clm/src/biogeophys/WaterStateType.F90 M clm/src/biogeophys/TotalWaterAndHeatMod.F90 M clm/src/biogeophys/CanopyFluxesMod.F90 M clm/src/biogeophys/SoilStateInitTimeConstMod.F90 M clm/src/biogeophys/RootBiophysMod.F90 M clm/src/biogeophys/OzoneMod.F90 M clm/src/main/paramUtilMod.F90 - new generic routines used by fates interface M clm/src/main/PatchType.F90 - new patch array to distinguish between fates patch and clm pft A clm/src/biogeophys/SoilWaterPlantSinkMod.F90 - from SoilWaterMovement M clm/src/biogeophys/SoilWaterMovementMod.F90 - moved to SoilWaterPlantSink M clm/src/biogeophys/WaterfluxType.F90 - new variable for fates M clm/src/biogeophys/SoilStateType.F90 - remove tsink =============================================================== =============================================================== Tag name: clm4_5_16_r237 Originator(s): erik (Erik Kluzek) Date: Tue May 9 15:05:51 MDT 2017 One-line Summary: Latest parameter file and some changes for CLM50, CLM45-transient changes answers also Purpose of changes ------------------ Adjustments to CLM50 climate from Keith Oleson, Peter Lawrence, Fang Li, Rosie Fisher, Danica Lombardozzi, Will Weider, Yaqiong Lu, Ben Sanderson, and David Lawrence. Some tweaks to the code and to parameters and the parameter file for CLM50. Use Medlyn 2011 stomotal conductance rather than Ball-Berry. Modify photosynthesis and leaf maintence respiration for crop. Limit irrigation if river flow is on and the storage drops below a threshold. Modify a long list of parameter values including 41 items on the parameter file. Only harvest up to 98% to allow for regrowth and divide by several terms rather than just deadstem. Add slash harvesting fields (debris left from harvesting). Add in a hack to read in FWS_TWS_A/B if they are on the surface dataset and use them to calculate finundation based on TWS rather than exponential function based on ZWT. For fires change an if statement from checking if atmosphere is non-freezing to a specific soil layer. If crops have hit peak LAI set leaf allocation to a small value. Make sure some retranslocation terms are positive. Let lnc in Photosynthesis max out at 10. Only vary baset for crops for spring wheat and sugarcane (irrigated and non) rather than all crops. Coupled crop cases will change since we now use the radiation time-step for fertilization rather than the model time-step. CLM45 cases remain the same with the exception for transient cases which change answers due to the change in how harvest is handled. Fix a few minor bugs that came in with clm4_5_15_r236 with specific datasets and finidat file handling in scripts (bugs 2448-2451). Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2452 --- Change to divisor for harvest and also allow for regrowth 2451 --- Bad comment marker in cropf10IC testmods 2450 --- Make sure finidat is reset after setting to coldstart 2449 --- Add no fail to init_interp_attributes and use_init_interp, needed for DV cases where interpolated IC files aren't there 2448 --- Misnamed dataset in CLM xml database Known bugs introduced in this tag (include bugzilla ID): 2457 --- Remove FINUN_HIST from history 2456 --- Remove extra settings of use_init_interp 2455 --- PGI specific error at initialization 2454 --- masterlist_addfld getting confused with names for species... 2453 --- Memory error with restart variable vcmx25t 2447 --- Floating overflow for CLM50CROP case on hobart_nag in Photosynthesis Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New slash harvest fields SLASH_HARVESTC, DWT_SLASH_CFLUX, C13_DWT_SLASH_CFLUX, C14_DWT_SLASH_CFLUX Changes made to namelist defaults (e.g., changed parameter values): Many adjustments for CLM50 modifyphoto_and_lmr_forcrop TRUE for CLM50 stomatalcond_method Medlyn2011 for CLM50 with PHS limit_irrigation_if_rof_enabled TRUE for CLM50 Following values modified for CLM50 cases modify baseflow_scalar, rh_low, bt_min, bt_max, cli_scale, boreal_peatfire_c, pot_hmn_ign_counts_alpha non_boreal_peatfire_c, cropfire_a1, lfuel, cmb_cmplt_fact, irrig_threshold_fraction, n_melt_glcmec, br_root, initial_Cstocks_depth Changes to the datasets (e.g., parameter, surface or initial files): CLM50 parameter file clm5_params.c170413.nc Update CLM50 paramsfile, fsurdat-f09-2000-crop, flanduse_timeseries-f45-hist/f10-hist Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): finundation changes are messy Changes to tests or testing: Fixed testmods for one tests Code reviewed by: self,dlawren,lawrence,oleson,rosiefisher,wweider,lifang Did you follow the steps in .CLMTrunkChecklist: Yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS (93 comparison tests fail to clm4_5_15_r236) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - PASS yellowstone_pgi - PASS yellowstone_gnu (clm45 only) - PASS hobart_nag --- OK hobart_pgi --- OK hobart_intel - PASS CLM tag used for the baseline comparisons: clm4_5_15_r236 Answer changes -------------- Changes answers relative to baseline: All CLM50 and CLM45-transient Summarize any changes to answers, i.e., - what code configurations: All CLM50 compsets and any transient cases with CLM45 Also coupled crop cases - what platforms/compilers: All - nature of change: Adjusted climate for CLM50 and changes to harvesting for CLM45 Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - modifyphoto_and_lmr_forcrop TRUE for CLM50 stomatalcond_method Medlyn2011 for CLM50 with PHS limit_irrigation_if_rof_enabled TRUE for CLM50 Following values modified for CLM50 cases modify baseflow_scalar, rh_low, bt_min, bt_max, cli_scale, boreal_peatfire_c, pot_hmn_ign_counts_alpha non_boreal_peatfire_c, cropfire_a1, lfuel, cmb_cmplt_fact, irrig_threshold_fraction, n_melt_glcmec, br_root, initial_Cstocks_depth Update CLM50 paramsfile, fsurdat-f09-2000-crop, flanduse_timeseries-f45-hist/f10-hist M components/clm/bld/CLMBuildNamelist.pm ------------------------- Make sure local finidat defined for cold start allow nofail for use_init_interp and init_interp_attributes M components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC/user_nl_clm - Use \! for comment M components/clm/src/biogeochem/CNPhenologyMod.F90 ----------- Use radiation time-step rather than dt for Fertilization M components/clm/src/biogeochem/CNFireBaseMod.F90 ------------ Change ufuel default M components/clm/src/biogeochem/dynHarvestMod.F90 ------------ Only harvest up to 98% and divide by sum of several terms rather than just deadstem (Peter Lawrence) M components/clm/src/biogeochem/ch4FInundatedStreamType.F90 -- Modify endrun call M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 ----- Slash flux to litter and CWD M components/clm/src/biogeochem/CNFireLi2016Mod.F90 ---------- Change if statement from check on forcing temp non-freezing to tsoi17 non-freezing (From Fang Li) M components/clm/src/biogeochem/ch4Mod.F90 ------------------- Add ability to read in FWS_TWS_A/B terms from surface dataset and use for calculation of finundated based on TWS rather than use exponential inversion on ZWT M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 ------ Add slash harvest flux fields (Peter Lawrence) M components/clm/src/biogeochem/CropType.F90 ----------------- Only do baset_map_latvary for spring wheat and sugarcane both irrigated and non. M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 - If crops have hit peaklai, then set leaf allocation to small value make sure leafn_to_retransn, frootn_to_retransn and livestemn_to_retransn is positive M components/clm/src/biogeophys/PhotosynthesisMod.F90 -------- Let lnc max out at 10 =============================================================== =============================================================== Tag name: clm4_5_15_r236 Originator(s): erik (Erik Kluzek) Date: Mon May 1 01:42:23 MDT 2017 One-line Summary: New surface datasets, and landuse.timeseries files, as well as interpolating initial conditions for almost all cases Purpose of changes ------------------ New fsurdat and landuse.timeseries datasets for almost all resolutions. New initial conditions datasets pointed to as much as possible for all resolutions (have ability to turn off to COLDSTART for some tests). Normally model will now die, if it can't find initial conditions to interpolate from. Reasonable f09 and f19 PE layout for cheyenne. Fix another issue on Cheyenne. Add new pgi/intel tests for hobart. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2445 -- ED not working with new datasets 2444 -- dimensional change in URBAN_AC/URBAN_HEAT for interpolation of restart files 2441 -- fire emissions on cheyenne 2291 -- fire emissions with gnu compiler sometimes NaN, sometimes change answers for trivial changes 2239 -- need new initial conditions files with extra metadata on them Known bugs introduced in this tag (include bugzilla ID): 2447 -- Floating overflow for CLM50CROP case on hobart_nag in Photosynthesis #1215 -- (in this version of cime) TestStatus sometimes needs to have write permission added #? -- manage_case doesn't work #? -- ./xmlquery --list-all doesn't work Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): All initial conditions have to be interpolated now, but there are ones that can be interpolated from for standard cases NOTE NOW IF YOU WANT TO DO A COLD START USING ARBITRARY INITIAL CONDITIONS SET CLM_FORCE_COLDSTART=on Also the model will normally fail if it can't find initial conditions close enough to interpolate from Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): A close finidat file will almost always be set Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files): New fsurdat/finidat/landuse.timeseris Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Changes to tests or testing: New pgi/intel tests for hobart Expected fails removed: bug 2239, 2291, and added SMS_Ld5_D_P24x1.f10_f10.IHISTCLM50BGC.hobart_nag.clm-monthl for bug 2447 Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - PASS yellowstone_pgi - PASS yellowstone_gnu (clm45 only) - PASS hobart_nag --- OK hobart_pgi --- OK hobart_intel - OK CLM tag used for the baseline comparisons: clm4_5_15_r235 Answer changes -------------- Changes answers relative to baseline: Yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Almost all CLM45 and CLM50 configurations because of new datasets - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): similar climate Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: ------ No longer need vrtlay_interp testmods, nor user_nl_clm for some D components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/README D components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp D components/clm/cime_config/testdefs/testmods_dirs/clm/cropColdStart/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/monthly_noinitial/user_nl_clm ------- Reduce size of files D components/clm/tools/mkprocdata_map/clm4054_ne30g16_I2000.clm2.h0.2000-01_c121107.nc D components/clm/tools/mkprocdata_map/clm4054_f19g16_I2000.clm2.h0.2000-01_c121107.nc List all files added and what they do: ------------ New test-mods with specific initial conditions A components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/defaultf09IC A components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/cropf10IC A components/clm/tools/mkprocdata_map/clm4054_ne30g16_I2000.clm2.h0.2000-01_c170430.nc A components/clm/tools/mkprocdata_map/clm4054_f19g16_I2000.clm2.h0.2000-01_c170430.nc A components/clm/cime_config/testdefs/testmods_dirs/clm/allActive/shell_commands - CLM_FORCE_COLDSTART=on A components/clm/cime_config/testdefs/testmods_dirs/clm/cropColdStart/shell_commands - CLM_FORCE_COLDSTART=on A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly_noinitial/shell_commands - CLM_FORCE_COLDSTART=on ------------ New unit test for surfrdUtilsMod subroutines A components/clm/src/main/test/surfrdUtils_test/test_surfrdUtils.pf A components/clm/src/main/test/surfrdUtils_test/CMakeLists.txt List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm ---------------------------- Add better handling of finding finidat files either with an exact match or interpolated from M components/clm/bld/unit_testers/build-namelist_test.pl ------------ Add new failure tests, fix ed-test M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- Add init_interp_how_close and init_interp_sim_years M components/clm/bld/namelist_files/namelist_defaults_overall.xml --- Start type is startup (requires finidat) for CLM45/CLM50 unless use_cndv is on M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ---- init_interp_how_close=75 and init_interp_sim_years=1850,2000 new 2-deg IC files IGM1850GSWCLM50BGCCROP, I1850CRUCLM45BGC, IGM2000GSWP3CLM50BGCCROPIRR, ICRUCLM45SP redone 2-deg ICLM45BGCCROP create_crop_landunit on unless ED is on updated: fsurdat, flanduse.timeseries for all but 1x1_smallville_IA and 0.125x0.125 M components/clm/cime_config/testdefs/ExpectedTestFails.xml -- ICLM45VIC.yellowstone now PASS SMS_Lm25.f19_g16.ICLM45BGCCROP.yellowstone_gnu now PASS, new fail SMS_Ld5_D_P24x1.f10_f10.IHISTCLM50BGC.hobart_nag.clm-monthly M components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/user_nl_clm - remove finidat and init_interp_fill_missing_with_natveg M components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/user_nl_clm - set fsurdat, create_crop_landunit M components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/user_nl_clm ------ create_crop_landunit = .false. M components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsROA/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/edTest/user_nl_clm ------ remove finidat='' M components/clm/cime_config/testdefs/testmods_dirs/clm/allActive/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLA/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLB/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/user_nl_clm - new finidat M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/README M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/user_nl_clm - remove finidat M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/README M components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB/shell_commands - CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm - remove finidat, init_interp_fill_missing_with_natveg M components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/shell_commands- CLM_FORCE_COLDSTART=on M components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/README M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm - new finidat M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm - remove finidat M components/clm/cime_config/testdefs/testlist_clm.xml -- new pgi/intel hobart tests, vrtlay_interp->vrtlay, use defaultf09IC cropf10IC for two LII tests M components/clm/cime_config/user_nl_clm ----------- Fix documentation of irrigate, add note to use CLM_FORCE_COLDSTART rather than finidat=' ' M components/clm/cime_config/config_component.xml - HIST and 2000 compsets for BGCCROP turn irrigation on M components/clm/cime_config/config_pes.xml -- Add f09 and f19 PE layouts for cheyenne M components/clm/src/biogeochem/FireEmisFactorsMod.F90 - Copy crop factors from generic crop CFT's M components/clm/src/main/test/CMakeLists.txt ---------- Add new unit test M components/clm/src/main/surfrdUtilsMod.F90 ----------- Add renormalize and convert_cft_to_pft functions lighten tolerence for sums to equal 1 from 1.e-14 to 1.e-13 M components/clm/src/main/surfrdMod.F90 ---------------- Add surfrd_pftformat, surfrd_cftformat to read in old/new formats for CFT or PFT, also handle conversion of new CFT format to PFT format for ED M components/clm/src/biogeophys/EnergyFluxType.F90 ----- For CLM50 urban add _L to URBAN_AC/HEAT to restart files, so CLM45 won't be confused with CLM50 =============================================================== =============================================================== Tag name: clm4_5_15_r235 Originator(s): sacks (Bill Sacks) Date: Thu Apr 27 07:47:50 MDT 2017 One-line Summary: Zero dynbal fluxes if namelist variable is set Purpose of changes ------------------ Some fully-active system tests were dying due to too-large sensible heat fluxes (thousands or 10s of thousands of W m-2) generated from the dynbal adjustments. These large fluxes arose because these tests trigger daily dynamics of CISM (via the test_coupling namelist option), rather than the typical annual coupling. The annual dribblers see these mid-year adjustments and apply them all in a single time step, rather than spreading them throughout the year. The changes in this tag allow you to set a namelist variable (for_testing_zero_dynbal_fluxes) to force the dynbal fluxes to zero. Typically, this namelist variable should only be used for testing: setting it to .true. breaks water and energy conservation. We will set this to true in the fully-active tests that use daily CISM dynamics. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - new namelist variable, for_testing_zero_dynbal_fluxes; should only be used for testing Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_15_r234 Answer changes -------------- Changes answers relative to baseline: NO - bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Changes as described above M components/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml =============================================================== =============================================================== Tag name: clm4_5_15_r234 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Apr 14 08:34:22 MDT 2017 One-line Summary: Update rtm and mosart externals Purpose of changes ------------------ Update externals to rtm1_0_61 and mosart1_0_23. Includes change to python build-namelist (rtm, mosart), histfile bugfix (rtm, mosart), and science bugfix / changes (mosart). Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2434 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): python build namelist for rtm will no longer automatically point rtm to a clm restart file. rtm should be configured to generate and read it's own restart files. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) : none --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: rtm was not outputing a restart file for the ref run in SSP tests, but was attempting to read one during the second run because of build-namelist changes in rtm1_0_60. All SSP tests in aux_clm45 and prebeta were changed to use a new test mod, rtmColdSSP, that forces rtm to do a cold start when coming out of spinup. Old RCP/clm40 tests that were not compatible with new rtm build-namelist behavior were removed. Code reviewed by: self, discussion with Erik Kluzek. Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - n/a, not run unit-tests (components/clm/src): yellowstone - n/a, not run tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok All tests passed bit for bit except: * expected failures * Tests with mosart passed functionality, answer changing as expected. * SSP tests passed functionality, but don't have baselines because of test-mod change. Manually ran cprc to compare with clm4_5_15_r233, files were reported as identical. e.g. cprnc clm4_5_15_r233/SSP_D_Ld10.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default/clm2.h0.0001-01-06-00000.nc clm4_5_15_r234-prelim/SSP_D_Ld10.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-rtmColdSSP/clm2.h0.0001-01-06-00000.nc CLM tag used for the baseline comparisons: clm4_5_15_r233 Answer changes -------------- Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all runs with mosart - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): unknown, assumed same climate. If bitwise differences were observed, how did you show they were no worse than roundoff? If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: mosart1_0_23 was reviewed by Sean Swenson and Dave Lawrence in a long coupled run and approved. URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): rtm1_0_61, mosart1_0_23 List all files eliminated: none List all files added and what they do: New test mod so rtm will do a cold start during SSP tests. A clm/cime_config/testdefs/testmods_dirs/clm/rtmColdSSP A clm/cime_config/testdefs/testmods_dirs/clm/rtmColdSSP/include_user_mods A clm/cime_config/testdefs/testmods_dirs/clm/rtmColdSSP/user_nl_rtm List all existing files that have been modified, and describe the changes: Remove old RCP/clm40 tests that were not compatible with new rtm behavior. Update SSP tests in aux_clm45 and prebeta to use new rtmColdSSP testmod. M clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_15_r233 Originator(s): erik (Erik Kluzek) Date: Tue Apr 11 23:30:45 MDT 2017 One-line Summary: Pull out some constants and options to namelist Purpose of changes ------------------ Pull out a list of constants and options to namelist control. Decrease a few error tolderances. Add in reseed_dead_plants namelist option, so can reseed dead plants for BGC on a restart. Reenable LND_TUNING_MODE, although currently set the same for all options. Add some new compsets in. Fix a few bugs, including Jim Edward's fix for vector output. Start adding in the stream option for finundated. Add option to turn Medlyn photosynthesis on. Replace & with & in namelist_definition_clm4_5.xml Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2440 Remove use of mct_gsmap_op and replace with mct_gsMap_orderedPoints 2430 Problem with units on new landuse.timeseries datasets 2184 vector (1d) history output crashes for some fields Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Added new parameters to namelist New namelists: cnvegcarbonstate and cn_general New namelist items: modifyphoto_and_lmr_forcrop, stomatalcond_method, lfuel, ufuel, cmb_complt_fact, initial_Cstocks_depth, ndep_taxmode, ndep_varlist, init_interp_attributes, reseed_dead_plants, initial_vegC, finundation_method New output history variables: Remove RSCANOPY unless use_ed add: VCMX25T, JMX25T, TPU25T New compsets I1850GSWCLM50BGCCROP, IM1850GSWCLM50BGCCROP, IGM1850GSWCLM50BGCCROP, IGMHISTGSWCLM50BGCCROP Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): params files Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (183 test fail comparison to previous namelists) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_15_r232 Answer changes -------------- Changes answers relative to baseline: None bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: -------- Start adding ability to read finundated streams data (currently disabled) A components/clm/src/biogeochem/ch4FInundatedStreamType.F90 List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm --- Add setup_logic_cnvegcarbonstate, new namelists: cnvegcarbonstate and cn_general, refactor how finidat files were being found to a better more general methodology that allows a second go to find finidat file to interpolate from (currently disabled) M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ---- New settings: modifyphoto_and_lmr_forcrop, stomatalcond_method, use_init_interp, init_interp_attributes initial_Cstocks_depth, initial_vegC, and new params file with medlynslope/intercept M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- New namelist items: modifyphoto_and_lmr_forcrop, stomatalcond_method, lfuel, ufuel, cmb_complt_fact, initial_Cstocks_depth, ndep_taxmode, ndep_varlist, init_interp_attributes, reseed_dead_plants, initial_vegC, finundation_method M components/clm/bld/unit_testers/build-namelist_test.pl -- Test that use_init_interp=.true. with finidat blank is fails M components/clm/cime_config/config_component.xml - Set LND_TUNING_MODE by compset M components/clm/cime_config/config_compsets.xml -- Add: I1850GSWCLM50BGCCROP, IM1850GSWCLM50BGCCROP, IGM1850GSWCLM50BGCCROP, IGMHISTGSWCLM50BGCCROP compsets M components/clm/src/biogeochem/CNBalanceCheckMod.F90 --- Decrease carbon error tolerance from 1xe-8 to 1xe-7 and nitrogen error tolerance from 1xe-4 to 1xe-3 M components/clm/src/biogeochem/CNFireBaseMod.F90 ---------- Add lfuel, ufuel, cmb_cmplt_fact to namelist M components/clm/src/biogeochem/CNVegCarbonStateType.F90 --- Add InitReadNML and read in initial_vegC in new cnvegcarbonstate namelist, handle reseeding with reseed_patch filter M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 - Handle reseeding with reseed_patch filter for restart read M components/clm/src/biogeochem/CNVegStateType.F90 --------- Handle reseeding with reseed_patch filter for M components/clm/src/biogeochem/CNVegetationFacade.F90 ----- Add namelist read CNReadNML for cn_general and reseed_dead_plants option, pass it and/or reseed filter down as appropriate M components/clm/src/biogeochem/CropType.F90 ------ Check validity of baset_mapping, set latbaset_patch to nan if not used, switch logic of how activated M components/clm/src/biogeochem/ch4varcon.F90 ----- Make default private, explicitly make items public as needed add finundation_mtd and valid settings, read finundation_method in namelist and check it M components/clm/src/biogeochem/dynHarvestMod.F90 - Validate units of harvest fields abort if not right M components/clm/src/biogeophys/CanopyStateType.F90 --- Only send RSCANOPY to history if use_ed is on M components/clm/src/biogeophys/PhotosynthesisMod.F90 - Add: stomatalcond_mtd, modifyphoto_and_lmr_forcrop namelist items, add extra history output: VCMX25T, JMX25T, TPU25T M components/clm/src/main/GetGlobalValuesMod.F90 - Replace mct_gsMap_orderedPoints for mct_gsMap_OP M components/clm/src/main/histFileMod.F90 -------- Fix for bug 2184 from Jim Edwards M components/clm/src/main/init_hydrology.F90 ----- Call init_root_moist_stress M components/clm/src/main/ncdio_pio.F90.in ------- Replace mct_gsMap_orderedPoints for mct_gsMap_OP M components/clm/src/main/ndepStreamMod.F90 ------ Add: ndep_taxmode, ndep_varlist to namelist M components/clm/src/main/pftconMod.F90 ---------- Add: medlynslope/intercept to data/alloc/read/clean M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 ------ Use decomp_cascade_con%initial_stock_soildepth in place of hardcoded 0.3 M components/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 ------- Increase error tolerence allowed for endrun for too much NH4 uptake predicted by FUN M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 -- Add initial_Cstocks_depth to namelist M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeConType.F90 - Add initial_stock_soildepth set to 0.3 M components/clm/src/unit_test_stubs/csm_share/mct_mod_stub.F90 - Replace mct_gsMap_OP with mct_gsMap_orderedPoints M components/clm/src/utils/quadraticMod.F90 - Allow for situations where a sqrt of a slightly negative number would crash the code. =============================================================== =============================================================== Tag name: clm4_5_15_r232 Originator(s): erik (Erik Kluzek) Date: Wed Apr 5 23:04:37 MDT 2017 One-line Summary: Some small changes that affect both clm4_5 and clm5_0 answers (can be considered bug fixes to clm4_5) Purpose of changes ------------------ A set of some small changes that affect answers for both CLM45 and CLM50. A fix to a rare problem where soil albedo could go negative from Keith Oleson and Sean Swenson, by making sure a calculated value could not go below a small value. A change from Will Weider to lower leaf maintence respiration (LMR) for low LAI. Always calculate btran2 even for low temperatures and/or low liquid water this lowers fires in boreal forests. For crop at peak LAI set hui to max of hui and huigrain. Change the handling of GDDPLANT accumulator field from Keith Oleson. Get rid of the use_voc logic that just prevented MEGAN from activating for crop, since now MEGAN handles the crop case (since clm4_5_13_r204). Bugs fixed or introduced ------------------------ 2437 -- lower leaf maintence respiration for low LAI 2436 -- some adjustments to crop 2435 -- always calculate btran to lower fires in boreal forests 2431 -- Rare condition under which soil albedo goes negative 2368 -- Remove use_voc as MEGAN can now work with crop Bugs fixed (include bugzilla ID): Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self, oleson, dll, dlawren, wweider, fangli Did you follow the steps in .CLMTrunkChecklist: Yes CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_14_r231 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: - what platforms/compilers: - nature of change (roundoff; larger than roundoff/same climate; new climate): Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SurfaceAlbedoMod.F90 --- Fix bug 2431 where soil albedo can got negative make sure a calculated value doesn't go too small M components/clm/src/biogeophys/PhotosynthesisMod.F90 -- Add check for divide by zero fo lnc, reduce lmr for low LAI M components/clm/src/biogeophys/SoilMoistStressMod.F90 - Always calculate btran2 even for low temperature and/or low liquid water this lowers fires in boreal forests M components/clm/src/biogeochem/CNPhenologyMod.F90 ----- For crop at peak lai set hui to max of hui and huigrain M components/clm/src/biogeochem/CropType.F90 ----------- Change handling of GDDPLANT ----------- Remove use_voc logic M components/clm/src/cpl/clm_cpl_indices.F90 ----------- Remove use_voc M components/clm/src/main/clm_instMod.F90 -------------- Remove use_voc M components/clm/src/main/controlMod.F90 --------------- Remove use_voc M components/clm/src/main/clm_varctl.F90 --------------- Remove use_voc M components/clm/src/main/clm_driver.F90 --------------- Remove use_voc M components/clm/src/main/lnd2atmMod.F90 --------------- Remove use_voc =============================================================== =============================================================== Tag name: clm4_5_14_r231 Originator(s): sacks (Bill Sacks) Date: Sun Apr 2 19:36:02 MDT 2017 One-line Summary: Add option to reset snow pack over non-glacier columns Purpose of changes ------------------ When going from a CLM spinup under one set of atmospheric forcings to a different set of atmospheric forcings, there is the possibility that columns that reached large snow depths in the spinup should really have only seasonal snow cover under the new atmosphere. This is a particular concern for the LIWG in some areas of Greenland that have only seasonal snow cover: there is a concern that the offline spinup can bake in too-high snow cover in some regions. This can especially be a problem because of feedbacks: starting out with a too-high snow cover can cool the overlying atmosphere, leading to persistence of the snow pack. This is also a problem because, once the snow pack becomes very deep, it seems that its refreezing can make it less likely that the snow pack will ever melt. To deal with these issues, this tag introduces a new namelist item, reset_snow. This is typically .false. If you set it to .true., then all snow in non-glacier columns is reset to 35 mm SWE over the first few time steps of the run. This resetting is accomplished via the snow capping mechanism. This implementation was chosen because it is challenging to reset the snow pack directly, since there are so many snow-related variables that must be kept in sync. The snow capping mechanism already handles this properly, so it's easiest to reuse this mechanism. However, unlike standard snow capping, the flux of snow lost through this resetting is sent into a new flux that simply disappears into nothingness - rather than being sent to glc and/or rof. Thus, note that - by design! - the flux from this snow resetting breaks conservation: We do not want this lost snow to affect ocean circulation, since this is essentially just a mechanism for adjusting an initial conditions file. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Specifying reset_snow = .true. will lead to a loss of conservation for the first few time steps of the run Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variable: reset_snow Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Leo van Kampenhout reviewed an initial prototype Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok tests pass; namelists differ from baseline, as expected unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, NLCOMP failures as expected CLM tag used for the baseline comparisons: clm4_5_14_r230 Answer changes -------------- Changes answers relative to baseline: no - bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: List all files added and what they do: ========= Unit test of new logic A components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_SnowCappingExcess.pf List all existing files that have been modified, and describe the changes: ========= Changes as described above M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/test/SnowHydrology_test/CMakeLists.txt =============================================================== =============================================================== Tag name: clm4_5_14_r230 Originator(s): sacks (Bill Sacks) Date: Fri Mar 31 16:25:49 MDT 2017 One-line Summary: Fix dynamic landunits water and energy conservation Purpose of changes ------------------ Fixes water and energy conservation with dynamic landunits. This conservation works by summing the water and energy contents of each grid cell before and after the dynamic landunits transitions, then dealing with the difference through runoff and sensible heat fluxes. However, the code for counting the water and energy contents was wrong - slightly wrong for water, and substantially wrong for energy. The energy counts were 6 orders of magnitude too small, and also were missing some terms. This tag fixes these counts, and reconciles the dynamic landunits water and energy accounting with the accounting done for the balance checks. It also refactors some of the code used in the balance checks to facilitate code sharing between various places. Also: adds wa_col (water in the unconfined aquifer) to the list of fields copied to newly-initializing columns, to avoid spurious dynlu adjustment fluxes due to changes in wa_col. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] - https://github.com/NCAR/CLM/issues/2 - Count energy of water in lakes in total gridcell heat content - https://github.com/NCAR/CLM/issues/3 - Perform gridcell-level water balance checks bracketing the entire run loop Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - Some renamed history variables: see below Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: not investigated, but no major changes expected Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Dave Lawrence and Keith Oleson reviewed the new code conceptually for scientifc correctness; Ryan Knox reviewed a slightly older version of the new code for counting water and energy contents Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, answers change as expected CLM tag used for the baseline comparisons: clm4_5_14_r229 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): (1) Roundoff-level changes for all cases (2) Large changes in diagnostics of total gridcell water and total gridcell heat, due to various fixes (3) Dynamic landunit water adjustments change substantially for glacier transitions, and somewhat for transient crops (4) Dynamic landunit heat adjustments change substantially for all runs with dynamic landunits (transient glacier, transient crop): change by about 6 orders of magnitude (5) Changes in transient crop cases due to copying wa_col to newly-initiating columns (though answer changes may be minor for clm5 cases?) (6) Possibly other changes that I'm forgetting? If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Moved code that sums water and energy contents to here, to facilitate code sharing; added unit tests for *some* of it A components/clm/src/biogeophys/TotalWaterAndHeatMod.F90 A components/clm/src/biogeophys/test/TotalWaterAndHeat_test/test_total_water_and_heat.pf A components/clm/src/biogeophys/test/TotalWaterAndHeat_test/CMakeLists.txt A components/clm/src/biogeophys/test/TotalWaterAndHeat_test List all existing files that have been modified, and describe the changes: ========= Major rework: Use new shared routines for calculating heat and water contents; adjust dynlu adjustment heat flux based on heat content of dynlu adjustment runoff flux M components/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 ========= Copy wa_col (water in the unconfined aquifer), to avoid large adjustments M components/clm/src/dyn_subgrid/dynInitColumnsMod.F90 M components/clm/src/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf ========= Use new routines for water content M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 ========= New constant for baseline wa_col (water in the unconfined aquifer), replacing hard-coded constants M components/clm/src/main/clm_varcon.F90 M components/clm/src/biogeophys/SoilHydrologyMod.F90 M components/clm/src/biogeophys/SoilHydrologyInitTimeConstMod.F90 ========= For performance: introduce col%hydrologically_active, used in place of is_hydrologically_active function call M components/clm/src/main/ColumnType.F90 M components/clm/src/main/filterMod.F90 M components/clm/src/main/initSubgridMod.F90 M components/clm/src/main/glcBehaviorMod.F90 ========= Changes to diagnostic fields: Remove incorrect HCSOI, HC; rename GC_HEAT1 and GC_HEAT2 to HEAT_CONTENT1 and HEAT_CONTENT2; add HEAT_CONTENT1_VEG; add LIQUID_WATER_TEMP1 M components/clm/src/biogeophys/TemperatureType.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/LakeTemperatureMod.F90 ========= Changes to diagnostic fields: Rename GC_LIQ1/2 to LIQUID_CONTENT1/2, and similar for ICE; also remove some now-unused fields M components/clm/src/biogeophys/WaterStateType.F90 ========= Minor changes to some calls M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/main/clm_driver.F90 ========= New stub routines M components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90.in ========= New unit tests M components/clm/src/biogeophys/CMakeLists.txt M components/clm/src/biogeophys/test/CMakeLists.txt ========= Add some comments M components/clm/src/main/subgridAveMod.F90 M components/clm/src/main/column_varcon.F90 ========= Remove unneeded use statements and associates M components/clm/src/biogeophys/UrbanParamsType.F90 M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/main/initVerticalMod.F90 =============================================================== =============================================================== Tag name: clm4_5_14_r229 Originator(s): erik (Erik Kluzek) Date: Fri Mar 10 11:32:20 MST 2017 One-line Summary: Update cime and make some changes to drv_flds_in namelist, fix some minor bugs Purpose of changes ------------------ Update cime to a version with an updated top level directory structure. It also fixes a problem with a CLM test. The new cime version has a new directory structure, and updates the XML schema for namelist files from verison 1 to version 2. Currently the perl namelist code can NOT read version 2. Hence, a local copy of the version 1 files for the drv_flds_in fields was added to CLM. Also the use of the datm namelist files for some tools was dropped -- I think that's OK though. Fix two minor bugs. The new cime has a few improvements. It now adjusts the PIO namelist when the PE layout changes. There's a new option to case.setup "--no-adjust-pio" to not do that. --mpilib option was added to create_test. You can remove your bld directory area without having to rerun case.setup. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2433 -- format error in HumanIndexMod.F90 2432 -- intent mixing in dynVarMod.F90 CIME Issues fixed (include issue #): 1074 -- ERP_D_Ld5.hcru_hcru.ICRUCLM45BGC.yellowstone_pgi fails with PIO error Known bugs introduced in this tag (include bugzilla ID): (recently found, but predate this tag) 1217 -- Driver allows a contradiction in drv_flds_in between CAM and CLM 1220 -- missing mpi-serial mpirun case on hobart Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): case.setup has a new option --no-adjust-pio The user now can rm -rf their build area and rebuild without having to re-run case.setup. Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): CIME now normally adjusts PIO layout if you change the PE layout. Also puts a copy of namelist definition and defaults for drv_flds_in in CLM. This copy needs to be doubly maintained with the versions in the driver and in CAM. The advantage of having a copy is because each copy is using a different version of the xml schema, and potentially the driver might have additional fields that interact with other components besides CLM. Currently the only drv_flds_in fields are all shared between only CAM and CLM. But, the fact that there are duplicated copies does mean changes in any one of those components will need to go into the other two. Also removes the datm namelist files read in tools such as queryDefaultNamelist.pl. I think this is OK, as this connection may no longer be needed. It does make the listDefaultNamelist.pl less useful, but I think that's OK as well. Changes to tests or testing: add --mpilib option to create_test Code reviewed by: self,mvertens Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: regular, tools [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - OK PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_14_r226 Answer changes No bit-for-bit -------------- Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, cime_config Update cime to cime5.3.0-alpha.02 Add new cime_config at cime_config0.0.1.alpha.03 List all files eliminated: None List all files added and what they do: These are version 1 of the files, the ones in the driver are now version 2. A components/clm/bld/namelist_files/namelist_definition_drv.xml ------- Copy from driver A components/clm/bld/namelist_files/namelist_definition_drv_flds.xml -- Copy from driver A components/clm/bld/namelist_files/namelist_defaults_drv.xml --------- Copy from driver List all existing files that have been modified, and describe the changes: ------------ Point to new local copy of drv_flds namelist files, and bunch of whitespace changes ------------ Remove pointing to datm namelist files as they are now version 2, and the perl code ------------ currently can't read them M components/clm/bld/queryDefaultNamelist.pl ------ Point to local location of namelist drv files and remove pointing to datm namelist files M components/clm/bld/CLMBuildNamelist.pm ---------- Point to local location of namelist drv files M components/clm/bld/listDefaultNamelist.pl ------- Point to local location of namelist drv files and remove pointing to datm namelist files M components/clm/cime_config/testdefs/ExpectedTestFails.xml Remove hcr_hcru test that is now passing ------------ Get unit tests working with the new cime directory structure M components/clm/src/CMakeLists.txt ----------------------------- New cime directory structure M components/clm/src/unit_test_stubs/csm_share/seq_comm_mct.F90 - Just change comment as to where the "real" file exists ------------ Fix a couple of bugs M components/clm/src/biogeophys/HumanIndexMod.F90 --- Fix bug 2433 format error M components/clm/src/dyn_subgrid/dynVarMod.F90.in --- Fix bug 2432 intent issue =============================================================== =============================================================== Tag name: clm4_5_14_r228 Originator(s): erik (Erik Kluzek) Date: Tue Mar 7 14:17:07 MST 2017 One-line Summary: Bring in changes that were accidentally dropped as part of the clm4_5_14_r226 tag Purpose of changes ------------------ Bring in changes that were included for the testing of clm4_5_14_r226, but accidentally dropped when moved to the trunk tag. Only a limited amount of testing was done, since the testing was done in the clm4_5_14_r226 tag. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: None Changes to tests or testing: None Code reviewed by: self Did you follow the steps in .CLMTrunkChecklist: yes CLM testing: short, since full testing was done in clm4_5_14_r226 CLM tag used for the baseline comparisons: clm4_5_14_r227 Answer changes -------------- Changes answers relative to baseline: No bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ---------- Remove --phys option to PTCLM testing M components/clm/test/tools/nl_files/PTCLM_USUMB_clm4_5 M components/clm/test/tools/nl_files/PTCLM_USUMB_Cycle_clm4_5 M components/clm/test/tools/nl_files/PTCLM_USUMB_Global_clm4_5 M components/clm/cime_config/config_component.xml ------------ Remove unnneeded "" M components/clm/cime_config/testdefs/ExpectedTestFails.xml -- Add note about hcru_hcru failing test M components/clm/cime_config/testdefs/testlist_clm.xml ------- Add a test with MOSART on hobart_nag M components/clm/src/biogeophys/SnowSnicarMod.F90 ------------ Fix bug 2479 workaround for Intel 16/17 compiler M components/clm/src/main/initGridCellsMod.F90 --------------- Fix bug 2417 for create_crop_landunit =============================================================== =============================================================== Tag name: clm4_5_14_r227 Originator(s): sacks (Bill Sacks) Date: Mon Mar 6 09:42:41 MST 2017 One-line Summary: Improve performance of dynamic landunits code Purpose of changes ------------------ Main change is to improve the performance of the column state updates portion of the dynamic landunits code. In optimized, global runs, this piece of the code (dyn_cnbal_col) was taking 3 - 5% of total CLM runtime. Its cost was greater in global debug runs (~ 10 - 15%), and even greater in single-point debug runs (30 - 40%). Here I have introduced a variable that tracks whether there are any changes in area in this time step for each clump; if not, then it avoids doing most of the work in dyn_cnbal_col. This tag also fixes the fortran unit tests with the new cime version: the necessary changes were accidentally dropped from the last tag. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: 3-4% performance improvement for global, optimized runs; greater performance improvement for debug and single-point runs Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_14_r226 Answer changes -------------- Changes answers relative to baseline: NO - bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as noted above. Primary changes are in dynColumnStateUpdater, but other modules needed to be changed in order to pass the clump index down. Eventually, we should be able to get rid of this clump index tracking, instead having a separate object for each clump - but that is waiting on a major refactor of how we handle threading in CLM. M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M components/clm/src/dyn_subgrid/test/dynColumnStateUpdater_test/test_column_state_updater.pf M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/dyn_subgrid/dynColumnStateUpdaterMod.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/ch4Mod.F90 ========= Fix unit tests M components/clm/src/CMakeLists.txt M components/clm/src/README.unit_testing ========= Note expected failure that started failing in r226 M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_14_r226 Originator(s): erik (Erik Kluzek) Date: Thu Mar 2 15:12:50 MST 2017 One-line Summary: Update cime, mosart and cism to a version that is cesm2_0_beta05 or part of beta06 Purpose of changes ------------------ Update cime version and cism to version that similar to that in cesm2_0_beta05, but with many fixes in cime for clm. Also update mosart to fix a bug on hobart. The update of cism does change answers as it makes CISM2 the default with the new 4km greenland grid. There are some user improvements in the latest cime version. Long term archiving issues are fixed. A file called CASEROOT is added to your case's $RUNDIR so that you can easily see where your source code is. A few options were added to create_newcase and create_clone (see details below). NOTE: You will need to use the "--run-unsupported" to create_newcase for compsets/grids that aren't tested or supported!!! Fix a few other existing bugs. Fix createcrop_landunit to work with the new CLM surface datasets. Bring in a workaround for a compiler bug with intel 17 and 18. Bring in a version of PTCLM that works with the latest cime. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2417 -- fix createcrop_landunit for new datasets with 16 PFTs 2418 -- fix grid match for config_compsets.xml (cime issue 1079 still there) 2424 -- Update mosart to 1_0_20 to fix the nag bug 2479 -- workaround for intel compiler bug 2357 -- PTCLM broken in cime5 CIME Issues fixed (include issue #): archiving, unit-tests, grid issues, gnu compiler, cime issues fixed: 1075, 1053, 1051, 1073, 1072, 1089 Known bugs introduced in this tag (include bugzilla ID): 1074 -- hcru_hcru grid no longer works gives PIO error 1079 -- mask for grid matching Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Add support for cheyenne CASEROOT file added to $RUNDIR to show path to source Add "--run-unsupported" option to create_newcase Add "--cime-output-root" option to create_clone Fix long term achiving issues Changes made to namelist defaults (e.g., changed parameter values): drv namelist changes PIO namelists have different defaults some grids are null and maps are idmap when they aren't needed clm40 removed from all active compsets Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: Seems to run faster (~20%)! (we don't understand why) Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Refactoring in cime for unit-testing, add schema files for cime xml files buildnml faster config_grids and xmlquery was refactored Changes to tests or testing: Add test with mosart on hobart Code reviewed by: self, sacks CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_14_r225 Answer changes -------------- Changes answers relative to baseline: Yes. Compsets with CISM Summarize any changes to answers, i.e., - what code configurations: All with CISM - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): compsets are updated to default to CISM2 and the new 4km greenland grid Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, cism, mosart, PTCLM cime to cime5.2.0-alpha25 mosart to mosart1_0_20 cism to cism2_1_29 PTCLM to PTCLM2_170302 List all files eliminated: D components/clm/tools/shared List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/cime_config/config_compsets.xml - Workaround in 2418 for cime issue 1079 M components/clm/test/tools/TSMscript_tools.sh --- Fix arguments sent to TCB scrip =============================================================== =============================================================== Tag name: clm4_5_14_r225 Originator(s): sacks (Bill Sacks) Date: Sun Feb 19 14:42:13 MST 2017 One-line Summary: Fix GDD accumulation in new crop columns Purpose of changes ------------------ For newly-initiating crop columns in the middle of a run, the 20-year-average GDD variables (gdd*20) were not being initialized correctly. They were starting as spval, and then taking essentially forever to come down to some reasonable value. This caused major problems for both dynamic landunits and when running init_interp from one crop run to another, where the new run had some new crop types - such as going from a run without irrigation to a run with irrigation. This tag fixes these problems. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - If you do a CROP run, setting use_init_interp to .true., and pointing to a previous initial conditions file generated with a crop run (thus having restyear > 0 on the restart file), then GDD accumulation will get reset. This means that the first year of your run won't have any crop growth. After this initial "burn-in" period, it will run fine. This is only an issue when using initial conditions generated prior to this tag, and is only an issue when using init_interp. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: New initial conditions file for tests that use clm50cropIrrigMonth_interp or irrig_spunup testmods - see details below. Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, some expected baseline failures: Baseline comparisons fail because GDD accumulation is now fixed for transient crops: ERS_Ly3_Mmpi-serial.1x1_smallvilleIA.IHISTCLM50BGCCROP.yellowstone_gnu.clm-cropMonthOutput ERS_Ly6_Mmpi-serial.1x1_smallvilleIA.IHISTCLM50BGCCROP.yellowstone_gnu.clm-cropMonthOutput SMS_D_Ly6_Mmpi-serial.1x1_smallvilleIA.IHISTCLM45BGCCROP.yellowstone_pgi.clm-cropMonthOutput These tests point to a previously-spun-up crop file, and use init_interp. I have adjusted this file to add the new patch-level nyrs_crop_active variable, which I set to 30 for all active crop PFTs (in agreement with the old restyear). However, this initial conditions file does not have any non-irrigated crop PFTs! So crop behavior was wrong in the baseline run, because non-irrigated crops started with spval gdd*20 values. SMS_D_P24x1_Ld5.f10_f10.ICRUCLM50BGCCROP.hobart_nag.clm-irrig_spunup ERP_D_P15x2_Ld5.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-irrig_spunup ERP_P60x2_Lm36.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-clm50cropIrrigMonth_interp CLM tag used for the baseline comparisons: clm4_5_14_r224 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: - Transient crop cases - Any crop case for which initial conditions are init_interp'ed from a previous crop case: nyrs_crop_active is reset to 0 in this case - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Potentially new climate for transient crop cases If bitwise differences were observed, how did you show they were no worse than roundoff? n/a If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: n/a URL for LMWG diagnostics output used to validate new climate: n/a Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as noted above M components/clm/src/main/clm_driver.F90 M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CropType.F90 ========= Point to new version of initial conditions file - same as old, but hacked to have nyrs_crop_active patch-level variable: set to 30 for active crop PFTs, 0 elsewhere. This hack is needed since we run init_interp for these tests: this allows crops to grow initially. M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_14_r224 Originator(s): sacks (Bill Sacks) Date: Fri Feb 10 06:27:53 MST 2017 One-line Summary: Tweaks for glaciers and snow Purpose of changes ------------------ A number of answer-changing tweaks for glaciers and snow, mainly with the intent of improving glacial melt over Greenland in CESM2. (1) New longwave downscaling parameterization, based on Van Tricht et al. (2016, TC) Figure 6, doi:10.5194/tc-10-2379-2016. This uses a simple linear lapse rate for LW radiation, which seems better supported by observations than the old approach based on blackbody radiation relationships. Two new namelist parameters control this behavior: lapse_rate_longwave and longwave_downscaling_limit. This changes answers for all cases with multiple elevation classes - i.e., all cases with CISM (IG, etc.), just over the regions with downscaling (Greenland and Antarctica). This change is from Leo van Kampenhout. (2) Modified the parameterization of snow cover fraction during melt to improve glacial melt over Greenland in CESM2: (a) New value for n_melt (parameter controlling snow cover fraction as snow melts) over glc_mec columns for CLM5. The previous value of 10 resulted in near-100% snow cover fraction over glaciers. The new value of 1 allows snow cover fractions more similar to other landcover types. n_melt_glcmec is a new namelist parameter. This changes answers for all CLM5 cases with glc_mec (i.e., with CISM), just over glc_mec landunits. (b) For all landcover types, maintain snow cover fraction at 100% when h2osno exceeds 2m SWE in CLM5. This is important in conjunction with (a), in order to avoid snow cover fraction dropping significantly below 100% when very deep (e.g., 10m) snow packs begin to melt. This is controlled by a new namelist parameter, int_snow_max, which can be set to a very large value to retrieve the old behavior. This changes answers for all CLM5 cases; in principle, it changes answersfor all landunits, but in practice, this should only matter where we have a very deep snow pack - typically, just in regions where the snow persists for more than a year. These two changes were proposed in a discussion with Leo van Kampenhout, Bill Lipscomb, Dave Lawrence and Sean Swenson. (3) Set the ddz3 term controlling snow compaction for glaciers and wetlands in the same way as for soil and crop landunits. This should have been included in clm4_5_14_r222 (which brought in consistency in the treatment of snow cover fraction in glacier & wetland landunits), but was overlooked in that tag. This changes answers for all cases, just over glacier, glacier_mec and wetland columns. This change was reviewed by Sean Swenson and Leo van Kampenhout. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variables: - lapse_rate: the temperature lapse rate for downscaling over ice sheets (default kept the same as before, though there may be roundoff-level differences due to the precision of the namelist parameter) - lapse_rate_longwave: the longwave radiation lapse rate for downscaling over ice sheets - longwave_downscaling_limit: the relative limit capping longwave radiation downscaling - n_melt_glcmec: the value of n_melt (parameter controlling snow cover fraction as snow melts) over glc_mec columns - int_snow_max: limit applied to integrated snowfall when determining changes in snow-covered fraction during melt - INT_SNOW diagnostic now inactive by default Changes made to namelist defaults (e.g., changed parameter values): Some of the new namelist variables have defaults different from the previous hard-coded values, but no changes to existing namelist defaults. Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Removed a bunch of no-longer-used test lists: aux_clm_ys_intel, aux_clm_ys_pgi, aux_science, aux_scripts, aux_testsystem, csltiming, aux_rasm Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok tests pass, differences as expected unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, but differences from baselines and namelist differences as expected CLM tag used for the baseline comparisons: clm4_5_14_r223 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff. Expect at least small regional climate differences - specifically over Greenland and Antarctica. Answer changes are from: (1) New longwave downscaling parameterization. This changes answers for all cases with multiple elevation classes - i.e., all cases with CISM (IG, etc.), just over the regions with downscaling (Greenland and Antarctica). (2) New value for n_melt over glc_mec columns for CLM5. This changes answers for all CLM5 cases with glc_mec (i.e., with CISM), just over glc_mec landunits. (3) Modified parameterization of snow cover fraction during melt to maintain snow cover fraction at 100% when h2osno exceeds 2m SWE in CLM5. This changes answers for all CLM5 cases; in principle, it changes answersfor all landunits, but in practice, this should only matter where we have a very deep snow pack - typically, just in regions where the snow persists for more than a year. (4) Set the ddz3 term controlling snow compaction for glaciers and wetlands in the same way as for soil and crop landunits. This changes answers for all cases, just over glacier, glacier_mec and wetland columns. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= New longwave radiation downscaling; make lapse rate a namelist parameter; refactor unit tests a bit M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/atm2lndType.F90 M components/clm/src/main/test/atm2lnd_test/test_downscale_forcings.pf M components/clm/src/main/test/atm2lnd_test/test_partition_precip.pf ========= Apply int_snow_max; treat ddz3 consistently for glaciers and wetlands M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 ========= New int_snow_max and n_melt_glcmec namelist variables M components/clm/src/main/controlMod.F90 M components/clm/src/main/clm_varcon.F90 M components/clm/src/main/initVerticalMod.F90 ========= Add some inactive history fields that just apply over glaciers M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/WaterStateType.F90 ========= New setup_landunit_ncols utility method M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 ========= New namelist parameters, as documented above M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Remove unneeded use statements M components/clm/src/main/clm_driver.F90 M components/clm/src/main/clm_instMod.F90 ========= Remove no-longer-used testlists: aux_clm_ys_intel, aux_clm_ys_pgi, aux_science, aux_scripts, aux_testsystem, csltiming, aux_rasm. Main motivation is that lists starting with 'aux_*' now have special meaning, indicating what compset/grid combinations we test regularly. M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_14_r223 Originator(s): sacks (Bill Sacks) Date: Fri Feb 3 10:07:24 MST 2017 One-line Summary: Handle patch-level C&N variables correctly with dynamic landunits Purpose of changes ------------------ The infrastructure for handling conservation of patch-level C&N variables when subgrid weights changed was only considering changes in patch weights on the COLUMN. For dynamic landunits - where whole columns and landunits can change in area - this needs to consider changes in patch weights on the GRIDCELL. This tag makes this change. In addition, this makes the fire code work correctly with changing column / landunit areas: Similarly to the other C&N variables, the variable that triggers landcover-related fires was previously only considering changes in patch weights on the column. Also, treat patch-level crop-specific variables (grain pools and cropseed[c/n]_deficit) in DynamicPatchAdjustments routines. C/N in the grain pool is sent to the crop product pool via a new flux (dwt_crop_productc_gain_patch / dwt_crop_productn_gain_patch). Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2106: CNFireMod code should be re-evaluated in the presence of dynamic landunits - Addresses point 14 in 1679: problems with new CNFireMod Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - Removed history variables: DWT_CLOSS / DWT_CLOSS_DRIBBLED (and similarly for N). These can be seen with DWT_CONV_CFLUX / DWT_CONV_CFLUX_DRIBBLED (previously dwt_closs was simply set equal to dwt_conv_cflux). Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - Changes ciso_bombspike1963 to use monthly output, because this testmod is used in a long-running test Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, changes answers for some tests as noted below CLM tag used for the baseline comparisons: clm4_5_14_r222 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with transient landcover (HIST, RCP, DV, IG) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): - For transient crop and transient glacier: larger than roundoff, may change climate slightly - Without crop: expected to be only roundoff-level changes (I confirmed this in various ways fo some of the changes; for other changes, I did not do a careful confirmation that changes were truly just roundoff-level) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= For some tests that have daily changes in glacier area, bypass the error-checks that ensure that we only have changes in landcover on the year boundary. A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase/user_nl_clm List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/src/dyn_subgrid/test/dynPatchStateUpdater_test/test_patch_state_updater.pf M components/clm/src/dyn_subgrid/dynPatchStateUpdaterMod.F90 M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 M components/clm/src/dyn_subgrid/dynPriorWeightsMod.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 - also, treat initialization of cropseedc_deficit consistently with grain pools (only initialize if use_crop is .true.) M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 - also, treat initialization of cropseedc_deficit consistently with grain pools (only initialize if use_crop is .true.) M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNDriverMod.F90 M components/clm/src/biogeochem/CNProductsMod.F90 M components/clm/src/biogeochem/CNFireLi2016Mod.F90 M components/clm/src/biogeochem/CNFireLi2014Mod.F90 M components/clm/src/biogeochem/CNFireBaseMod.F90 ========= Unrelated change: Since we now only initialize cropseedc_deficit when use_crop is true, make sure we only adjust it over prognostic crop patches M components/clm/src/biogeochem/CNC14DecayMod.F90 ========= Unrelated change: use YEAR_POSITION_START_OF_TIMESTEP for consistency with other dyn file stuff. This doesn't make any difference for harvest, now that it's applied once per year at the beginning of the year, but it seems like we might as well be consistent. M components/clm/src/biogeochem/dynHarvestMod.F90 ========= Add create_vegetated_landunit_n_patches to aid unit testing M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 ========= For some tests that have daily changes in glacier area, bypass the error-checks that ensure that we only have changes in landcover on the year boundary. M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/user_nl_clm M components/clm/src/biogeochem/CNVegStateType.F90 ========= Change output to monthly to speed up this long-running test M components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_bombspike1963/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_14_r222 Originator(s): sacks (Bill Sacks) Date: Fri Jan 27 06:14:19 MST 2017 One-line Summary: Fix energy budget bug with fractional snow on glacier and wetland columns Purpose of changes ------------------ Leo van Kampenhout discovered a bug in the surface energy budget when there is fractional snow cover on glacier and wetland columns: frac_sno_eff was being treated inconsistently for these landcover types. Note that the old code did not conserve energy due to this bug! The fix slightly increases melt rates over Greenland, and has other minor effects on the simulation, but not a huge effect. More detailed notes on the problem, from Leo, are below. However, note that we ended up fixing the problem differently from what he suggested - based on Sean Swenson's suggestion: We fix this by treating glaciers and wetlands the same as soil with respect to frac_sno_eff. (Note that we still treat lakes differently [these are handled by different code] and urban differently.) Here are the more detailed notes from Leo: Background: - CLM columns are allowed to have fractional snow cover. - The first application of fractional snow cover is to the albedo calculation. This is done for every landcover type in the same fashion. This fractional snow cover is referred to as 'frac_sno'. - The second type of fractional snow cover is 'frac_sno_eff', which adjusts surface fluxes other than shortwave radiation, such as sensible and latent heat. The frac_sno_eff has only been implemented for soil and crop landcover types. - For all other landcovers, such as glaciers, frac_sno_eff =1.0 - As such, glaciers can have frac_sno < frac_sno_eff whereas those are otherwise identical Problem: - Frac_sno is used to compute absorbed radiation in the soil (sabg_soil) and snow (sabg_snow). When frac_sno < 1.0, this will lead to a positive value of sabg_soil at the expense of sabg_snow - In other radiation code, notably SoilTemperatureMod.F90, the other definition of fractional snow cover is used (frac_sno_eff) - For glaciers (frac_sno_eff), this means that no radiation is allowed to enter the soil. To my current understanding, this energy is totally lost: eflx_gnet_soil = sabg_soil(p) + dlrad(p) & + (1-frac_veg_nosno(p))*emg(c)*forc_lwrad(c) - lwrad_emit_soil(c) & - (eflx_sh_soil(p)+qflx_ev_soil(p)*htvp(c)) real(r8) , intent(out) :: hs_soil( bounds%begc: ) ! heat flux on soil [W/m2] ! separate surface fluxes for soil/snow hs_soil(c) = hs_soil(c) + eflx_gnet_soil * patch%wtcol(p) ! this is the snow/soil interface layer rt(c,j) = t_soisno(c,j) + fact(c,j) & *((1._r8-frac_sno_eff(c))*(hs_soil(c) - dhsdT(c)*t_soisno(c,j)) & + cnfac*(fn(c,j) - frac_sno_eff(c) * fn(c,j-1))) BalanceCheckMod: - The missing energy should have been detected by BalanceCheckMod, specifically the check on surface energy balance: ! Surface energy balance ! Changed to using (eflx_lwrad_net) here instead of (forc_lwrad - eflx_lwrad_out) because ! there are longwave interactions between urban columns (and therefore patches). ! For surfaces other than urban, (eflx_lwrad_net) equals (forc_lwrad - eflx_lwrad_out), ! and a separate check is done above for these terms. if (.not. lun%urbpoi(l)) then errseb(p) = sabv(p) + sabg_chk(p) + forc_lwrad(c) - eflx_lwrad_out(p) & - eflx_sh_tot(p) - eflx_lh_tot(p) - eflx_soil_grnd(p) - Here, the total absorbed solar radiation is FSA has been been replaced by sabg_chk. This hides the problem, because sabg_chk itself is also flawed: ! save sabg for balancecheck, in case frac_sno is set to zero later sabg_chk(p) = frac_sno_eff(c) * sabg_snow(p) + (1._r8 - frac_sno_eff(c) ) * sabg_soil(p) - With frac_sno_eff = 1.0, this simplifies to sabg_snow. So, sabg_soil is not accounted for in the surface energy balance check. Workaround: 1. Let all solar soil radiation be absorbed by snow: SurfaceRadiationMod: if (frac_sno(c) < frac_sno_eff(c)) then ! LvK: workaround, absorb all soil radiation in snow sabg_snow(p) = fsa(p) sabg_soil(p) = 0.0_r8 end if 2. Recommended: let the surface energy balance use FSA : errseb(p) = sabv(p) + fsa(p) + forc_lwrad(c) - eflx_lwrad_out(p) & - eflx_sh_tot(p) - eflx_lh_tot(p) - eflx_soil_grnd(p) Long-term solution: - Thought needs to be given on how SW radiation should be treated on partially covered glacier columns - Allowing radiation to penetrate the soil, yet keeping frac_sno_eff = 1, is a possibility. This will heat the snowpack from below. - Alternatively, we could look at implementing glacier fractional snow cover for surface fluxes also. Needs discussion about the physical basis & how to evaluate this. Again, note that the final implementation differs from Leo's suggestion - I'm just giving Leo's original notes for completeness. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Sean Swenson, Leo van Kampenhout, myself CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, but baseline comparisons fail as expected CLM tag used for the baseline comparisons: clm4_5_14_r221 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45/CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; expected to be similar climate (effects are relatively small) - although there are non-negligible effects on glacier surface mass balance (SMB). If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 =============================================================== =============================================================== Tag name: clm4_5_14_r221 Originator(s): sacks (Bill Sacks) Date: Tue Jan 24 05:12:14 MST 2017 One-line Summary: Change rain-snow partitioning and ice albedo for glacier columns Purpose of changes ------------------ Main change is to change the rain-snow partitioning ramp for glacier columns, when using clm5 physics: Rather than assuming all snow below 0 C and all rain above 2 C, for glacier columns we now assume all snow below -2 C and all rain above 0 C. The main motivation for this change is to increase the otherwise-too-low glacier melt we have been seeing in recent CESM2 test runs. The physical justification is: The old partitioning makes sense when you have a positive atmospheric lapse rate - so slightly-above-freezing temperatures at the surface are associated with below-freezing temperatures at the cloud level. However, for glaciers, it is relatively common to have a temperature inversion. Thus, we would expect a different relationship between near-surface temperatures and phase of precipitation. This new relationship was proposed and agreed upon by LIWG leadership, particularly Bill Lipscomb, Jan Lenaerts and Leo van Kampenhout. To implement this change, I have reimplemented the (simple) rain-snow partitioning code inline, rather than using the csm_share code. In addition, I have introduced four new namelist parameters controlling its behavior: - precip_repartition_glc_all_snow_t - precip_repartition_glc_all_rain_t - precip_repartition_nonglc_all_snow_t - precip_repartition_nonglc_all_rain_t Also changes ice albedo (albice) for glacier columns from 0.6/0.4 (vis/nir) to 0.5/0.3, when using clm5 physics. This is based on suggestions by Jan Lenaerts, and agreed upon by the LIWG leadership. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Four new namelist parameters: - precip_repartition_glc_all_snow_t - precip_repartition_glc_all_rain_t - precip_repartition_nonglc_all_snow_t - precip_repartition_nonglc_all_rain_t Changes made to namelist defaults (e.g., changed parameter values): - Changed defaults for albice with CLM50 Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - Duplicated code for rain-snow partitioning: in CLM and csm_share (the csm_share version is used by datm). However, this is just a couple of lines of code, and it's possible that the CLM version will change in the future, so I felt that duplication was not a bad thing here. Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok (pass, but differences from baseline for clm45/clm50 as expected) unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, some baseline comparisons fail as noted below CLM tag used for the baseline comparisons: clm4_5_14_r220 Answer changes -------------- Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM50 cases - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; expected to be new climate regionally, but may not change global climate significantly. I have confirmed via one test that differences are only over glacier columns. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml MM components/clm/bld/CLMBuildNamelist.pm M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/atm2lndType.F90 M components/clm/src/main/clm_instMod.F90 MM components/clm/src/main/controlMod.F90 M components/clm/src/main/test/atm2lnd_test/test_downscale_forcings.pf M components/clm/src/main/test/atm2lnd_test/test_partition_precip.pf M components/clm/src/main/clm_varctl.F90 ========= Fix unit tests for new interface for initializing atm2lnd M components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_newSnowBulkDensity.pf =============================================================== =============================================================== Tag name: clm4_5_14_r220 Originator(s): sacks (Bill Sacks) Date: Sun Jan 22 14:05:50 MST 2017 One-line Summary: Fix crop seeding Purpose of changes ------------------ The implementation of crop seeding made a few months ago didn't handle the transfers back into the "seed deficit" (cropseedc) pool quite right. In addition, that implementation didn't treat nitrogen. This tag fixes these problems. In addition, it separates the seed (deficit) pools into crop seed deficit and dwt seed deficit, and moves the latter to gridcell level (which is needed for some upcoming work). In the new implementation, when crops are seeded, this is tracked via cropseedc_deficit_patch and cropseedn_deficit_patch. At the end of the growing season, these deficits are repaid out of grain c/n, if there is enough available grain c/n (if there is not enough available, the deficits accumulate until they can eventually be repaid). Thus, this repayment decreases the fluxes of C & N to the crop product pools. This new implementation was co-developed with Danica Lombardozzi. In addition, this tag does some other minor cleanup, as noted below. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Seed pools won't restart properly from old initial conditions files. However, since these seed pools were previously just used as a diagnostic, this won't affect the evolution of anything except for the seed pools and related summary variables. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - Removed a few unused diagnostic history variables Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: nothing substantial Code reviewed by: Conceptual implementation co-developed with Danica Lombardozzi (although she has not reviewed the final code) CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; baseline failures as noted below CLM tag used for the baseline comparisons: clm4_5_14_r219 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Many CLM45/CLM50, but significant changes only for CROP - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff / same climate Details: There are differences from baseline in: - Crop product pools and related fluxes, including summary fluxes (LAND_USE_FLUX, NBP) - FCO2 - PFT_CTRUNC - changed due to fix in treatment of crop seed (previously, substantial grain was being truncated) - SEEDC / SEEDN - STORVEGC / STORVEGN in crop tests - Various other summary C/N state variables - LITFALL for CLM50 with CROP Other than TOTCOLC / TOTCOLN, greater-than-roundoff diffs just in CROP tests and tests that have non-blank finidat (latter just in SEEDC & SEEDN, because we don't read in the current value of SEEDC/SEEDN). Summary variables have diffs in some CLM45 crop tests as well as CLM50, because of diffs in which summary variables include seed c/n. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes to crop seed deficit tracking, and moving dwt seed pools to gridcell level M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 M components/clm/src/biogeochem/CNNStateUpdate1Mod.F90 M components/clm/src/biogeochem/CNC14DecayMod.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 - Also remove grainc_to_food from LITFALL diagnostic when sending grain to crop product pool M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 ========= Refactor some loops to support upcoming changes, and remove no-longer-needed code due to removed variabes M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 ========= Change arguments used in calling changed routines M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNDriverMod.F90 ========= Remove unused lf_conv_cflux_col variable and LF_CONV_CFLUX diagnostic field M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 M components/clm/src/main/clm_driver.F90 ========= Just add some comments M components/clm/src/dyn_subgrid/dynPatchStateUpdaterMod.F90 ========= Change a field that is output, because I have removed the old field M components/clm/cime_config/testdefs/testmods_dirs/clm/crop/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_14_r219 Originator(s): erik (Erik Kluzek) Date: Sat Jan 21 23:36:23 MST 2017 One-line Summary: Answer changes for CLM50 for several issues that lead to dead crops Purpose of changes ------------------ Fix several problems that led to dead crops. latbaset was set in InitAllocate and then reset to spval in InitHist. Set lnca to interp rather than skip for interpolation on startup. Remove rootfr from restart unless use_dynroots is on. Fix FERTNITRO and add BTRAN2 to history. Turn LUNA on by default for all CLM50 cases including SP. Change from Fang Li to remove fb from the calculation of baf_crop. Fix LUNA in SP mode by setting PARi10 to a small value rather than zero. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): [If none, remove this line] bug 2407 -- latbaset is set in InitAllocate and then reset to spval in InitHist bug 2409 -- set lnca to interp rather than skip on restart bug 2408 -- remove rootfr from restart unless use_dynroots is on bug 2402 -- FERTNITRO bug 2403 -- BTRAN2 bug 2413 -- luna in SP mode Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] bug 2414 LATBASET is always output to history even if not used and even though it's a constant Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self, oleson, dlawren CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_14_r218 Answer changes -------------- Changes answers relative to baseline: YES! CLM50 with crop CLM45 BGCCROP cases show an answer change just because fill value change of LATBASET even though it's unused Summarize any changes to answers, i.e., - what code configurations: CLM50 with CROP - what platforms/compilers: All - nature of change: new climate with crops growing Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- LUNA on by default for ALL CLM50 cases included SP mode M components/clm/src/biogeochem/CNFireLi2016Mod.F90 --- Remove fb from calculation of baf_crop M components/clm/src/biogeochem/CropType.F90 ---------- Move initialization of latbaset_patch to InitCold, fix FERTNITRO on history file M components/clm/src/biogeophys/EnergyFluxType.F90 ---- Save BTRAN2 to history files M components/clm/src/biogeophys/LunaMod.F90 ----------- Set PARi10 to 0.01 rather than zero if LAI is zero M components/clm/src/biogeophys/PhotosynthesisMod.F90 - Set lnca_patch to interp rather than skip M components/clm/src/biogeophys/SoilStateType.F90 ----- Only save rootfr if use_dynroots on. =============================================================== =============================================================== Tag name: clm4_5_14_r218 Originator(s): sacks (Bill Sacks) Date: Fri Jan 13 19:46:27 MST 2017 One-line Summary: For newly initiating columns: do not copy snow state Purpose of changes ------------------ Keith Oleson found that, in some circumstances, the model was dying with a snow balance error in newly initiating crop columns. We tracked this down to an incomplete copying of the snow state from the 'template' column to the newly initiating crop column. Our two options for solving this problem were (1) copy the snow state more completely; or (2) do not try to copy any part of the snow state - leaving the snow in cold start conditions for the newly initiating column. (1) is challenging because there are so many snow variables, so we have chosen to go with (2) for now. Thus, when a column initiates (for now, this just applies to crop columns), we (Dave Lawrence, Keith Oleson and Bill Sacks) have decided that we should copy only the belowground temperature and water state. Also, some minor, unrelated changes: (1) Fix broken unit tests (2) Abort run if try to run use_ed with transient crops (3) Add two single-point exact restart tests for transient crop, because we did not have good coverage of exact restart tests for transient crop before this Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Added two tests: ERS_Ly3_Mmpi-serial.1x1_smallvilleIA.IHISTCLM50BGCCROP.yellowstone_gnu.clm-cropMonthOutput # restart is right before the transition from 100% nat veg to 100% crop ERS_Ly6_Mmpi-serial.1x1_smallvilleIA.IHISTCLM50BGCCROP.yellowstone_gnu.clm-cropMonthOutput # restart is right before increasing natural veg to > 0 while also shifting PCT_CFT Code reviewed by: Main change reviewed by Keith Oleson and myself CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run On my Mac - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass In principle, answers could change for runs with transient crops, but I did not observe any answer changes in the test suite. CLM tag used for the baseline comparisons: clm4_5_14_r217 Answer changes -------------- Changes answers relative to baseline: YES, in principle - but not in the test suite If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with transient crops - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Assumed to be larger than roundoff, but only slightly, and only rare answer changes I did not observe any answer changes in the test suite. But in principle, this could change answers for any transient crop run, as evidenced by the change in behavior seen in Keith Oleson's transient crop run with and without this fix. It just seems like the conditions in which this matters only occur rarely. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main change: only copy belowground state to new columns M components/clm/src/dyn_subgrid/dynInitColumnsMod.F90 M components/clm/src/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf ========= Abort run if try to run use_ed with transient crops M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 ========= Fix a comment M components/clm/src/dyn_subgrid/dynVarMod.F90.in ========= Fix unit tests, apparently broken in r215 M components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90.in ========= Add two tests (see above) M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_14_r217 Originator(s): sacks (Bill Sacks) Date: Fri Jan 13 05:36:28 MST 2017 One-line Summary: Bug fix for snow radiation absorption Purpose of changes ------------------ Bug fix from Leo van Kampenhout regarding snow radiation absorption: Remove multipliers of (1.-albsnd(c,ib)) or (1.-albsni(c,ib)) in the calculation of flx_absdv, flx_absiv, flx_absdn, flx_absin. This stops this conditional in SurfaceRadiationMod from being triggered every time step: if (abs(sabg_snl_sum-sabg_snow(p)) > 0.00001_r8) then The triggering of this conditional led to an unintended constant partitioning of radiation through the snowpack: 75% of radiation was absorbed in the top layer, 25% in the bottom layer. In the old code, Leo found that sabg_snow(p) is over three times as big as sabg_snl_sum , so these are certainly not roundoff problems. With the fix, the conditional is no longer triggered at every time step. Possible explanation: the factors flx_absd_snw and flx_absi_snw are already accounted for albedo, because they are "absorbed flux per unit incident flux at top of snowpack" according to SnowSnicarMod.F90. Sean Swenson tested this, and says: I agree with leo; in fact the sum of the flx_ terms should be the total absorption, or 1-albedo. When I remove the albedo modification leo suggested, I see most of the cases where the if-block is triggered go away. Those that remain are much smaller (the abs() in the if block is ~1.e--4). I don't know if this is a precision issue, or an indication of a remaining bug. Or it could be those cases for which the if-block was meant to correct originally. Sean later added: a couple more pieces of information: 1) the if-statement is still needed. If it is removed, a balance check error will abort the model. 2) in SurfaceAlbedoMod, I tested the equality of sum_over_layers(flx_absdn) and (1.0 - albsnd). For direct beam, they were equal, but occassionally for ib=2 (near infrared), they were slightly different. These were the same points that were then triggered by the if block in SurfaceRadiation. This may point to something within the snicar_rt subroutine, but I don't plan on following this issue further. Mark Flanner added: I will note that there is an energy conservation check in the snicarRT code which should ensure that the sum of absorption in all snow layers never differs substantially (>0.00001) from what is expected from albedo and transmittance. If it does, the model aborts. This is true, however, only at the snow column level, so I suspect the issue you identified ultimately relates either to (1) fractional snow cover and merging of sub-grid energy fluxes to the gridcell level and/or (2) a change in the number of snow layers in between the calculations of snow albedo and absorbed energy flux. I think (2) is what necessitated the patch in the first place. I could be wrong. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Sean Swenson CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, baseline comparisons fail as expected CLM tag used for the baseline comparisons: clm4_5_14_r216 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45/CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Based on Sean Swenson's analysis: larger than roundoff; same or only slightly changed climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Change described above M components/clm/src/biogeophys/SurfaceAlbedoMod.F90 =============================================================== =============================================================== Tag name: clm4_5_14_r216 Originator(s): erik (Erik Kluzek,UCAR/TSS,303-497-1326) Date: Thu Jan 12 16:12:45 MST 2017 One-line Summary: Fix previous broken tag Purpose of changes ------------------ Three files were missing from the merge to transcrop_n08_clm4_5_14_r214 on the trunk. I don't know why they were missing, but they were. I have on very rare occasions seen subversion fail with this problem, and it maybe due to a rare bug in subversion. Manytimes I double check to make sure this doesn't happen, but I didn't this time. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): 2400 2400 --- Problem with mksurfdata_map for smallville landuse.timeseries (found in clm4_5_14_r215 after it's creation) Notes of particular relevance for users --------------------------------------- Caveats for users None Changes to CLM's user interface: None Changes made to namelist defaults: None Changes to the datasets: None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: None Changes to tests or testing: None Code reviewed by: self CLM testing: short [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: None unit-tests (components/clm/src): None tools-tests (components/clm/test/tools): None PTCLM testing (components/clm/tools/shared/PTCLM/test): None regular tests: clm_short_50 yellowstone_intel - PASS CLM tag used for the baseline comparisons: none Answer changes None -------------- Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: >>>>>> Bring in files that were modified on the branch transcrop_n08_clm4_5_14_r214, but for some reason were dropped on >>>>>> the merge to the trunk M components/clm/src/dyn_subgrid/dyncropFileMod.F90 M components/clm/src/main/clm_initializeMod.F90 M parse_cime.cs.status =============================================================== =============================================================== Tag name: clm4_5_14_r215 Originator(s): erik (Erik Kluzek) Date: Thu Jan 12 10:29:52 MST 2017 One-line Summary: Update mksurfdata_map to work with new updated datasets from Peter, and use hand-edited surfdata/landuse.timeseries datasets for f09/f19 Purpose of changes ------------------ Use new rawdatasets from Peter Lawrence that have crop landunit changing in size as well as specific crop and natural vegetation types. Modify mksurfdata_map to use new fields for fertilizer for each crop type. Add in harvest fields to both surface dataset as well as landuse.timeseries dataset. Also add in two new fields: UNREPRESENTED_PFT_LULCC, UNREPRESENTED_CFT_LULCC for the impact of landuse changes for natural and crop vegetation types. Currently these two fields are zeroed out. But, they will be added in with a new update to the datasets. Implement crop fertilizer for each crop type into CLM. The fertilizer is also moved from one crop type to another by the mapping system between the different crop types (to handle unrepresented crop types by mapping them to one that is represented). The new datasets also have harvest fields in units by mass rather than area. This requires new surface and landuse.timeseries datasets. We have hand-edited versions from Peter Lawrence for f09 and f19. The older datasets for other resolutions still work, but they don't have the new fertilizer field so anthropogenic fertilizer is zeroed out. NOTE: Capability to create rcp scenarios for landuse.timeseries was removed from mksurfdata_map. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2386 -- buildnamelist unit tester can't compare to itself 2389 -- make nitrogen balance threshold higher 2382 -- soil depth dataset into subversion 2369 -- Remove hacks to change PFT's in tropics in mksurfdata_map 2097 -- mksurfdata_map changes Known bugs introduced in this tag (include bugzilla ID): 2399 -- Problem with urban points with latest mksurfdata_map Notes of particular relevance for users --------------------------------------- Caveats for users: mksurfdata.pl "-crop" option changed to "-no_crop" so the default is to create datasets with crop on it. Also default for -glc_nec is now 10 so default is to create datasets with glacier Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New optional history field: FERTNITRO', 'gN/m2/yr', 'Nitrogen fertilizer for each crop Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files): New f09/f19 surfdata/landuse.timeseries Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: Changes to tests or testing: Removed clm4_0 tools tests Code reviewed by: self, lawrence CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - OK regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_14_r214 Answer changes -------------- Changes answers relative to baseline: f09 and f19 CLM45/CLM50 cases change answers because of new datasets Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 for f09 and f19 - what platforms/compilers: all - nature of change: adjusted surface datasets Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: Remove clm4_0 tools, rename clm4_5/shared directories under CLM tools D components/clm/tools/clm4_0/* D components/clm/tools/clm4_5/mksurfdata_map/* D components/clm/tools/clm4_5/refactorTools/* D components/clm/tools/shared/mkmapdata/* D components/clm/tools/shared/mkmapgrids/* D components/clm/tools/shared/mkprocdata_map/* D components/clm/tools/shared/ncl_scripts/* D components/clm/tools/clm4_5/mksurfdata_map/landuse_timeseries_hist_16pfts_simyr1850-2005.txt List all files added and what they do: Renamed from clm4_5 or shared subdirectories, unit tests for mkharvest A components/clm/tools/mksurfdata_map/* A components/clm/tools/refactorTools/* A components/clm/tools/mkmapdata/* A components/clm/tools/mkmapgrids/* A components/clm/tools/mkprocdata_map/* A components/clm/tools/ncl_scripts/* A components/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkharvest.F90 A components/clm/tools/clm4_5/mksurfdata_map/unit_testers/inputs/harvestfieldsold.nc A components/clm/tools/clm4_5/mksurfdata_map/unit_testers/inputs/harvestfields.nc A components/clm/tools/clm4_5/mksurfdata_map/landuse_timeseries_hist_78pfts_simyr1850-2015.txt List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNBalanceCheckMod.F90 - Increase threshold for Nitrogen balance check M components/clm/src/biogeochem/CropType.F90 ---------- Add fertnitro to fert array M components/clm/src/biogeochem/CNPhenologyMod.F90 ---- Add fertnitro_patch and InitCold for it Add FERTNITRO as optional output to history files (default inactive) M components/clm/src/biogeochem/dynHarvestMod.F90 ----- Read in harvest units, divide by deadstem if new units of gC/m2/yr, otherwise take directly M components/clm/src/main/surfrdMod.F90 --------------- Read in CONST_FERTNITRO_CFT if on file, otherwise zero it out, pass fert_cft to collapse_crop_types M components/clm/src/main/clm_varsur.F90 -------------- Add fert_cft as temporary read in from dataset M components/clm/src/main/clm_driver.F90 -------------- Pass crop_inst to dynSubgrid_driver M components/clm/src/main/surfrdUtilsMod.F90 ---------- Handle fert_cft in collapse_crop_types if crops are being merged also move the fertilizer M components/clm/src/main/clm_initializeMod.F90 ------- Allocate/dealloc fert_cft, pass crop_inst to dynsubgrid_init M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ------ Pass crop_inst in and down to dyncrop_interp M components/clm/src/dyn_subgrid/dynVarMod.F90.in ------------- Add allow_nodata and data_on_file flags and get_att method, if allow_nodata is true, read_variable will NOT abort, but zero out the variable M components/clm/src/dyn_subgrid/dynVarTimeUninterpMod.F90.in - Pass allow_nodata as optional argument down M components/clm/src/dyn_subgrid/dyncropFileMod.F90 ----------- Add fertcft dyn_var_time_uninterp_type initialize it, pass crop_inst to dyncrop_interp, get current data from fertcft%get_current_data and pass to collapse_crop_types M components/clm/bld/listDefaultNamelist.pl -------------- Remove phys option maxpft is now 78 M components/clm/bld/unit_testers/build-namelist_test.pl - Fix bug 2386 for comparing to itself for user_nl_clm_real_parameters files M components/clm/bld/unit_testers/NMLTest/CompFiles.pm - - Fix bug 2386 for comparing to itself, change double quotes in filenames to a "+" like other special characters M components/clm/bld/README ------------------------------ Remove note on clm4_0_tools file M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - Updated (hand-edited) surface and historical landuse.timeseries datasets for f09 and f19 M components/clm/bld/namelist_files/namelist_defaults_clm4_0_tools.xml - Remove contents (keep file around as queryDefaultNamelist.pl assumes it's existance) M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Add "hrv" hgrid/lmask option for 0.9x1.25/GDRC, add "hrv" mksrf_filename called mksrf_fhrvtyp use lnd/clm2/rawdata/mksf_soilthk_5x5min_ORNL_SOILS_simyr1900-2015_c150701.nc file in place of file pointing to Sean Swenson's personal disk space, remove: mksrf_fvegtyp for test years, crop=off, and all rcp's as well as before 1850, but add historical through 2015 Add harvest datasets at f09 resolution M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add mksrf_fhrvtyp M components/clm/cime_config/config_compsets.xml -- Add IGMHISTCRUCLM50BGCCROP compset M components/clm/tools/mksurfdata_map/Makefile.data ------------- Change default list of files created so crop is output for all resolutions M components/clm/tools/mksurfdata_map/src/mkharvestMod.F90 ------ Add harvestDataType and it's methods: init, get1DFieldPtr, get2DFieldPtr, getFieldsIdx, getFieldsDim, isField1D, isField2D, num1DFields num2DFields, and clean. Increase by three harvest fields, have two names for them (one for surfdace dataset, and one for landuse.timeseries datasets), mkharvest subroutine sends harvestDataType as it's output M components/clm/tools/mksurfdata_map/src/mkncdio.F90 ----------- Add varexists option to check_ret, make nf_enotatt and nf_strerror public M components/clm/tools/mksurfdata_map/src/mkfileMod.F90 --------- Send harvdata to mkfile use it's methods to define and output the harvest data M components/clm/tools/mksurfdata_map/src/mkpftMod.F90 ---------- Always put CFT's on seperate land-unit new variables for: natural-veg-percent, crop-percent, fraction for each CFT and PFT handles oldformat and newformat (oldformat only used when PFT's are specified in namelist) remove missing_crops logic as crop is now required M components/clm/tools/mksurfdata_map/src/mkvarctl.F90 ---------- Add mksrf_fhrvtyp M components/clm/tools/mksurfdata_map/src/mkpctPftTypeMod.F90 --- Add new constructor constructor_pong to initialize a new pct_pft_type object with all PFT's on the gridcell M components/clm/tools/mksurfdata_map/src/mksurfdat.F90 --------- Update mkharvestMod calls and add harvestDataType, add read of harvest dataset names, handle harvest 1D and 2D data. Remove hack to set pfts 7 and 10 to 6 in the tropics because of bad LAI's (2369) M components/clm/tools/mksurfdata_map/unit_testers/Srcfiles ---------------- Add new test files M components/clm/tools/mksurfdata_map/unit_testers/test_mksurfdata_map.F90 - add mkharvest tests and test_get_nonexisting_var for test_mkncdio M components/clm/tools/mksurfdata_map/unit_testers/test_mkncdio.F90 -------- Add test_get_nonexisting_var M components/clm/tools/mksurfdata_map/mksurfdata_map.namelist - Updated files, crop and glc_nec=10 is on, does historical range of: 1850-2005 M components/clm/tools/mksurfdata_map/mksurfdata.pl ----------- -crop=>-no_crop glc_nec default to 10, default is to create datasets with crop, rawdatasets always have crop=on for them, harvest resolution different, remove -rcp option (no longer available), year range 1850-2000 actually does 1850-2015 now. =============================================================== =============================================================== Tag name: clm4_5_14_r214 Originator(s): sacks (Bill Sacks) Date: Tue Jan 3 12:55:28 MST 2017 One-line Summary: Update column states based on dwt terms between dyn_cnbal_patch and dyn_cnbal_col, and minor fixes Purpose of changes ------------------ Main change is to move the update of column states based on dwt terms (which come from shrinking patches) to earlier in the time step - specifically, between dyn_cnbal_patch and dyn_cnbal_col. This includes updating the seed pools and decomp_[c/n]pools_vr_col. This change is important when there are changing column areas: In order to conserve C & N, we need to transfer the mass from the dwt terms to the column-level pools before executing the code to conserve the column-level state variables. For example, if the natural vegetated column shrinks (with a corresponding shrinkage of each patch on the column), and is replaced by crop: Then the dwt terms from the shrinking patches (assuming this was handled correctly - fix coming in an upcoming tag...) add some C & N to the below-ground pools. But some of this C & N needs to end up in the growing crop column. Since the column-level conservation operates on state variables (not fluxes), this only works properly if the mass from the dwt fluxes is added to the states before dyn_cnbal_col. In addition, this tag fixes a number of minor problems: (1) Separates the crop uses of seed(c/n)_to_leaf from the dwt uses. This is needed for the main change to work correctly. (2) Allows longer history field names in hist_fincl and hist_fexcl (3) Fixes SNOINTABS diagnostic field (bug 2321) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2321: SNOINTABS diagnostic field incorrect for columns with more than one patch Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Changes meaning of DWT_SEEDC_TO_LEAF andDWT_SEEDN_TO_LEAF: No longer includes crop seed terms Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Minor tweak to crop testmod: Add a history field that uses SUM, in order to test the SUM history field infrastructure Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, baseline failures as expected, as noted below CLM tag used for the baseline comparisons: clm4_5_14_r213 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 in these configurations: - All transient cases (HIST, RCP, DV): changes due to the main changes in this tag - All CROP cases: changes answers in DWT_SEEDC_TO_LEAF and DWT_SEEDN_TO_LEAF diagnostic fields - Many cases: differences in SNOINTABS diagnostic field. From one test (LII_D_Ld3.f19_g16_gl5.IG1850CLM45.yellowstone_intel.clm-glcMEC_spunup_1way), there are diffs of up to 0.06. No diffs over Antarctica or most of Greenland (though there are some diffs on the southern tip of Greenland). Lots of other diffs in the northern hemisphere mid-high latitudes. This is as expected: expect diffs in vegetated areas, but not in glacier. - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; not investigated as to whether this is new climate, but expected not to be If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Move update of column states based on dwt terms to earlier in the time step M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 M components/clm/src/biogeochem/CNNStateUpdate1Mod.F90 M components/clm/src/biogeochem/CNBalanceCheckMod.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Separate the crop uses of seed(c/n)_to_leaf from the dwt uses M components/clm/src/biogeochem/CNPhenologyMod.F90 ========= Change sub_surf_abs_SW to a patch-level variable, rather than column-level (fixes bug 2321) M components/clm/src/biogeophys/SurfaceRadiationMod.F90 M components/clm/src/biogeophys/SolarAbsorbedType.F90 ========= Allow longer history field names in hist_fincl and hist_fexcl M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ========= Add a history field that uses SUM, in order to test the SUM history field infrastructure M components/clm/cime_config/testdefs/testmods_dirs/clm/crop/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_14_r213 Originator(s): sacks (Bill Sacks) Date: Fri Dec 30 08:48:44 MST 2016 One-line Summary: Convert buildnml to python Purpose of changes ------------------ Changes from Mariana Vertenstein Main purpose is to convert buildnml to python; also some minor tweaks to other related files, which seem to be moving towards the removal of CLM4. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - pass I did NOT compare against baseline, because this is currently broken (bug 2386) unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_13_r211 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: A components/clm/cime_config/buildcpp List all existing files that have been modified, and describe the changes: M components/clm/cime_config/buildlib M components/clm/cime_config/buildnml M components/clm/cime_config/config_component.xml =============================================================== =============================================================== Tag name: clm4_5_14_r212 Originator(s): sacks (Bill Sacks) Date: Thu Dec 22 16:06:09 MST 2016 One-line Summary: Bump science version number due to changes in clm4_5_13_r211 Purpose of changes ------------------ Bump science version number only. This is the same code base as clm4_5_13_r211. No additional testing was done. clm4_5_13_r211 has a significant change in behavior for transient cases, for CLM45 as well as CLM50 cases. Thus, we are incrementing the science version number. =============================================================== =============================================================== Tag name: clm4_5_13_r211 Originator(s): sacks (Bill Sacks) Date: Thu Dec 22 13:46:52 MST 2016 One-line Summary: Apply transient land cover and harvest all at once in first time step of year Purpose of changes ------------------ Scientists have found it problematic to have land cover change and harvest interpolated throughout the year: By having just a bit of deforestation done in each time step, the forest can regrow, so that you never see the full, appropriate impacts of land cover change. This tag changes the behavior of both land cover change (transient PFTs) and harvest so that they are applied all at once in the first time step of a given year. This new behavior is consistent with the pre-existing timing of transient crops and transient glaciers: Now all land cover changes happen all at once on the first time step of the year. For the fire code, we pass a dribbled version of the land cover change - i.e., a version that smooths the area changes over the year, similarly to what was done before. This is important because the fire code involves interactions between the current time step's change in area and the current time step's climate. So, if the fire code only saw area changes on Jan 1, it could only generated deforestation fires on Jan 1 (I think). Note there is some inconsistency now, since the above-ground C/N biomass is all adjusted at once on Jan 1, so the fire code operates on the already-adjusted biomass. This could be problematic in principle if there were relatively large changes in area in a given year, but Fang Li believes this won't cause major problems in practice, since (1) area changes are typically small, (2) the carbon fluxes from fire are just diagnostic, and (3) the bigger impact of deforestation fires is as a trigger for other fires. In addition, for fluxes passed to the atmosphere from land cover change and harvest, we dribble these fluxes over the year, avoiding pulsed fluxes on Jan. 1. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - Partially addresses 2106: CNFireMod code should be re-evaluated in the presence of dynamic landunits Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - Moved some transient tests (HIST/RCP) from CLM45 to CLM50 Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; some baseline comparisons fail as noted below NOTE: Most testing done on dynpft_and_harvest_jan1_n07_clm4_5_13_r210. dynpft_and_harvest_jan1_n08_clm4_5_13_r210 fixed an error check for CNDV cases. I reran the CNDV tests on the n08 tag, along with the following additional tests: SMS_Ld5_D_P24x1.f10_f10.IHISTCLM50BGC.hobart_nag.clm-monthly ERP_D_P24x1.f10_f10.IHISTCLM50BGC.hobart_nag.clm-decStart ERP_D_Ld5.f19_g16.IRCP85CLM45BGC.yellowstone_pgi.clm-drydepnomegan CLM tag used for the baseline comparisons: clm4_5_13_r210 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All transient cases (HIST/RCP) with CLM45/CLM50; also DV cases - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): For transient cases: Larger than roundoff; unsure whether this is climate changing For DV cases, just roundoff-level changes in FCO2, LAND_USE_FLUX, NBP - I think from the dribbling of fluxes that are generated on the first time step of each year. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Do land cover change (transient PFTs) and harvest all at once in the first time step of the year, rather than gradually throughout the year; dribble the fluxes to the atmosphere, and dribble the weight changes to the fire code. M components/clm/src/dyn_subgrid/dynpftFileMod.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/biogeochem/dynHarvestMod.F90 M components/clm/src/biogeochem/CNFireLi2014Mod.F90 M components/clm/src/biogeochem/CNFireLi2016Mod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNVegStateType.F90 ========= Just add some documentation for history fields M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 ========= Just remove unneeded use statements M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Introduce ability for dribblers to work at any subgrid level, and some other feature additions M components/clm/src/utils/test/annual_flux_dribbler_test/test_annual_flux_dribbler.pf M components/clm/src/utils/AnnualFluxDribbler.F90 M components/clm/src/main/decompMod.F90 M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/WaterfluxType.F90 ========= Move some transient tests to CLM50 M components/clm/cime_config/config_compsets.xml M components/clm/cime_config/testdefs/testlist_clm.xml ========= Unrelated change: add a bit of documentation M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml =============================================================== =============================================================== Tag name: clm4_5_13_r210 Originator(s): sacks (Bill Sacks) Date: Sat Dec 17 05:43:34 MST 2016 One-line Summary: Update cime to cesm2_0_beta04 version Purpose of changes ------------------ Update cime to version used in cesm2_0_beta04 Main motivation is to get the fix for the problem where manual modifications to user_nl files in a test case get overwritten. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - pass ran build-namelist tests without compare or generate - all passed ran prove test_build_namelist.pl - pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass (NLCOMP changes for IG tests; this was expected) CLM tag used for the baseline comparisons: clm4_5_13_r209 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime: cime5.2.0-alpha.7 -> cime5.2.0-alpha.9 List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Update cime from cime5.2.0-alpha.7 to cime5.2.0-alpha.9 M SVN_EXTERNAL_DIRECTORIES =============================================================== =============================================================== Tag name: clm4_5_13_r209 Originator(s): sacks (Bill Sacks) Date: Sun Dec 11 06:09:22 MST 2016 One-line Summary: Change definition of NBP; bring back some r205 changes that were lost in r207 Purpose of changes ------------------ Changes definition of NBP to match the definition used by CMIP: Now NBP includes all terms that contribute to the net carbon exchange between land and atmosphere. Also, moves both NBP and NEE diagnostics to the gridcell level, rather than the column level - so you can no longer see column-by-column values for these diagnostics. The move of NBP to gridcell-level is needed because it now includes some gridcell-level quantities; the move of NEE to gridcell-level is needed to support some upcoming changes. Removes NET_CARBON_EXCHANGE diagnostic output because this is now redundant with NBP. Also, brings back some r205 changes that were lost in r207 - most notably, the fix for the MEGAN crash due to negative intracellular co2. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2371: Change definition of NBP Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - Changes definition of NBP diagnostic output; also, NBP is now defined at the gridcell level - Removes NET_CARBON_EXCHANGE diagnostic output - NEE diagnostic output is now defined at the gridcell level Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self; but change in definition of NBP was reviewed conceptually by Dave Lawrence CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; baseline comparison failures as noted below CLM tag used for the baseline comparisons: clm4_5_13_r208 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 change answers as noted below - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Most changes are roundoff-level changes in diagnostics and in FCO2 sent to atmosphere; larger than roundoff-level changes in the diagnostic field NBP; potentially larger-than-roundoff level changes in MEG_isoprene Roundoff-level diffs in FCO2 (sent to atmosphere), NEE, C13_NEE, C14_NEE NBP diagnostic field: roundoff-level change for many cases; larger than roundoff for transient, crop, or year-2000 using initial conditions from a transient case (due to non-zero product pools) Small diffs in MEG_isoprene in SMS_Lm1.f09_g16_gl4.IG1850CRUCLM50BGC.yellowstone_intel.clm-clm50KitchenSink, due to bringing back the fix for Megan. (In principle, this could show up in other cases as well.) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Change definition of NBP, and other changes to carbon flux diagnostics as noted above M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 ========= Bring back fix for Megan crash (prevent negative intracellular CO2) that was added in r205 and accidentally dropped in r207; also, change 1. to 1._r8 M components/clm/src/biogeophys/PhotosynthesisMod.F90 ========= Bring back SNO_TK_ICE diagnostic field that was added in r205 and accidentally dropped in r207 M components/clm/src/biogeophys/SoilStateType.F90 =============================================================== =============================================================== Tag name: clm4_5_13_r208 Originator(s): sacks (Bill Sacks) Date: Sat Dec 10 15:00:21 MST 2016 One-line Summary: Rework irrigation; remove glc_smb logical Purpose of changes ------------------ Main change is a complete rework of the irrigation algorithm. This reworks the trigger and target for irrigation to be based on soil physical properties, removing the dependence on btran and smpso (which were problematic in conjunction with PHS). With the current out-of-the-box parameter values, irrigation is substantially lower than before. In addition, this reworks the river volume-based limitation. One change to the volr-based limitation is that we now apply the limit to the deficit computed in calcIrrigationNeeded - rather than applying the limit in applyIrrigation. The point of this is to avoid problems that arise due to evolving VOLR - e.g., if we had calculated irrigation demand such that VOLR was just barely big enough, then we apply irrigation for some time steps, then we get an updated VOLR which is lower due to these irrigation withdrawals - the old implementation would impose an unnecessary limit on the irrigation to be applied for the rest of this day. It's hard to see a clean way around this problem, given that VOLR is updated in some but not all CLM time steps. So we're going to impose the limit on irrigation deficit, rather than on the time step-by-time-step flux. The downside here is that it's somewhat more likely that irrigation would try to draw volr negative, if volr has been evolving for other reasons. Also, this tag pulls out the parameters controlling irrigation to namelist. Also reworks the limit_irrigation namelist parameter, changing it to limit_irrigation_if_rof_enabled. This was a problem before, in that we needed to remember to set it to false whenever we ran without an active ROF model (e.g., during spinup); now runtime logic handles that automatically. (We considered trying to have limit_irrigation set properly by build-namelist, but that turns out to be tricky because there are a number of ways in which you can end up with an inactive ROF model.) Also removes the glc_smb logical: this has been unused for quite a while (we have always assumed it's true). Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variables for irrigation - limit_irrigation changed to limit_irrigation_if_rof_enabled - Remove glc_smb namelist variable, both for clm4_0 and clm4_5/clm5_0 - Renamed ICE_MASK history variable to ICE_MODEL_FRACTION Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: minor tweaks to tests involving irrigation, as described below Code reviewed by: Irrigation changes reviewed by Sean Swenson CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok Many failures due to bug 2386; other than that, expected differences from baseline unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - ok ok means pass or xfail; expected baseline failures as noted below CLM tag used for the baseline comparisons: clm4_5_13_r207 Answer changes -------------- Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All cases with irrigation - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated closely, but potentially new climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): List all files eliminated: ========= Merge irrigation unit tests into a single module D components/clm/src/biogeophys/test/Irrigation_test/test_irrigation_singlepatch.pf D components/clm/src/biogeophys/test/Irrigation_test/test_irrigation_deficit.pf D components/clm/src/biogeophys/test/Irrigation_test/test_irrigation_multipatch.pf D components/clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 ========= No longer needed because we turn on irrigation river limitation in some other irrigation tests D components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw List all files added and what they do: ========= Merge irrigation unit tests into a single module, and add some tests A components/clm/src/biogeophys/test/Irrigation_test/test_irrigation.pf List all existing files that have been modified, and describe the changes: ========= Rework irrigation: both the core algorithm and the volr limitation M components/clm/src/biogeophys/IrrigationMod.F90 M components/clm/src/main/clm_driver.F90 ========= Determine if rof_enabled is true, needed now that we have changed limit_irrigation to limit_irrigation_if_rof_enabled M components/clm/src/cpl/lnd_comp_mct.F90 ========= Add new irrigation namelist variables, in their own group; also remove glc_smb namelist item M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm M components/clm/src/main/clm_instMod.F90 ========= Remove glc_smb logical: we have been assuming it's true for quite a while M components/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/glacierMEC_pd.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M components/clm/bld/namelist_files/namelist_definition_clm4_0.xml M components/clm/bld/namelist_files/namelist_defaults_overall.xml M components/clm/cime_config/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/fire_emis/user_nl_clm M components/clm/bld/unit_testers/build-namelist_test.pl M components/clm/cime_config/buildnml M components/clm/doc/UsersGuide/preface.xml M components/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M components/clm/src_clm40/main/clm_glclnd.F90 M components/clm/src_clm40/main/controlMod.F90 M components/clm/src_clm40/main/clm_varctl.F90 ========= Remove glc_smb and move limit_irrigation logic into IrrigationMod M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/controlMod.F90 ========= Rework some of the column filter creators to have a clearer interface M components/clm/src/main/filterColMod.F90 M components/clm/src/main/test/filter_test/test_filter_col.pf M components/clm/src/main/TopoMod.F90 M components/clm/src/main/test/atm2lnd_test/test_downscale_forcings.pf ========= Put some extra metadata on the restart file: whether we have create_crop_landunit and irrigate, since these are helpful for understanding when init_interp is needed M components/clm/src/main/restFileMod.F90 ========= Rename ICE_MASK to ICE_MODEL_FRACTION, based on suggestion from Marcus Lofverstrom M components/clm/src/main/glc2lndMod.F90 ========= Merge irrigation unit tests into a single module M components/clm/src/biogeophys/test/Irrigation_test/CMakeLists.txt ========= Fix for cime update M components/clm/bld/test_build_namelist/test_build_namelist.pl ========= Stop turning off limit_irrigation, since it's working now; also include QIRRIG_DEMAND in output M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/user_nl_clm ========= Rename limit_irrigation to limit_irrigation_if_rof_enabled M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm ========= Include QIRRIG_DEMAND in output M components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/user_nl_clm ========= Remove limit_river_withdraw test - now we test that option via other irrigation tests M components/clm/cime_config/testdefs/testlist_clm.xml ========= Add a stub routine M components/clm/src/unit_test_stubs/utils/restUtilMod_stub.F90.in ========= Add a couple of tests to the xfail list - bug 2388 M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_13_r207 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Dec 8 10:34:56 MST 2016 One-line Summary: PHS - Calculate separate root water uptake resistances for soil and root pathways Purpose of changes ------------------ PHS changes from Sean Swenson. Root water uptake resistances calculated separately for soil and root pathways. These resistances are added in series to give a soil to root resistance that replaces the ad hoc expression used in the original code. History variables added for soil and root resistances, as well as a variable 'QPHSNEG' obtained by summing negative transpiration terms in each soil layer. Bugs fixed or introduced ----------------------- none Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown Notes of particular relevance for developers: (including Code reviews and testing) none --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Sean Swenson, David Lawrence, Keith Oleson CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means - all functionality tests passed. All baseline comparisons passed except: - all clm5 compsets : expected, tag is answer changing for clm5 - SMS_D_Ld5.f10_f10.ICLM45ED.hobart_nag.clm-edTest : baseline failed with RMS difference in a single field: RMS NBP 3.3589E-12 NORMALIZED 4.4201E-05 SMS_D_Ld5.f10_f10.ICLM45ED.yellowstone_{gnu|intel|pgi} tests are all bit for bit with the baseline. Just documenting the discrepancy and ignoring. CLM tag used for the baseline comparisons: clm4_5_13_r206 Answer changes -------------- Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all clm5 compsets with PHS on - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): unknown, assumed larger than roundoff same climate. If bitwise differences were observed, how did you show they were no worse than roundoff? n/a If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: clm/src/biogeochem/CNVegetationFacade.F90 - new routines to calculate froot and croot carbon. clm/src/main/histFileMod.F90 - add nlevsoi as 2d output option clm/src/main/clm_driver.F90 - add calls froot and croot carbon calculations clm/src/main/pftconMod.F90 - new root desnity and root radius variables clm/src/biogeophys/WaterfluxType.F90 - new qflx_phs_neg_col variable clm/src/biogeophys/WaterStateType.F90 - remove smp_l_col variable clm/src/biogeophys/PhotosynthesisMod.F90 - calculate root-soil interface conductance clm/src/biogeophys/SoilStateType.F90 - new variables k_soil_root_patch, soil_conductance_patch, root_conductance_patch clm/src/biogeophys/CanopyFluxesMod.F90 - pass froot and croot carbon through to photosynthesis. clm/src/biogeophys/SoilWaterMovementMod.F90 - calculate qflx_phs_neg_col =============================================================== =============================================================== Tag name: clm4_5_13_r206 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue Nov 29 10:46:24 MST 2016 One-line Summary: Bugfix for Luna causing arctic to die Purpose of changes ------------------ Bugfix from Keith Oleson to prevent Luna from causing the arctic to die. Bug 2366 prevented a divide by zero in Luna, but it sets Jmax/Vcmax to zero. This causes plants in the arctic to never start growing and hence never get any leaf nitrogen. The fix tested by Keith Oleson is to set them to the default values that Luna uses when it can't converge. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2379 Bugs introduced: re-opening 2291. An old problem with fire emissions fields and gnu has resurfaced. This is a two line science change in clm5-bgc-luna that is unlikely to caused this regression. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none expected Notes of particular relevance for developers: (including Code reviews and testing): none --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: K. Oleson, Erik Kluzek CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - unit tests - fail, can't import perl module: 'Streams/TemplateGeneric.pm'. Not related to current changes, must have come in a previous tag. unit-tests (components/clm/src): yellowstone - all tests pass tools-tests (components/clm/test/tools): yellowstone - no related changes, not run. PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - no related changes, not run. regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40 : ok, 45 : ok yellowstone_pgi - 40 : ok, 45 : ok yellowstone_gnu (clm45 only) - ok, see note below. hobart_nag - ok ok means - all test functionality passes - baselines pass except for clm5 bgc (answer changing) - expected failures ignored - tput, memleak, memcomp, nlcomp ignored. - bug 2291 has resurfaced? Also appears to have been an issue in clm4_5_13_r205. FAIL SMS_Lm25.f19_g16.ICLM45BGCCROP.yellowstone_gnu.clm-cropMonthOutput BASELINE $ grep -e RMS TestStatus.log RMS l2x_Fall_fire001 1.9723E-17 NORMALIZED 8.4685E-06 RMS l2x_Fall_fire002 2.7612E-17 NORMALIZED 1.2077E-06 RMS l2x_Fall_fire003 1.9723E-17 NORMALIZED 5.5636E-06 CLM tag used for the baseline comparisons: clm4_5_13_r205 Answer changes -------------- Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm5-bgc with luna on. - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Unknown, larger than roundoff to prevent die-off of the arctic. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: unknown, Keith Oleson ran a longer simulation and verbally confirmend he was happy with the results. URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: clm/src/biogeophys/LunaMod.F90 - bugfix for 2379 set vcmx25 and jmx25 to non-zero values. clm/cime_config/testdefs/ExpectedTestFails.xml - add xfail issue 2291 =============================================================== =============================================================== Tag name: clm4_5_13_r205 Originator(s): sacks (Bill Sacks) Date: Wed Nov 16 10:38:15 MST 2016 One-line Summary: Add diagnostics for snow and glacier; fix crash in Megan Purpose of changes ------------------ Most changes are to add new diagnostics fields related to the snow pack, and diagnostic fields that are averaged just over istice/istice_mec landunits (fields ending with _ICE). These changes are from from Leo van Kampenhout (l.vankampenhout@uu.nl). In addition: (1) Fix to prevent negative (or very small positive) values of intracellular leaf CO2. This just affects Megan emissions, and only has a small effect. This change is needed to prevent a crash in Megan due to negative values of these variables. This fix is from Keith Oleson. (2) Ensure that we don't try to do any downscaling over urban points; this is not currently allowed (see bug 2377). This change is needed in case there is an overlap between active urban points and the CISM domain: without this change, urban points in the CISM domain would be inside the downscaling filter, which would then cause an abort in check_downscale_consistency. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - pass yellowstone_gnu (clm45 only) - ok hobart_nag - pass ok means tests pass; some answer changes as noted below CLM tag used for the baseline comparisons: clm4_5_13_r204 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: some clm45/clm50; answer changes show up rarely - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): seem to be roundoff or only slightly greater Changes are due to the fix for Megan, in PhotosynthesisMod. Answer changes only appeared for three tests in the test suite: SMS_Lm1.f09_g16_gl4.IG1850CRUCLM50BGC.yellowstone_intel.clm-clm50KitchenSink ERP_P15x2_Lm13.f10_f10.IHISTCLM50BGC.yellowstone_intel.clm-monthly Small diffs in MEG_isoprene SMS_Lm25.f19_g16.ICLM45BGCCROP.yellowstone_gnu.clm-cropMonthOutput NaN values in a few fields in old; non-NaN in new: RMS l2x_Fall_fire001 NaN NORMALIZED 0.0000E+00 RMS l2x_Fall_fire002 NaN NORMALIZED 0.0000E+00 RMS l2x_Fall_fire003 NaN NORMALIZED 0.0000E+00 If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Do not downscale over urban points M components/clm/src/main/glc2lndMod.F90 M components/clm/src/main/atm2lndMod.F90 ========= Fix Megan crash: prevent negative intracellular CO2 M components/clm/src/biogeophys/PhotosynthesisMod.F90 ========= New snow / glacier diagnostic fields M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/biogeophys/SoilStateType.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/SolarAbsorbedType.F90 M components/clm/src/biogeophys/WaterStateType.F90 M components/clm/src/biogeophys/LakeTemperatureMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/FrictionVelocityMod.F90 M components/clm/src/biogeophys/TemperatureType.F90 =============================================================== =============================================================== Tag name: clm4_5_13_r204 Originator(s): erik (Erik Kluzek) Date: Tue Nov 15 13:55:16 MST 2016 One-line Summary: A few answer changing bug fixes Purpose of changes ------------------ This brings in some simple answer changing bug fixes. And some non-answer changing bug fixes for mksurfdata.pl and the build-namelist unit tester. In order to calculate infiltration the fraction of ground covered in surface water is used, before it is updated to account for snow. Also mulitply k_wet by a factor of 1x10^-4 both of these will makes a difference in surface runoff. This change is essentially a bug in the original formulation and so we've applied to both CLM45 and CLM50 answers. Since this is hydrology it will affect answers in all CLM45/CLM%0 cases whether SP, CN or BGC modes. For MEGAN we were using a new updated dataset that had the addition of all 78 PFT's including the new crop ones. Unfortunately it also had bare-soil, and this moved the indexing off by one. So we have a new dataset that removes the bare-soil and hence works correctly now. A simple addition of setting lake energy flux (eflx_grnd_lake) for the case when there isn't one layer os snow. For CLM45 we found that the use of the Zheng rooting profile for Carbon resulted in 1850 spinup simulations that couldn't spinup the tropics. Using the Jackson-1996 profile alleviates this problem. There were also some simple problems in using the queryNamelistDefaults.pl tool with the new cime tag brought in with clm4_5_12_r202. Fixing this allows mksurfdata.pl and build-namelist-unit_tester to work. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2378 Zheng profile for carbon results in tropics that doesn't spin up 2376 mksurfdata tests fail in clm4_5_12_r202 with new cime version 2375 84 tests are removed from build-namelist unit_tester 2374 Use fraction of ground covered in surface water used in infiltration before it's updated to account for snow 2370 Megan was giving incorrect answers 1717 Lake hydrology issue Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): carbon rooting is now Jackson-1996 rooting_profile_method_carbon for CLM$5 was Zheng method, now changed to Jackson-1996 Changes to the datasets (e.g., parameter, surface or initial files): New megan dataset without bare-soil Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: None Code reviewed by: self, swensosc, fvitt CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS (176 compare tests fail because of bug unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_12_r203 Answer changes -------------- Changes answers relative to baseline: YES! Both CLM45 and CLM50 Summarize any changes to answers, i.e., - what code configurations: CLM45BGC/CLM50 - what platforms/compilers: all - nature of change: surface runoff for CLM45/CLM50 changes when snow exists, small change for short simulations MEGAN answers are corrected, important for isoprene Change in fill mask for lake energy flux CLM45BGC changes in an important way Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/CanopyHydrologyMod.F90 --- Save frac_h2osfc_nosnow M components/clm/src/biogeophys/SoilHydrologyMod.F90 ----- Use frac_h2osfc_nosnow to determine frac_infclust (bug 2374) Also change k_wet by a factor of e-4 M components/clm/src/biogeophys/LakeHydrologyMod.F90 ----- Save off eflx_grnd_lake as an "else" in one more conditional when snl is not 1 layer (bug 1717) M components/clm/src/biogeophys/WaterStateType.F90 ------- Write out FH2OSFC_NOSNOW to history M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 -- Add commented out change in NBP definition M components/clm/src/biogeochem/MEGANFactorsMod.F90 ------ Check of n_pft's on MEGAN file changes by 1 M components/clm/bld/listDefaultNamelist.pl -------------- Can't use version 2 format datm namelist_definition for datm or drv but use the version 1 drv file. M components/clm/bld/queryDefaultNamelist.pl ------------- Can't use version 2 format datm namelist_definition for datm or drv but use the version 1 drv file. M components/clm/bld/unit_testers/build-namelist_test.pl - Make sure glc_present is added to tests for glc command line options M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - rooting_profile_method_carbon==1 for clm45 M components/clm/bld/namelist_files/namelist_defaults_drydep.xml - New 78 pft file for megan that doesn't have bare soil (bug 2370) =============================================================== =============================================================== Tag name: clm4_5_12_r203 Originator(s): sacks (Bill Sacks) Date: Thu Nov 10 12:14:13 MST 2016 One-line Summary: Bug fixes for snow-on-lake Purpose of changes ------------------ Fix a few bugs with the calculation of melt fluxes arising from snow-on-lake. These appear to mainly just affect diagnostic terms, although in theory it looks like this fixes a possible balance check error in rare conditions. Changes are from Leo van Kampenhout (l.vankampenhout@uu.nl) Specific changes are: (1) qflx_sl_top_soil is no longer updated in LakeHydrology. Terms that used to go into that variable now go into qflx_snow_drain. (2) For the case where we have snow above an unfrozen lake: Changed calculation of qflx_snomelt and eflx_snomelt: Old: qflx_snomelt(c) = qflx_snomelt(c) + h2osno(c)/dtime eflx_snomelt(c) = eflx_snomelt(c) + h2osno(c)*hfus/dtime New: qflx_snomelt(c) = qflx_snomelt(c) + sumsnowice(c)/dtime eflx_snomelt(c) = eflx_snomelt(c) + sumsnowice(c)*hfus/dtime So note that this now just considers sumsnowice rather than the full h2osno - i.e., it doesn't remelt the liquid in the snow pack. (3) For the case where we have snow above an unfrozen lake: The update of qflx_snow_drain is changed; and again, note that qflx_sl_top_soil is gone. Old: ! update drainage from snow pack for this case qflx_snow_drain(c) = qflx_snow_drain(c) + qflx_snomelt(c) qflx_sl_top_soil(c) = qflx_sl_top_soil(c) + h2osno(c) New: ! update incidental drainage from snow pack for this case qflx_snow_drain(c) = qflx_snow_drain(c) + h2osno(c)/dtime The old version had three apparent problems: (a) This drainage term contributed to both qflx_snow_drain and qflx_sl_top_soil, which later get added together (in BalanceCheck); as noted in (1), the uses of qflx_sl_top_soil and qflx_snow_drain have been merged into just qflx_snow_drain in this routine. (b) The qflx_sl_top_soil addition adds h2osno, without dividing by dtime (c) The qflx_snow_drain addition adds the full qflx_snomelt, which may include other terms that were set prior to this code. Note that, in the new version, we want to drain the *whole* snow pack in this case - liquid as well as solid. (4) qflx_top_soil is changed: Old: qflx_top_soil(c) = qflx_prec_grnd_rain(p) + qflx_snomelt(c) New: qflx_top_soil(c) = qflx_prec_grnd_rain(p) + qflx_snow_drain(c) What the soil receives should equal what leaves the snow, i.e. the drainage term. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: Bill Sacks, Sean Swenson CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - pass yellowstone_gnu (clm45 only) - ok hobart_nag - pass ok means tests pass; baseline comparison failures as noted below CLM tag used for the baseline comparisons: clm4_5_12_r202 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Potentially any clm45/clm50 case - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Appears to only change diagnostic fields Answer changes only show up in a few cases, presumably because this case in the code is relatively uncommon. Specifically, these tests have answer changes: FAIL LII_D_Ld3.f09_g16.ICLM45BGC.yellowstone_gnu.clm-default BASELINE FAIL ERI_D_Ld9.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay BASELINE FAIL ERP_Ld5.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default BASELINE FAIL ERP_P180x2_D_Ld5.f19_g16.I1850CLM45.yellowstone_intel.clm-default BASELINE FAIL ERS_D_Ld3.f09_g16_gl4.IG1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KitchenSink BASELINE If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/LakeHydrologyMod.F90 =============================================================== =============================================================== Tag name: clm4_5_12_r202 Originator(s): sacks (Bill Sacks) Date: Wed Nov 9 12:02:41 MST 2016 One-line Summary: Update to cime version used in cesm2_0_beta03 Purpose of changes ------------------ Update to cime version used in cesm2_0_beta03, which fixes some bugs and has some new features. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): - Unexpected answer change with Build namelist2 PR (#640) (https://github.com/ESMCI/cime/issues/766) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - preview_namelists is now first run by case.build - no longer by case.setup Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - ERI and ERP tests now pass on hobart (previously these were failing silently) Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - ok yellowstone_gnu (clm45 only) - pass hobart_nag - ok ok means tests pass; answer changes in these two tests: FAIL ERP_Ld5.f19_g16.ICNCROP.yellowstone_pgi.clm-40default BASELINE - testing-specific change, due to the fact that we now build this ERP test non-threaded FAIL SMS_D_Ld1_Mmpi-serial.1x1_mexicocityMEX.I.hobart_nag BASELINE - unknown reason for diff: see https://github.com/ESMCI/cime/issues/766 CLM tag used for the baseline comparisons: clm4_5_12_r201 Answer changes -------------- Changes answers relative to baseline: YES - but only in the two tests noted above If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: just the two tests noted above - what platforms/compilers: just the two tests noted above - nature of change (roundoff; larger than roundoff/same climate; new climate): not investigated closely If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime: clm4_5_12_r198_cime5.2.0-alpha.1-9-gd321e4b -> cime5.2.0-alpha.7 List all files eliminated: ========= Just rename xmlchange_cmnds to shell_commands, because xmlchange_cmnds is being deprecated D components/clm/cime_config/testdefs/testmods_dirs/clm/pts/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLA/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLB/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/decStart/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/40pts/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsROA/xmlchange_cmnds D components/clm/cime_config/testdefs/testmods_dirs/clm/drydepnomegan/xmlchange_cmnds List all files added and what they do: ========= Just rename xmlchange_cmnds to shell_commands, because xmlchange_cmnds is being deprecated A components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsROA/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/drydepnomegan/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/pts/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLA/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLB/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/40pts/shell_commands List all existing files that have been modified, and describe the changes: none =============================================================== =============================================================== Tag name: clm4_5_12_r201 Originator(s): sacks (Bill Sacks) Date: Wed Nov 9 09:18:54 MST 2016 One-line Summary: Rework initialization of transient weights - important with use_init_interp Purpose of changes ------------------ Two main purposes: (1) When using use_init_interp in a case with CISM, avoid doing dyn landcover adjustments in the first time step after initialization. This is important in order to avoid fictitious adjustments simply due to the fact that we don't have updated glacier areas until the run loop (as in the cold start case). (2) Update prescribed transient weights (for PFTs and crops) in initialization, rather than weighting for the run loop. This undoes part of clm4_5_1_r101. This can be important if you're running use_init_interp, so that the appropriate points are active on the output dataset. More details on (1): This was achieved by initializing glacier areas at the start of the run loop for cold start or interpolated start. Rationale: In the first time step of a run that used cold start or init_interp, glacier areas will start at whatever is specified on the surface dataset, because coupling fields from GLC aren't received until the run loop. Thus, CLM will see a potentially large, fictitious glacier area change in the first time step after cold start or init_interp. We don't want this fictitious area change to result in any state or flux adjustments. Thus, we apply this area change at the start of the driver loop, so that in dynSubgrid_driver, it will look like there is no glacier area change in the first time step. This needs to happen very early in the run loop, before any balance checks are initialized, because - by design - this doesn't conserve mass at the grid cell level. (The whole point of this new code block in the driver is that we adjust areas without doing the typical state or flux adjustments that need to accompany those area changes for conservation.) This accomplishes approximately the same effect that we would get if we were able to update glacier areas in initialization. The one difference - and minor, theoretical problem - that could arise from this start-of-run-loop update is: If the first time step of the CESM run loop looked like: (1) GLC runs and updates glacier area (i.e., glacier area changes in the first time step compared with what was set in initialization); (2) coupler passes new glacier area to CLM; (3) CLM runs. Then the code here would mean that the true change in glacier area between initialization and the first time step would be ignored as far as state and flux adjustments are concerned. But this is unlikely to be an issue in practice: Currently GLC doesn't update this frequently, and even if it did, the change in glacier area in a single time step would typically be very small. If we are ever able to change the CESM initialization sequence so that GLC fields are passed to CLM in initialization, then this new code block in the driver can be removed. Also: Remove handling of first step since cold start in dynSubgrid_driver. This is no longer needed now that (a) we update prescribed transient weights in initialization, and (b) we have special-purpose code for glacier in the main driver run loop. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; a few answer changes as noted below CLM tag used for the baseline comparisons: clm4_5_12_r200 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with CISM, with use_init_interp = .true. ALSO: Possible answer changes in other transient configurations - see note below - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): not investigated carefully; could have climate changing impacts in a coupled run Baseline comparisons failed for the following three tests, as expected: FAIL ERS_D_Ld3.f09_g16_gl4.IG1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KitchenSink COMPARE_baseline FAIL ERS_D_Ld12.T31_g37_gl4.IG1850CRUCLM50BGC.yellowstone_intel.clm-glcMEC_spunup_inc_dec_bgc COMPARE_baseline FAIL SMS_Lm1.f09_g16_gl4.IG1850CRUCLM50BGC.yellowstone_intel.clm-clm50KitchenSink COMPARE_baseline In addition, it's possible that answer changes would appear in a coupled, transient run with cold start or init_interp, due to changes in the fields sent in initialization (or the first time step???) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Set prescribed transient weights in initialization. Get rid of special handling of the first step since cold start; this is no longer needed now that we (1) handle the glacier case specially in the driver, and (2) initialize prescribed transient weights in initialization M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Handle glacier area specially in the first time step since cold start or interpolated start M components/clm/src/main/clm_driver.F90 ========= Get rid of special handling of the first step since cold start; this is no longer needed now that we (1) handle the glacier case specially in the driver, and (2) initialize prescribed transient weights in initialization M components/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 ========= Add a flag to determine if we're doing an interpolated start M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/clm_initializeMod.F90 ========= Just add some comments M components/clm/src/main/subgridWeightsMod.F90 M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/init_interp/initInterpMindist.F90 =============================================================== =============================================================== Tag name: clm4_5_12_r200 Originator(s): erik (Erik Kluzek) Date: Tue Nov 8 00:56:00 MST 2016 One-line Summary: Latest parameter updates from simulations, ignore surface soil layer for plant hydraulic stress for greater variability in surface soil moisture Purpose of changes ------------------ Ignore surface soil layer for plant hydraulic stress which allows for greater variability in surface soil moisture. Output some additional diagnostic fields to the history files. Rename fertnitro to manunitro, both in code and in params files. Add options to vary baset by latitude for tropical crops and subtract crop seed carbon. Don't call precision control for frootc for CLM50. For CLM50 we bring in the latest set of parameter updates to the parameter file as well as namelist items. Many of these were manually tuned, and many were tuned with the automatic calibration process of Rosie Fisher and Ben Sanderson. Fix several bugs, a couple possible divide by zero errors, only run freelivingfixation when FUN on. Limit array operation by number of layers. Add option to put list of namelist parameters to an output file. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): bug 2349 Run freelivingfixation only if FUN is on bug 2362 Possible divide by zero for fert when fertilizer off bug 2364 maxval check goes over whole range of array when should just go to nlayers bug 2366 fix divide by zero in Luna bug 2367 Add a list of real parameters as part of what CLM build-namelist does (partial) Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] bug 2368 Remove old "use_voc" logic that was put in as a hack to turn off megan when prognostic crop was on (This existed before, but we just spotted it as a problem) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist items: baset_mapping, baset_latvary_slope, baset_latvary_intercept, initial_seed_at_planting new clm namelists: cnphenology, crop New history fields: LNC, BTRANMN, CPHASE, LATBASET, CROPSEEDC all default active Changes made to namelist defaults (e.g., changed parameter values): New parameter calibrations For CLM50: baseflow_scalar=0.0090932, zetamaxstable=0.5, pot_hmn_ign_counts_alpha=0.0079611 Jmaxb1=0.093563, initial_seed_at_planting=3.0, decomp_depth_efolding=10.0 baset_mapping=varytropicsbylat, baset_latvary_intercept=12, baset_latvary_slope=0.4 br_root=1.1678d-06, initial_Cstocks=3*200.0 For CLM45 old defaults used: Changes to the datasets (e.g., parameter, surface or initial files): New parameter dataset with latest calibration changing following fields: mergetoclmpft modified to remove generic crops and merge to more existing crops mxmat,arooti,lfemerg,min_planting_temp,bfact,fleafi,manunitro manually calibrated by Yaqiong Lu for crop soilpsi_off,soilpsi_on manually calibrated FUN_fracfixers,fun_cn_flex_a,fun_cn_flex_b,fsr_pft,grperc,dsladlai,ekn_active,froot_leaf,kmax,leaf_long,leafcn,mbbopt,psi50,slatop,stem_leaf modified by automatic calibration scripts maxpsi_hr hand calibrated by Charlie Koven i_vcad,s_vcad hand calibrated for crop FUN_fracfixers,dsladlai,ekn_active,froot_leaf,kmax,krmax,leaf_long,leafcn,mbbopt,psi50,slatop hand calibrated akn_active,croot_stem,ekc_active,kc_nonmyc,kn_nonmyc modified slightly leafcn,slato from TRY database Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): crop phase should be changed into integer with parameter indices for the different types as well as outputting flag_meanings and flag_values attributes. Remove LATBASET from history as constant. There is some complication for clm4_5 to preserve answers for latbarybaset and subtract_cropseed for fert that should be removed. Will change answers but shouldn't matter. Changes to tests or testing: None Code reviewed by: self, oleson, dll, yaqiong CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS (254 comparison tests are different because of namelist changes) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_12_r199 Answer changes -------------- Changes answers relative to baseline: YES! for CLM50 (identical for CLM45) Summarize any changes to answers, i.e., - what code configurations: CLM50 - what platforms/compilers: ALL - nature of change: calibrated climate If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) clm5n04r193phs_2deg_calibrateparams_v3_crop_fert10perc_medshrb_1850ADspin b.e20.B1850.f09_g16.pi_control.all.123 (neither of these are identical to this tag for different reasons) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M parse_cime.cs.status -- Update to work with newer cime being used now M components/clm/bld/unit_testers/build-namelist_test.pl ----------- Add tests for user_nl_clm_real_parameters and baset_mapping and Jmaxb1 M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - New namelist items: baset_mapping, baset_latvary_slope, baset_latvary_intercept, initial_seed_at_planting M components/clm/bld/namelist_files/namelist_defaults_clm4_0.xml --- Ran xmllint on it. M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Use latest calibrated parameters for clm5_0: baseflow_scalar=0.0090932, zetamaxstable=0.5, pot_hmn_ign_counts_alpha=0.0079611 Jmaxb1=0.093563, initial_seed_at_planting=3.0, decomp_depth_efolding=10.0 baset_mapping=varytropicsbylat, baset_latvary_intercept=12, baset_latvary_slope=0.4 br_root=1.1678d-06, initial_Cstocks=3*200.0 New paramfiles with fertnitro renamed to manunitro, clm50 version with updated parameters M components/clm/bld/CLMBuildNamelist.pm --- Add output_reals command line option, new namelist subroutines: setup_logic_crop, setup_logic_friction_vel, setup_logic_century_soilbgcdecompcascade, new general subroutine: error_if_set to abort on error if a list of namelist items is set new namelists: cnphenology, crop M components/clm/src/biogeochem/CNPhenologyMod.F90 ------- Add CNPhenologyReadNML, and namelist items: initial_seed_at_planting, and subtract_cropseed, rename fertnitro to manunitro, and add cropseedc, and cphase, if fertilizer off don't add in manunitro, subtract cropseed if subtract_cropseed is set. M components/clm/src/biogeochem/CNVegCarbonStateType.F90 - Add cropseed and CROPSEEDC to history and restart files M components/clm/src/biogeochem/CNDriverMod.F90 -- Only call CNFreeLivingFixation if use_fun M components/clm/src/biogeochem/CropType.F90 ----- Add ReadNML, baset_mapping/latvary_intercept/latvary_slope cphase_patch, add CPHASE and LATBASET to history files, cphase to restart M components/clm/src/biogeochem/CNPrecisionControlMod.F90 -- Set prec_control_for_froot to opposite of use_nguardrail so turned on for clm5_0 and off for clm4_5 (currently needed to get c4 crops to grow) M components/clm/src/main/clm_initializeMod.F90 --- call energyflux_inst%initAccVars M components/clm/src/main/controlMod.F90 ---------- call CNPhenologyReadNML M components/clm/src/main/clm_driver.F90 ---------- call energyflux_inst%UpdateAccVars M components/clm/src/main/pftconMod.F90 ----------- change fertnitro to manunitro M components/clm/src/main/clm_instMod.F90 --------- call call crop_inst%ReadNML and energyflux_inst%InitAccBuffer M components/clm/src/biogeophys/BalanceCheckMod.F90 ------ Set indexg when found M components/clm/src/biogeophys/SoilWaterMovementMod.F90 - Remove local krmax, and use param_inst version, set rai to zero, limit array operation to :nlayers as nlayers change M components/clm/src/biogeophys/EnergyFluxType.F90 ------- Add btran_min, and btran_min_inst, and interfaces: InitAccBuffer, InitAccVars, UpdateAccVars, add BTRANMN to history, BTRAN_MIN, BTRAN_MIN_INST to restart M components/clm/src/biogeophys/FrictionVelocityMod.F90 -- Initialize zetamaxstable M components/clm/src/biogeophys/LunaMod.F90 -------------- Initialize Jmaxb1 = 0.1_r8, check lnc>0 as well as tlai, otherwise set: jmax25_z/vcmx25_z==0 M components/clm/src/biogeophys/PhotosynthesisMod.F90 ---- Add LNC to history and lnca to restart, rai(j=1)==0 for PHS M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 - initial_Cstocks(:)==200 =============================================================== =============================================================== Tag name: clm4_5_12_r199 Originator(s): sacks (Bill Sacks) Date: Sat Oct 29 14:20:11 MDT 2016 One-line Summary: Use CISM2 by default Purpose of changes ------------------ For compsets that use CISM (IG compsets), use CISM2 by default. Also update CISM external to a version with an updated input file for CISM2 at 4 km resolution. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Meaning of IG compset aliases changes - now means CISM2 rather than CISM1 - This results in slightly changed glacier cover over Greenland - If you have a restart file from an old IG case (CLM45/CLM50) and plug it into an IG case with the same compset alias in this tag, there will be an initial shock to the system as CLM responds to the new glacier area. Thus, this change should be made in an offline run, and run for a year or more before plugging the new restart file into a coupled run. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - All IG tests will BFAIL because they use a new resolution Changes to tests or testing: - Removed the ICLM45GLCMEC compset that had been used for testing; replaced its test with IGCLM45BGC - Added an ERP IG test, replacing the old PEM IG test Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; IG tests BFAIL because of new CISM resolution CLM tag used for the baseline comparisons: clm4_5_12_r198 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Out-of-the-box IG compsets change answers due to the new definition of these compsets (CISM2 rather than CISM1) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated closely; expected to be larger than roundoff/same climate for most applications Also note that, for the sake of testing, IG tests will BFAIL If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cism: cism2_1_21 -> cism2_1_24 List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/cime_config/config_compsets.xml M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_12_r198 Originator(s): Sean Swenson, Bill Sacks Date: Thu Oct 20 20:39:56 MDT 2016 One-line Summary: Rework mapping of irrigation from CLM to ROF Purpose of changes ------------------ The combination of the CLM changes and the changes to externals (RTM, MOSART, CIME) change the way irrigation is mapped from CLM to ROF: Rather than being mapped like other fluxes, it is mapped with normalization by VOLR. This means that, if there are multiple ROF cells underlying a single CLM cell, then irrigation with be withdrawn from each in an amount proportional to the VOLR in each. So, for example, if a CLM cell overlaps two ROF cells, one with VOLR = 100 and one with VOLR = 200, then twice as much water will be withdrawn from the second as from the first. The main motivation for this change is to decrease the frequency at which irrigation exceeds available river volume in a given ROF cell. This will especially be improved once the volr-limited irrigation is working (this will be fixed in an upcoming tag). Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - If this CLM tag is plugged in to a CESM tag, note that it depends on rtm1_0_59, mosart1_0_19, and the cime changes from https://github.com/ESMCI/cime/pull/681 Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: - Added some irrigation tests - Removed some tests from the xfail list which are now passing Code reviewed by: Bill Sacks Note that most changes are originally from Sean Swenson; Bill Sacks reviewed and tested the changes CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; some baseline comparison failures as noted below; also, many IG tests (i.e., with CISM) showed baseline comparison failures, but this is simply because, with the new CISM external, CISM now does not produce a history file in short tests, whereas it did before. CLM tag used for the baseline comparisons: clm4_5_12_r197 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 with irrigation, and all tests with MOSART - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): not investigated carefully, but expected to be larger than roundoff/same climate Irrigation changes are due to the new mapping MOSART changes are just in x2l_Flrr_volr, r2x_Flrr_vol,r TWS & VOLR - due to a fixed calculation of VOLR in the MOSART external If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): - cime: 5.1.1 -> clm4_5_12_r198_cime5.2.0-alpha.1-9-gd321e4b This is a cime branch tag, based off of baseline version d321e4b (this baseline version is 9 commits past cime5.2.0-alpha.1). The changes on this branch are expected to be available by cime5.2.0-alpha.5 - cism: 2_1_18 -> 2_1_21 (unrelated to the other changes in this tag) - rtm: 1_0_58 -> 1_0_59 - mosart: 1_0_18 -> 1_0_19 List all files eliminated: ========= Renamed to include irrigation D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/README D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp List all files added and what they do: ========= Added more tests with irrigation turned on A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/README A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_spunup A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/README A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropIrrigMonth_interp List all existing files that have been modified, and describe the changes: ========= Main changes for routing irrigation as a separate field M components/clm/src/main/lnd2atmType.F90 M components/clm/src/main/lnd2atmMod.F90 M components/clm/src/cpl/clm_cpl_indices.F90 M components/clm/src/cpl/lnd_import_export.F90 ========= Use volrmch rather than volr for limiting irrigation M components/clm/src/main/clm_driver.F90 ========= Add 'target' attributes to some variables that need them M components/clm/src/init_interp/initInterpMultilevelContainer.F90 ========= Add a deallocation statement M components/clm/src/main/atm2lndType.F90 ========= Rename a variable M components/clm/src/biogeophys/IrrigationMod.F90 ========= Add a new diagnostic field M components/clm/src/biogeophys/WaterfluxType.F90 ========= Added two debug tests with irrigation; changed the long clm50 crop test to include irrigation M components/clm/cime_config/testdefs/testlist_clm.xml ========= Two ERI tests now pass due to fixes in CIME and CISM M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_12_r197 Originator(s): erik (Erik Kluzek) Date: Fri Oct 14 03:17:41 MDT 2016 One-line Summary: Some cleanup for plant hydraulic stress and adding more namelist parameters Purpose of changes ------------------ Some cleanup and documentation of Plant Hydraulic Stress (PHS) from Keith Oleson and Daniel Kennedy. Setup some common indices as paraameters for PHS rather than hard-coded magic numbers. Add several new namelist parameters and -light_res option to use high resolution lightnng dataset. Work with bombspike option and add a test for it. Remove ESMF interfaces. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2361 Setting of glc_nec in a clm aux test 2359 build-namelist_test.pl has 20 fails in clm4_5_12_r195 2351 1x1_tropicAtl case issues 2350 problems with bombspike case 2339 don't write crootfr and rootfr 2333 build-namelist_test.pl should label the mode in a more useful way 2332 fire_emis test is unable to add fire emission-specific history fields 2325 limit scope of mct vars in decompInitMod and lnd_comp_mod 2293 better handling of bombspike case Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): new namelist items: initial_Cstocks, freelivfix_slope_wET, freelivfix_intercept, zetamaxstable, Jmaxb1 new build-namelist option (CLM_NAMELIST_OPTS): light_res Change DJK history field to TSINK Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files): Option for a half-degree lightning dataset rather than just the default T62 dataset Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): SoilWaterMovementMod.F90 duplicates the soil and root indices that are in PhotosynthesisMod Changes to tests or testing: Add bombspike test Code reviewed by: self, oleson, djk CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS (310 comparisons fail, because of namelist changes) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_12_r196 Answer changes -------------- Changes answers relative to baseline: no bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: Remove ESMF interfaces no longer supported D components/clm/src/cpl/lnd_comp_esmf.F90 D components/clm/src_clm40/main/lnd_comp_esmf.F90 List all files added and what they do: -------------- Add new test for c14 bombspike A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_bombspike1963/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_bombspike1963/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_bombspike1963/shell_commands List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm --------------------------- Add -light_res option, clarify help, set cnfireson when lightning/pop defns needed M components/clm/bld/unit_testers/build-namelist_test.pl ----------- set mode to options to configure, add new tests for the new namelist options M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- set spinup_state, use_dynroot, paramfile for use_ed/clm50 correct year range for flanduse_timeseries 1x1_tropicAtl, set light_res defaults, add half-degree lightning file, M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add new namelist items: initial_Cstocks, freelivfix_slope_wET, freelivfix_intercept, zetamaxstable, Jmaxb1, and light_res option --------------- Remove unneeded ndepsrc="stream", add cnfireson for pop-dens and lightning M components/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/2000_control.xml M components/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M components/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/user_nl_clm - turn hydrstess off M components/clm/src/utils/spmdGathScatMod.F90 ------- Add only clause to use statements (bug 2325) M components/clm/src/biogeochem/CNFireLi2016Mod.F90 -- Add need_lightning_and_popdens, remove note on compiler bug M components/clm/src/biogeochem/CNNDynamicsMod.F90 --- Add CNNDynamicsReadNML for freelivfix_intercept, freelivfix_slope_wET M components/clm/src/biogeochem/CNFireMethodMod.F90 -- Add need_lightning_and_popdens M components/clm/src/biogeochem/CNFireLi2014Mod.F90 -- Set need_lightning_and_popdens=.true. M components/clm/src/biogeochem/CNFireNoFireMod.F90 -- Set need_lightning_and_popdens=.false. M components/clm/src/biogeochem/CNFireBaseMod.F90 ---- Add need_lightning_and_popdens, only do them if true M components/clm/src/main/clm_varpar.F90 ------------- Make i_met_lit, i_cel_lit, i_lig_lit parameters, set i_cwd appropriately M components/clm/src/main/decompInitMod.F90 ---------- Add only clause to use statements (bug 2325) M components/clm/src/main/ndepStreamMod.F90 ---------- Add only clause to use statements (bug 2325) M components/clm/src/main/GetGlobalValuesMod.F90 ----- Add only clause to use statements (bug 2325) M components/clm/src/main/controlMod.F90 ------------- Add calls to: LunaReadNML, FrictionVelReadNML, and CNNDynamicsReadNML, DecompCascadeBGCreadNML M components/clm/src/main/clm_instMod.F90 ------------ Shorten statement M components/clm/src/main/ncdio_pio.F90.in ----------- Have pio bcast error, so terminate here rather than deep in pio M components/clm/src/biogeophys/SoilWaterMovementMod.F90 - Change djk to tsink, work on documentation set soil=1,root=4 as parameters M components/clm/src/biogeophys/EnergyFluxType.F90 ------- Always allocate bsun_patch/bsha regardless of phs on/off M components/clm/src/biogeophys/FrictionVelocityMod.F90 -- Add FrictionVelReadNML for zetamaxstable M components/clm/src/biogeophys/UrbanFluxesMod.F90 ------- Use zetamaxstable from frictionvel_parms_inst M components/clm/src/biogeophys/LunaMod.F90 -------------- Add LunaReadNML for Jmaxb1 M components/clm/src/biogeophys/PhotosynthesisMod.F90 ---- Set sun,sha,xyl,root,veg,soil parameters and use them, always allocate regardless of phs, ensure nvegwcs==4, remove commented out lines, add back call to bombspike, remove some unused arrays, work on comments, M components/clm/src/biogeophys/LakeFluxesMod.F90 -------- Use zetamaxstable from frictionvel_parms_inst M components/clm/src/biogeophys/SoilStateType.F90 -------- Change DJK to TSINK, always alloc tsink, don't output CROOTFR/ROOTFR M components/clm/src/biogeophys/BareGroundFluxesMod.F90 -- Use zetamaxstable from frictionvel_parms_inst M components/clm/src/biogeophys/CanopyFluxesMod.F90 ------ Use zetamaxstable from frictionvel_parms_inst, work on comments for PHS M components/clm/src/biogeophys/CanopyStateType.F90 ------ Always allocate vegwp_patch don't worry about PHS M components/clm/src/biogeophys/test/HumanStress_test/test_humanstress.pf - Speed up test M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 - Add DecompCascadeBGCreadNML for initial_Cstocks, make som pools indices parameters =============================================================== =============================================================== Tag name: clm4_5_12_r196 Originator(s): erik (Erik Kluzek) Date: Tue Sep 27 02:46:29 MDT 2016 One-line Summary: Bring new cime5 to clm trunk Purpose of changes ------------------ Update from cime4 version to cime5.1.1. This has some changes to the user interface on setting up cases. create_newcase largely works the same, but other scripts have subtle to more extensive changes to them. cime5 has scripts all done in object oriented Python. create_newcase changes some options to use double "--" signs. Standard options can be done with either one or two. --user-compset If set, then the --compset argument is treated as a user specified compset.This assumes that all of the compset settings in thecompset along name have been defined for all of its componentsIf the compset name is found as a supported compset, then it will be treated as such. --pesfile PESFILE Only used and required for --user-compset argument.Full pathname of the pes specification fileThis argument is required if --user-compset is True --user-grid If set, then the -grid argument is treated as a user specified grid.As such, the additional argument --gridfile must also be specifiedand all of the grid settings in the --res argument name (which must be a grid longnamehave been defined in the grid file pointed to by --gridfile --gridfile GRIDFILE Full pathname of config grid file to useThis should be a copy of cime_config/config_grids.xmlwith the new user grid changes added to itThis argument is required if --user-grid is True create_newcase adds the following new options ( --ninst NINST Specify number of component instancesSet the number of component instances in the case. --srcroot SRCROOT Alternative path for source root directory. By default this is set to cimeroot/../ --walltime WALLTIME Set the wallclock limit for this case. Can use env var CIME_GLOBAL_WALLTIME to set this. -q QUEUE, --queue QUEUE Force batch system to use a certain queue New files created in your case: Several files are softlinked into the case directory (such as case.setup, case.run, etc.) software_environment.txt -- your environment variables PESetupHist --- Directory for processor history changes case.clean_build is removed, and you use the --clean option to case.build xmlchange adds several options such as -subgroup xmlquery operation is a bit different. list capability is done via --list-all option. Format of *.xml files in your case is a bit different than before PIO versions were updated and PIO now catches some errors that it let go before. userdefinedmachine is upgraded so that you create a ~/.cime directory and put config_machines.xml for your machine (as well as config_compilers.xml, config_batch.xml, and config_build.xml if needed) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2354 New streams file for urban is in NetCDF4 classic format and new cime5 gnu/pnetcdf/pio1 combination can't read it Known bugs introduced in this tag (include bugzilla ID): 2359 build-namelist_test.pl has 20 fails in clm4_5_12_r195 2358 test_driver.sh has problems with running gen_domain with clm4_5_12_r19620 fails in 2357 PTCLMmkdata broken in clm4_5_12_r196 due to changes in cime5 2356 Test hangs on hobart 2353 areas wrong on mosart input file cime5 issues: 589 Add back ability of ascii test lists to have comments... 580 ERI test with REFCASE prestages file into wrong run directory 506 make 1 processor cases by default mpi-serial (won't fix) 294 Need an "all" option to --xml-* options 293 cs.status looses some of the functionality that was in cimeteststatus Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): No longer need to provide machine name in create_newcase New options to create_newcase: --queue, --walltime, --ninst Some less used options were removed. Single processor cases, no longer by default use mpi-serial. You have to manually change MPILIB! manage_* scripts command line options changed use --help option to query usage. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): Urban streams file changed from NetCDF4 format to NetCDF3 so new pio1 and pnetcdf could read it Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): Changes to tests or testing: interface to create_test is changed. Standard operation is like this... ./create_test --xml-category aux_clm45 --compare clm4_5_12_r195 --generate clm4_5_12_r196 \ --xml-machine yellowstone--xml-compiler intel Currently in this version cime5.1.1 you can't leave off --xml-compiler as it will run everything with the default compiler Code reviewed by: self, sacks, mvertens CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK 20 tests fail, and 104 namelists are different unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - OK (bug 2358) PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - FAIL (bug 2357) regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_12_r195 Answer changes -------------- Changes answers relative to baseline: No bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime, rtm, mosart, cism cime to cime5.1.1 rtm to rtm1_0_58 mosart to mosart1_0_18 cism to cism2_1_18 --- Removes ESMF tests add output history frequency "time_period_freq" to history files List all files eliminated: Move to name with clm50 in it D components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp D components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/README List all files added and what they do: A parse_cime.cs.status ---- Script to parse cs.status output like cime4 cimeteststatus usage: ../../parse_cime.cs.status cs.status.* Scripts to take a baseline directory in cime4 format to cime5 format A components/clm/tools/clm4_5/refactorTools/esmciport/convertbaselines4esmci A components/clm/tools/clm4_5/refactorTools/esmciport/linkhistorytofileswdate.ncl A components/clm/tools/clm4_5/refactorTools/esmciport Have clm50 in the name, and add new CLM50 crop finidat file A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/README A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50cropMonth_interp List all existing files that have been modified, and describe the changes: Initial changes to get gen_domain build/run working with cime5 M components/clm/test/tools/test_driver.sh -- update module settings, M components/clm/test/tools/TCBCFGtools.sh -- update how configure works, and run .env_mach_specific.sh, don't need to softlink Macros.make M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- Update urban streams files so in NetCDF3 format rather than NetCDF4 so cime5/pio can read it M components/clm/cime_config/config_compsets.xml - whitespace changes, add CAM55 M components/clm/cime_config/testdefs/testmods_dirs/clm/default/shell_commands ---- no mult-instance extra copies, don't set PIO rearranger type M components/clm/cime_config/testdefs/testmods_dirs/clm/40pts/xmlchange_cmnds ----- don't set TOTALPES, but set ESP M components/clm/cime_config/testdefs/testmods_dirs/clm/pts/xmlchange_cmnds ------- don't set TOTALPES, but set ESP M components/clm/cime_config/testdefs/testmods_dirs/clm/40default/shell_commands -- no mult-instance extra copies M components/clm/cime_config/testdefs/testlist_clm.xml - Make all single processor tests with _Mmpi-serial, change cropMonth_interp to clm50cropMonth_interp change some PEM tests to ERP M components/clm/cime_config/buildlib ------------------ Changed to python M components/clm/cime_config/config_component.xml ------ Additive option changed to cime5 Remove BARRIER_N/OPTION/ATM_PIO_TYPENAME/LND_PIO_REARRANGER moved to cime5 format changed slightly M components/clm/src/biogeochem/MEGANFactorsMod.F90 -- MEGAN comp_names is 40 rather than 32, to match size on megan emissions file M components/clm/src/main/histFileMod.F90 --- Add time_period_freq MM components/clm/src/main/clm_varcon.F90 ---- Add isecsphr, and isecspmin =============================================================== =============================================================== Tag name: clm4_5_12_r195 Originator(s): sacks (Bill Sacks) Date: Sun Sep 18 07:23:25 MDT 2016 One-line Summary: Remove dependence on column weights in fire code Purpose of changes ------------------ When calculating the fractions of BET and BDT, calculate these as fractions of the column rather than as fractions of the grid cell. This is needed for consistency with all other code in CLM, where the behavior of a column shouldn't depend on that column's weight on the grid cell. The earlier code was problematic in this way: If you did two runs - one with default areas, and one where the natural veg landunit area was larger or smaller - then the evolution of the natural veg landunit could change; this is not supposed to happen in CLM. This can change answers for the fire code, but should generally only change answers by a small amount, unless there are tropical grid cells where the natural vegetation fraction is substantially less than 1. Fang Li is okay with this change. She points out that the original formulation was based on the fact that literataure typically cites 60% of (BET+BDT) on the grid cell as a cutoff for being classified as tropical closed forest for the sake of fire determinations, so the earlier code was consistent with that definition. She also points out that this change could lead to a slightly greater prevalence of deforestation fires and (after tuning) a consequently smaller prevalence of non-deforestation fires - which can be undesirable. But overall she is okay with this change given its likely small impact. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): partial fix of item #14 in bugz 1679 Still need to address dtrotr_col, but the fix there is different, and is tied in with other changes for dynamic landunits. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass; some baseline comparisons fail CLM tag used for the baseline comparisons: clm4_5_12_r194 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CN/BGC with CLM45/CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Greater than roundoff for at least some tests; from some spot-checks, seems to be same climate. I have spot-checked a few tests that have greater than roundoff-level diffs (ERP_D_Ld5.f19_g16.IRCP85CLM45BGC.yellowstone_pgi.clm-drydepnomegan and ERI_D_Ld9.f09_g16.ICLM45BGC.yellowstone_intel.clm-default). These tests have diffs in just a few grid cells in the tropics. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNFireBaseMod.F90 M components/clm/src/biogeochem/CNFireLi2016Mod.F90 M components/clm/src/biogeochem/CNVegStateType.F90 M components/clm/src/biogeochem/CNFireLi2014Mod.F90 =============================================================== =============================================================== Tag name: clm4_5_12_r194 Originator(s): sacks (Bill Sacks) Date: Thu Sep 1 10:17:56 MDT 2016 One-line Summary: Dribble annual dyn landunit adjustment fluxes; initialize water state for new crop columns Purpose of changes ------------------ Two main changes in this tag: (1) Add infrastructure for dribbling fluxes that are generated once per year (on the first time step of the year), so that an equal flux is produced throughout the year. Apply this infrastructure to the correction fluxes that are generated to conserve water and energy with dynamic columns / landunits. (2) When a new column comes into existence (currently, this means a new crop column), copy h2osoi_liq, h2osoi_ice and h2osoi_vol from the "template" column (typically a natural veg column on that gridcell), in addition to the pre-existing copy of t_soisno. Also, a minor change that touched many files: Replace inline uses of __FILE__ with a constant ('sourcefile') defined in each source file where this is used. This is needed because lines with __FILE__ were sometimes exceeding pgi's line limit (since pgi enforces the line limit after expanding macros). (See bugz 2049.) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2049 - __FILE__ expands to be too long sometimes Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): - New uses of __FILE__ on outstanding branches should be replaced with sourcefile. Note that I have only added 'sourcefile' in files that currently needed it; other files may need to have this added. Changes to tests or testing: - Added this test in order to test mid-year restarts with dynamic landunits (needed to test restartability of annual flux dribbling): ERS_D_Ld7.1x1_smallvilleIA.IHISTCLM45BGCCROP.yellowstone_intel.clm-decStart1851_noinitial Note that the "noinitial" aspect of this can be removed once ignore_ic_date is made compatible with crop. Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass or xfail; expected baseline failures as noted below CLM tag used for the baseline comparisons: clm4_5_12_r193 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 transient (HIST, RCP, DV or IG) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but expected to be new climate for configurations that have transient column areas (IG or transient crop, which involve large transitions at the starat of the year), but only marginally different (approx. roundoff-level) for transient natural PFTs (which involve gradual transitions throughout the year). I have verified that the answer changes due to the new annual flux dribbler infrastructure are reduced to 0 (i.e., bit-for-bit) if the flux on the first time step of the year is treated as before - i.e., other than the dribbling that is applied to this start-of-year flux, this infrastructure is bit-for-bit with the earlier code in dynConsBiogeophys. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= New infrastructure for dribbling fluxes that are generated on the first time step of the year - spreading these fluxes throughout the year A components/clm/src/utils/AnnualFluxDribbler.F90 A components/clm/src/utils/test/annual_flux_dribbler_test/test_annual_flux_dribbler.pf A components/clm/src/utils/test/annual_flux_dribbler_test/CMakeLists.txt A components/clm/src/utils/test/annual_flux_dribbler_test ========= Test mods for new test: tests mid-year restarts with dynamic landunits A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial/README A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart1851_noinitial List all existing files that have been modified, and describe the changes: ========= Use new infrastructure for dribbling fluxes in order to dribble fluxes generated to conserve water and energy with changing column areas M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 ========= New time manager routines needed for dribbling infrastructure M components/clm/src/utils/test/clm_time_manager_test/test_clm_time_manager.pf M components/clm/src/utils/clm_time_manager.F90 M components/clm/src/unit_test_shr/unittestTimeManagerMod.F90 ========= When a new column comes into existence, also copy some soil water variables from the template column M components/clm/src/dyn_subgrid/dynInitColumnsMod.F90 M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf ========= Add tests of new dribbling infrastructure M components/clm/src/utils/CMakeLists.txt M components/clm/src/utils/test/CMakeLists.txt ========= New test: tests mid-year restarts with dynamic landunits M components/clm/cime_config/testdefs/testlist_clm.xml ========= Replace inline uses of __FILE__ with a constant ('sourcefile') M components/clm/src/utils/quadraticMod.F90 M components/clm/src/utils/clmfates_interfaceMod.F90 M components/clm/src/utils/restUtilMod.F90.in M components/clm/src/utils/SimpleMathMod.F90 M components/clm/src/biogeochem/NutrientCompetitionMethodMod.F90 M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNDVType.F90 M components/clm/src/biogeochem/VOCEmissionMod.F90 M components/clm/src/biogeochem/CNFireFactoryMod.F90 M components/clm/src/biogeochem/CNVegComputeSeedMod.F90 M components/clm/src/biogeochem/C14BompbSpikeMod.F90 M components/clm/src/biogeochem/CNBalanceCheckMod.F90 M components/clm/src/biogeochem/DryDepVelocity.F90 M components/clm/src/biogeochem/CNDVDriverMod.F90 M components/clm/src/biogeochem/MEGANFactorsMod.F90 M components/clm/src/biogeochem/SpeciesNonIsotopeType.F90 M components/clm/src/biogeochem/SatellitePhenologyMod.F90 M components/clm/src/biogeochem/NutrientCompetitionFactoryMod.F90 M components/clm/src/biogeochem/CNDVEstablishmentMod.F90 M components/clm/src/biogeochem/DUSTMod.F90 M components/clm/src/biogeochem/CNProductsMod.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/biogeochem/CNFireLi2014Mod.F90 M components/clm/src/biogeochem/CNVegStateType.F90 M components/clm/src/biogeochem/CNMRespMod.F90 M components/clm/src/biogeochem/CNFUNMod.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/ch4Mod.F90 M components/clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 M components/clm/src/biogeochem/CNGapMortalityMod.F90 M components/clm/src/biogeochem/SpeciesIsotopeType.F90 M components/clm/src/biogeochem/CNFireBaseMod.F90 M components/clm/src/biogeochem/CropType.F90 M components/clm/src/biogeochem/CNCIsoFluxMod.F90 M components/clm/src/biogeochem/CNPrecisionControlMod.F90 M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 M components/clm/src/biogeochem/dynHarvestMod.F90 M components/clm/src/biogeochem/ch4varcon.F90 M components/clm/src/biogeochem/CNSharedParamsMod.F90 M components/clm/src/biogeochem/CNFireLi2016Mod.F90 M components/clm/src/init_interp/initInterpBounds.F90 M components/clm/src/init_interp/initInterpMultilevelCopy.F90 M components/clm/src/init_interp/initInterp1dData.F90 M components/clm/src/init_interp/initInterpMultilevelSplit.F90 M components/clm/src/init_interp/initInterp.F90 M components/clm/src/init_interp/initInterp2dvar.F90.in M components/clm/src/init_interp/initInterpMultilevelInterp.F90 M components/clm/src/init_interp/initInterpMultilevelContainer.F90 M components/clm/src/init_interp/initInterpMindist.F90 M components/clm/src/init_interp/initInterpMultilevelSnow.F90 M components/clm/src/main/subgridAveMod.F90 M components/clm/src/main/initGridCellsMod.F90 M components/clm/src/main/initVerticalMod.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/lnd2atmMod.F90 M components/clm/src/main/surfrdMod.F90 M components/clm/src/main/ncdio_pio.F90.in M components/clm/src/main/reweightMod.F90 M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/main/subgridWeightsMod.F90 M components/clm/src/main/decompInitMod.F90 M components/clm/src/main/subgridRestMod.F90 M components/clm/src/main/subgridMod.F90 M components/clm/src/main/ndepStreamMod.F90 M components/clm/src/main/atm2lndType.F90 M components/clm/src/main/column_varcon.F90 M components/clm/src/main/histFileMod.F90 M components/clm/src/main/restFileMod.F90 M components/clm/src/main/controlMod.F90 M components/clm/src/main/GetGlobalValuesMod.F90 M components/clm/src/main/initSubgridMod.F90 M components/clm/src/main/filterColMod.F90 M components/clm/src/main/glc2lndMod.F90 M components/clm/src/main/filterMod.F90 M components/clm/src/main/lnd2glcMod.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/surfrdUtilsMod.F90 M components/clm/src/main/pftconMod.F90 M components/clm/src/ED/biogeochem/EDCanopyStructureMod.F90 M components/clm/src/ED/biogeochem/EDSharedParamsMod.F90 M components/clm/src/ED/main/EDCLMLinkMod.F90 M components/clm/src/ED/main/EDRestVectorMod.F90 M components/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 M components/clm/src/biogeophys/OzoneOffMod.F90 M components/clm/src/biogeophys/SnowSnicarMod.F90 M components/clm/src/biogeophys/SoilMoistStressMod.F90 M components/clm/src/biogeophys/OzoneMod.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/FrictionVelocityMod.F90 M components/clm/src/biogeophys/UrbanFluxesMod.F90 M components/clm/src/biogeophys/LunaMod.F90 M components/clm/src/biogeophys/SurfaceAlbedoMod.F90 M components/clm/src/biogeophys/TemperatureType.F90 M components/clm/src/biogeophys/HumanIndexMod.F90 M components/clm/src/biogeophys/SurfaceAlbedoType.F90 M components/clm/src/biogeophys/PhotosynthesisMod.F90 M components/clm/src/biogeophys/SoilHydrologyMod.F90 M components/clm/src/biogeophys/GlacierSurfaceMassBalanceMod.F90 M components/clm/src/biogeophys/SoilHydrologyType.F90 M components/clm/src/biogeophys/CanopyFluxesMod.F90 M components/clm/src/biogeophys/SoilHydrologyInitTimeConstMod.F90 M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 M components/clm/src/biogeophys/SurfaceRadiationMod.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/WaterStateType.F90 M components/clm/src/biogeophys/UrbBuildTempOleson2015Mod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/LakeTemperatureMod.F90 M components/clm/src/biogeophys/CanopyStateType.F90 M components/clm/src/biogeophys/TridiagonalMod.F90 M components/clm/src/biogeophys/SoilWaterRetentionCurveFactoryMod.F90 M components/clm/src/biogeophys/UrbanTimeVarType.F90 M components/clm/src/biogeophys/UrbanRadiationMod.F90 M components/clm/src/biogeophys/BandDiagonalMod.F90 M components/clm/src/biogeophys/AerosolMod.F90 M components/clm/src/biogeophys/IrrigationMod.F90 M components/clm/src/biogeophys/SurfaceResistanceMod.F90 M components/clm/src/biogeophys/UrbanParamsType.F90 M components/clm/src/biogeophys/UrbanAlbedoMod.F90 M components/clm/src/biogeophys/RootBiophysMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/SoilWaterMovementMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemDecompMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrifDenitrifMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNLeachingMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeCNMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemPotentialMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenUptakeMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 M components/clm/src/dyn_subgrid/dynTimeInfoMod.F90 M components/clm/src/dyn_subgrid/dynFileMod.F90 M components/clm/src/dyn_subgrid/dynColumnStateUpdaterMod.F90 M components/clm/src/dyn_subgrid/dynPatchStateUpdaterMod.F90 M components/clm/src/dyn_subgrid/dynColumnTemplateMod.F90 M components/clm/src/dyn_subgrid/dynpftFileMod.F90 M components/clm/src/dyn_subgrid/dynVarMod.F90.in M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 M components/clm/src/dyn_subgrid/dyncropFileMod.F90 M components/clm/src/dyn_subgrid/dynLandunitAreaMod.F90 M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 =============================================================== =============================================================== Tag name: clm4_5_12_r193 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue Aug 30 21:08:39 MDT 2016 One-line Summary: Bugfix in BalanceCheckMod: revert tolerance change and removal of endrun call introduced in r192 Purpose of changes ------------------ Revert the tolerance change and removal of an endrun call in BalanceCheckMod that were introduced in r192. Bugs fixed or introduced ------------------------ see above. Notes of particular relevance for users --------------------------------------- none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - n/a, not run unit-tests (components/clm/src): yellowstone - n/a, not run tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok CLM tag used for the baseline comparisons: clm4_5_12_r192 Answer changes -------------- Changes answers relative to baseline: Test suite bit for bit with r191 and r192. Long runs may change from r192 if the tolerance or endrun condition are encountered, but should agree with r191. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: components/clm/src/biogeophys/BalanceCheckMod.F90 - revert tolerance change and reinstate endrun call. =============================================================== =============================================================== Tag name: clm4_5_12_r192 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Sat Aug 27 21:32:54 MDT 2016 One-line Summary: ED bugfixes and initial refactoring for a standalone library Purpose of changes ------------------ NGEE-T development work on ED. Numerous science and software improvements including: * carbon linkages between ed and subsurface soil biogeochemistry * improvements to patch fusion * litter flux outputs * phenology is self contained * sites now align to clm columns * ed is now a bgc mode * check for incompatible physics * better error checking of compatible name list options * fixed memory corruption during restarts * bit for bit restart for single point and global simulations works * numerous threading related issues fixed * numerous improvements to history and diagnostics Start renaming some routines and modules to use the new name "FATES". Initial development work to isolate ED from CLM and create a standalone library, including: * start defining an API for external drivers to interact with fates. * start moving interaction between clm and fates into interface routines * start work on thread safe fates objects * data structures to clearly identify and isolate input and output boundary conditions for some fates routines. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): BalanceCheckMod.F90 was changed to loosen tolerances, not call endrun if error in energy balance. This change isn't changing answers in tests and wasn't caught until late in merging. It should probably be reverted. ED is not thread safe, and is not expected to work or restart with multiple threads. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): ed is now a bgc mode. Changes made to namelist defaults (e.g., changed parameter values): Changes to the datasets (e.g., parameter, surface or initial files): updated ed parameters file. Substantial timing or memory changes: unknown Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Code reviewed by: Ryan Knox, Charlie Koven, Rosie Fisher CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ED tests have been added to the clm test suite on yellowstone and hobart. These will BFAIL because they do not have baselines in r191. CLM tag used for the baseline comparisons: clm4_5_12_r191 ed test suite - baseline 9c008bf - yellowstone gnu, intel, pgi - pass Answer changes -------------- Changes answers relative to baseline: bit for bit for clm40, clm45 and clm50. Changes answers for ED compsets, but these were not supported or expected to work in the trunk. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: components/clm/src/ED/biogeochem/EDPhenologyType.F90 List all files added and what they do: new mods test for ed components/clm/cime_config/testdefs/testmods_dirs/clm/edNoFire components/clm/cime_config/testdefs/testmods_dirs/clm/edNoFire/user_nl_clm components/clm/cime_config/testdefs/testmods_dirs/clm/edNoFire/include_user_mods components/clm/cime_config/testdefs/testmods_dirs/clm/edNoFire/shell_commands new unit tests for ED build namelist options components/clm/bld/test_build_namelist/t/test_ed_mode.pm new clm-ed interface: components/clm/src/utils/clmfates_interfaceMod.F90 - components/clm/src/ED/main/FatesInterfaceMod.F90 components/clm/src/ED/main/FatesGlobals.F90 call subsurface biogeochemistry driven by ED components/clm/src/biogeochem/EDBGCDynMod.F90 List all existing files that have been modified, and describe the changes: =============================================================== ED as a bgc mode, better testing of namelist options related to ed and checks for compatibility. components/clm/cime_config/config_component.xml components/clm/bld/CLMBuildNamelist.pm components/clm/bld/namelist_files/namelist_definition_clm4_5.xml components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml components/clm/bld/unit_testers/build-namelist_test.pl components/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml components/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml components/clm/bld/test_build_namelist/t/test_vichydro.pm components/clm/bld/test_build_namelist/t/template_test_XXX.pm Update ED test mods and add ED tests to clm test suite components/clm/cime_config/testdefs/testmods_dirs/clm/edTest/user_nl_clm components/clm/cime_config/testdefs/testlist_clm.xml clm-ed interaction components/clm/src/main/clm_varpar.F90 - ed compatibility with century decomp components/clm/src/main/subgridWeightsMod.F90 - ed bugfix with pft diags components/clm/src/main/decompInitMod.F90 - cohorts aligned with columns components/clm/src/main/clm_initializeMod.F90 - use ed interfaces, ed as bgc mode components/clm/src/main/subgridMod.F90 - ed subgrid info updated components/clm/src/main/ColumnType.F90 - whitespace components/clm/src/main/histFileMod.F90 - ed specific output components/clm/src/main/clm_driver.F90 - call ed interfaces components/clm/src/main/initGridCellsMod.F90 - ed cohorts align with columns components/clm/src/main/initVerticalMod.F90 - ed levels components/clm/src/main/surfrdMod.F90 - bug work around, ed related? components/clm/src/main/clm_instMod.F90 - use ed interfaces, ed as bgc mode components/clm/src/main/readParamsMod.F90 - ed as bgc mode components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 - add ed to literfall components/clm/src/biogeochem/CNVegetationFacade.F90 - ed requires veg data structures for subsurface bgc components/clm/src/biogeochem/CNDriverMod.F90 - remove canopy state components/clm/src/biogeophys/BalanceCheckMod.F90 - loosen tolerances, don't endrun if error. components/clm/src/biogeophys/TemperatureType.F90 - remove ed_gdd0 components/clm/src/biogeophys/LunaMod.F90 - remove ed variables components/clm/src/biogeophys/PhotosynthesisMod.F90 - remove ed variables components/clm/src/biogeophys/CanopyFluxesMod.F90 - interface to get ed canopy fluxes components/clm/src/biogeophys/SurfaceRadiationMod.F90 - remove ed code components/clm/src/biogeophys/CanopyStateType.F90 - remove gccanopy_patch components/clm/src/biogeophys/SurfaceAlbedoMod.F90 - call ED routine through interface components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 - ed needs hydrology parameters components/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 - fates c litr fluxes components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 - don't do some liter pools if ed on. ed as bgc mode, work with subsurface bgc components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 components/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeCNMod.F90 components/clm/src/soilbiogeochem/SoilBiogeochemPotentialMod.F90 components/clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 components/clm/src/soilbiogeochem/SoilBiogeochemDecompMod.F90 ed refactoring for interfaces, science and software big fixes components/clm/src/ED/biogeochem/EDCanopyStructureMod.F90 components/clm/src/ED/biogeochem/EDGrowthFunctionsMod.F90 components/clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 components/clm/src/ED/biogeochem/EDPhysiologyMod.F90 components/clm/src/ED/biogeochem/EDPatchDynamicsMod.F90 components/clm/src/ED/main/EDCLMLinkMod.F90 components/clm/src/ED/main/EDVecCohortType.F90 components/clm/src/ED/main/EDRestVectorMod.F90 components/clm/src/ED/main/EDPftvarcon.F90 components/clm/src/ED/main/EDInitMod.F90 components/clm/src/ED/main/EDParamsMod.F90 components/clm/src/ED/main/EDMainMod.F90 components/clm/src/ED/main/EDTypesMod.F90 components/clm/src/ED/main/CMakeLists.txt components/clm/src/ED/fire/SFMainMod.F90 components/clm/src/ED/biogeophys/EDAccumulateFluxesMod.F90 components/clm/src/ED/biogeophys/EDSurfaceAlbedoMod.F90 components/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 components/clm/src/ED/biogeophys/EDBtranMod.F90 =============================================================== =============================================================== Tag name: clm4_5_12_r191 Originator(s): sacks (Bill Sacks) Date: Sun Aug 21 06:18:04 MDT 2016 One-line Summary: New parameterizations for snow overburden compaction and wind drift compaction Purpose of changes ------------------ Introduces new parameterizations for snow overburden compaction and wind drift compaction. By default, CLM45 uses the old snow overburden compaction formula and no wind drift compaction, whereas CLM5 uses the new parameterizations. Also changes the CLM50 default for the snow compaction parameter upplim_destruct_metamorph Scientific changes are from Leo van Kampenhout (l.vankampenhout@uu.nl) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist option: snow_overburden_compaction_method Changes made to namelist defaults (e.g., changed parameter values): - Changed clm5_0 default for upplim_destruct_metamorph - upplim_destruct_metamorph appears in namelist for clm4_5 as well as clm5_0 - upplim_destruct_metamorph default changed for clm5_0 (175 rather than 150) - overburden_compress_tfactor appears in namelist for clm4_5; does not appear when using the new Vionnet2012 snow overburden compaction method (so does not appear by default with clm5_0); no longer has a different default value for clm5_0 (since it doesn't appear by default in clm5_0) Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none The new parameterizations are more complex, so have a moderate (~ 20%) impact on SnowCompaction timing, but this is a very small part of the overall CLM run time. The timing impacts of extracting routines to compute some of the snow compaction terms were minimal (only increased the SnowCompaction time by a few percent). Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self; Dave Lawrence also reviewed Leo's initial implementation CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok tests pass; differences from baseline for CLM45 and CLM50 as expected unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass (or xfail); differences from baseline for CLM50 as expected CLM tag used for the baseline comparisons: clm4_5_12_r190 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; I haven't investigated this closely, but suspect these changes result in new climate, at least regionally If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: ========= Remove unit tests for a function that has been removed D components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_windCompactionFactor.pf List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= (a) Add namelist option controlling which snow overburden compaction method to use; (b) make upplim_destruct_metamorph appear in the namelist for clm4_5 as well as clm5_0; (c) change default value for upplim_destruct_metamorph; (d) remove clm5_0 default for overburden_compress_tfactor (because it is not used by default in clm5_0), and make sure this namelist item appears with clm4_5 but doesn't appear when using the new Vionnet2012 snow overburden compaction method M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Unrelated change: Remove spaces from timer names for easier automated processing of timing output (requested by Sean Santos) M components/clm/src/main/clm_driver.F90 ========= Unrelated changes: Add more informative error messages M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/restFileMod.F90 ========= Remove unit tests for a function that has been removed M components/clm/src/biogeophys/test/SnowHydrology_test/CMakeLists.txt =============================================================== =============================================================== Tag name: clm4_5_12_r190 Originator(s): sacks (Bill Sacks) Date: Sat Aug 20 05:53:49 MDT 2016 One-line Summary: Fix snow capping bug Purpose of changes ------------------ When snow capping was triggered, the bottom snow layer was decreased in mass, but dz remained unchanged, leading to problematic changes in density. This tag fixes that problem, so that dz is decreased appropriately. These changes are from Leo van Kampenhout (l.vankampenhout@uu.nl) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2343 (new snow capping scheme does not conserve density) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: Changed ERP_P15x2_Lm36.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-cropMonth_interp to ERP_P30x2_Lm36.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-cropMonth_interp so that it could reliably finish within 2 hours. Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - pass ok means tests pass, some baseline comparisons fail as expected CLM tag used for the baseline comparisons: clm4_5_11_r189 Answer changes -------------- Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: potentially all clm45 / clm50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff. I have not investigated the size of the change, suspect it is the same climate with respect to land-atmosphere fluxes, but potentially different climate with respect to snow pack properties and SMB sent to CISM. (This changes the density of the bottom layer of deep snow packs.) If bitwise differences were observed, how did you show they were no worse than roundoff? n/a If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: n/a URL for LMWG diagnostics output used to validate new climate: n/a Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Change a test, as noted above M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_11_r189 Originator(s): erik (Erik Kluzek) Date: Thu Aug 18 14:52:36 MDT 2016 One-line Summary: Bring in time varying streams for urban data (currently triggers AC use) Purpose of changes ------------------ Instead of reading in maximum building temperature from surface dataset, read it in from a time-varying urban streams file. This allows the urban Air Conditioning use to change in time as well as spatially. Other spatially and temporally changing parameters can be added (or moved from the surface dataset) to this stream file as well in the future. The new streams file has the same kind of controls as other streams file for years to run over, filename, and mapping algorithm. The time interpolation is assumed to be linear, and it uses taxmode=extend so it will use the first value before the first year and the last value after the last year. The current dataset is at one degree resolution (f09_g16). Because of this CLM45 compsets will change answers to some extent (because interpolation methods for streams are slightly different) -- but answers will be very similar. CLM50 compsets will change because there is variation in the max building temperature both spatially and temporally. Also fix a problem with the humanstress module for GSWP3 forcing because of temperatures that are too high for the WetBulb method to converge. WetBulb previously only allowed two iterations and now it allows more (10,000). If the method fails to converge it will exit with a warning. This was worked on by Jonathan Buzan with Keith Oleson. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2348 PTCLMmkdata not able to run for clm5_0 2305 Nan for wet bulb temperature Known bugs introduced in this tag (include bugzilla ID): 2346 Unnecessary arguments to WetBulb subroutine (WONTFIX) Notes of particular relevance for users --------------------------------------- Caveats for users: No longer reads t_building_max from surface dataset reads it in from the urban time-varying streams Changes to CLM's user interface: New namelist variables Add: stream_year_first_urbantv, stream_year_last_urbantv model_year_align_urbantv, stream_fldfilename_urbantv, urbantvmapalgo Changes made to namelist defaults: None Changes to the datasets: new urban time-varying streams datasets Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: None Changes to tests or testing: None Code reviewed by: self, oleson CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (all tests PASS and 64 compare tests are different because of namelist changes) unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_11_r188 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: clm4_5 (to a minor degree) and clm5_0 with urban - what platforms/compilers: All - nature of change: same climate CLM45 answers with urban areas may change to a minor degree because of the use of the urban time varying stream file to get the max indoor building temperature rather than reading from the surface dataset. Shorter tests are fairly close and some may be identical (all CLM45 single point cases are identical). Tests of a month or longer likely diverge. Five of the 22 CLM45 and CLM45BGC tests with f10_f10 on hobart are different, others are the same. Half of the 46 CLM45 CLM45 tests with f10_f10 on yellowstone are the same. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): PTCLM update PTCLM to PTCLM2_160818 List all files eliminated: None List all files added and what they do: Object to handle time-varying urban data A components/clm/src/biogeophys/UrbanTimeVarType.F90 -- Read max building temperature from urban time-varying stream for three urban classes (Tall Building Disctrict-TBD, High Density-HD, and Medium Density-MD) add TBUILD_MAX as possible output to history List all existing files that have been modified, and describe the changes: M components/clm/src/main/clm_instMod.F90 -- Setup urbantv_type and initialize it M components/clm/src/main/clm_driver.F90 --- Interpolate urban-time-varying data and pass it down M components/clm/src/biogeophys/HumanIndexMod.F90 ------------- Remove duplicated wbt_ref2m_u_patch, output urban stull wet bulb temp, new convergence routine for Wet-Bulb (increase max iterations from 2 to 10000) M components/clm/src/biogeophys/SoilTemperatureMod.F90 -------- Pass urbantv down, get t_building_max from it M components/clm/src/biogeophys/UrbBuildTempOleson2015Mod.F90 - Pass urbantv down, get t_building_max from it M components/clm/src/biogeophys/UrbanParamsType.F90 ----------- Remove t_building_max and no longer read T_BUILDING_MAX from surface dataset M components/clm/src/biogeophys/test/HumanStress_test/test_humanstress.pf -- Test new WetBulb subroutine Add in stream_year_ first/last/align for urbantv for clm5_0 M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/2000_control.xml M components/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M components/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_transient.xml M components/clm/bld/namelist_files/use_cases/20thC_transient.xml M components/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M components/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add: stream_year_first_urbantv, stream_year_last_urbantv model_year_align_urbantv, stream_fldfilename_urbantv, urbantvmapalgo M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Defaults for above list. clm4_5 has first, last and align set to 2000 whereas clm5_0 sets it appropriatly by sim_year and sim_year_range there is also a CLM45 urbantv streams file that approximates the previous method of reading from the surface dataset the urbantv streams files are at 0.9x1.25 resolution M components/clm/bld/CLMBuildNamelist.pm - Add setup_logic_urbantv_streams, call it and add urbantv_streams namelist Remove T_BUILDING_MAX from clm45/clm50 surface datasets M components/clm/test/tools/test_driver.sh --- Make CLM_SOFF default FALSE M components/clm/test/tools/README ----------- Show example usage on yellowstone and update note about CLM_SOFF M components/clm/tools/clm4_5/mksurfdata_map/src/mkfileMod.F90 ----- Don't write T_BUILDING_MAX M components/clm/tools/clm4_5/mksurfdata_map/src/mkurbanparMod.F90 - Remove T_BUILDING_MAX as urban parameter M components/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl --------- Remove T_BUILDING_MAX from copy for urban single point datasets =============================================================== =============================================================== Tag name: clm4_5_11_r188 Originator(s): erik (Erik Kluzek) Date: Fri Aug 5 13:53:01 MDT 2016 One-line Summary: A set of small answer changing issues for both clm45 and clm50 Purpose of changes ------------------ Changes to CLM45 and CLM50: Some adjustments to methane output (for both CLM45BGC) and CLM50BGC from Xiyan Xu. This calculates finundated from before snow, and rather than multiplying m_tiller by LAI it multiples by a constant four. Update Li2014 FireFluxes subroutine for spinupstate=2 capability (so can be shared). Changes to roundoff, while also allowing spinupstate==2 and fixing bug 2243 to deal with deadstem for CLM45 from Charlie Koven. Carbon and Nitrogen state missing accounting of of live to dead from fire in StateUpdate3. Nitrogen uptake code error from Xiaojuan Yang and Charlie Koven. This is dependent on use_nitrif_denitrif=T so will only affect CLM45BGC/CLM50BGC, but not SP or CN alone simualtions. This is for the 2nd pass at Nitrogen uptake. Issue with carbon Isotopes. Brett Raczka and Henrique Duarte at Univ. Utah have identified a bug in the CLM4.5 C isotope code, in which isotope ratios are not stabilizing when fire turns on. Turning off the fire code produces reasonable isotope ratios. We've identified that the problem is that the variables for the fire fluxes associated with the new fire code were never copied over into CNCIsoFluxMod.F90. Fix from Charlie Koven, and Brett verified it worked in his version. Also update the C14 half-life value. Used to be the Libbey value, now use the updated one. Changes to CLM50 only: Remove discontiniues in new CLM50 snow density. Also if wind is zero (< 0.1m/s), wind_dependent_snowdensity is increase is also zero. Some changes to fire injection height for fire-emissions from Simone Tilmes, and a unit-conversion error fix from Fang Li. Summary: ciso errors bug 2116 (isotopes both clm45/clm5) Bug 2224 N uptake code clm45/clm50BGC Transfer of live to dead bug 2264 clm45/clm5) 2326 for shade part clm5 snow density discontinuities (2328) clm5 Methane change on mechanges branch clm45/clm5 2338 2340 Fire emission change from Simone clm5 Units conversion error in CNfire-emissions bug 2342 clm5 Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2116 C isotope (13C and 14C) errors in fire code 2141 Precision of constants in SoilTemperature.F90 is not double 2224 plant N uptake second pass error 2243 Deadstem not taken into account in fire-fluxes for clm4_5 (koven) (FOR CLM45 SIDE) 2264 Fire is missing accounting of C/N for transfer of live to dead 2326 Bad multiplication of btran for soybean or nbrdlf_dcd_tmp_shrub with CNDV in CanopyFluxes 2328 Discontinuities in new CLM50 snow density formulation 2338 Use the pre-snow values to calculate finundated, and use constant for m_tiller rather than elai 2340 Adjustments to fire-emission from Simone Tilmes 2342 Units conversion error in CNfire-emissions Known bugs introduced in this tag (include bugzilla ID): 2332 fire_emiss tests don't work (cime issue) Notes of particular relevance for users --------------------------------------- Caveats for users: None Changes to CLM's user interface: None Changes made to namelist defaults: None Changes to the datasets: None Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: None Changes to tests or testing: None Code reviewed by: self,fangli,slater,stilmes,koven CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_10_r187 Answer changes -------------- Changes answers relative to baseline: Yes! for clm4_5 and clm5_0 Summarize any changes to answers, i.e., - what code configurations: clm4_5 BGC and CN compsets and Carbon isotopes clm5_0 all (BGC, CN, SP) any fire-emission - what platforms/compilers: All - nature of change: similar climate Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): cime Add changes to cime branch tag to be able to run on hobart cime to clm4_5_11_r188_cime4.5.14 List all files eliminated: None List all files added and what they do: Add unit tests for new snow bulk density M components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_newSnowBulkDensity.pf List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNNStateUpdate3Mod.F90 - Subtract live to dead from fire Nitrogen for live/dead and stem/root (both CLM45CN/BGC and CLM50CN/BGC) M components/clm/src/biogeochem/CNC14DecayMod.F90 ------ Adjust C14 half-life from Libby 5568 to 5730 a more recent value used (this affects both clm45 and clm50 when use_c14 is on) M components/clm/src/biogeochem/ch4Mod.F90 ------------- Calculate finundated from before snowfall, and don't multiply m_tiller by LAI, but by 4.0. M components/clm/src/biogeochem/CNCStateUpdate3Mod.F90 - Subtract live to dead from fire Nitrogen for live/dead and stem/root (both CLM45CN/BGC and CLM50CN/BGC) M components/clm/src/biogeochem/CNCIsoFluxMod.F90 ------ Missing isotope flux calculations (changes for CLM45/50 with c13/c14) M components/clm/src/biogeochem/CNFireEmissionsMod.F90 - Changes from Simone Tilmes adjusting the fire injection heights for different PFTS and also use indices rather than hard-coded magic integers M components/clm/src/biogeochem/CNFireLi2014Mod.F90 ---- Use same version of FireFluxes as in Li2016, so can use spinup_state==2 and apply a previous bug-fix (). Outside of that answers are roundoff different. And this allows the base version of the subroutine to be used for both Li2014 and Li2016 to reduce code duplication. M components/clm/src/biogeophys/CanopyFluxesMod.F90 ---- Remove 2nd adjustment to btran for soybean and nbrdlf_dcd_tmp_shrub for CNDV (both clm45crop/cndv and clm50crop/cndv) M components/clm/src/biogeophys/SnowHydrologyMod.F90 --- Make new snow bulk density continuous (only CLM50) add unit-test function: SnowHydrologySetControlForTesting. Apply new wind dependent density increase only for wind > 0.1m/sec (so no density increase at zero wind) M components/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 --- Bug 2224 M components/clm/src/biogeophys/test/SnowHydrology_test/CMakeLists.txt - add tests for snow density =============================================================== =============================================================== Tag name: clm4_5_10_r187 Originator(s): erik (Erik Kluzek) Date: Tue Jul 26 21:45:51 MDT 2016 One-line Summary: Some CLM45/50 changes: seperate carbon and water root profile, turn off undercanopystability, really turn PHS on, LUNA predict sunlit fraction N fractionation Purpose of changes ------------------ Seperate the carbon and water fine root profiles and have namelist control over them. This removes the duplicated code in soilbiogeochem that replicated the Jackson-1996 profile. Also remove several soilbiogeochem specific namelist control items over the root profile. CLM50 changes: Change water fine root profile to Jackson-1996 with rootprof_beta adjusted for water (dlawren) Turn off under canopy stability term that decreases csoilc by Sakaguchi&Zeng, 2008 (swensosc) When fun on -- add cpool_to_leafc to leafc_to_litter and cpool_to_frootc to frootc_to_litter (rfisher) Really turn Plant Hydrallic Stress on (erik) Have LUNA predict the scaling of Vcmax/Jmax/tpu-25 for canopy sunlight/shaded average in Photosynthesis (rfisher) CLM45 changes: Water fine root profile was used rather than carbon profile (used in soilbiogeochem) in CNMResp and in ch4Mod The profile used was zeng_2001, it's now Jackson_1996. You can duplicate the old answers if you use zeng-2001 for CNMResp and ch4Mod, but Jackson-1996 for soilbiogeochem. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2337, 2334, 2331, 2311 2337 Methane and maintenance respiration using water root profile rather than carbon root profile 2334 Plant hydraulic stress not turned on by default in clm4_5_9_r184 2331 many build-namelist tests fail when comparing a tag against itself 2311 Move mbbopt to params file Known bugs introduced in this tag (include bugzilla ID): 2339 2339 croofr and rootfr profiles are time constant but default output to history Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Namelist items added and removed Add: rooting_profile_method_water, rooting_profile_method_carbon, rooting_profile_varindex_water, rooting_profile_varindex_carbon, and use_undercanopy_stability Remove: pftspecific_rootingprofile, exponential_rooting_profile, and rootprof_exp Changes made to namelist defaults (e.g., changed parameter values): set: rooting_profile_method_carbon = Jackson1996 (both clm50 and clm45) rooting_profile_method_water = Zeng_2001 (clm45), Jackson1996 (clm50) use_undercanopy_stability = .true. (clm45), .false. (clm50) Changes to the datasets (e.g., parameter, surface or initial files): New params files (with 2D rootprof_beta) Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): None Changes to tests or testing: tests with dynroots need to explicitly turn PHS off Code reviewed by: self, dlawren, rfisher CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (83 compare tests FAIL, because of namelist changes) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_9_r186 Answer changes -------------- Changes answers relative to baseline: CLM45BGC and all CLM50 Summarize any changes to answers, i.e., - what code configurations: CLM45BGC and all CLM50 compsets - what platforms/compilers: All - nature of change: CLM45 CLM50 If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: rfisher respmods_n31_clm4_5_9_r185_00 Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/unit_testers/NMLTest/CompFiles.pm --- Fix bug comparison to same version (2331) M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- Remove: rooting_profile_method Add: rooting_profile_method_water, rooting_profile_method_carbon, rooting_profile_varindex_water, rooting_profile_varindex_carbon, and use_undercanopy_stability Remove: pftspecific_rootingprofile, exponential_rooting_profile, and rootprof_exp M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- new param files, set: rooting_profile_method_carbon, rooting_profile_method_water, and use_undercanopy_stability M components/clm/bld/namelist_files/namelist_defaults_drydep.xml -- Update to use 79 PFT megan factors file M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50dynroots/user_nl_clm --- add use_hydrstress = .false. M components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/user_nl_clm - add use_hydrstress = .false. M components/clm/bld/CLMBuildNamelist.pm --- Pass use_ed to use_hydrstress default (2334), add setup_logic_canopyfluxes namelist, change rooting_profile_method to rooting_profile_method_water/carbon, canopyfluxes_inparm namelist group. M components/clm/src/biogeochem/CNMRespMod.F90 ------ Use carbon fine root profile (crootfr_patch) rather than water profile (2337) M components/clm/src/biogeochem/CNFUNMod.F90 -------- Use carbon fine root profile (crootfr_patch) rather than water profile M components/clm/src/biogeochem/CNDriverMod.F90 ----- Use carbon fine root profile (crootfr_patch) rather than water profile M components/clm/src/biogeochem/ch4Mod.F90 ---------- Use carbon fine root profile (crootfr_patch) rather than water profile (2337) M components/clm/src/biogeochem/CNPhenologyMod.F90 -- When fun on -- add cpool_to_leafc to leafc_to_litter and cpool_to_frootc to frootc_to_litter from Rosie Fisher M components/clm/src/biogeochem/VOCEmissionMod.F90 -- Have factors handle new PFT size M components/clm/src/biogeochem/MEGANFactorsMod.F90 - Read mxpft PFT's from emissions file, check that number PFT's is as expected M components/clm/src/main/controlMod.F90 ------------ Remove exponential_rooting_profile, rootprof_exp, surfprof_exp, pftspecific_rootingprofile Add CanopyFluxesReadNML call M components/clm/src/main/pftconMod.F90 ------------- rootprof_beta is 2D with nvariants as 2nd dimension M components/clm/src/main/clm_varpar.F90 ------------ Add nvariants=2 for variants of PFT constants M components/clm/src/biogeophys/LunaMod.F90 --------- Make LUNA predict sunlit fraction N fractionation, scale in PhotosynthesisMod M components/clm/src/biogeophys/PhotosynthesisMod.F90 --------- Implement scaling of Vcmax/Jmax/tpu/25 from sunlit average to shaded canopy average value (both PHS and non-PHS versions), read mbbopt from params file in PHS section (2311) M components/clm/src/biogeophys/SoilStateType.F90 ------------- Add crootfr_patch for carbon and initialize. rootfr_patch is for water write out crootfr and rootfr to history (this should be cleaned up to be only for first history file as it's an unchanging constant profile) M components/clm/src/biogeophys/CanopyFluxesMod.F90 ----------- Add CanopyFluxesReadNML with canopyfluxes_inparm namelist and set use_undercanopy_stability in it, decrease csoilc as in in Sakaguchi&Zeng, 2008 if on M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 - Initialize crootfr M components/clm/src/biogeophys/RootBiophysMod.F90 ------------ Add water and carbon versions of rooting_profile_method and add water and carbon rooting_profile_varindx for variant index of rootprof_beta to use from params file, make m_to_cm and beta double precision, pass 'water' or 'carbon' to init_vegrootfr, varindx for carbon is 2 and 1 for water. default rooting method is zen_2001 for water and jackson_1996 for carbon. M components/clm/src/biogeophys/SurfaceRadiationMod.F90 ------- Remove unused single precision reals M components/clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 - Use crootfr_patch to set cinput_rootfr, remove version of jackson_1996 carbon root profile that was in here. Also remove: exponential_rooting_profile, pftspecific_rootingprofile, and rootprof_exp =============================================================== =============================================================== Tag name: clm4_5_9_r186 Originator(s): sacks (Bill Sacks) Date: Wed Jul 13 07:17:04 MDT 2016 One-line Summary: Initialize biogeochem balance checks AFTER column area updates Purpose of changes ------------------ Reworking the conservation of patch-level CN variables with dynamic landunits will make the column-level CN balance checks very tricky to determine, as they were written before - where the starting point for the comparison was determined before all of the area adjustments (i.e., the dyn subgrid stuff). This becomes much more straightforward if the initialization of the balance checks is done after the dyn subgrid stuff is complete. This tag makes that change. In order to accomplish that, we need to re-compute the CN state summaries after the dyn subgrid stuff is complete - near the beginning of the time step. We still need the summary routine at the end of the time step for the sake of diagnostics (e.g., history output). So now we call the summary routines twice for states. We do NOT need to call the flux summary routines at the beginning of the time step - and doing so doesn't really make sense (how do you summarize fluxes that haven't yet been computed) - so I have split the top-level summary routine into a SummarizeStates and SummarizeFluxes. I also reworked the ch4 balance check similarly, in order to be consistent with the CN balance checks. Longer-term, we'll add another set of balance checks that operate on gridcell-level quantities, which WILL cover the dyn subgrid stuff. So the overall sequence will look like: Initialize gridcell-level balance checks Dyn subgrid stuff (adjust patch, column, landunit areas, and do associated state updates) Initialize column-level balance checks Science routines Check column-level balance Check gridcell-level balance Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, some answers change as noted below CLM tag used for the baseline comparisons: clm4_5_9_r185 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 runs with BGC that involve changes in subgrid areas (including HIST, RCP, DV and IG cases) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated closely; expected to be larger than roundoff / same climate Answer changes come from the fact that some code (e.g., the fire code) uses the summarized state variables. Previously, this code was using the state summaries from the previous time step. With this change, the code is now using the updated state summaries. These updated state summaries can differ from the ones computed at the end of the last time step if subgrid areas have changed. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Move initialization of CN and ch4 balance checks to AFTER dyn subgrid M components/clm/src/main/clm_driver.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 ========= Split the CNDriverSummary into separate SummarizeStates and SummarizeFluxes: Only SummarizeStates needs to be done near the start of the time step. Also move calls to PrecisionControl to a higher level, since these aren't needed in the initial call to SummarizeStates. M components/clm/src/biogeochem/CNDriverMod.F90 ========= Change which terms (from transient landcover) are included in the balance checks, due to new placement in the driver sequence M components/clm/src/biogeochem/CNBalanceCheckMod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 ========= Rework timing of ch4 balance check initialization, too, for consistency M components/clm/src/biogeochem/ch4Mod.F90 ========= Remove some variables from Restart, now that we call the Summary routines early in the time step. Also no longer update summary diagnostic variables in DynamicPatchAdjustments M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 ========= No longer need to do p2c of totvegc in fire code: this is done in the initial call to the summary routines M components/clm/src/biogeochem/CNFireLi2014Mod.F90 M components/clm/src/biogeochem/CNFireLi2016Mod.F90 M components/clm/src/biogeochem/CNFireNoFireMod.F90 ========= Replace __FILE__ with filename parameter to avoid too-long lines with pgi M components/clm/src/soilbiogeochem/SoilBiogeochemNitrifDenitrifMod.F90 ========= Remove SMS_Lm25.f19_g16.ICLM45BGCCROP.yellowstone_gnu.clm-cropMonthOutput: baseline comparisons have been passing for me recently M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_9_r185 Originator(s): sacks (Bill Sacks) Date: Thu Jul 7 13:54:09 MDT 2016 One-line Summary: Revert glacial melt behavior to the old treatment for mountain glaciers Purpose of changes ------------------ The primary change in this tag is reverting glacial melt behavior to the old scheme (i.e., the scheme used in standard glacier landunits) for mountain glaciers: The old scheme involves glacial melt remaining in place until it refreezes (thus, not generating any runoff), whereas the new scheme involves glacial melt running off and being replaced by ice from below (generating positive liquid runoff and negative ice runoff). The main motivation for this change is that mountain glaciers were generating substantial negative ice runoff in coupled cases when using the new scheme. This is presumably because CAM's topographic height is significantly lower than the real-world topographic height in mountainous regions, so mountain glaciers were seeing too-warm temperatures. The LIWG, in consultation with members of the LMWG, decided to revert mountain glaciers to their old behavior to avoid detrimentally impacting a coupled simulation. Because the LIWG is not currently concerned with the SMB of mountain glaciers - and because mountain glacier SMB cannot be realistically simulated yet anyway - this reversion was not considered a problem from their perspective. To accomplish this, this tag introduces a new namelist variable, glacier_region_melt_behavior. Like glacier_region_behavior, this keys off of the GLACIER_REGION field on the surface dataset. Thus, different melt behavior can be used for each glacier region. By default, Greenland and Antarctica use the new (glc_mec) behavior ('replaced_by_ice'), whereas other regions use the old behavior ('remains_in_place'). For glacier columns that use the 'remains_in_place' melt behavior, we do NOT compute surface mass balance (SMB). This is because the negative component of SMB is non-sensical in those columns. To make this clear, the QICE, QICE_FRZ and QICE_MELT diagnostic fields are filled with missing values in the 'remains_in_place' region. (However, the qice fields sent to the coupler, along with QICE_FORC, are 0 in that region - giving the values that are actually sent to the coupler.) Note that this means that different glacier regions now have different physics. And, in order to compute SMB, you would need to change the melt behavior, thus significantly changing the behavior of mountain glaciers. This tag also refactors the code that computes glacier surface mass balance so that this is centralized in one module: GlacierSurfaceMassBalanceMod.F90. Other behavior changes in this tag are: (1) The subgrid snow parameterization (subgrid topographic variability) now differentiates between 'single_at_atm_topo' istice_mec columns and other istice_mec columns. Previously, istice_mec columns were treated differently from others because they already have some representation of subgrid variability. However, 'single_at_atm_topo' columns should be treated more like other columns, because they do not allow for this elevation class variability; this has been fixed here. (2) Vegetated SMB is computed for all grid cells where we compute glacier SMB. Previously vegetated SMB was only generated in the 'virtual' glacier region, but that added complexity to the code logic without having a good reason (as far as I can tell). This change just affects the lnd -> glc fields for elevation class 0 (changing where these are valid, rather than being set to 0): it does not affect the evolution of CLM. (3) Adds some error checking for the glacier_region_behavior namelist item as well as for the new glacier_region_melt_behavior namelist item. Between the primary change in this tag and the change in (1), it is now possible to make the evolution of glacier_mec landunits exactly match the evolution of standard glacier landunits - by using a glacier_region_behavior of 'single_at_atm_topo' and a glacier_region_melt_behavior of 'remains_in_place'. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Surface mass balance (SMB) is no longer computed for mountain glaciers. In order to compute SMB, you would need to change the glacier_region_melt_behavior, which will change behavior significantly for mountain glaciers. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variable, glacier_region_melt_behavior - as described above. This variable and glacier_region_behavior both appear in the default namelist for all clm45/clm50 cases, but only affect the operation of CLM if running with glc_mec (i.e., a case with CISM). Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: none Changes to tests or testing: none Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - ok (tests pass, expected differences from baseline; but see bug 2331) unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, some answer changes as noted below In addition, ran 100-year runs to confirm that glc_mec (istice_mec) landunits can now be made to operate exactly the same as standard glacier (istice) with the appropriate namelist settings: IG run: 1850_DATM%QIA_CLM50%BGC_SICE_SOCN_RTM_CISM1_SWAV I run: 1850_DATM%QIA_CLM50%BGC_SICE_SOCN_RTM_SGLC_SWAV For the IG run, put the following in user_nl_clm: glacier_region_behavior = 'single_at_atm_topo','single_at_atm_topo','single_at_atm_topo' glacier_region_melt_behavior = 'remains_in_place','remains_in_place','remains_in_place' and with the following sourceMods to allow running an IG compset with single_at_atm_topo & remains_in_place behavior over Greenland (because we currently cannot run with SGLC with glc_mec): --- ../../../components/clm/src/cpl/lnd_import_export.F90 2016-07-01 06:11:13.117799391 -0600 +++ SourceMods/src.clm/lnd_import_export.F90 2016-07-04 09:58:37.130458468 -0600 @@ -228,8 +228,8 @@ glc2lnd_inst%topo_grc(g,num) = x2l(index_x2l_Sg_topo(num),i) glc2lnd_inst%hflx_grc(g,num) = x2l(index_x2l_Flgg_hflx(num),i) end do - glc2lnd_inst%icemask_grc(g) = x2l(index_x2l_Sg_icemask,i) - glc2lnd_inst%icemask_coupled_fluxes_grc(g) = x2l(index_x2l_Sg_icemask_coupled_fluxes,i) + glc2lnd_inst%icemask_grc(g) = 0._r8 + glc2lnd_inst%icemask_coupled_fluxes_grc(g) = 0._r8 end if end do Both run at T31_g37 for 100 years; compared last CLM and RTM history files; these were bit-for-bit except for expected differences in PCT_LANDUNIT Also ran similar sets of runs, but with all temperatures increased by 10 deg (along with setting initial glacier h2osno to 0) or decreased by 10 deg (along with setting initial glacier h2osno to 10000) to try to make sure we're exercising both glacier increases and decreases (though I didn't confirm that we're definitely exercising increases and decreases in those). Ran those hot & cold runs with both CLM50 and CLM45. CLM tag used for the baseline comparisons: clm4_5_9_r184 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with glc_mec (i.e., compsets with CISM) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): expected to be new climate, at least regionally - but not investigated closely Answers change over glaciers outside of Greenland and Antarctica (i.e., mountain glaciers). In addition, fields sent to coupler are now present for elevation class 0 (bare ground) over Antarctica. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Consolidate code that computes glacier surface mass balance. This code was primarily taken from SoilTemperatureMod and HydrologyDrainageMod, with some other code also removed from BalanceCheckMod and clm_driver. Relevant fluxes have also been moved here (from WaterfluxType) A components/clm/src/biogeophys/GlacierSurfaceMassBalanceMod.F90 List all existing files that have been modified, and describe the changes: ========= Add new glacier_region_melt_behavior namelist item; always add glacier_region_behavior and glacier_region_melt_behavior to namelist, so error checks pass (these currently aren't needed in cases without CISM, but will always be needed soon, once glc_mec becomes the only glacier option - so it seemed reasonable to go ahead and always add them to the namelist) M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ========= Add differentiation on glacier_region_melt_behavior, add an 'allow_multiple_columns_grc' flag that can be used elsewhere, improve error checking, and some other tweaks M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/main/test/glcBehavior_test/test_glcBehavior.pf ========= Move code out, into new GlacierSurfaceMassBalanceMod M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 ========= Call new routines, move initialization to 0 out of clm_drv_init into new GlacierSurfaceMassBalanceMod M components/clm/src/main/clm_driver.F90 ========= Change do_smb filter: (1) only includes points where melt_replaced_by_ice is true; (2) includes vegetated points for the same grid cells as glacier points, rather than using the has_virtual_columns mask M components/clm/src/main/filterMod.F90 ========= For glc_mec columns: Subgrid topographic variability depends on whether a gridcell allows multiple elevation classes M components/clm/src/main/initVerticalMod.F90 ========= Add new glacier_smb_inst, and pass it where needed M components/clm/src/main/clm_instMod.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/lnd2glcMod.F90 ========= Do some additional error checking M components/clm/src/main/glc2lndMod.F90 ========= Build new code for unit tests M components/clm/src/biogeophys/CMakeLists.txt ========= Unrelated change: do necessary teardown M components/clm/src/dyn_subgrid/test/dynPatchStateUpdater_test/test_patch_state_updater.pf ========= Just remove unneeded use statements M components/clm/src/biogeochem/DUSTMod.F90 M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/histFileMod.F90 ========= Remove a bunch of tests that no longer fail (due to changes in r184 or earlier) or have been removed from the test list; also change bug associated with one test M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_9_r184 Originator(s): erik (Erik Kluzek) Date: Wed Jul 6 14:55:45 MDT 2016 One-line Summary: CLM50 changes: Atkin form of lmr changed, new parameters and files, turn PHS on by default, new cold snow density, reseed at exit AD spinup, fix fire bug, fix npool and cpool errors Purpose of changes ------------------ The CLM50 Atkin method of leaf maintenence respiration was changed by Dr. Rosie Fisher according to communication with Atkin. New parameter file for CLM50 with new parameters. Plant Hydraulic Stress is now on by default for CLM50. New formulation for snow density below -15C from Dr. Andrew Slater and Dr. David Lawrence. If plants are dead at the exit AD spinup stage, reseed. Fix a fire bug which allowed negative nitrogen pools. Fix some npool and cpool errors. Work on CNFUNMod, correct indentation, and headers and comments. Remove unused arrays. Initialize arrays in filter loops. Do some changes to increase performance. Make logic for CNPrecisionControl into three subroutines, when especially negative abort. -- Changes from Rosie Fisher to remove leaf_acc term, and implement changes to the Atkin form of lmr25top based on personal communication with Atkin. -- new params file -- Read in PHS parameters from PFT params file. -- CNPrecisionControl modularization -- new method for snow density and a change to how wind_dependent_snow_density is done for clm5.0. upplim_destruct_metamorph and -- overburden_compress_tfactor for clm5 -- Put mbbopt on params file and use it -- removed deciduous retranslocation adjustment to plant_ndemand to fix cpool error -- Checkin Keith's changes to reseed Carbon for forests if in spinup mode and returning to regular mode, this should help in making it so -- that you can always use spunup initial conditions rather than always having to do a cold start -- Some FUN optimization. -- Set barriers, ATM_PIO_TYPENAME -- removed deciduous retranslocation adjustment to plant_ndemand to fix cpool error -- If leafn small set leaf-CN and storage_leaf-CN to missing value in NutrientCompetitionFlexibleCNMod -- add in free N flux direct to npool -- Move update of grainn_to_food side by side to grainc_to_food with the same equation, from Rosie and Keith (bug 2327) -- fire (bug 2314) -- soybean CNDV shrub (bug 2326) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2255, 2314, 2326, 2327 2327 Change location of where grainn_to_food is updated for CROP 2326 Bad multiplication of btran for soybean or nbrdlf_dcd_tmp_shrub with CNDV in CanopyFluxes (partial see below) 2314 Negative carbon pools due to issue in fire model 2255 bug in soil resistance, likely causes crop soil resistance to remain unset Known bugs introduced in this tag (include bugzilla ID): 2326, 2328, 2330 2330 Issue with frootc/frootn for C4 crops 2328 Discontinuities in new CLM50 snow density formulation The issue that created the need for bug 2326 was removed for both CLM45 and CLM50 in this tag (The same code was left in the CLM40 version) Notes of particular relevance for users --------------------------------------- Caveats for users: None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist items: ncrit, ccrit, nnegcrit, cnegcrit, baseflow_scalar, denitrif_respiration_coefficient/exponent (and nitrconc), lotmp_snowdensity_method, upplim_destruct_metamorph, overburden_compress_Tfactor, min_wind_snowcompact, remove leaf_acc Changes made to namelist defaults: Set new namelist items, update CLM50 params file Changes to the datasets: CLM50 params file with updated parameters Substantial timing or memory changes: CLM50BGC is a good bit faster Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: Now reading some mxpft dimensioned arrays in local readParams SnowHydrologyMod parameters should go into a params_type Changes to tests or testing: None Code reviewed by: self,oleson,rfisher,dlawren,aslater@cu.edu,lifang CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (78 compare tests FAIL because of namelist changes) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_9_r183 Answer changes -------------- Changes answers relative to baseline: Yes for CLM50 and CLM45CROP Summarize any changes to answers, i.e., - what code configurations: CLM50 or CLM45CROP - what platforms/compilers: All - nature of change: CLM50 ------ Modified climate CLM45CROP -- soybean results and grainn_to_food (all CFTS) will be different If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: rfisher: respmods_n30_clm4_5_8_r182_4x5_bareground_AD_1 (has some other SourceMods too) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: D components/clm/bld/namelist_files/namelist_definition_drv_flds.xml List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/main/clm_initializeMod.F90 ------------ Always call readParameters M components/clm/src/main/controlMod.F90 ------------------- Add calls to: soilHydReadNML, nitrifReadNML, CNPrecisionControlReadNML M components/clm/src/main/clm_driver.F90 ------------------- Work with timing calls M components/clm/src/main/pftconMod.F90 -------------------- Read in mbbopt M components/clm/src/main/readParamsMod.F90 ---------------- Add some comments, pass in photosyns_inst and call it's readParams M components/clm/src/biogeophys/SoilWaterMovementMod.F90 --- Use Photosyns_inst->params_inst%psi_soil_ref in place of -50000 M components/clm/src/biogeophys/SoilHydrologyMod.F90 ------- Add soilHydReadNML, read in baseflow_scalar M components/clm/src/biogeophys/SnowHydrologyMod.F90 ------- New namelist items: lotmp_snowdensity_method, upplim_destruct_metamorph, overburden_compress_Tfactor, min_wind_snowcompact, new_snowdensity method has lower density for < -15C, change to wind_dependent_snow_density M components/clm/src/biogeophys/PhotosynthesisMod.F90 ------ New parameters for Plant Hydralic Stress: krmax, kmax, psi50, psi_soil_ref, lmr_intercept_atkin, remove leaf_acc. Add: allocParams, readParams methods remove mbbopt and move into pftcon. Atkin method of leaf respiration modified per communication with Atkin. M components/clm/src/biogeophys/SurfaceResistanceMod.F90 --- Properly use land-unit rather than column (bug 2255) M components/clm/src/biogeophys/CanopyFluxesMod.F90 -------- Remove multiplication of btran for soybean or nbrdlf_dcd_tmp_shrub for CNDV, inside the ITERATION loop (bug 2326). This means the reason that brought this code is no longer addressed. M components/clm/src/soilbiogeochem/SoilBiogeochemNitrifDenitrifMod.F90 --- Add nitrifReadNML, read in: denitrif_respiration_coefficient, denitrif_respiration_exponent, denitrif_nitrateconc_coefficient, denitrif_nitrateconc_exponent. Remove k_nitr_max M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 - Remove hardwired minpsi M components/clm/src/biogeochem/CNFireBaseMod.F90 --------- Multiply several terms by (1-cc_other) (bug 2314) M components/clm/src/biogeochem/CNPrecisionControlMod.F90 - Add CNPrecisionControlReadNML, cnegcrit, nnegcrit do actual truncation inside new subroutines: TruncateCandNStates, TruncateCState, TruncateNState Die if C or N is very negative < cnegcrit or < nnegcrit (currently set to very large negative numbers -60 and -6) M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 Set actual_storage_leafcn to missing value if leafn_storage is zero. actual_leafcn to missing if leafn is < n_min, remove modification of plant_ndemand if FUN off M components/clm/src/biogeochem/CNVegCarbonStateType.F90 - Reseed if plants are dead and going from Accel. Decomp. to regular mode M components/clm/src/biogeochem/CNNStateUpdate1Mod.F90 --- Add handling of npool and retransn M components/clm/src/biogeochem/CNFUNMod.F90 ------------- Add more comments, correct indentation, speed up, and remove extra garbage, use filters to zero out arrays add free_Nretrans and fraction of N retranslocation which is automatic/free M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 Add free_retransn_to_npool_patch, write out to history, write LEAFN_TO_LITTER to history by default M components/clm/src/biogeochem/CNDriverMod.F90 ----------- Pass bounds to CNPrecisionControl M components/clm/src/biogeochem/CNPhenologyMod.F90 -------- grainn_to_food modified to be similar to Carbon version (bug 2327) for FUN leafn_to_retransn is paid_retransn_to_npool plus free_retransn_to_npool M components/clm/bld/unit_testers/build-namelist_test.pl ----------- Add in new tests for new parameters M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - New namelist items: ncrit, ccrit, nnegcrit, cnegcrit, baseflow_scalar, denitrif_respiration_coefficient/exponent (and nitrconc), lotmp_snowdensity_method, upplim_destruct_metamorph, overburden_compress_Tfactor, min_wind_snowcompact, remove leaf_acc M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Add CPOOL to history for clm_accelerated_spinup remove leaf_acc, set: lotmp_snowdensity_method, upplim_destruct_metamorph, overburden_compress_tfactor, new params file for CLM50, set ncrit, turn use_hydrstress=T for CLM50 M components/clm/bld/CLMBuildNamelist.pm --------------------------- Add setup_logic_cnprec, call setup_logic_nitrif_params, setup_logic_hydrology_params, add in cnprecision_inparm namelist M components/clm/cime_config/buildnml ------------ Don't need to copy user_nl_clm to Buildconf use CASEROOT version be sure to set $tuning options for build-namelist M components/clm/cime_config/config_component.xml Add BARRIER_N/OPTION, ATM_PIO_TYPENAME settings =============================================================== =============================================================== Tag name: clm4_5_9_r183 Originator(s): erik (Erik Kluzek) Date: Sat Jul 2 18:23:09 MDT 2016 One-line Summary: Decomposition is not water limited after reaching a field capacity parameter (maxpsi_hr) rather than only at soil saturation Purpose of changes ------------------ Replaces the hard-coded assumption present in prior versions of the model that soil respiration was water-limited whenever it fell below saturation. Now it's no longer water-limited from maxpsi_hr up to saturation. This was due to a wrong interpretation/implementation of the following paper. Andren, O., and K. Paustian, 1987. Barley straw decomposition in the field: a comparison of models. Ecology, 68(5):1190-1200 Also see notes in.. http://bb.cgd.ucar.edu/cesm12z-clm45-important-bad-interpretation-maximum-rate-constant-soil-water-content Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2324 Bad interpretation of maximum rate constant for soil water content in SoilbiogeochemDecompCascade Notes of particular relevance for users --------------------------------------- Caveats for users: None Changes to CLM's user interface: None Changes made to namelist defaults: new params files Changes to the datasets: parameter files Substantial timing or memory changes: None Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers (e.g., code that is duplicated that requires double maintenance): maxpsi_hr is a scalar constant but read in on the parameter file (should change to namelist) Changes to tests or testing: Removed RTM ESMF test Code reviewed by: self, ckoven, dlawren CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (70 comparision tests fail, because of the new params files) unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_8_r182 Answer changes -------------- Changes answers relative to baseline: Yes, for CLM4_5/CLM5_0 for BGC and CN cases Summarize any changes to answers: - what code configurations: clm4_5 and clm5_0 with BGC or CN - what platforms/compilers: all - nature of change: should be similar climate Charlie Koven did some work with sensitivity analysis to find a good value for maxpsi_hr Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): rtm rtm to rtm1_0_57 -- remove ESMF test List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- new params file with maxpsi_hr on them M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 - Use maxpsi_hr from params file rather than max soil saturation M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeCNMod.F90 -- Use maxpsi_hr from params file rather than max soil saturation =============================================================== =============================================================== Tag name: clm4_5_8_r182 Originator(s): erik (Erik Kluzek) Date: Fri Jun 24 23:33:36 MDT 2016 One-line Summary: Bring in option for plant hydraulic stress for clm50 Purpose of changes ------------------ Add in option for plant hydraulic stress (PHS). Does Leaf stomatal resistance and leaf photosynthesis simultaneous solution of sunlit/shaded per Pierre Gentine/Daniel Kennedy PHS method. Vegetation water potential is calculated for four conductance segments: sun, shade, xylem, and root. The main impact in the code is in CanopyFluxes where either the PHS or non-PHS version of Photosynthesis is called and the sun and shade segments are solved simultaneously. Then it impacts soil water movement for transpiration. PHS is turned on by adding use_hydrstress=.true. in your user_nl_clm file for your case. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): 2323 (duplication in Photosynthesis) Notes of particular relevance for users --------------------------------------- Caveats for users: use_dynroots and use_hydrstress can NOT both be on at the same time! Changes to CLM's user interface: None Changes made to namelist defaults: Add use_hydrstress=.false. Changes to the datasets: None Substantial timing or memory changes: Maybe 4% slower with PHS, memory only increases by 0.1MB Notes of particular relevance for developers: (including Code reviews and testing) --------------------------------------------- Caveats for developers: PhotosynthesisMod is duplicated for PHS and non-PHS PhotosynthesisMod more than doubles in size from 2k to 4.5k. Most of it is duplicated. Thus changes to one part of the code, have to be duplicated in the other part, causing double maintence. Also the effects of 6 namelist control items are duplicated meaning those items have to be tested in both branches to ensure they are working correctly. (bug 2323) Changes to tests or testing: Add use_hydrstress=T to KitchenSink tests add clm50dynroots test directory Code reviewed by: self,oleson,Daniel Kennedy CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (78 compare tests fail compared to r181 because of addition of use_hydrstress=.false.) unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - No PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - No regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_8_r181 Answer changes -------------- Changes answers relative to baseline: No (only changes answers if use_hydrstress=.true.) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, mosart, cism, etc.): None List all files eliminated: None List all files added and what they do: ::::::::::: Add specific tests for dynamic roots A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50dynroots/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50dynroots/include_user_mods List all existing files that have been modified, and describe the changes: ::::::::::: Add in use_hydrstress namelist item, add in vegetation water columns mainly handled in Photosynthesis and ::::::::::: then impacts soil water transpiration in soil water movement. M components/clm/src/main/controlMod.F90 ------------ Read in use_hydrstress, check it M components/clm/src/main/clm_varctl.F90 ------------ Add use_hydrstress M components/clm/src/main/clm_driver.F90 ------------ Pass canopystate to HydrologyDrainage M components/clm/src/main/histFileMod.F90 ----------- Add ability to handle nvegwcs dimension M components/clm/src/main/restFileMod.F90 ----------- Write out nvegwcs dimension if use_hydrstress M components/clm/src/main/clm_varpar.F90 ------------ Set nvegwcs number of vegetation water conductance segments to 4 M components/clm/src/biogeophys/PhotosynthesisMod.F90 ------ Add PHS versions of subroutines M components/clm/src/biogeophys/SoilStateType.F90 ---------- Add djk column soil transpiration sink by layer, add SMP and HK to restart file. Set SMP to -1000 for cold start. M components/clm/src/biogeophys/CanopyFluxesMod.F90 -------- Call PHS or regular version of Photosynthesis don't limit or adjust qflx_trans_veg by btran0 for PHS M components/clm/src/biogeophys/SoilWaterMovementMod.F90 --- Add Compute_VertTranSink_PHS, pass canopystate, energyfluxstate down to soilwater zengdecker and moisture_form, for PHS rootr_col is zeroed only for pervious road. Compute_VertTranSink_PHS apply transpiration as a sink condition that is vertically distributed over the soil column. M components/clm/src/biogeophys/SoilMoistStressMod.F90 ----- btran only set if NOT PHS M components/clm/src/biogeophys/EnergyFluxType.F90 --------- Add bsun/bsha canopy transpiration wetness factor for PHS M components/clm/src/biogeophys/CanopyStateType.F90 -------- Add vegwp_patch for PHS, add to restart and history (vegetation water matric potential for sun/sha canopy,xyl,root segments) M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 - Pass in canopystate and pass canopystate, energystate to SoilWater ::::::::::: Handle use_hydrstress in namelist M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Define use_hydrstress M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Set default use_hydrstress to .false. M components/clm/bld/CLMBuildNamelist.pm --------------------------- Add setup_logic_hydrstress, use_hydrstress and use_dynroots cannot both be on at the same time. M components/clm/bld/unit_testers/build-namelist_test.pl ----------- Add checks that can't use PHS with ED or dynroot ::::::::::: Add dynroot tests and turn on use_hydrstress for KitchenSink tests M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm - Add use_hydrstress=T, remove use_dynroots=T M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KSinkMOut/user_nl_clm --- Add init_interp_fill_missing_with_natveg=T so test will pass as SMP is now required. M components/clm/cime_config/testdefs/testlist_clm.xml - Add 3 clm50dynroots testing, remove ESMF tests =============================================================== =============================================================== Tag name: clm4_5_8_r181 Originator(s): erik (Erik Kluzek) Date: Fri Jun 17 16:39:22 MDT 2016 One-line Summary: Update cime version which fixes several issues and changes answers Purpose of changes ------------------ Update for CIME: This includes several key important changes and bug-fixes - Now there is a CLMGSWP3 option for DATM_MODE - When running with datm driver has a barrier each day (so MP_EAGER_LIMIT is no longer needed) - PIO1 is now the default (PIO2 still exists as an option, but now both exist and you can choose which one) - Driver now merges CLM's drv_flds_in file with CAM's and aborts on a conflict - The drv_flds_in namelists are defined in the driver where the source exists - The drv_flds_in file is recreated just like other namelist files (no longer have to delete earlier one) - Changes answers because of shr_orb (see below) Changes to tools to include Caspian Sea as a CLM Lake on surface datasets Add LND_TUNING_MODE as an env_run.xml variable. The default setting is determined by which major physics version is being used (clm4_0, clm4_5, or clm5_0). We intend to use this in order to tune parameters for clm5_0 offline with GSWP3 forcing versus clm5_0 with CAM5.5. CLM_ACCELERATED_SPINUP="on" now cuts back I/O and is much faster (only outputs a few fields yearly). It also applies for both with and without CN. If CN is on, it will by default turn Accellerated Decomposition on. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2308 prebeta clm40 compsets are using the wrong testmods directories 2307 MOSART assumes no-leap calendar 2303 user_compset gives different settings than expected for options defined in config_compsets.xml 2280 Need to set barriers for I compsets with PE layouts where datm runs concurrently to CLM 2256 Problems with early versions of PIO2 in CLM 2252 Changing megan_specifier for I cases doesn't actually change the megan_specifier in the resolved namelist 1674 RTM assumes no-leap calendar CIME Issues fixed (include issue #): #315 Add CLMGSWP3 #324 Merge drv_flds_in in driver from different components #56 SSP test treats ref1 case as the main case Known bugs introduced in this tag (include bugzilla ID): cime issue #453 (ESMF tests fail) cime issue #455 (have to delete Macros file if switch to mpi-serial) Notes of particular relevance for users --------------------------------------- Caveats for users: CLM_ACCELERATED_SPINUP mean is different (see below) Changes to CLM's user interface: Add LND_TUNING_MODE, CLM_ACCELERATED_SPINUP changed. LND_TUNING_MODE added in with defaults for each CLM physics option (clm4_0, clm4_5, and clm5_0) clm4_0 wasn't tuned for a specific forcing, clm4_5 was tuned for CRUNCEP forcing, clm5_0 is currently tuned for running with CAM5.5. We eventually expect to have tunings for clm5_0 for running offline and also with CAM5.5. Also CLM_ACCELERATED_SPINUP is now changed and available whether CN is on or not. For a SP (Satelite Phenology) case this mode reduces output in order to speed up the simulation. For a CN case it does that as well as turning on spinup_state to an accelerated decomposition mode. If you don't want CN to be in accelerated decomposition set spinup_state=0 in your user_nl_clm (for final spinup after AD mode). Changes made to namelist defaults: nsegspc is now 35 instead of 20 Changes to the datasets: None Changes for tools: Files for surface dataset creation now set Caspian sea as a Lake new maps for mksurfdata with Caspian Sea Substantial timing or memory changes: The change in nsegspc should improve performance slightly Running with CLM_ACCELERATED_SPINUP="on" will make a good 20% improvement in performance from less history output and turning megan off Code reviews and testing ------------------------ Code reviewed by: self CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - OK (All PASS, but 121 tests are different that r180, because nsegspc changes, new namelists, and spinup_state) unit-tests (components/clm/src): yellowstone - PASS tools-tests (components/clm/test/tools): yellowstone - OK (Some mksurfdata compare tests fail because of Caspian as lake) PTCLM compare tests for clm4_5 fail because of new mapping files being used) PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK (all but one PASS and 44 compare tests are different as expected) U-UMB ------- Testing this resolution works regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_8_r180 Answer changes -------------- Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: all including clm4_0 and clm4_5/clm5_0 - what platforms/compilers: all - nature of change: same climate The cime update includes a change to shr_orb that changes answers for all simulations. There is also a change in vertical_gradient calculations for glacier model that changes answers when CISM is active (IG compsets). Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): cime, rtm, mosart cime to clm4_5_8_r180_cime4.5.14 (This includes option for DATM=CLMGSWP3) rtm to rtm1_0_56 mosart to mosart1_0_16 List all files eliminated: Remove NCL script to compare files (can now use -m option to FORTRAN cprnc program) D components/clm/tools/shared/ncl_scripts/cprnc.pl --- use FORTRAN cprnc D components/clm/tools/shared/ncl_scripts/cprnc.ncl -- use FORTRAN cprnc D components/clm/bld/namelist_files/namelist_definition_drv_flds.xml --- Now in driver List all files added and what they do: None List all existing files that have been modified, and describe the changes: ::::::::::::::::::: Replace setting of CPRNC_EXE with CPRNC_OPT="-m" M components/clm/test/tools/CLM_compare.sh ----- Add use of CPRNC_OPT M components/clm/test/tools/test_driver.sh ----- Export CPRNC_EXE and set default CPRNC_OPT to blank M components/clm/test/tools/TBLCFGtools.sh ----- Replace CRPNC_EXE with CPRNC_OPT="-m" M components/clm/test/tools/TOPtools.sh -------- Replace CRPNC_EXE with CPRNC_OPT="-m" M components/clm/test/tools/TBLscript_tools.sh - Replace CRPNC_EXE with CPRNC_OPT="-m" M components/clm/test/tools/TBLtools.sh -------- Replace CRPNC_EXE with CPRNC_OPT="-m" M components/clm/test/tools/TSMscript_tools.sh - Use $CSMDATA/lnd/clm2/PTCLMmydatafiles.c160208 directory rather than older directory M components/clm/test/tools/nl_files/PTCLM_USUMB_clm4_5 -------- Update map date to 160208 M components/clm/test/tools/nl_files/PTCLM_USUMB_Cycle_clm4_5 -- Update map date to 160208 M components/clm/test/tools/nl_files/PTCLM_USUMB_Global_clm4_5 - Update map date to 160208 M components/clm/tools/shared/ncl_scripts/README - Remove mention of cprnc.pl/cprnc.ncl M components/clm/tools/shared/mkmapdata/mkmapdata.sh -------- Create map for 3x3min_MODIS-wCsp instead of 3x3min_MODIS M components/clm/tools/clm4_5/mksurfdata_map/Makefile.data -- By default also create f45 1850 and transient datasets M components/clm/bld/unit_testers/build-namelist_test.pl --------------- Add a ton more testing M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ------- Add lnd_tuning_mod, clm_accelerated_spinup, megan, atm_c14_filename use_c14. Set spinup_state by clm_accelerated_spinup, set hist settings by clm_accelerated_spinup cutting back history output drastically. Set pot_hmn_ign_counts_alpha by lnd_tuning_mode. Set nsegspc=35. Change maps from 3x3min_MODIS to 3x3min_MODIS_wCsp to add Caspian Sea to lake mask M components/clm/bld/namelist_files/namelist_definition_clm4_0.xml ----- Add lnd_tuning_mod, clm_accelerated_spinup, megan. M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Add SCRIP grid file for 0.125x0.125 resolution. add SCRIP grid for 3x3min_MODIS-wCsp, lake mask is now MODIS-wCsp, update mksrf_flakwat to include Caspian Sea M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ----- Add some commented out namelist variables that will be added later. Correct or clarify some definitions. Add lnd_tuning_mode, clm_accelerated_spinup and megan. Remove MEGAN, drydep, and fire-emission namelist items as in drv namelist_definition now. M components/clm/bld/namelist_files/namelist_defaults_clm4_0.xml ------- Set lnd_tuning_mod, clm_accelerated_spinup, and megan. Also use history options to cut back history output drastically for clm_accelerated_spinup="on". M components/clm/bld/CLMBuildNamelist.pm ----- Add clm_accelerated_spinup, lnd_tuning_mode, remove bgc_spinup. Read in driver namelist-definition for drv_flds_in. Add nitrif_inparm, and soilhydrology_inparm although empty. Always write out (and overwrite) the drv_flds_in namelist even if empty. Driver now reads clm's version and merges it with CAM's version and terminates on a contradiction. M components/clm/cime_config/testdefs/testlist_clm.xml - Work on test lists. M components/clm/cime_config/user_nl_clm --------------- Remove note about drv_flds_in file only be written out if certain build-namelist options are on as no longer true. M components/clm/cime_config/buildnml ------------------ Turn megan off if NOT datm. Send LND_TUNING_MODE to build-namelist. Remove setting of bgc_spinup based on CLM_ACCELERATED_SPINUP (build-namelist itself figures this out). Leave drv_flds_in in clmconf directory and let the driver merge it's contents with CAM and create a version in RUNDIR. Copy CASEROOT/Buildconf/clmconf/user_nl_clm${inst_string} to clmconf (doesn't really need to be done). M components/clm/cime_config/config_component.xml ------ Add LND_TUNING_MODE (clm4_0_default,clm5_0_cam5.5,clm5_0_GSW3P,clm4_5_CRUNCEP) Right now default for CLM5.0 is clm5_0_cam5.5 for all cases. For clm4_0 or clm4_5 the one given default is used. =============================================================== =============================================================== Tag name: clm4_5_8_r180 Originator(s): sacks (Bill Sacks) Date: Mon Jun 6 06:13:57 MDT 2016 One-line Summary: Refactor dyn_cnbal_patch Purpose of changes ------------------ There was a lot of duplicated code, as well as almost-but-not-quite-duplicated code in dyn_cnbal_patch (the code used for updating patch-level BGC variables when patch weights change). This tag consolidates this duplicated code into some new shared infrastructure, similar to what is done for column-level variables. In addition, do not zero states when patch weights go to zero. (I expected this to change answers for 1-d history files, but it didn't.) Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2317: incorrect handling of isotopes with transient PFTs starting in clm4_5_1_r097 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: small timing increase: about 10% increase in dyn_cnbal_patch, amounting to a 0.3% increase in total CLM runtime (for both transient and non-transient runs). Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] NOTE: testing done on refactor_seed_calculation_n01_maintain_state_zero_weight_n01_dynlu_conserve_cn3_n03_clm4_5_8_r179; very minor changes since then tested with SMS_Ld5_D_P24x1.f10_f10.IRCP45CLM45BGC.hobart_nag.clm-decStart build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass, expected baseline failures as noted below CLM tag used for the baseline comparisons: clm4_5_8_r179 CLM40 compared with clm4_5_8_r178 (baselines missing for r179) Answer changes -------------- Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 / CLM50 transient and CNDV runs - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Roundoff-level changes Greater than roundoff-level for transient runs with isotopes, due to fixing bug 2317 If bitwise differences were observed, how did you show they were no worse than roundoff? Created a branch (dynlu_conserve_cn3_oneoff) that was as similar as possible to the trunk, with just those changes needed to get bit-for-bit behavior with my new code (e.g., reordering some calculations). Confirmed that my new code is bit-for-bit with that. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= New, shared infrastructure for updating patch-level variables A components/clm/src/dyn_subgrid/dynPatchStateUpdaterMod.F90 A components/clm/src/dyn_subgrid/test/dynPatchStateUpdater_test/CMakeLists.txt A components/clm/src/dyn_subgrid/test/dynPatchStateUpdater_test/test_patch_state_updater.pf A components/clm/src/dyn_subgrid/test/dynPatchStateUpdater_test ========= Extract code common to carbon and nitrogen, for computing seed amounts A components/clm/src/biogeochem/CNVegComputeSeedMod.F90 A components/clm/src/biogeochem/test/CNVegComputeSeed_test/CMakeLists.txt A components/clm/src/biogeochem/test/CNVegComputeSeed_test/test_ComputeSeedAmounts.pf A components/clm/src/biogeochem/test/CNVegComputeSeed_test ========= Define constants for identifying whether we're operating on c12, c13, c14 or n (used in CNVegComputeSeedMod) A components/clm/src/biogeochem/CNSpeciesMod.F90 List all existing files that have been modified, and describe the changes: ========= Remove a bunch of code - moved to new infrastructure M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 ========= Make calls to new patch state updater for each necessary variable M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 ========= Set up patch state updater M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Other minor changes needed to use the new infrastructure M components/clm/src/biogeochem/CNVegetationFacade.F90 ========= Add some unit tests, and some related unit testing infrastructure M components/clm/src/biogeochem/test/CMakeLists.txt M components/clm/src/main/pftconMod.F90 M components/clm/src/biogeochem/CMakeLists.txt M components/clm/src/dyn_subgrid/CMakeLists.txt M components/clm/src/dyn_subgrid/test/CMakeLists.txt M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 ========= Remove a bit of unused code M components/clm/src/dyn_subgrid/dynColumnStateUpdaterMod.F90 ========= Remove a test that has been passing for a while, and one that no longer exists M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_8_r179 Originator(s): sacks (Bill Sacks) Date: Fri May 27 10:39:15 MDT 2016 One-line Summary: Update column-level BGC state variables with dynamic landunits Purpose of changes ------------------ Update column-level BGC state variables as column areas change due to dynamic landunits. This does NOT yet update column-level biogeophysical variables, NOR does it handle patch-level variables appropriately. Note that some variables that used to only exist on vegetated landunits (soil & crop landunits) now exist on all landunits. When columns in special landunits expand, they take on the carbon, nitrogen and methane state variables from the columns they expanded into. These C / N / CH4 quantities essentially get trapped underneath the glacier (or lake / urban / whatever), staying fixed until that column eventually shrinks. If that special column shrinks and is replaced by a vegetated column, then the vegetated column takes on the state that had been trapped under the glacier (or whatever). In this way, we avoid destroying or creating mass when special landunits grow and shrink. This change explains why summary variables that used to be computed for just soil & crop columns are now computed for all columns. Note: In some testing (particularly in the 6-year Smallville test case), I found that there was a huge decrease (order of magnitude or more) in methane concentrations in the unsaturated portion of the growing column after column area adjustments. As far as I can tell, the state variable is being updated correctly due to the changing column areas, but then within the timestep evolves rapidly to a different value. From talking with Dave Lawrence, we assume this is due to the adjustment of carbon content in the column, since the carbon state can have a big effect on methane fluxes. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): - 2283: dynamic landunits: handle methane with changing lake area [enhancement] Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): When using use_init_interp in cases with CISM (IG/JG/FG/BG) compsets, there is a large adjustment in Greenland glacier area in the first timestep of the run, even if glacier evolution is off. (This is because, in initialization, glacier area comes from CLM's surface dataset; then glacier area is updated to match CISM in the first timestep, even if glacier evolution is off.) This results in big adjustments in column-level carbon, nitrogen and methane content in the first timestep, because it looks to CLM like some glacier columns have grown significantly (trapping C / N / CH4), while others have shrunk significantly, causing a significant decrease in area-average C / N / CH4 in the growing vegetated columns. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests all pass or xfail; expected baseline comparison failures as noted below CLM tag used for the baseline comparisons: clm4_5_8_r178 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with CN/BGC, either with CISM (IG) or transient crops (historical / RCP compsets with CROP) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; potentially new climate The evolution of C / N / CH4 state variables is now completely different with transient glacier or crop areas, causing significant answer changes. These transient crop tests change answers: ERS_Ly5_P60x1.f10_f10.IHISTCLM45BGCCROP.yellowstone_intel.clm-cropMonthOutput SMS_D_Ly6.1x1_smallvilleIA.IHISTCLM45BGCCROP.yellowstone_pgi.clm-cropMonthOutput For glacier, typically changes will only be seen if ice evolution is turned on, but changes will be seen even with ice evolution off - or ice evolution on in a test too short to actually see any ice evolution - if use_init_interp is used, as noted in the "caveats" section above. This affects these tests: ERS_D_Ld3.f09_g16_gl5.IG1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KitchenSink SMS_Lm1.f09_g16_gl5.IG1850CRUCLM50BGC.yellowstone_intel.clm-clm50KitchenSink In addition, there are FILLDIFFs in a few methane variables. The new code seems correct, in that the fill pattern (over special landunits) is the same for all vertical levels: the old code had different fill patterns in lower soil layers. This affects these variables: CONC_O2_SAT, CONC_O2_UNSAT, CONC_CH4_SAT, CONC_CH4_UNSAT, LAYER_SAT_LAG. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): cism2_1_10 -> cism2_1_17: Update to latest CISM trunk tag (main purpose is to bring in new capabilities for overriding glacier areas for testing purposes) List all files eliminated: none List all files added and what they do: ========= Add a test that has both increases and decreases in glacier together with BGC A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/user_nl_cism A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_inc_dec_bgc List all existing files that have been modified, and describe the changes: ========= Add capabilities to the column state updater: fractional areas (needed for methane) and keeping track of the adjustments made to each variable M components/clm/src/dyn_subgrid/dynColumnStateUpdaterMod.F90 M components/clm/src/dyn_subgrid/test/dynColumnStateUpdater_test/test_column_state_updater.pf ========= Update BGC column-level state variables when column areas change M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNDriverMod.F90 M components/clm/src/biogeochem/CNBalanceCheckMod.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/main/clm_driver.F90 ========= Update ch4 column-level state variables when column areas change. Also did a bunch of refactoring that was either needed or nice to have to support these changes, especially related to (1) determining whether it's the first timestep where ch4 is included; (2) how variables that keep track of states in the previous timestep are set (this needed to be reworked to simplify what we need to adjust in the column conservation); (3) initializing conserved state variables to 0 (rather than spval) over special landunits (this required putting in place backwards compatibility with old restart files in subroutine Restart); (4) where start-of-timestep balance check initialization is done M components/clm/src/biogeochem/ch4Mod.F90 ========= Add new routine to support restart file backwards compatibility: set_missing_vals_to_constant M components/clm/src/utils/restUtilMod.F90.in ========= Add a new history averaging option: SUM. Also refactored some handling of avgflag to remove duplication, and allowed l2g_scale_type in more contexts M components/clm/src/main/histFileMod.F90 ========= Add a new l2g_scale_type: veg_plus_lake (needed for some methane variables) M components/clm/src/main/subgridAveMod.F90 ========= Add a new fiter: allc - all (active) columns M components/clm/src/main/filterMod.F90 ========= Add col_filter_from_lunflags, change col_filter_from_ltype to col_filter_from_ltypes (allowing multiple landunit types) M components/clm/src/main/filterColMod.F90 M components/clm/src/main/test/filter_test/test_filter_col.pf M components/clm/src/main/test/topo_test/test_topo.pf ========= Updates for modified interface in new CISM M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease/user_nl_cism M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase/user_nl_cism ========= Add a test that has both increases and decreases in glacier together with BGC; remove a long test for which we have sufficient coverage with a similar, shorter test M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_8_r178 Originator(s): sacks (Bill Sacks) Date: Sun Apr 17 19:28:55 MDT 2016 One-line Summary: Remove some consistency checks, and merge crop_prog with use_crop in code Purpose of changes ------------------ (1) Remove some consistency checks that were doing more harm than good: (a) Make init_interp_fill_missing_with_natveg unnecessary for natural veg and crop output points. This is mainly to support the common use case of interpolating from non-crop to crop, without requiring the user to set this flag. (b) Remove check_finidat_fsurdat_consistency - now never do this check. This consistency check was causing more problems than it solved, both for users and developers (e.g., requiring us to change initial conditions files whenever we made new versions of the surface dataset). (2) Remove crop_prog logical variable from the code. This was identical to use_crop, and so it was confusing to have these two separate variables. Uses of crop_prog have been replaced by uses of use_crop. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - No longer need to set init_interp_fill_missing_with_natveg when interpolating from non-crop to crop Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - pass (ran bld/unit_testers tests) unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_8_r177 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Make init_interp_fill_missing_with_natveg unnecessary for natural veg and crop output points D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/README D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Make init_interp_fill_missing_with_natveg unnecessary for natural veg and crop output points M components/clm/src/init_interp/initInterpMindist.F90 M components/clm/src/init_interp/test/initInterpMindist_test/test_init_interp_mindist.pf M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm M components/clm/cime_config/testdefs/testlist_clm.xml ========= Remove check_finidat_fsurdat_consistency - now never do this check M components/clm/src/main/restFileMod.F90 M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml ========= Remove crop_prog - use use_crop in its place M components/clm/src/ED/main/EDCLMLinkMod.F90 M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/biogeophys/TemperatureType.F90 M components/clm/src/biogeophys/WaterStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNStateUpdate1Mod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 M components/clm/src/unit_test_stubs/main/histFileMod_stub.F90 M components/clm/src/biogeochem/CNFUNMod.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/CNPrecisionControlMod.F90 M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CropType.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/CNDriverMod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/biogeochem/CNNStateUpdate1Mod.F90 M components/clm/src/biogeochem/CNBalanceCheckMod.F90 M components/clm/src/biogeochem/CNVegStateType.F90 M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/initGridCellsMod.F90 M components/clm/src/main/histFileMod.F90 M components/clm/src/main/clm_varpar.F90 M components/clm/src/main/pftconMod.F90 M components/clm/src/main/surfrdMod.F90 M components/clm/src/main/clm_instMod.F90 =============================================================== =============================================================== Tag name: clm4_5_8_r177 Originator(s): sacks (Bill Sacks) Date: Thu Apr 14 14:00:25 MDT 2016 One-line Summary: Move CN product pools to gridcell level Purpose of changes ------------------ Move CN product pools from column-level to gridcell-level. This is needed for C & N conservation with dynamic landunits, because there was no obvious way to handle column-level product pools. This also required moving some of the related fluxes to the gridcell-level. Also added capability for init_interp to handle gridcell-level fields. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - If you run an old restart file through init_interp, product pools will be zeroed out. This is an issue for restart files that come from a transient run or from a clm5 crop run. Solution: run your old file through the new code for at least one time step without init_interp, then you can init_interp that file. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): - clmi.I2000CLM45CRUBGC.2000-01-01.0.9x1.25_gx1v6_simyr2000_c160127.nc replaced by clmi.I2000CLM45CRUBGC.2000-01-01.0.9x1.25_gx1v6_simyr2000_c160413.nc. This is basically the same, but run for one extra time step using compset ICLM45BGC with this branch. Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass or xfail, some baseline comparisons fail, as noted below CLM tag used for the baseline comparisons: clm4_5_8_r176 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all CLM45/CLM5 with CN/BGC - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): All CLM45/CLM5 CN/BGC cases have roundoff-level changes in product pools and related fluxes. In addition: (1) Bigger answer changes for CN/BGC cases using the f09 year-2000 initial conditions file: this file has been run through the new code for one additional time step. Differences were larger than I expected for this seemingly-minor change; I did not spend time tracking down the cause for those differences. (2) Cases doing interpolation (init_interp) from some other out-of-the-box initial conditions files have normalized RMS differences greater than roundoff. This is because some other initial conditions files (including some 1850 files) have non-zero product pools, which are zeroed out by init_interp: see the caveat for users above. However, in an absolute sense, RMS differences are small, because the product pools on these files are very small. (3) Greater than roundoff-level changes in product pools and related fluxes in transient crop cases. This shows up in the test ERS_Ly5_P60x1.f10_f10.IHISTCLM45BGCCROP.yellowstone_intel.clm-cropMonthOutput. This is expected: in the old code, the product pools grew or shrank as the columns they were on grew or shrank. This was essentially the motivation for this tag. If bitwise differences were observed, how did you show they were no worse than roundoff? Based on RMS errors in baseline comparisons If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Move product pools from column to gridcell-level, as described above M components/clm/src/biogeochem/CNProductsMod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNDriverMod.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/lnd2atmMod.F90 ========= Allow suffix on restart field name M components/clm/src/biogeochem/SpeciesBaseType.F90 M components/clm/src/biogeochem/SpeciesIsotopeType.F90 M components/clm/src/biogeochem/SpeciesNonIsotopeType.F90 M components/clm/src/biogeochem/test/Species_test/test_SpeciesNonIsotope.pf M components/clm/src/biogeochem/test/Species_test/test_SpeciesIsotope.pf ========= Add a routine to assist with backwards compatibility: set_grc_field_from_col_field: Set a gridcell-level field from a column-level field on the restart file M components/clm/src/utils/restUtilMod.F90.in ========= Add capability for init_interp to handle gridcell-level fields. This is needed for the new gridcell-level product pools. M components/clm/src/init_interp/test/initInterpMindist_test/test_init_interp_mindist.pf M components/clm/src/init_interp/initInterp.F90 M components/clm/src/init_interp/initInterpMindist.F90 M components/clm/src/init_interp/initInterpBounds.F90 ========= Point to new year-2000 initial conditions file. This is basically the same as the old one, but has been run for one timestep using the new code, in order to have gridcell-level product pools. This is needed for LII testing. M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Replace a year-1850 LII test with a year-2000 LII test to exercise the gridcell-level init_interp code (need non-zero product pools to test that code fully) M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_8_r176 Originator(s): sacks (Bill Sacks) Date: Wed Apr 13 09:32:56 MDT 2016 One-line Summary: Calculate active flags and filters earlier in initialization Purpose of changes ------------------ Main purpose is to compute the active flags and filters (via reweight_wrapup) earlier in initialization. These are now computed initially before the restart file is read, and tehn updated as soon as possible in the midst of the restart file read. This is useful in case any of the clm_instMod restart routines want to use the active flags or filters, directly or indirectly (e.g., via a call to subgridAveMod) - e.g., for the sake of backwards compatibility. This is something that's needed in an upcoming tag. Also: change some CN product history fields, so that they are more consistent - so totals just give totals for wood: - Renamed TOTPROD to TOT_WOODPROD - Removed PRODUCT_CLOSS and similar fields, added TOT_WOODPRODC_LOSS and similar fields: now just include losses from wood product pools Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - renamed CN product history fields: see above Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass or xfail, plus expected failures in CLM5 ERP tests with threading, just in the NPP_BURNEDOFF field, due to bug 2292 (leading to differences both in the baseline comparison and in the comparison of the restart to the baseline) CLM tag used for the baseline comparisons: clm4_5_8_r175 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Compute active flags and filters earlier in initialization M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/restFileMod.F90 ========= Change CN product history fields (see above for details) M components/clm/src/biogeochem/CNProductsMod.F90 ========= Unrelated change: fix argument intent M components/clm/src/biogeochem/CNPhenologyMod.F90 =============================================================== =============================================================== Tag name: clm4_5_8_r175 Originator(s): sacks (Bill Sacks) Date: Sat Apr 9 15:48:47 MDT 2016 One-line Summary: For glc_mec, change downscaling and where SMB is computed Purpose of changes ------------------ Two main purposes: (1) Change some aspects of the downscaling of atmospheric fields that is done when running with glc_mec. These change answers for glc_mec cases (i.e., cases with CISM): (a) Change the downscaling equations for forc_th (potential temperature) and forc_rho (atmospheric density): Previously, these gave different answers from the gridcell value even if the column's topographic height exactly matched the atmosphere's topographic height. The new equations are based on deltas, ensuring that the downscaled values match the atmospheric forcings if topographic heights match. Now all downscaled fields should match the atmosphere forcings if the topographic heights match, *except* rain and snow (potentially large differences, by design), and longwave radiation (roundoff-level differences can be introduced by the normalization). (b) Perform downscaling over all column types inside the icemask. Previously, downscaling was done only for istice_mec and istsoil columns (because those were the only ones that computed SMB). I realized that this could result in some undesirable inconsistencies between the different columns in a gridcell, which could be especially problematic if a glacier advanced over (e.g.) a lake column. In practice, this seems to affect a few lake points inside Greenland. (c) Do *not* perform downscaling over glacier columns with the single_at_atm_topo behavior. This is for consistency with other column types whose topographic height (implicitly) matches the atmosphere's. In practice, this shouldn't effect the simulation much, if at all (due to the change in (a)). (d) Extracted a function to compute forc_rho, and pulled out some magic numbers from this calculation (2) When running with glc_mec, compute SMB for all vegetated columns in the 'virtual' glacier region, rather than just in the icemask. This affects the CLM fields sent to the coupler for elevation class 0, but otherwise has no direct effect on CLM evolution. This is important if you want coupler forcings that can drive a later TG run: Typically you will want these forcings in the same place where you wanted virtual columns (for the same reason: so that forcings are available in every elevation class, including elevation class 0). In order to accomplish these changes, I have separated the logic for where to compute SMB from the logic for where to do downscaling. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: Bill Lipscomb reviewed the new equations for forc_th and forc_rho. Other changes have not been reviewed. CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass or xfail, except for expected baseline comparison failures, as noted below. Also: There were failures in CLM5 ERP tests with threading, just in the NPP_BURNEDOFF field, due to bug 2292, which was introduced in the previous tag. I confirmed that the same tests failed in a fresh checkout of r174. CLM tag used for the baseline comparisons: clm4_5_8_r174 Answer changes -------------- Changes answers relative to baseline: YES, for CLM45/CLM5 with CISM (IG compsets) If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM5 with CISM (IG compsets) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated closely, but expected to be larger than roundoff/same climate Differences are due to: (1) New downscaling equations for forc_th and forc_rho. I have confirmed that these differences are relatively small in one test run (one-day average differences < 0.31 deg for forc_th and < 1e-10 for forc_rho). (2) Downscaling now done over all column types within the icemask (including lakes) (3) Small change in constants used in calculation of forc_rho downscaling, to be consistent with shr_const_mod (4) Now compute SMB for all vegetated columns in the 'virtual' glacier region, rather than just columns in the icemask. This affects fields sent from CLM to the coupler in elevation class 0. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Changed to a .in file processed by genf90 D components/clm/src/unit_test_shr/unittestArrayMod.F90 List all files added and what they do: ========= New module to handle coordination of updating topographic heights, as well as determining which columns need downscaling A components/clm/src/main/TopoMod.F90 ========= New unit tests A components/clm/src/main/test/atm2lnd_test/test_downscale_forcings.pf A components/clm/src/main/test/topo_test/test_topo.pf A components/clm/src/main/test/topo_test/CMakeLists.txt A components/clm/src/main/test/topo_test ========= Changed to a .in file processed by genf90 A components/clm/src/unit_test_shr/unittestArrayMod.F90.in ========= Routines that assist unit tests working with glc_mec landunits A components/clm/src/unit_test_shr/unittestGlcMec.F90 List all existing files that have been modified, and describe the changes: ========= Main changes, as described above. Some of these changes have to do with the creation of TopoMod: topo_type, and the move of glc_topo into there - now called topo_col M components/clm/src/main/restFileMod.F90 M components/clm/src/main/clm_varcon.F90 M components/clm/src/main/filterMod.F90 M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/main/subgridRestMod.F90 M components/clm/src/main/ColumnType.F90 M components/clm/src/main/clm_instMod.F90 M components/clm/src/main/reweightMod.F90 M components/clm/src/main/glc2lndMod.F90 M components/clm/src/main/lnd2glcMod.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= New routines for creating filters - some for production code, some for unit tests M components/clm/src/main/filterColMod.F90 ========= Add some unit tests and some unit test helper functionality M components/clm/src/main/CMakeLists.txt M components/clm/src/main/test/filter_test/test_filter_col.pf M components/clm/src/main/test/glcBehavior_test/test_glcBehavior.pf M components/clm/src/main/test/CMakeLists.txt M components/clm/src/main/test/atm2lnd_test/CMakeLists.txt M components/clm/src/unit_test_shr/unittestFilterBuilderMod.F90 M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 M components/clm/src/unit_test_shr/unittestSubgridMod.F90 M components/clm/src/unit_test_shr/CMakeLists.txt M components/clm/src/unit_test_shr/test/unittestArray_test/test_unittestArray.pf ========= Add a bit of documentation M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml =============================================================== =============================================================== Tag name: clm4_5_8_r174 Originator(s): erik (Erik Kluzek) Date: Fri Mar 25 01:57:24 MDT 2016 One-line Summary: Bring Rosie's respiration and FUN changes to the trunk Purpose of changes ------------------ Changes to respiration from Rosie Fisher. Leaf respiration option (CLM50 default is Atkin 2015) as well as option for light inhibition. Also reworking of FUN so it calculates both fixers and non fixers and then figures out the impact of both based on a fixed percentage of plants that fix by PFT. Add burned-off carbon to FUN (Carbon that can not be used for N uptake). Add new term to directly account for Ben Houlton's temperature response function. New defaults for CLM5.0. Add some new history fields including LEAFCN and LEAFCN_STORAGE CN ratios. Default for snow depth is now lowered to 10m from 20m. If you startup from an initial condition file with 20m of snow, a big flow of ice-runoff will flow out of the system (takes a few days to clear it all out). In general you should NOT allow this for fully coupled simulations with an active ocean model. For other simulations, you should be aware of this sudden ice runoff that will happen. Remove some unused unworking or unrefined options for flexible-CN. Check for small C or N values in CNPrecisionControl after each StateUpdate section rather than just after summary. This is turned on for CLM50 with use_nguardrail (will turn this on for CLM45 as well, and remove later). Improve some of the documentation of namelist items. Also work on some of the dependencies: use_flexiblcCN requires CN, carbon_resp_opt=1 only if CN on and FUN off. use_luna on only if CN and nitrif_denitrif on. lnc_opt depends on if CN on or off. FUN on depends on both CN and nitrig_dentrif FUN is very expensive in this version and causes model to be significantly slower. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2302 Some missing nh4 arrays set to zero. 2298 Init method NOT called for NutrientCompetetionFlexibleCN 2296 Missing meta-data for new data added to params file 2273 Log output without a "if masterproc " around it... 2268 Irrigation unit tests broken because of irrigation limiter 2253 Nitrogen Deposition accidentally turned off when FUN on 2249 Low LAI over Amazon Known bugs introduced in this tag (include bugzilla ID): 2295 ch4 conservation error in ne30 1850 case 2269 log(0) possibility Notes of particular relevance for users --------------------------------------- Caveats for users: This changes the default for snow capping from 20m to 10m. If you use an initial condition file with 20m of snow it will cut the snow depth to 10m, and send 10m of snow as ice runoff. This sudden influx of ice runoff could be problematic for fully coupled simulations to an ocean model. Changes to CLM's user interface: New namelist items: leafresp_method, light_inhibit, leaf_acc, rootstem_acc, use_nguardrail Remove namelist : carbon_excess_opt, carbon_storage_excess_opt, dynamic_plant_alloc_opt Remove vcmax_opt options 1 and 2 new namelist group: photosyns_inparm carbon_resp_opt should be 0 when use_fun=.true. New history fields: Vcmx25Z, Jcmx25Z, LEAFCN, LEAFCN_STORAGE COST_NFIX, COST_NACTIVE, COST_NRETRANS, NUPTAKE_NPP_FRACTION add IGM1850CRUCLM50BGCCROP and IGM1850CRUCLM50BGC compsets Changes made to namelist defaults: New defaults: CLM50: leaf_acc, light_inhibit, use_bedrock, use_nguardrail all on leaf_resp=2 h2osno_max=10m use_flexibleCN (if CN on) carbon_resp_opt=1 (only if use_cn ON and FUN off) use_luna (only if use_cn and use_nitrif_denitrif on) lnc_opt depends on CN FUN on depends on CN and nitrif_denitrif Change leaf_mr_vcm phys default value Changes to the datasets: New params file for clm5_0 with metadata on new fields. Remove some uneeded parameters. Substantial timing or memory changes: Approximately 20% slower with FUN on Code reviews and testing ------------------------ Code reviewed by: self,rfisher CLM testing: regular [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - PASS CLM tag used for the baseline comparisons: clm4_5_8_r173 Answer changes Yes! for CLM50, clm45 and clm40 bit-for-bit -------------- Changes answers relative to baseline: Summarize any changes to answers, i.e., - what code configurations: CLM50 - what platforms/compilers: All - nature of change: new climate - casename: /home/oleson/respn14r172_1850spin Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: None List all files added and what they do: A components/clm/doc/clm3_0_ChangeLog -- split out CLM3.0 section A components/clm/doc/clm4_0_ChangeLog -- split out CLM4.0 section List all existing files that have been modified, and describe the changes: M components/clm/cime_config/config_component.xml formatting change, removing extra spaces M components/clm/cime_config/config_compsets.xml add IGM1850CRUCLM50BGCCROP and IGM1850CRUCLM50BGC compsets M components/clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 Changes from Ben Andre to fix irrigation unit test issue bug 2268 M components/clm/src/biogeophys/test/Irrigation_test/test_irrigation_singlepatch.pf Add new test irrigation_for_limiting_volr M components/clm/src/biogeophys/SoilWaterMovementMod.F90 Remove warning for h2osoi_liq negative. Limit dLow, diag, dUpp, and rhs arrays by nlayers. M components/clm/src/biogeophys/IrrigationMod.F90 Make sure a constant is double (_r8) M components/clm/src/biogeophys/UrbanParamsType.F90 Add if_masterproc MM components/clm/src/biogeophys/SnowHydrologyMod.F90 Add min_snow_to_keep, fix bug 2270 M components/clm/src/biogeophys/LunaMod.F90 Work on warnings and error conditions. M components/clm/src/biogeophys/PhotosynthesisMod.F90 Make Vcmx25Z, Jcmx25Z default inactive on history. Add ReadNML method add leafresp_method, light_inhibit, leaf_acc, rootstem_acc to namelist. Remove vcmax_opt=1/2 change to if,else if, else. M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 Add if_masterproc M components/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 Add timer around FUN. Add some new terms when FUN on. Use local version of use_fun (so optimizer can know it won't change). Write to iulog rather than *. M components/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 Comment out warning M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 Remove unneeded use_fun dependence M components/clm/src/biogeochem/CNGapMortalityMod.F90 Remove: carbon_excess_opt, carbon_storage_excess_opt M components/clm/src/biogeochem/CNGRespMod.F90 Remove: carbon_excess_opt and carbon_storage_excess_opt calcs M components/clm/src/biogeochem/CNBalanceCheckMod.F90 Lower carbon balance warning from e-10 to e-9, use iulog rather than * M components/clm/src/biogeochem/CNMRespMod.F90 Add rootstem_acc option which Rosie says "will increase R and decrease productivity in boreal forests, A LOT. :)" M components/clm/src/biogeochem/CNFUNMod.F90 Extensive changes from Rosie. Some formatting changes. Add burnedoff_carbon. and timer around main FUN calculations. Fix bug 2302 (missing setting of some nh4 arrays to zero). Base if fixer on c3psn. Add bit about unmet_demand. Loop over both fixers and non-fixers, params file has percentage of each for each PFT type. Set some things to spval for night. Add new term to directly account for Ben Houlton's temperature response function M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 Add if masterproc around writes M components/clm/src/biogeochem/NutrientCompetitionFactoryMod.F90 Pass in bounds and add call to init method M components/clm/src/biogeochem/CNPrecisionControlMod.F90 Make ccrit and ncrit parameters and add n_min (for calculation of CN ratio) M components/clm/src/biogeochem/NutrientCompetitionMethodMod.F90 Add init method to abstract class and add readParams as completed method M components/clm/src/biogeochem/CNPhenologyMod.F90 Add some use_fun changes, add logical for additional_onset_condition M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 Remove cpool imbalance warning M components/clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 Add init method, move readParams to base class. Check if aleaf is nan in calc_plant_nutrient_competition. Set frootn/livestemn "to_retransn" to zero for fun M components/clm/src/biogeochem/CNNDynamicsMod.F90 Remove setting of ndep_to_sminn(c) to zero for FUN, do set nfix_to_sminn to zero M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 Add new history fields: COST_NFIX, COST_NACTIVE, COST_NRETRANS, NUPTAKE_NPP_FRACTION all default on. M components/clm/src/biogeochem/CNDriverMod.F90 Add some more timers. Call CNPrecisionControl after each set of updates (if use_nguardrail is on). Move CNPhenology/CNFUNInit call to before nutrient_demand, so that croplive didn't change half way through crop N cycle M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 Add actual_storage_leafcn and actual_leafcn CN ratios and output to history and always calculate them. Remove readParams, and add init method. Remove unused dynamic_plant_alloc. Remove "if(use_fun)" and just rely on carbon_resp_opt to be 0 when FUN on. Use n_min as parameter for minimun N used for CN ratio. M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 Add NPP_BURNEDOFF to history file, change ordering M components/clm/src/biogeochem/CNVegCarbonStateType.F90 Add if(masterproc) around write statements M components/clm/src/main/clm_varpar.F90 Add if_masterproc MM components/clm/src/main/clm_initializeMod.F90 Pass bounds_proc to create_nutrient_competition_method M components/clm/src/main/findHistFields.pl Get working again with all F90 files. MM components/clm/src/main/controlMod.F90 Remove carbon_excess_opt, carbon_storage_excess_opt, dynamic_plant_alloc_opt namelist variables. Add use_nguardrail (to preserve answers for clm45 will remove later). M components/clm/src/main/clm_driver.F90 Add some timers. M components/clm/src/main/clm_varctl.F90 Remove carbon_excess_opt, carbon_storage_excess_opt, dynamic_plant_alloc_opt Add use_nguardrail. M components/clm/src/main/pftconMod.F90 Add some more FUN parameters and remove some uneeded ones. M components/clm/src/main/clm_instMod.F90 Add call to photsyns ReadNML M components/clm/src/main/ncdio_pio.F90.in Add some if_masterproc M components/clm/bld/unit_testers/build-namelist_test.pl Add some new tests for new variables M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml Add rootstem_acc, leaf_acc, light_inhibit, leafresp_method use_nguardrail. Improve documentation for some flex-CN options M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml New defaults: CLM50: leaf_acc=T, light_inhibit=T,leaf_resp=2, use_bedrock=T h2osno_max=10m, use_flexibleCN (if use_cn=T) carbon_resp_opt=1 (only if use_cn ON and FUN off) use_luna (only is use_cn and use_nitrif_denitrif on) lnc_opt depends on use_cn, use_fun depends on use_cn=T AND use_nitrif_dentrif=T Change leaf_mr_vcm phys default value use_nguardrail=T M components/clm/bld/CLMBuildNamelist.pm New photsyns_inparm namelist. carbon_resp_opt should NOT be 1 when FUN on. lnc_opt only matters when luna on or vcmax_opt=3/4. leafresp_method MUST be 0 if CN off. =============================================================== =============================================================== Tag name: clm4_5_8_r173 Originator(s): sacks (Bill Sacks) Date: Thu Mar 17 20:24:45 MDT 2016 One-line Summary: Major refactor of CN Products Purpose of changes ------------------ Variables and code related to CN product pools and fluxes were scattered among 12 modules. In addition, CNWoodProducts was difficult to understand and modify, because the same basic logic was repeated four times: once for carbon, once for c13, once for c14 and once for nitrogen. This tag does a major overhaul of CNWoodProducts - now renamed to CNProducts, since it includes grain as well as wood products. This is now an object-oriented class. The major benefits of this overhaul are: - Only 2 modules now contain any information about the number of product pools and their lifetimes: CNProducts and pftconMod. Previously, 12 modules had some information about the fact that there were 10 & 100-year product pools. This makes it easier to change the number or lifetimes of product pools in the future. Code outside CNProducts (and pftconMod) knows that there is some flux to wood and crop product pools in general, but doesn't need to know anything about how this is partitioned into the 10 and 100-year product pools - or even the fact that there are 10 and 100-year product pools. - Because this product pool information is now self-contained, with a clear interface, CNProductsMod could be used by ED, if this were desired. - Changes to CNProductsMod no longer need to be replicated four times for the four different species. Instead, there are four instances of cn_products_type declared at a higher level. This tag also introduces a "species" class (with subclasses for isotope and non-isotope species), which can be used to remove duplication in the setting of history and restart variables in classes for which we have one instance for each species. This is currently used in the new CNProductsMod; eventually it can also be used in CNVegCarbonStateType, CNVegCarbonFluxType, and the similar SoilBiogeochem Types (at which point this should probably be moved out of the biogeochem directory into somewhere like 'utils'). This tag also introduces a utility method for setting one restart field (not found on the restart file) from another (which was present). This is used in a few places in CNProductsMod, and should be used in other places as well (e.g., see bug 2282). Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2297 (code looks wrong for initializing carbon isotopes from a restart file without isotope information) Known bugs introduced in this tag (include bugzilla ID): - 2256 (Problems with early versions of PIO2 in CLM) -- shows up in more (hobart) tests now, for unclear reasons Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - ok (pass except irrigation expected failure) tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means tests pass or xfail, some baseline comparisons failed as noted below Two new expected failures appear to be PIO2 issues: RUN ERP_Ly5.1x1_numaIA.ICRUCLM50BGCCROP.hobart_nag.clm-monthly RUN ERP_Ld5_P24x1.f10_f10.I1850CLM45BGC.hobart_nag.clm-default CLM tag used for the baseline comparisons: clm4_5_8_r172 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 / CLM50 - anything other than 1850 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? This is complicated; ready? Let's work backwards: Branch tag product_pools_gridcell2_n03_clm4_5_8_r172 (which is becoming this trunk tag) was bit-for-bit with branch tag product_pool_total_fluxes_n01_clm4_5_8_r172 (from branch product_pool_total_fluxes), except for roundoff-level changes in c13 / c14 product pools / fluxes for transient cases with ciso. (Those changes remained roundoff-level throughout the simulation, so their roundoff-level nature was obvious from the cprnc output.) Determining that the changes in product_pool_total_fluxes_n01_clm4_5_8_r172 were no greater than roundoff-level was tricky, because these differences propagated over time. To do this, I created another branch: product_total_roundoff, with branch tag product_total_roundoff_n01_product_pool_total_fluxes_n01_clm4_5_8_r172. This was essentially the same as product_pool_total_fluxes_n01_clm4_5_8_r172, except that hrv_deadstemc_to_prod10c and hrv_deadstemc_to_prod100c (and same for nitrogen) were computed twice: once the old way and once the new way. The updates of deadstemc/n in CStateUpdate2h and NStateUpdate2h (which were the uses that caused differenes to propagate to greater than roundoff-level) were also done twice: once using the "old" hrv_deadstem values, and once with the "new" values. The "old" version was used in the rest of the code, but the two versions were compared for every point in every time step to ensure that they were never more than roundoff-level different. I ran the full test suite on product_total_roundoff_n01_product_pool_total_fluxes_n01_clm4_5_8_r172 and ensured that it differed by no more than roundoff-level from trunk tag clm4_5_8_r172. To summarize: I ensured that the only differences greater than roundoff-level in the test suite were attributable to differences in the order of operations when computing the hrv_deadstem*_to_prod* variables, and I also ensured that the new and old calculations of the hrv_deadstem*_to_prod* variables never differed by more than roundoff. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Renamed to CNProductsMod.F90 D components/clm/src/biogeochem/CNWoodProductsMod.F90 List all files added and what they do: ========= Main change - this used to be CNWoodProductsMod.F90, but has been completely rewritten, as described above A components/clm/src/biogeochem/CNProductsMod.F90 ========= Utility classes to hold metadata for classes that can have one instance for each species (carbon, nitrogen, c13, c14) - removing duplication in the calls to hist_addfld and restartvar A components/clm/src/biogeochem/SpeciesNonIsotopeType.F90 A components/clm/src/biogeochem/SpeciesBaseType.F90 A components/clm/src/biogeochem/SpeciesIsotopeType.F90 A components/clm/src/biogeochem/test/Species_test/test_SpeciesNonIsotope.pf A components/clm/src/biogeochem/test/Species_test/CMakeLists.txt A components/clm/src/biogeochem/test/Species_test/test_SpeciesIsotope.pf A components/clm/src/biogeochem/test/Species_test A components/clm/src/biogeochem/test/CMakeLists.txt A components/clm/src/biogeochem/test ========= Add a short ciso transient test (eventually can remove the long ciso transient test and just keep this one) A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_decStart/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_decStart List all existing files that have been modified, and describe the changes: ========= Main changes in this tag: remove information about the different product pools that are scattered throughout the code, and add some objects of cn_products_type M components/clm/src/biogeochem/CNCIsoFluxMod.F90 M components/clm/src/biogeochem/dynHarvestMod.F90 M components/clm/src/biogeochem/CNCStateUpdate2Mod.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/biogeochem/dynConsBiogeochemMod.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/CNDriverMod.F90 M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNNStateUpdate2Mod.F90 M components/clm/src/biogeochem/CNBalanceCheckMod.F90 ========= Add utility method for setting an absent restart field from one that was present on the initial conditions file M components/clm/src/utils/restUtilMod.F90.in ========= Move duplicate code into clm_varcon, and fix bug 2297 M components/clm/src/main/clm_varcon.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 ========= Add a short ciso transient test (eventually can remove the long ciso transient test and just keep this one) M components/clm/cime_config/testdefs/testlist_clm.xml ========= Add unit tests of new Species classes M components/clm/src/CMakeLists.txt M components/clm/src/biogeochem/CMakeLists.txt ========= Two new expected failures on hobart; appear to be PIO2-related M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_8_r172 Originator(s): sacks (Bill Sacks) Date: Tue Mar 8 13:10:19 MST 2016 One-line Summary: Exclude land use and product pools / fluxes from CN summary diagnostics Purpose of changes ------------------ Changed TOTCOLC, TOTECOSYSC, TOTCOLN and TOTECOSYSN (and same for C isotopes) to no longer include product pools. Also changed NBP and NEE to no longer include land use or product fluxes (dwt_closs and product_closs). Removed LAND_UPTAKE diagnostic, because NEE now gives this quantity. Added NET_CARBON_EXCHANGE, which gives the sum of NEE and the land use flux - i.e., what NEE used to be. Also, changed TOTPRODN to be consistent with TOTPRODC: no longer includes crop product pool. The main motivation for this change is the upcoming move of product pools to the grid cell level. However, this also addresses inconsistencies between CLM's definition of NEE, etc., and what would be more typically considered in the literature. These changes to the meaning of these diagnostics were all proposed and/or approved by Peter Lawrence. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Changes meaning of a number of diagnostics, as noted above Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - ok All pass except expected failure in irrigation tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok All pass or xfail; baseline failures as noted below CLM tag used for the baseline comparisons: clm4_5_8_r171, but with mods to produce FCO2 history field (in order to ensure that this tag hasn't changed the co2 flux sent to the atmosphere). Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: many CLM45 / CLM50 - see below for details - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff for prognostic fields, larger than roundoff for some diagnostic fields Significant changes in the diagnostic fields noted under "purpose of changes", above, for transient, CNDV and crop runs, and for year-2000 runs that used initial conditions from a transient run (so presumably had non-zero product pools). Changes in these diagnostic fields (if any) are roundoff-level for non-crop, non-CNDV, non-crop 1850 runs. Roundoff-level changes in CO2 flux sent to atmosphere and new FCO2 diagnostic field for some runs. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes as described above. Some files just changed because some uses of nee have been changed to net_carbon_exchange. M components/clm/src/cpl/lnd_import_export.F90 M components/clm/src/biogeochem/CNVegetationFacade.F90 M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/lnd2atmMod.F90 ========= Since total column C / N no longer includes the product pools, the CN balance checks needed to be changed so that outputs from the system include fluxes into the product pools, but NOT fluxes out of the product pools (the product pools are outside of the system as far as the balance checks are concerned) M components/clm/src/biogeochem/CNBalanceCheckMod.F90 ========= Fluxes into the product pools now need to be saved for the sake of the balance checks M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNWoodProductsMod.F90 M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 ========= Output FCO2 (co2 flux to atmosphere), since otherwise this isn't covered by any of our tests - since our tests don't turn on this coupling field M components/clm/src/main/lnd2atmType.F90 M components/clm/cime_config/testdefs/testmods_dirs/clm/default/user_nl_clm ========= Add expected baseline failure (bug 2291) M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_8_r171 Originator(s): erik (Erik Kluzek) Date: Wed Mar 2 23:53:40 MST 2016 One-line Summary: Backout COMP_RUN_BARRIERS setting, and remove all ED tests from testlist (as they all fail) Purpose of changes ------------------ Backout the setting of COMP_RUN_BARRIERS. It has a nasty side effect of disabling the automatic setting of CONTINUE_RUN=TRUE when RESUBMIT>1. See bug 2280. Now production simulations with I compsets will fail because datm is allowed to get ahead of CLM. Also remove all of the ED tests from the testlists as they currently all fail. Known bugs introduced in this tag (include bugzilla ID): bug 2280 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: Code reviews and testing ------------------------ Code reviewed by: self CLM testing: short [PASS means all tests PASS and OK means tests PASS other than expected fails.] regular tests: yellowstone_intel - clm_short_45 -- PASS CLM tag used for the baseline comparisons: clm4_5_8_r170 Answer changes -------------- Changes answers relative to baseline: No, bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/cime_config/config_component.xml ------- Remove setting of COMP_RUN_BARRIERS M components/clm/cime_config/testdefs/testlist_clm.xml -- remove all ED tests =============================================================== =============================================================== Tag name: clm4_5_8_r170 Originator(s): sacks (Bill Sacks) Date: Sun Feb 28 20:33:37 MST 2016 One-line Summary: Add a wrapper for CN Vegetation Purpose of changes ------------------ This tag introduces a wrapper (or "facade" in software design terms) for the CN vegetation types and modules. My main motivations were: (1) Insulating the rest of the code from changes in the CNVeg implementation. For example, I'm about to introduce a class to hold variables related to the product pools. With the old implementation, I would have to touch a number of modules outside of the biogeochem directory in order to do this; now I won't. (2) Simplifying interfaces to subroutines – reducing the number of arguments. (3) Starting to put in place a structure that could be leveraged to support CN and ED (and SP) side-by-side, polymorphically, so that we don't need 'if use_cn' / 'if use_ed' conditionals scattered throughout the code or direct "use" statements of types in CN / ED / etc. The changes here are a big first step towards these goals, although this isn't yet complete. There are some notes in comments in CNVegetationFacade.F90 listing some things that should still be done. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2276: Need ne30_g16 surface dataset with crop for CLM45/CLM50 - 2291: fire emissions NaN with gnu compiler Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: a bit of review by Erik Kluzek and Ben Andre CLM testing: [PASS means all tests PASS and OK means tests PASS other than expected fails.] build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - ok ok means: all tests pass except expected irrigation failure tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok There was one unexpected baseline failure: SMS_Lm25.f19_g16.ICLM45BGCCROP.yellowstone_gnu.clm-cropMonthOutput Upon investigation, it appeared that this was a problem in the previous tag (and possibly earlier): fire emissions differed between the trunk and this branch, and were often NaN in the trunk version (but no longer NaN on this branch). This seems most likely to be a compiler-related problem, since it seems to just appear with gnu. See bug 2291. CLM tag used for the baseline comparisons: clm4_5_8_r169 Answer changes -------------- Changes answers relative to baseline: NO (However, see above notes about the baseline failure due to bug 2291.) Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Moved some modules from dyn_subgrid into biogeochem (no changes to contents) D components/clm/src/dyn_subgrid/dynCNDVMod.F90 D components/clm/src/dyn_subgrid/dynHarvestMod.F90 D components/clm/src/dyn_subgrid/dynConsBiogeochemMod.F90 List all files added and what they do: ========= New module, providing a wrapper / facade for the CNVeg stuff; this is the crux of the change A components/clm/src/biogeochem/CNVegetationFacade.F90 ========= Moved some modules from dyn_subgrid into biogeochem (no changes to contents) A components/clm/src/biogeochem/dynCNDVMod.F90 A components/clm/src/biogeochem/dynHarvestMod.F90 A components/clm/src/biogeochem/dynConsBiogeochemMod.F90 List all existing files that have been modified, and describe the changes: ========= Moved stuff out of these places into the new CNVeg, and simplified some interfaces M components/clm/src/main/clm_driver.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/clm_instMod.F90 M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Rather than passing CNVeg derived types, instead directly pass the small number of arrays that are needed from CNVeg (removes dependence on the CNVeg types, which is needed if we ever want to make CNVeg polymorphic; also makes the connection between CNVeg and other parts of the model more explicit) M components/clm/src/biogeophys/CanopyFluxesMod.F90 M components/clm/src/biogeophys/PhotosynthesisMod.F90 M components/clm/src/main/lnd2atmMod.F90 M components/clm/src/biogeochem/ch4Mod.F90 ========= Added a bit of time manager functionality, and some unit tests M components/clm/src/utils/clm_time_manager.F90 M components/clm/src/unit_test_shr/unittestTimeManagerMod.F90 M components/clm/src/utils/test/clm_time_manager_test/test_clm_time_manager.pf ========= Moved some crop-specific variables out of CNVegStateType into CropType (but there are still more that need to be moved) M components/clm/src/biogeochem/CNVegStateType.F90 M components/clm/src/biogeochem/CropType.F90 M components/clm/src/biogeochem/CNPhenologyMod.F90 M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 M components/clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 M components/clm/src/biogeochem/NutrientCompetitionMethodMod.F90 M components/clm/src/biogeochem/CNNDynamicsMod.F90 M components/clm/src/biogeochem/CNVegStructUpdateMod.F90 M components/clm/src/biogeochem/CNRootDynMod.F90 M components/clm/src/biogeochem/CNDriverMod.F90 ========= Unrelated change: fix xml entries for ne30 and ne120 surface datasets (bug 2276) M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Restore expected failure that seems to have been deleted accidentally M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_8_r169 Originator(s): erik (Erik Kluzek) Date: Tue Feb 23 14:35:02 MST 2016 One-line Summary: Add fire several fire constants to namelist, add some new clm45/50 surface datasets Purpose of changes ------------------ Add a bunch of constants to the namelist for the CN Li fire parameterization. They are added in the new lifire_inparm namelist. Work on user interface a bit. * New lifire_inparm namelist (see above) * Add CLM_ACCELERATED_SPINNUP env_run variable * Turn COMP_RUN_BARRIERS for f09 and f19 resolutions * New compsets: ICLM50BGCDVCROP, IHISTCLM50BGC * Model dies if CLM namelist isn't found * Models dies if use_bedrock=.true. and zbedrock NOT on surface dataset * Remove use_nofire move to fire_method as "nofire" option Add some new clm45/50 surface datasets for crop at 1850 for f45, ne30, and ne120 Add back in the USUMB PTCLMmkdata dataset directory for testing. Work on testing a bit. Remove CLM50 testmods directories just there because of missing CLM50 compsets. Add set of complete clm40 testmod directories. Add nofire test, update USUMB test. Remove use_dynroot from all but Kitchen sink tests. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): [If none, remove this line] 2285 Problem creating ne120 surface dataset 2280 (partial) need to set barriers for concurrent PE layouts with I compsets 2279 Need new USUMB PTCLMmkdata directory 2276 New ne30, ne120, f45 surface datasets for CLM45/50 1850 CROP Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): Introduce new env_run variable: CLM_ACCELERATED_SPINUP (It can be "on" or "off") When "on" it sets the simulation up in such a way for it to run in an accelerated fashion (currently all it does is turn "-bgc_spinup on" if NOTE: Turns on COMP_RUN_BARRIERS by default for f09 and f19 resolutions as they are usually run concurrent (f09 is by default). Add new CLM namelist: lifire_inparm Add new compsets: ICLM50BGCDVCROP, IHISTCLM50BGC NOTE: Model will now die if you set use_bedrock=.true., and your surface dataset doesn't have zbedrock on it. WARNING: Now if a CLM namelist is NOT found -- the model will abort with an error. Changes made to namelist defaults (e.g., changed parameter values): Many new namelist items New CN fire namelist items: rh_low, rh_hgh, bt_min, bt_max, cli_scale, boreal_peatfire_c, pot_hmn_ign_counts_alpha, non_boreal_peatfire_c, cropfire_a1, and occur_hi_gdp_tree They are set by default according to either fire_method=li2014qianfrc or li2016crufrc use_nofire removed and "nofire" added as an option to fire_method Changes to the datasets (e.g., parameter, surface or initial files): A few new surface datasets New 1850 clm45/50 surface datasets for: f45, ne30, ne120 Alos new PTCLMmkdata USUMB directory Substantial timing or memory changes: None known Code reviews and testing ------------------------ Code reviewed by: self CLM testing: regular build-namelist tests: yellowstone - OK (36 comparison tests fail because of new namelist) unit-tests (components/clm/src): yellowstone - OK tools-tests (components/clm/test/tools): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK [PASS means all tests PASS and OK means tests PASS other than expected fails.] Two comparision tests (ro clm4_5_8_r168) fail, because use_dynroot no longer on. CLM tag used for the baseline comparisons: clm4_5_8_r168 Answer changes -------------- Changes answers relative to baseline: No bit-for-bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: Eliminate testmods directories just needed for CLM50 compsets Now that compsets are added they are unneeded. D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCCROPmonthly/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCCROPmonthly/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCDVCROPmonthly/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCDVCROPmonthly/shell_commands D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCmonthly/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCmonthly/shell_commands List all files added and what they do: A + components/clm/src/biogeochem/CNFireNoFireMod.F90 -- No fire option New testmods for nofire, clm40 tests, and USUMB PTCLMmkdata directory A + components/clm/cime_config/testdefs/testmods_dirs/clm/nofire A + components/clm/cime_config/testdefs/testmods_dirs/clm/40crop A + components/clm/cime_config/testdefs/testmods_dirs/clm/40pts A + components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLA A + components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsRLB A + components/clm/cime_config/testdefs/testmods_dirs/clm/40ptsROA A + components/clm/cime_config/testdefs/testmods_dirs/clm/40reduceOutput A + components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/shell_commands List all existing files that have been modified, and describe the changes: M components/clm/bld/CLMBuildNamelist.pm --- Add new fire constants and lifire_inparm namelist. Allow spinup state to be 1 or 2, and document better M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- When -bgc_spinup on, spinup_state=1 for clm45 and 2 for clm50. Add new fire constants set to values for either lifire2014qianfrc or lifire2016crufrc new f45, ne30, ne120 datasets (bug 2276) M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- Add nofire option to fire_method, remove use_nofire, and add new fire constants M components/clm/cime_config/buildnml --- Add CLM_ACCELERATED_SPINUP set "-bgc_spinup on" if CLM_ACCELERATED_SPINUP="on" M components/clm/cime_config/config_component.xml -- Set COMP_RUN_BARRIERS (bug 2280) to TRUE for f09 and f19 resolutions. Define CLM_ACCELERATED_SPINUP. M components/clm/cime_config/config_compsets.xml -- Add ICLM50BGCDVCROP and IHISTCLM50BGC compsets M components/clm/cime_config/testdefs/testlist_clm.xml --- move all clm40 tests to have 40 testmods directories, add f45 nofire test, remove clm50 testmods just use new CLM50 compsets M components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/user_nl_clm -- update (bug 2279) M components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/xmlchange_cmnds -- update M components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_flexCN_FUN/user_nl_clm -- remove use_dynroot M components/clm/cime_config/testdefs/testmods_dirs/clm/crop/include_user_mods -- point to default NOT 40default M components/clm/cime_config/testdefs/testmods_dirs/clm/flexCN_FUN/user_nl_clm -- remove use_dynroot M components/clm/cime_config/testdefs/testmods_dirs/clm/pts/include_user_mods -- point to default not 40default M components/clm/cime_config/testdefs/testmods_dirs/clm/reduceOutput/shell_commands -- remove -no-fire_emis M components/clm/src/biogeochem/CNFireBaseMod.F90 ----- add cnfire_const object and CNFireReadNML method, put li2016 version of FireFluxes in place (can be shared by li2014 and li2016). M components/clm/src/biogeochem/CNFireFactoryMod.F90 -- Abort if namelist not found, and add nofire case for fire_method. M components/clm/src/biogeochem/CNFireLi2014Mod.F90 --- Use constants from cnfire_const object in CNFireBaseMod, remove use_nofire option M components/clm/src/biogeochem/CNFireLi2016Mod.F90 --- Use constants from cnfire_const object in CNFireBaseMod, remove use_nofire option, remove CNFireFluxes M components/clm/src/biogeochem/CNFireMethodMod.F90 --- Add CNFireReadNML_interface M components/clm/src/biogeochem/CNMRespMod.F90 -------- abort if can't find namelist M components/clm/src/biogeochem/CNSharedParamsMod.F90 ----- abort if can't find namelist M components/clm/src/biogeochem/SatellitePhenologyMod.F90 - abort if can't find namelist M components/clm/src/biogeochem/VOCEmissionMod.F90 M components/clm/src/biogeochem/ch4varcon.F90 ---------- abort if can't find namelist M components/clm/src/biogeophys/CanopyHydrologyMod.F90 - abort if can't find namelist M components/clm/src/biogeophys/CanopyStateType.F90 ---- abort if can't find namelist M components/clm/src/biogeophys/HumanIndexMod.F90 ------ abort if can't find namelist M components/clm/src/biogeophys/RootBiophysMod.F90 ----- abort if can't find namelist M components/clm/src/biogeophys/SnowHydrologyMod.F90 --- abort if can't find namelist M components/clm/src/biogeophys/SoilHydrologyType.F90 -- abort if can't find namelist M components/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 - abort if can't find namelist M components/clm/src/biogeophys/SoilWaterMovementMod.F90 - abort if can't find namelist M components/clm/src/biogeophys/SurfaceResistanceMod.F90 - abort if can't find namelist M components/clm/src/biogeophys/UrbanParamsType.F90 ------ abort if can't find namelist M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 abort if can't find namelist M components/clm/src/dyn_subgrid/dynpftFileMod.F90 ------- abort if can't find namelist M components/clm/src/init_interp/initInterp.F90 ---------- abort if can't find namelist M components/clm/src/main/clm_initializeMod.F90 ---------- Pass NL_Filename to create_cnfire_method M components/clm/src/main/clm_varcon.F90 ----------------- Add secsphr M components/clm/src/main/clm_varctl.F90 ----------------- Remove use_nofire M components/clm/src/main/controlMod.F90 ----------------- Remove use_nofire, abort if can't find namelists M components/clm/src/main/glcBehaviorMod.F90 ------------- abort if can't find namelist M components/clm/src/main/initVerticalMod.F90 ------------ abort if can't find namelist and abort if use_bedrock and zbedrock NOT on surface dataset M components/clm/src/main/ndepStreamMod.F90 -------------- abort if can't find namelist M components/clm/src/main/restFileMod.F90 ---------------- abort if can't find namelist M components/clm/tools/clm4_5/mksurfdata_map/Makefile.data -- by default create crop 1850 f45, ne30, ne120 datasets M components/clm/tools/clm4_5/mksurfdata_map/src/mkpftMod.F90 - Lower tolerance to relerr = 0.0001 (bug 2285) =============================================================== =============================================================== Tag name: clm4_5_8_r168 Originator(s): Keith Oleson; brought to trunk by Bill Sacks Date: Sat Feb 20 20:55:19 MST 2016 One-line Summary: Fix zero methane production Purpose of changes ------------------ Methane production has been zero ever since the hydrology reordering in clm4_5_13 (in June, 2013). The order of subroutine calls appears to be a problem: 1. CNDriverNoLeaching is called in clm_driver, which calls SetValues, which zeros out the carbon fluxes (e.g., somhr and lithr) 2. ch4 is called which uses the zero-ed out fluxes 3. Summary is called which sums up the soil layer som and litter heterotrophic respiration terms which should be used by ch4, but are not. This tag puts the call to ch4 back where it was in clm4_5_12 and earlier. Keith Oleson found and fixed this bug; Bill Sacks brought it to the trunk. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2225: CH4PROD=0 - 2287: Methane production is zero Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - Changes answers far more than you might immediately suspect; this was deemed to be reasonable at the 02/18/2016 TSS meeting Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: Keith Oleson, Bill Sacks CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok 'ok' means: tests pass or xfail, some baseline comparisons fail as noted below CLM tag used for the baseline comparisons: clm4_5_7_r167 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 & CLM50 with BGC turned on - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: clm5_respn05r162_2degGSWP3_1850spin_allN_v01_ch4A_785_799 URL for LMWG diagnostics output used to validate new climate: http://www.cgd.ucar.edu/tss/clm/diagnostics/clm5_dev/clm5_respn05r162_2degGSWP3_1850spin_allN_v01_ch4A_785_799-clm5_respn05r162_2degGSWP3_1850spin_allN_v01_785_799/setsIndex.html Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Move ch4 call to where it was in clm4_5_12 and earlier M components/clm/src/main/clm_driver.F90 =============================================================== =============================================================== Tag name: clm4_5_7_r167 Originator(s): sacks (Bill Sacks) Date: Sat Feb 20 06:24:52 MST 2016 One-line Summary: Fix carbon isotopes in transient runs Purpose of changes ------------------ In CNWoodProducts, code that was supposed to operate on c13 / c14 variables was actually operating on the standard c12 variable. This led to incorrect results and carbon balance errors if running with carbon isotopes in transient runs. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2207: bug in carbon isotopes in CNWoodProducts Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - not run tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_7_r166 Answer changes -------------- Changes answers relative to baseline: NO In principle, changes answers for transient runs with carbon isotopes enabled, but since those runs were previously crashing (due to carbon balance errors), I am listing this as no answer changes. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Add a long, transient test with carbon isotopes A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_monthly/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_monthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_monthly List all existing files that have been modified, and describe the changes: ========= Main change M components/clm/src/biogeochem/CNWoodProductsMod.F90 ========= Add a long, transient test with carbon isotopes M components/clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_7_r166 Originator(s): sacks (Bill Sacks) Date: Tue Feb 2 04:58:45 MST 2016 One-line Summary: Infrastructure for carbon and nitrogen conservation with dynamic landunits Purpose of changes ------------------ Puts in place the infrastructure that will be used for carbon and nitrogen conservation, and possibly other column-level conservation, with dynamic landunits. This does NOT yet implement any of teh actual carbon and nitrogen conservation - that will be done in an upcoming tag via uses of this new infrastructure. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: High-level (design) review by Erik Kluzek and Ben Andre CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_7_r165 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): List all files eliminated: List all files added and what they do: ========= Main new code: class for adjusting column-level state variables due to transient column areas A components/clm/src/dyn_subgrid/dynColumnStateUpdaterMod.F90 A components/clm/src/dyn_subgrid/test/dynColumnStateUpdater_test/CMakeLists.txt A components/clm/src/dyn_subgrid/test/dynColumnStateUpdater_test/test_column_state_updater.pf A components/clm/src/dyn_subgrid/test/dynColumnStateUpdater_test ========= Find a template column to use for the state variables on some other column of interest. Used by both dynColumnStateUpdater and dynInitColumns. A components/clm/src/dyn_subgrid/dynColumnTemplateMod.F90 A components/clm/src/dyn_subgrid/test/dynColumnTemplate_test/test_column_template.pf A components/clm/src/dyn_subgrid/test/dynColumnTemplate_test/CMakeLists.txt A components/clm/src/dyn_subgrid/test/dynColumnTemplate_test ========= Add test of get_ltype_special A components/clm/src/unit_test_shr/test/unittestSubgrid_test/test_unittestSubgrid.pf A components/clm/src/unit_test_shr/test/unittestSubgrid_test/CMakeLists.txt A components/clm/src/unit_test_shr/test/unittestSubgrid_test List all existing files that have been modified, and describe the changes: ========= Initialize a column state updater object, and set the old and new weights in each time step M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Extract initial_template_col and some related code into a new class (dynColumnTemplateMod) M components/clm/src/dyn_subgrid/dynInitColumnsMod.F90 ========= New function: landunit_is_special M components/clm/src/main/landunit_varcon.F90 M components/clm/src/main/initSubgridMod.F90 ========= Just add a comment M components/clm/src/dyn_subgrid/dynPriorWeightsMod.F90 ========= Add get_ltype_special function M components/clm/src/unit_test_shr/unittestSubgridMod.F90 ========= Move test of extracted code to new directory M components/clm/src/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf ========= Add files for unit test build M components/clm/src/dyn_subgrid/CMakeLists.txt M components/clm/src/unit_test_shr/test/CMakeLists.txt M components/clm/src/dyn_subgrid/test/CMakeLists.txt =============================================================== =============================================================== Tag name: clm4_5_7_r165 Originator(s): sacks (Bill Sacks) Date: Fri Jan 29 05:28:46 MST 2016 One-line Summary: Remove fglcmask, recreate all surface datasets Purpose of changes ------------------ Remove fglcmask input file. This specified the area where we need virtual glacier elevation classes for coupling to CISM. There were two main problems with this: - There was a separate file for each CLM resolution at which we can couple to CISM - currently this is just f09, f19, T31. There was no easy way to generate these files for all CLM resolutions - which is something that is needed in order to turn GLC on by default in all CESM runs. - This did not provide the flexibility needed to treat three regions differently: Greenland (virtual columns), Antarctica (multiple elevation classes, but no virtual columns), and mountain glaciers (single elevation class). This tag removes the fglcmask file, replacing it with an integer-valued GLACIER_REGION field on the surface dataset, together with a glacier_region_behavior namelist option. I have generated all new surface datasets. In addition to adding the GLACIER_REGION field, these: - add zbedrock (most out-of-the-box surface datasets did not yet have this field) - remove unused TOPO field There are also new initial conditions files, simply for the sake of updating the surface_dataset metadata. Also changed mksurfdata_map's Makefile.data and perl script (including adding a -no_surfdata flag) so that the necessary out-of-the-box surface datasets are all created correctly (with correct names). Also did some other cleanup of mksurfdata.pl. Also removed kludgey Greenland and Antarctica areas from history file, which weren't being used. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - New surface datasets have zbedrock, in contrast to many old datasets. This means that runs with use_bedrock = .true. will actually use bedrock now, changing answers - For any run with glc_mec (compsets with CISM): will need to run initial conditions through init_interp - For any initial conditions file of your own that was generated from an earlier CLM tag (i.e., not an out-of-the-box initial conditions file): will need to set check_finidat_fsurdat_consistency to .false., because surface datasets have changed - 1/8 deg surface dataset hasn't been updated, so won't work with this tag Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - New namelist variable: glacier_region_behavior Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): - all new surface datasets; these are the same as before, but have: - GLACIER_REGION added - TOPO removed - zbedrock added (wasn't yet on most default surface datasets) - new initial conditions files: just updated surface_dataset metadata to point to new surface datasets Substantial timing or memory changes: NONE Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - pass 010 blg54 failed due to problems / differences in the baseline; when I fixed those, it passed PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok ok means: all tests pass or xfail. Baseline comparisons fail for IG tests and ERS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_intel.clm-deepsoil_bedrock, as noted below. CLM tag used for the baseline comparisons: clm4_5_7_r164 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: - configurations with glc_mec (compsets with CISM) - configurations with use_bedrock = .true. for which zbedrock wasn't previously on the surface dataset - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; not investigated closely The answer changes in glc_mec runs are due to small changes in the set of grid cells in the single_at_atm_topo zone. I confirmed that, if I use a kludged-in GLACIER_REGION field that exactly matches the old fglcmask, then answers are bit-for-bit. The answer changes in use_bedrock configurations are presumably due to zbedrock being on the surface datasets now. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): No updates to top-level externals; updated PTCLM to allow build on pronghorn (same as yellowstone / caldera) List all files eliminated: ========= Mostly-unrelated change: remove unused fields from history file (these specified Greenland and Antarctica areas in a kludgey way) D components/clm/src/main/glcDiagnosticsMod.F90 ========= Renamed to be consistent with new naming convention D components/clm/tools/clm4_5/mksurfdata_map/landuse_timeseries_hist_simyr1850-2005.txt List all files added and what they do: ========= Add GLACIER_REGION field to surface dataset A components/clm/tools/clm4_5/mksurfdata_map/src/mkglacierregionMod.F90 ========= Renamed to be consistent with new naming convention A components/clm/tools/clm4_5/mksurfdata_map/landuse_timeseries_hist_16pfts_simyr1850-2005.txt List all existing files that have been modified, and describe the changes: ========= Remove glcmask. Instead make glcBehavior based on GLACIER_REGION on the surface dataset M components/clm/src/utils/domainMod.F90 M components/clm/src/main/controlMod.F90 M components/clm/src/main/test/glcBehavior_test/test_glcBehavior.pf M components/clm/src/main/glc2lndMod.F90 M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/surfrdMod.F90 M components/clm/src/main/clm_instMod.F90 M components/clm/src/main/glcBehaviorMod.F90 M components/clm/src/main/clm_initializeMod.F90 ========= Add GLACIER_REGION to surface datasets. Point to all new surface datasets. Also new initial conditions files - identical to old ones except for surface_dataset metadata. Also, remove fglcmask namelist option and add glacier_region_behavior namelist option. M components/clm/tools/clm4_5/mksurfdata_map/mksurfdata_map.namelist M components/clm/tools/clm4_5/mksurfdata_map/README M components/clm/tools/clm4_5/mksurfdata_map/README.developers M components/clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 M components/clm/tools/clm4_5/mksurfdata_map/src/README.unit_testing M components/clm/tools/clm4_5/mksurfdata_map/src/Srcfiles M components/clm/tools/clm4_5/mksurfdata_map/src/mkvarctl.F90 M components/clm/tools/clm4_5/mksurfdata_map/src/mkfileMod.F90 M components/clm/tools/clm4_5/mksurfdata_map/unit_testers/Srcfiles M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/namelist_files/checkmapfiles.ncl M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm M components/clm/doc/UsersGuide/tools.xml M components/clm/doc/UsersGuide/adding_files.xml ========= Make new routine for printing diagnostics of index fields M components/clm/tools/clm4_5/mksurfdata_map/src/mkdiagnosticsMod.F90 M components/clm/tools/clm4_5/mksurfdata_map/src/mkagfirepkmonthMod.F90 M components/clm/tools/clm4_5/mksurfdata_map/src/mkurbanparMod.F90 ========= Fix makefile-based creation of surface datasets so that 'make all' gives correctly-named datasets for all needed out-of-the-box surface datasets M components/clm/tools/clm4_5/mksurfdata_map/Makefile.data M components/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl ========= Point to new PTCLM: allow testing on pronghorn (same as yellowstone) M components/clm/tools/SVN_EXTERNAL_DIRECTORIES ========= Add tests to xfail list. These all failed in clm4_5_7_r164 except for these new failures: + RUN ERS_Ly3.f10_f10.I1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KSinkMOut + RUN SMS_D_Lm1_Mmpi-serial.CLM_USRDAT.I1PTCLM45.yellowstone_pgi.clm-USUMB M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_7_r164 Originator(s): erik (Erik Kluzek) Date: Tue Jan 12 00:51:25 MST 2016 One-line Summary: Update CIME and MOSART Purpose of changes ------------------ Update CLM to a robust version of CIME and PIO2 (used in the cesm1_5_beta04 tag). Also fix an issue in MOSART that changes answers (qgwl_volume must be multiplied by area). Note, that this version of CIME removes the archive.locked directory. Change the rearranger that PIO uses from being hardcoded to using the one picked with the env_run.xml variable LND_PIO_REARRANGER (both clm40 and clm45). By default, as before the rearranger for clm40 is the new SUBSET option, and for clm45/clm50 it's the old BOX method, eventually we expect both to use the SUBSET method by default. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2265 Workaround for pio_syncfile issue 2257 Unit tests fail 2256 Issues with early versions of PIO2 Bugs found recently: 2264 Fire is missing accounting of C/N for transfer of live to dead Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): This version of CIME now creates cimeteststatus instead of cs.status when create_test used (use the summary option to cimeteststatus "-s"!) Remove the archive.locked functionality and the creation of the associated hardlinks Removed the option DOUT_S_SAVE_DATES_RESTART_FILE_SET Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): pio rearranger set the rearranger for PIO to use by the env_run variable LND_PIO_REARRANGER Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: PIO2 should have better performance and it allows two options for the rearranger to use (BOX and SUBSET). Code reviews and testing ------------------------ Code reviewed by: self CLM testing: regular build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - OK (all PASS except irrigation test with known issue) tools-tests (components/clm/test/tools): yellowstone - PASS PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - OK (only one expected FAIL) regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK OK means tests pass except those in the ExpectedFail list file CLM tag used for the baseline comparisons: clm4_5_7_r163 Answer changes -------------- Changes answers relative to baseline: only if MOSART used If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: with mosart - what platforms/compilers: all - nature of change: fixes bug where area wasn't multiplied in Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): cime, mosart cime to cime4.3.9 mosart to mosart1_0_15 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/README.unit_testing -- Add note about how to delete previous build M components/clm/src/init_interp/initInterp.F90 -- workaround for bug in PIO where you can't do two writes in a row. M components/clm/src/main/ncdio_pio.F90.in -- Take out specifier for rearranger so will use the setting from LND_PIO_REARRANGER (was hardcoded to BOX) M components/clm/src_clm40/main/ncdio_pio.F90.in so will use the setting from LND_PIO_REARRANGER (was hardcoded to SUBSET) M components/clm/cime_config/config_component.xml -------- Set LND_PIO_REARRANGER to BOX for clm45/clm50 and to SUBSET for clm40 M components/clm/cime_config/testdefs/testlist_clm.xml --- change goldbach tests for prealpha/prebeta to hobart M components/clm/cime_config/testdefs/testmods_dirs/clm/40default/shell_commands M components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/shell_commands M components/clm/cime_config/testdefs/testmods_dirs/clm/default/shell_commands M components/clm/cime_config/testdefs/testmods_dirs/clm/reduceOutput/shell_commands =============================================================== =============================================================== Tag name: clm4_5_7_r163 Originator(s): sacks (Bill Sacks) Date: Wed Dec 23 20:31:22 MST 2015 One-line Summary: Improve crop performance by allocating less memory Purpose of changes ------------------ Performance of crop runs was being significantly impacted by the fact that we allocate memory for any crop column that might ever need to come into existence with dynamic landunits. Most CLM code should not run over 0-weight points, so this likely reflects problems in memory/cache efficiency. This tag changes that behavior, so that for non-transient runs, we only allocate memory for the crops that actually exist in that run (i.e., have non-zero area). For transient crop runs, the behavior is currently the same as before. Note that this means you need to run init_interp when transitioning from a non-transient crop run to a transient crop run, or vice versa. For an f09 run without i/o (20-day run with compset 1850_DATM%QIA_CLM50%BGC-CROP_SICE_SOCN_RTM_SGLC_SWAV), this resulted in: - number of columns reduced from 674,634 to 277,728 - number of patches reduced from 968,284 to 571,378 - throughput increased from 3.821 seconds/mday to 2.404 seconds/mday Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to interpolate initial conditions): - need to run old crop initial conditions through init_interp (already done for out-of-the-box finidat files) - from now on, need to run init_interp when transitioning from a non-transient crop run to a transient crop run, or vice versa Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): new finidat files for crop: same as old, but run through init_interp Substantial timing or memory changes: substantial improvement in both timing and memory for crop cases Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - irrigation test fails, as in r162 tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): NOTE: On yellowstone, I tested with a cime version that had a one-off from cime4.3.1: changes to cprnc. This was done so that cprnc would indicate which tests had differences in dimension sizes in the h1 history files. yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok 'ok' means tests pass, baseline comparisons fail for some diagnostic fields for crop cases, as described below Also, this test failed the baseline comparison due to a problem in the baseline: ERS_D_Ld5.f10_f10.IMCRUCLM50BGC.yellowstone_intel.clm-default Some r162 baselines were missing on hobart; I reran these tests and verified that baseline comparisons passed. CLM tag used for the baseline comparisons: clm4_5_7_r162 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 with crop - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Changes in diagnostic fields only Changes are in: (1) PCT_CFT in grid cells that have 0% crop (2) dimension sizes differ for history files with dov2xy = .false. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/main/initGridCellsMod.F90 M components/clm/src/main/subgridMod.F90 ========= Initialize dynSubgridControl earlier in initialization. This is needed because it now needs to be initialized before we set up the subgrid stuff / bounds. M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 M components/clm/src/main/clm_initializeMod.F90 ========= Add more informative message if it looks like use_init_interp is needed M components/clm/src/main/restFileMod.F90 M components/clm/src/main/ncdio_pio.F90.in ========= Point to new finidat files for crop, which have been run through init_interp M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/user_nl_clm ========= Minor tweak: remove explicit reference to init_interp, to make it easier to search the ChangeLog for true references to init_interp M .ChangeLog_template =============================================================== =============================================================== Tag name: clm4_5_7_r162 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Dec 18 15:59:23 MST 2015 One-line Summary: negative runoff updates. Purpose of changes ------------------ Remove direct to ocean runoff from coupler fields. Negative runoff is now handled by mosart. Only limit irrigation if volume is below the threshold instead of lake evaporation and irrigation. Bugs fixed or introduced ------------------------ unknown Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): Not well tested after some last minute changes. Not recommended for use. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): limit_lake_evap_and_irrig phys changed to limit_irrigation Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown Code reviews and testing ------------------------ Code reviewed by: swenson CLM testing: minimal build-namelist tests yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - irrigation test fails tools-tests (components/clm/test/tools): yellowstone - n/a not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - clm50 kitchen sink looks ok yellowstone_pgi - not run prior to tag yellowstone_gnu (clm45 only) - not run prior to tag hobart_nag - not run prior to tag prior testing of branch indicated that it was bit for bit with the exception of coupler and histroy fields related to runoff: TWS VOLR VOLRMCH l2x_Flrl_rofgwl l2x_Flrl_rofdto x2l_Flrr_volr x2l_Flrr_volrmch r2x_Forr_rofl r2x_Flrr_volr r2x_Flrr_volrmch x2r_Flrl_rofgwl x2r_Flrl_rofdto several last minute changes were made and testing was not rerun before making the tag. No significant problems are expected, but... CLM tag used for the baseline comparisons: clm4_5_6_r161 Answer changes -------------- Changes answers relative to baseline: yes, only in clm and coupler runoff fields listed above. If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff. If bitwise differences were observed, how did you show they were no worse than roundoff? none If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): rtm1_0_55, mosart1_0_12 List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: change name of limit_irrigation namelist item clm/bld/CLMBuildNamelist.pm clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/namelist_files/namelist_defaults_clm4_5.xml clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 change negative runoff clm/src/main/lnd2atmType.F90 clm/src/main/controlMod.F90 clm/src/main/clm_varctl.F90 clm/src/main/lnd2atmMod.F90 clm/src/biogeophys/WaterfluxType.F90 clm/src/biogeophys/LakeFluxesMod.F90 clm/src/biogeophys/HydrologyDrainageMod.F90 clm/src/biogeophys/BandDiagonalMod.F90 clm/src/biogeophys/IrrigationMod.F90 clm/src/cpl/clm_cpl_indices.F90 clm/src/cpl/lnd_import_export.F90 =============================================================== =============================================================== Tag name: clm4_5_6_r161 Originator(s): sacks (Bill Sacks) Date: Thu Dec 17 14:04:29 MST 2015 One-line Summary: Collapse mountain glaciers to a single elevation class Purpose of changes ------------------ Mountain glaciers (i.e., glaciers outside of Greenland and Antarctica) have been problematic when running with glc_mec, because of the adjustments made to the atmospheric forcings (downscaling to columns). Mountain glaciers generally sit above the atmospheric forcing height, so this downscaling leads to discrepancies between the atmospheric forcings and what CLM sees, on the gridcell mean. This same problem can exist over ice sheets, but it can be more pronounced in mountain regions because of the steep topography. Furthermore, because no downscaling is done over the vegetated portion of the grid cell: Even if CLM's mean grid cell elevation agreed exactly with the atmosphere's topographic height, there would still be a discrepancy after downscaling, because the glaciers would see a cooler climate, whereas the rest of the grid cell would see exactly the climate sent by the atmosphere. Mountain glaciers are not currently a focus of the Land Ice Working Group (LIWG), and surface mass balance is known to have significant problems in mountain glacier regions. Therefore, because of the above atm-lnd discrepancy, the LIWG (especially Bill Lipscomb and Jeremy Fyke) decided it would be best to have only a single glacier elevation class in mountain glacier regions (outside of Greenland and Antarctica), whose elevation exactly matches the atmosphere's topographic height. Note that this match of the atmopshere's topographic height is the same as what is done (implicitly) in all other landunits - resulting in no adjustment to the atmospheric forcings. This tag implements the collapse of mountain glaciers to a single elevation class, whose topographic height matches the atmosphere's topographic height. It allows for the possibility that the atmosphere's topographic height could change in time - in which case the glacier's topographic height (and potentially its elevation class index) will change in time to follow the atmosphere. In addition, this tag puts in place infrastructure to make it easier to define different glacier regions that receive different treatments (whether virtual columns are enabled, and whether glacier should be collapsed to a single elevation class). The "frontend" user interface for this will be added in an upcoming tag. Currently, the definition of the "mountain glacier" region is done so that it includes everything except the region defined by fglcmask and the region south of 60 deg S (assumed to be Antarctica). Soon the setting of this region will be made more robust and more easily user-definable. Finally, this tag adds infrastructure to aid in creating and using local column-level filters. Eventually, this could be extended to cover all of CLM's filters. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): - For runs with glc_mec (IG, FG, BG), need to run initial conditions through init_interp, because there are now slightly fewer glc_mec columns. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): - just changed finidat for glcMEC_spunup_1way testmod Substantial timing or memory changes: - small decrease in memory for glc_mec runs, due to a decrease in the number of columns Code reviews and testing ------------------------ Code reviewed by: self; review by others will come later CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - pass 'ok' means: tests pass except for baseline comparison failures for IG tests, as noted below CLM tag used for the baseline comparisons: clm4_5_6_r160 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM50 with glc_mec (IG, FG, BG) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): not investigated closely, but expected to give new climate, at least regionally If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Verified that changes came just from collapsing mountain glaciers to a single elevation class, by running the affected tests with this change in glcBehaviorMod: - logical, parameter :: collapse_mountain_glaciers = .true. + logical, parameter :: collapse_mountain_glaciers = .false. -- these tests were bit-for-bit with the trunk, as expected. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Main addition in this tag. This module determines a number of aspects of the behavior of glacier_mec classes in each grid cell. A components/clm/src/main/glcBehaviorMod.F90 ========= Unit tests of glcBehaviorMod A components/clm/src/main/test/glcBehavior_test/CMakeLists.txt A components/clm/src/main/test/glcBehavior_test/test_glcBehavior.pf A components/clm/src/main/test/glcBehavior_test ========= Supports creation and use of local column-level filters. Eventually we may want to consider moving all filters to using infrastructure like this. A components/clm/src/main/filterColMod.F90 ========= Unit tests of filterColMod A components/clm/src/main/test/filter_test/CMakeLists.txt A components/clm/src/main/test/filter_test/test_filter_col.pf A components/clm/src/main/test/filter_test ========= Stub of module in driver_cpl/shr (needed by glc_elevclass_mod.F90) A components/clm/src/unit_test_stubs/csm_share/seq_comm_mct.F90 ========= Minor changes needed to support unit tests List all existing files that have been modified, and describe the changes: ========= Delegate glc_mec-related logic to glcBehaviorMod: this logic is now complex, and tied in with various aspects of glcBehaviorMod; also, delegating the logic to glcBehaviorMod allows us to reduce duplication between initGridCellsMod and subgridMod. Also split the monster routine in subgridMod (which had a bunch of optional arguments) into small routines with well-defined interfaces. M components/clm/src/main/subgridMod.F90 M components/clm/src/main/initGridCellsMod.F90 - also, put setting of ice_mec landunits in a separate routine, thus cleaning up set_landunit_wet_ice_lake; and remove no-longer-used setdata argument ========= Use glc_behavior instead of glcmask M components/clm/src/main/glc2lndMod.F90 - also call glc_behavior%update_collapsed_columns M components/clm/src/main/subgridWeightsMod.F90 ========= Track whether a column's type is dynamic M components/clm/src/main/ColumnType.F90 M components/clm/src/main/initSubgridMod.F90 ========= Update glc classes upon restart; also separate read and write routines (which now have different required arguments) M components/clm/src/main/subgridRestMod.F90 ========= Declare and initialize glc_behavior, and pass it where needed M components/clm/src/main/clm_instMod.F90 M components/clm/src/main/clm_initializeMod.F90 ========= Pass newly-needed arguments M components/clm/src/main/clm_driver.F90 M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M components/clm/src/main/decompInitMod.F90 M components/clm/src/main/restFileMod.F90 M components/clm/src/main/reweightMod.F90 ========= Make sure glcmask is always initialized to 0 M components/clm/src/main/surfrdMod.F90 ========= Add new history field M components/clm/src/main/atm2lndType.F90 ========= For unit tests: add glc_elevclass_mod.F90 from driver_cpl/shr M components/clm/src/CMakeLists.txt ========= Minor changes needed to support unit tests M components/clm/src/unit_test_stubs/csm_share/CMakeLists.txt M components/clm/src/unit_test_shr/unittestArrayMod.F90 M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 M components/clm/src/main/CMakeLists.txt M components/clm/src/main/test/CMakeLists.txt ========= Remove some costly CLM4 tests M components/clm/cime_config/testdefs/testlist_clm.xml M components/clm/cime_config/testdefs/ExpectedTestFails.xml ========= Point to new initial conditions file, init_interp'ed from the old one M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm =============================================================== =============================================================== Tag name: clm4_5_6_r160 Originator(s): sacks (Bill Sacks) Date: Fri Dec 11 10:51:21 MST 2015 One-line Summary: For glc_mec, get gridcell topographic height from atm Purpose of changes ------------------ Runs with glc_mec require knowledge of the atmosphere's topographic height. Up until now, this was specified via the flndtopo file for the three resolutions that glc_mec can run at: f09, f19, T31. However, this has three problems: (1) It made it hard to extend glc_mec to run at other CLM resolutions (2) If CAM's topographic height changes (as it did recently), CLM would need to have its flndtopo files updated to agree. (3) It assumes that datm's forcing heights match CAM's. In this tag, flndtopo has been removed, and CLM now gets the atmosphere's topographic height from the atmosphere, via the coupler. References to flndtopo / ldomain%topo have been replaced with atm2lnd_inst%forc_topo_grc. For now, datm is set up to provide a topographic height that matches the old f09 file - so offline (I compset) runs at f09 will be the same as before. But this will soon change. Also removed fatmtopo, which was in the code in a few places but was no longer being used. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): - removed flndtopo file for CLM45/CLM5 Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): - removed flndtopo file for CLM45/CLM5 Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - pass (both sets of tests) unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - pass 'ok' means pass except baseline comparison failures for IG tests, as noted below Note that most hobart_nag tests are now in the expected fail list. The only passing tests on hobart_nag are: SMS_D_Ld1_P24x1.f10_f10.ICRUCLM45.hobart_nag.clm-af_bias_v5 SMS_Ld1_Mmpi-serial.f45_f45.ICLM45.hobart_nag.clm-ptsRLA SMS_D_Ld1_Mmpi-serial.1x1_vancouverCAN.I1PTCLM50.hobart_nag.clm-default SMS_Ld1.f45_f45.ICLM45.hobart_nag.clm-ptsRLA SMS_D_Ld1_Mmpi-serial.1x1_vancouverCAN.I1PTCLM45.hobart_nag.clm-default SMS_D_Ld1_P24x1.f10_f10.ICLM45BGC.hobart_nag.clm-oldhyd ERS_D_Ld5_Mmpi-serial.1x1_mexicocityMEX.I1PTCLM45.hobart_nag.clm-default SMS_D_Ld1_Mmpi-serial.f45_f45.ICLM45.hobart_nag.clm-ptsRLA ERP_Ly5.1x1_numaIA.ICLM45BGCCROP.hobart_nag.clm-monthly SMS_Ld1.1x1_mexicocityMEX.I1PT.hobart_nag.clm-40default SMS_Ld5_P24x1.f10_f10.IRCP26CN.hobart_nag.clm-reduceOutput SMS_D_Ld1.1x1_mexicocityMEX.I.hobart_nag CLM tag used for the baseline comparisons: clm4_5_6_r159 compared ERS_Ly3.f10_f10.I1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KSinkMOut against baselines in directory clm4_5_r159 Answer changes -------------- Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45/CLM5 with glc_mec (IG, FG, BG) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Potentially new climate. Magnitude of change depends on how different atm's topography is from the flndtopo file that had been used before. For now, offline (IG) runs at f09 are the same as before, but all other offline runs change. Runs coupled to CAM may be the same (or close to the same), if CAM is using its old topography files, but will change for the new CAM topography. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M components/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/doc/UsersGuide/preface.xml M components/clm/doc/UsersGuide/adding_files.xml M components/clm/doc/UsersGuide/single_point.xml M components/clm/src/cpl/clm_cpl_indices.F90 M components/clm/src/cpl/lnd_import_export.F90 M components/clm/src/utils/domainMod.F90 M components/clm/src/main/clm_initializeMod.F90 M components/clm/src/main/atm2lndType.F90 M components/clm/src/main/controlMod.F90 M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/surfrdMod.F90 ========= Remove topo field that is defined but never set M components/clm/src/main/histFileMod.F90 ========= Add test to expected fail list; it seemed to pass for Erik, but fails for me even out of a fresh checkout of clm4_5_6_r159: ERP_Ly5.1x1_numaIA.ICRUCLM50BGCCROP.hobart_nag.clm-clm50BGCCROPmonthly M components/clm/cime_config/testdefs/ExpectedTestFails.xml =============================================================== =============================================================== Tag name: clm4_5_6_r159 Originator(s): erik (Erik Kluzek) Date: Thu Dec 10 15:46:47 MST 2015 One-line Summary: Update CIME version Purpose of changes ------------------ Update to very latest version of CIME, which we had held off doing for two months. The main user change is that CIME no longer names scripts with your CASENAME. This CIME version also updated PIO1 to PIO2 (pio2_0_27). These early versions of PIO2 do have issues with potential hangs. See bug 2256 below. As such we do not recommend use of this tag. Or if you do and run into problems, do the workaround in the bug report to move the PIO version to pio1_9_23. The behavior is that after the simulation finishes it times out when writing history and restart files. mpi-serial (single point) cases seem to work fine. Also fix two minor bugs. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2251 Bad NH max planting date of 6/31 2250 History filenames get truncated to 128 characters Bugs Introduced (include bugzilla ID): 2257 Unit tests fail in clm4_5_6_r159 2256 Problems with early versions of PIO2 in CLM Notes of particular relevance for users --------------------------------------- Caveats for users: CASE SCRIPTS NO LONGER NAMED $CASE.build, $CASE.run, $CASE.submit After ./create_newcase is run you have the scripts named: ./case.build ./case.run ./case.submit Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): New params file Substantial timing or memory changes: None Code reviews and testing ------------------------ Code reviewed by: self, fvitt CLM testing: regular build-namelist tests: yellowstone - PASS (several comparision tests to clm4_5_6_r158 FAIL because of the new params file) unit-tests (components/clm/src): yellowstone - FAIL tools-tests (components/clm/test/tools): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK OK means failures are in the ExpectedFails list CLM tag used for the baseline comparisons: clm4_5_6_r158 Answer changes -------------- Changes answers relative to baseline: Yes (but just for two diagnostic cpl fields) Summarize any changes to answers, i.e., - what code configurations: All - what platforms/compilers: All - nature of change: diagnostic change atm topo field For drydep TERPOOH (dd034) is different because of an update by Francis Vitt Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): cime to cime4.3.1 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: Increase max filename length from 128 to 199. M components/clm/src/main/ncdio_pio.F90.in -- increase tmp_string to 199 M components/clm/src/main/histFileMod.F90 --- max length filename and max_chars to 199 and dimension locfnhr by max_length_filename M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- new params file change 631 max planting date for NH to 701 =============================================================== =============================================================== Tag name: clm4_5_6_r158 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Dec 4 16:10:11 MST 2015 One-line Summary: misc clm 5.0 bugfixes Purpose of changes ------------------ Misc. bugfixes for CLM 5.0 * LUNA bugfix to endrun if simulation goes out of range. * Restore ability to run SP simulations on nag. * String length bugfix from Jim Edwards. * New long clm50 kitchen sink test. * endrun if history filenames exceed maximum allowed length. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2209, 2213 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): CLM will now endrun if the generated history filename exceeds the maximum allowed length. This will result in run time failures for people running the test suite with the default or longer test ids. CLM 5.0 with LUNA will now endrun if nitrogen computations go outside a valid range. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown, none expected Code reviews and testing ------------------------ Code reviewed by: self CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40 - pass; 45 - ok yellowstone_pgi - 40 - pass; 45 - pass yellowstone_gnu (clm45 only) - 45 - pass hobart_nag - ok CLM tag used for the baseline comparisons: clm4_5_6_r157 Notes: 'ok' means: * Fixing bug 2213 restores tests an hobart_nag. These tests BFAIL because they don't have baselines. * r157 yellowstone baseline does not include clm history files for two tests because of filename length issues: ERS_D_Ld3.f09_g16_gl5.IG1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KitchenSink ERP_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-clm50BGCDVCROPmonthly * New test does not have baseline: ERS_Ly3.f10_f10.I1850CRUCLM50BGCCROP.yellowstone_intel.clm-clm50KSinkMOut Answer changes -------------- Changes answers relative to baseline: bit for bit Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: new clm50 kitchen sink test with monthly output clm/cime_config/testdefs/testmods_dirs/clm/clm50KSinkMOut/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/clm50KSinkMOut/include_user_mods clm/cime_config/testdefs/testmods_dirs/clm/clm50KSinkMOut List all existing files that have been modified, and describe the changes: new clm50 compset w/o glc and test for longer kitchen sink test clm/cime_config/config_compsets.xml clm/cime_config/testdefs/testlist_clm.xml string length bugfix from Jim Edwards clm/src/biogeochem/FireEmisFactorsMod.F90 runtime error if constructed history filename exceeds max length clm/src/main/histFileMod.F90 bug 2209 clm/src/biogeophys/LunaMod.F90 - runtime error if LUNA nitrogen computations become unrealistic bug 2213 clm/src/biogeophys/PhotosynthesisMod.F90 - remove association of CN variable that may not be allocated for SP runs. clm/src/biogeophys/CanopyFluxesMod.F90 - remove association of CN variable that may not be allocated for SP runs. clm/cime_config/testdefs/ExpectedTestFails.xml - remove nag tests from xfails. =============================================================== =============================================================== Tag name: clm4_5_6_r157 Originator(s): sacks (Bill Sacks) Date: Fri Nov 27 05:52:17 MST 2015 One-line Summary: Reduce memory use for init_interp Purpose of changes ------------------ For high-resolution source files (e.g., f09 or higher), the new multilevel interpolation was using unacceptable amounts of memory, since it was reading all source points at all levels into memory at once - both for the actual data to interpolate, and for the metadata used in the multilevel interpolators. This tag reworks how 2-d interpolation is done so that only one level from the source is needed in memory at a time, thus greatly reducing memory use in some cases. In making this change, I extracted routines to do 1-d interpolation. In doing so, I made some minor changes to their behavior for consistency. The main change was to remove special handling of inactive points: Inactive points used to have their values set to spval, and I have now removed that logic, keeping inactive points at their values before interpolation. This results in some NaNs appearing in finidat_interp_dest (if values were NaN before interpolation); before spval was put in all inactive points. Also, some unrelated changes: - Remove a duplicate and unused entry in config_component.xml (found by Jim Edwards) - Remove redundant settings in CNZero (found by Dave Lawrence) - Fix cray compilation problems (found by Jim Edwards) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: - Big memory decrease in some cases when using init_interp Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass CLM tag used for the baseline comparisons: clm4_5_6_r156 Answer changes -------------- Changes answers relative to baseline: NO Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Extract interpolation routines so they can be called from a few places. Note that I have changed the behavior slightly, in ways that affect the produced restart file, but not a CLM run done with this restart file. In particular, inactive points are no longer set to spval. A components/clm/src/init_interp/initInterp1dData.F90 List all existing files that have been modified, and describe the changes: ========= Rework multi-level interpolation in order to reduce memory use M components/clm/src/init_interp/initInterp2dvar.F90.in M components/clm/src/init_interp/initInterpMultilevelInterp.F90 M components/clm/src/init_interp/initInterpMultilevelContainer.F90 M components/clm/src/init_interp/initInterpMultilevelSnow.F90 M components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_split.pf M components/clm/src/init_interp/test/initInterpMultilevel_test/initInterpMultilevelMock.pf M components/clm/src/init_interp/test/initInterpMultilevel_test/multilevel_interp_factory.F90 M components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_interp.pf M components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_snow.pf M components/clm/src/init_interp/initInterpMultilevelBase.F90 M components/clm/src/init_interp/initInterpMultilevelCopy.F90 M components/clm/src/init_interp/initInterpMultilevelSplit.F90 M components/clm/src/init_interp/initInterp.F90 ========= Unrelated change: remove an entry that duplicates an entry in the driver's config_component, and is unused here M components/clm/cime_config/config_component.xml ========= Unrelated change: remove redundant settings in CNZero M components/clm/src/biogeochem/CNDriverMod.F90 ========= Unrelated change: add helpful tip to output message M components/clm/src/init_interp/initInterpMindist.F90 ========= Unrelated change: add a comment M components/clm/src/main/clm_varcon.F90 =============================================================== =============================================================== Tag name: clm4_5_6_r156 Originator(s): sacks (Bill Sacks) Date: Tue Nov 24 10:19:07 MST 2015 One-line Summary: Fix QSNWCPICE and glacial inception Purpose of changes ------------------ Previously, qflx_snwcp_ice_col (output to history file as QSNWCPICE) had served two purposes: (1) the snow capping flux, which is used to specify surface mass balance to GLC; and (2) the ice runoff flux, sent to ROF. However, these two uses are not identical, which resulted in a number of problems. This tag separates these uses, so that qflx_snwcp_ice_col truly gives the snow capping flux, and the ice runoff flux is derived separately (via two new variables: qflx_ice_runoff_snwcp_col and qflx_ice_runoff_xs_col). This changes serves the following purposes: (1) QSNWCPICE becomes a useful history field, actually giving the amount of snow capping (2) Fixes snow balance check: bug 2238 (3) Fixes a problem where xs1 could initiate glacial inception, by contributing to qflx_glcice_frz (4) simplifies some logic in BalanceCheckMod Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): - 2238: Possible bug involving qflx_snwcp_ice in BalanceCheckMod.F90 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): QSNWCPICE_TO_COUPLER history field renamed to QRUNOFF_ICE_TO_COUPLER Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: Leo van Kampenhout CLM testing: NOTE: most testing done on r75327 (prior to the workaround for the intel compiler bug); after that, just ran: ERP_Lm40.1x1_numaIA.ICRUCLM50BGCCROP.yellowstone_gnu.clm-clm50BGCCROPmonthly (previously-failing test) ERI_D_Ld9_P24x1.f10_f10.ICRUCLM50BGC.hobart_nag.clm-reduceOutput (to run one nag test with the change) build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok 'ok' means tests pass, baseline comparisons fail as noted below Also BFAILs for: BFAIL SMS_Ld5_D_P24x1.f10_f10.IHISTCLM45BGC.hobart_nag.clm-clm50BGCmonthly.GC.1123-1949.45.n.nlcomp BFAIL ERP_Ly5.1x1_numaIA.ICRUCLM50BGCCROP.hobart_nag.clm-clm50BGCCROPmonthly.GC.1123-1949.45.n.nlcomp And for the newly-passing: ERP_Lm40.1x1_numaIA.ICRUCLM50BGCCROP.yellowstone_gnu.clm-clm50BGCCROPmonthly CLM tag used for the baseline comparisons: clm4_5_6_r155 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: diagnostic changes for all CLM45/CLM50; significant changes for IG compsets with CLM45/CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Changes for IG compsets are potentially climate changing Changes in non-IG compsets are in the following three fields: - QSNWCPICE - ERRH2OSNO - SNOW_SINKS (roundoff-level) Bigger changes in IG compsets are due to no longer including the xs1 term in qice (SMB) sent to GLC. Of the tests in the test suite, this was just observed for SMS_Lm37.T31_g37_gl5.IG1850CLM45.yellowstone_intel.clm-glcMEC_long. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/SoilHydrologyMod.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/main/lnd2atmType.F90 M components/clm/src/main/lnd2atmMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 - also add dtime multiplier for some log message output ========= Unrelated change: workaround for intel compiler bug M components/clm/src/biogeochem/CNFireLi2016Mod.F90 ========= Unrelated change: edit ChangeLog entry for clm4_5_6_r153 M ChangeLog M components/clm/doc/ChangeLog =============================================================== =============================================================== Tag name: clm4_5_6_r155 Originator(s): erik (Erik Kluzek) Date: etl xim 23 1:11:04 carra MST 2015 One-line Summary: Bring FUN (Fixation and Uptake of Nitrogen) code option to trunk from Mingjie, as well as latitude varying spinup from Charlie, and a fix for crop with dynamic roots from Beth Purpose of changes ------------------ FUN Model (Fixation and Uptake of Nitrogen): Mingjie Shi, Josh Fisher, and implmentation changes from Rosie Fisher FUN is the Fixation and Uptake of Nitrogen, and is designed to extract N from numerous sources (fixation, active uptake, retranslocation) in proportion to the inverse of the cost of uptake in term of C spent per unit N acquired. The FUN model developed by Fisher et al. 2010 and end Brzostek et al. 2014. Coded by Mingjie Shi 2015. Coding logic and structure altered by Rosie Fisher. October 2015. Critically, this removes the 'FUN-resistors' idea of Brzostek et al. 2014 and replaces it with uptake that is proportional to the N/C exchange rate. and adjusts the logic so that FUN does not depends upon the CLM4.0 'FPG' downregulation idea and instead it takes C spent on N uptake away from growth. The critical output so fthis code are sminn_to_plant_fun and npp_Nuptake, which are the N available to the plant for growth, and the C spent on obtaining it. When FUN is on and flex-CN is on, do NOT do the burn off of excess Carbon. Add in Kyla's rainfall trigger for stress decidious when FUN is on. Call FUN in the middle of SoilBiogeochemCompetition. Split CNPhenology into two sections for FUN. Get N available from N in NutrientCompetetion. Add annual average ET needed for CNFreeLivingFixation which is used in place of CNNFixation when FUN on. Dynamic Roots bug fix: Beth Drewniak Initialize huigrain to zero (rather than nan). And add a check for huigrain non zero to prevent a divide by zero. BGC Spinup Changes: Charlie Koven Add a latitude dependent factor for accellerated decomposition spinup that is about 1 at the equator goes to 26 at 60 degree's latitude (south and north) and then about 50 at the poles. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2237 2237 (dynamic root code divides by 0) Known bugs introduced in this tag (include bugzilla ID): 2249 2249 Simulations have low LAI in the amazon with FUN and Luna on Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): LAI is low in Amazon (see bug 2249) FUN requires use_nitrif_denitrif to be on. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): new CLM namelist variables Add br_root (maintenence respiration for roots) And leaf_mr_vcm (Scalar of leaf respiration to vcmax) And use_fun logical to turn the FUN model on Changes made to namelist defaults (e.g., changed parameter values): For CLM50 set leaf_mr_vcm to 0.032 (from Atkin 2015) For CLM50 with FUN on br_root=2.525e-06 Changes to the datasets (e.g., parameter, surface or initial files): New params files with FUN parameters on them Substantial timing or memory changes: Slower with FUN Will notices that simulations with FUN are slower (there are optimization issues that I saw that could be worked on) Code reviews and testing ------------------------ Code reviewed by: self, rfisher, koven, drewniak CLM testing: build-namelist tests: yellowstone - PASS unit-tests (components/clm/src): yellowstone - PASS regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK OK, means tests PASS or only expected fails do not. SSP comparision tests FAIL for CLM50, since spinup changes CLM tag used for the baseline comparisons: clm4_5_6_r154 Answer changes -------------- Changes answers relative to baseline: No Except when spinup on, or when FUN turned on. Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: =========== Remove flexibleCN test directory and add one for flexCN and FUN D components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN D components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN/include_user_mods List all files added and what they do: =========== Add new tests =========== NOTE: clm50* tests can be used with CLM45 compsets to create new CLM50 configurations A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCDVCROPmonthly/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCDVCROPmonthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCDVCROPmonthly A components/clm/cime_config/testdefs/testmods_dirs/clm/flexCN_FUN/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/flexCN_FUN/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/flexCN_FUN A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_flexCN_FUN/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_flexCN_FUN/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso_flexCN_FUN A components/clm/cime_config/testdefs/testmods_dirs/clm/noFUN_flexCN/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/noFUN_flexCN/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/noFUN_flexCN A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCCROPmonthly/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCCROPmonthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCCROPmonthly A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCmonthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCmonthly/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/clm50BGCmonthly =========== Add unit test for AC spinup A components/clm/src/soilbiogeochem/CMakeLists.txt A components/clm/src/soilbiogeochem/test/CMakeLists.txt A components/clm/src/soilbiogeochem/test/ACSpinup_test/test_acspinup.pf A components/clm/src/soilbiogeochem/test/ACSpinup_test/CMakeLists.txt A components/clm/src/soilbiogeochem/test/ACSpinup_test A components/clm/src/soilbiogeochem/test =========== New FUN module A components/clm/src/biogeochem/CNFUNMod.F90 List all existing files that have been modified, and describe the changes: M components/clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm - Remove fsurdat, new finidat file, turn: luna, flex-CN, dynroots, and FUN on, for interpolation fill missing with bare ground, limit_lake_evap to false set leaf_mr_vcm to 0.015 M components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/user_nl_clm - turn dynroots on M components/clm/bld/namelist_files/namelist_definition.xsl ------- Update to work again M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml -- New params file with FUN parameters c151024 and ED.c151023, turn dynroots off for bgc, add phys version to finidat, set use_fun br_root, leaf_mr_vcm M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml Add br_root, leaf_mr_vcm, use_fun M components/clm/bld/CLMBuildNamelist.pm -------------------------- Handle cnmresp_inparm, clm_canopy_inparm, send phys for finidat M components/clm/bld/unit_testers/build-namelist_test.pl ---------- Add tests for fun w/o CN, or w/o nitrif_denitrif, or flexCN w/o CN M components/clm/cime_config/config_compsets.xml - Change IGCRUCLM50BGCCROP to IG1850CRUCLM50BGCCROP add IG1850CRUCLM50BGC M components/clm/cime_config/testdefs/testlist_clm.xml ---- Add new tests (noFUN_flexCN, flexCN_FUN, ciso_flexCN_FUN, clm50KitchenSink w/o crop, clm50BGCDVCROPmonthly, transient clm50BGC, and new CROP-off clm50KitchenSink test add luna test on hobart, ICRUCLM50BGCCROP tests use clm50BGCCROPmonthly test-mods (with dynroots on, so change answers) M components/clm/cime_config/testdefs/testmods_dirs/clm/luna/user_nl_clm - Turn flexCN on M components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/user_nl_clm - Turn dynroot on (so these tests change answers) M components/clm/src/ED/fire/SFMainMod.F90 - Comment from Rosie M components/clm/src/biogeophys/WaterfluxType.F90 --- Add running annual mean of ET when FUN is on M components/clm/src/biogeophys/BalanceCheckMod.F90 - Increase tolerance for snow balance check (see bug 2238) M components/clm/src/biogeophys/LunaMod.F90 --------- Make 0.015 constant a parameter (later should use parameter in CanopyStateMod) M components/clm/src/biogeophys/PhotosynthesisMod.F90 - Use leaf_mr_vcm constant from canopystate_ins Remove (.not.use_cn) for recalculation of lmr25 when luna on, for natural C3 veg (see bug 2249) M components/clm/src/biogeophys/CanopyStateType.F90 - Make parameter for leaf_mr_vcm, add ReadNML sub M components/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 ---- Call FUN module M components/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 When FUN on, do not do precision control of low values for no3 and nh4 M components/clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 -- Increase transport by same factor as spinup acceleration by latitude M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 ----- Add soilc_change_col M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 ---- Take into account latitude varying spinup acceleration for exit and enter spinup states M components/clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 ---------- Add function for increase of spinup acceleration by latitude M components/clm/src/soilbiogeochem/SoilBiogeochemNStateUpdate1Mod.F90 --- For FUN N deposition and fixation put into NH4 use FUN version of sminn_to_plant_fun M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 --- Add several FUN nitrogen fixation terms: ffix_to_sminn_col, ffix_to_sminn_col, sminn_to_plant_fun_no3_vr_col, sminn_to_plant_fun_nh4_vr_col, sminn_to_plant_fun_vr_col M components/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 Changes from Charlie to make spinup acceleration higher as latitude increases. M components/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 -- decomp_npools_vr depends on spinup state and depends on latitude. M components/clm/src/CMakeLists.txt - Add soilbiogeochem directory for unit tests M components/clm/src/biogeochem/CNBalanceCheckMod.F90 --------- Add cbalance warning, for FUN add in free living fixation, decrease n-balance error from e-8 to e-6, add nbalance warn M components/clm/src/biogeochem/CNVegStateType.F90 ------------ Init huigrain to zero, add leafcn_offset and plantCN for FUN M components/clm/src/biogeochem/CNMRespMod.F90 ---------------- Add CNMRespReadNML, br_root M components/clm/src/biogeochem/CNVegNitrogenStateType.F90 ---- Add leafn_storage_xfer_acc and storage_ndemand for FUN, change two conditionals on spinup_state/restart_spinup_state to include 0 AND 1 MM components/clm/src/biogeochem/CNRootDynMod.F90 -------------- Only update root_depth if huigrain>0 M components/clm/src/biogeochem/CNPhenologyMod.F90 ------------ Add phase to CNPhenology (divides into two parts, needed for FUN), add Kyla's rainfall trigger (ONLY when FUN on) other changes when FUN on M components/clm/src/biogeochem/CNCStateUpdate1Mod.F90 -------- Add carbon spent on uptake respiration M components/clm/src/biogeochem/CNSharedParamsMod.F90 --------- Add use_fun M components/clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 If FUN on get N available from FUN turn off some sections when fun on M components/clm/src/biogeochem/CNFireLi2016Mod.F90 ----------- Changes for spinup from Charlie M components/clm/src/biogeochem/CNNDynamicsMod.F90 ------------ Add CNFreeLivingFixation subroutine set ndep_to_sminn to zero for FUN M components/clm/src/biogeochem/CNVegNitrogenFluxType.F90 ----- Add N demand and aquired arrays for FUN M components/clm/src/biogeochem/CNDriverMod.F90 --------------- If FUN on call CNFreeLivingFixation rather than CNNFixation. If FUN on split CNPhenology into two segments M components/clm/src/biogeochem/CNC14DecayMod.F90 ------------- Speed up radioactive decay by lat factor M components/clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 Changes for FUN. Burning off option not used for FUN, for FUN, growth respiration is not part of the allometry calculation M components/clm/src/biogeochem/CNVegCarbonFluxType.F90 ------- Carbon flux fields for FUN M components/clm/src/biogeochem/CNVegCarbonStateType.F90 ------ Add leafc_storage_xfer_acc and storage_cdemand for FUN, change some spinup state if statements from Charlie MM components/clm/src/main/clm_initializeMod.F90 - Call clm_instReadNML and waterflux_inst%initAccVar MM components/clm/src/main/clm_varcon.F90 -------- Add fun_period and small_value for FUN M components/clm/src/main/pftconMod.F90 --------- Add FUN pft-varying parameters M components/clm/src/main/clm_instMod.F90 ------- Add clm_instReadNML for namelists, calls canopystate_inst%ReadNML, add waterflux_inst%InitAccBuffer M components/clm/src/main/readParamsMod.F90 ----- Add readCNFUNParams MM components/clm/src/main/controlMod.F90 -------- Set use_fun and call CNMRespReadNML M components/clm/src/main/clm_driver.F90 -------- Call waterflux_inst%UpdateAccVars =============================================================== =============================================================== Tag name: clm4_5_6_r154 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Wed Nov 18 16:35:53 MST 2015 One-line Summary: one year grain product pool Purpose of changes ------------------ Introduce logic to have the harvested grain products decay over the course of one year. This is distinct from previous versions, in which the grain products decayed immediately. Including the one-year decay time changes soil carbon and nitrogen pools, and is on by default in the CLM5 crop model. (Danica Lombardozzi, Dave Lawrence) Add temporary test for all the clm50 functionality being used in the preliminary cesm2 coupled run: ERS_D_Ld3.f09_g16_gl5.IGCRUCLM50BGCCROP.yellowstone_intel.clm-clm50KitchenSink Note: Because of the brittle situation with the surface datasets/initial conditions files being inconsistent, we had to use the f09 grid for this test. The run length was kept short to keep the wall clock time ~30 min. This is too short exercise some new clm50 functionality, e.g. one year grain product pool. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): New namelist variables: use_fertilizer, use_grainproduct Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown, none expected Code reviews and testing ------------------------ Code reviewed by: Lombardozzi, Sacks CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 4.0 pass; 4.5 ok yellowstone_pgi - 4.0 pass; 4.5 ok yellowstone_gnu (clm45 only) - 4.5 ok hobart_nag - 4.5 ok 'ok' means: all tests passed functionality and baseline comparisons. only nlcomp failures. new test did not have a baseline. Note: Grain product pool is active by default for clm50 with crops, and will change answers. None of the tests are affected because they do not run long enough to trigger the new code. CLM tag used for the baseline comparisons: clm4_5_6_r153 Answer changes -------------- Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm50-crop - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): expected to be larger than roundoff, same climate. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: test mods for new test clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/clm50KitchenSink/include_user_mods List all existing files that have been modified, and describe the changes: one year grain product pool clm/src/biogeochem/CNPhenologyMod.F90 - fertilizer switch clm/src/biogeochem/CNVegNitrogenFluxType.F90 clm/src/biogeochem/CNDriverMod.F90 clm/src/biogeochem/CNVegCarbonFluxType.F90 clm/src/biogeochem/CNVegCarbonStateType.F90 clm/src/biogeochem/CNVegNitrogenStateType.F90 clm/src/biogeochem/CNWoodProductsMod.F90 new namelist/control variables for fertilizer and grain product pool clm/src/main/clm_varctl.F90 clm/src/main/controlMod.F90 clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/namelist_files/namelist_defaults_clm4_5.xml clm/bld/CLMBuildNamelist.pm new test clm/cime_config/config_compsets.xml clm/cime_config/testdefs/testlist_clm.xml =============================================================== =============================================================== Tag name: clm4_5_6_r153 Originator(s): Sean Swenson, Bill Sacks Date: Tue Nov 17 06:01:47 MST 2015 One-line Summary: Fix snow cover fraction bug Purpose of changes ------------------ The fix for infinite snow growth made in clm4_5_3_r144 turned out to have a bug of its own: It made snow cover fraction rarely reach ~ 1, instead maxing out around 0.8 or 0.9. This tag fixes the snow cover fraction bug while still (it appears) preventing infinite snow growth. This results in a potentially large change in model behavior, so the version number is bumped from clm4_5_5 to clm4_5_6. However, this is mainly due to backing out some of the (buggy) changes that were put in place in clm4_5_3_r144. That is, if the changes in this new tag were made on top of clm4_5_3_r143, the differences would likely be relatively small. Changes made by Sean Swenson, brought to trunk by Bill Sacks. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: Bill Sacks CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok 'ok' means: tests pass, but answers change for CLM45 / CLM50 cases CLM tag used for the baseline comparisons: clm4_5_5_r152 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45 / CLM50 cases - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate Main changes are due to backing out some of the (buggy) changes that were put in place in clm4_5_3_r144. That is, if the changes in this new tag were made on top of clm4_5_3_r143, the differences would likely be relatively small. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes. The change to CanopyHydrologyMod backs out a change that was made in r144. The change to SnowHydrologyMod is new. M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Unrelated change: get rid of a write statement that was resulting in lots of messages to the log file M components/clm/src/biogeophys/SoilHydrologyMod.F90 =============================================================== =============================================================== Tag name: clm4_5_5_r152 Originator(s): sacks (Bill Sacks) Date: Tue Nov 17 05:22:58 MST 2015 One-line Summary: Fix glc_mec energy conservation bug Purpose of changes ------------------ Fix an energy conservation bug that appears when running glc_mec (i.e., in runs with CISM). Energy loss due to ice melt was essentially being double-counted: First, all snow capping was being sent as ice runoff, and thus melted by the ocean; in addition, glacier ice melt removed energy from CLM. This removal of energy due to both accumulation (snow capping) and melt amounts to double-counting. This is fixed by: For each unit of ice melt, remove one unit of ice runoff (possibly resulting in negative ice runoff), adding one unit of liquid runoff. You can think of this as: each unit of ice melt indicates one unit of snow capping that should not have made it to the ocean, but instead melted en route. Note that we currently adjust qflx_snwcp_ice in order to achieve this ice runoff adjustment. This affects the QSNWCPICE history field. (In an upcoming tag, I plan to return qflx_snwcp_ice to its original values (i.e., truly indicating snow capping), by introducing a new field that explicitly gives ice runoff.) This tag also fixes a minor bug in the calculation of negative runoff for glc_mec columns with glc_dyn_runoff_routing=T. In addition, it reorders some code in HydrologyDrainageMod in order to remove some complexity. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - pass 'ok' means: pass except for baseline failures in IG tests (IG compsets plus ICLM45GLCMEC) Also, unexpected baseline failures in the following tests; however, these same baseline failures were observed when comparing a fresh checkout of r151 against the r150 baselines (this new branch is bit-for-bit with a fresh checkout of r151 for these tests): NCK_Ld5.f10_f10.I.yellowstone_pgi.clm-40default RMS x2l_Flrr_volr 8.8324E-18 NORMALIZED 2.2700E-15 RMS x2l_Flrr_volrmch 8.8324E-18 NORMALIZED 2.2700E-15 SMS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-default SMS_Ld5.f19_g16.IRCP45CLM45BGC.yellowstone_pgi.clm-decStart FILLDIFF topo NCK_Ld1.f10_f10.ICRUCLM45.yellowstone_intel.clm-default RMS VOLR 8.9604E-05 NORMALIZED 1.2335E-14 RMS VOLRMCH 8.9604E-05 NORMALIZED 1.2335E-14 RMS x2l_Flrr_volr 2.8268E-17 NORMALIZED 6.8542E-15 RMS x2l_Flrr_volrmch 2.8268E-17 NORMALIZED 6.8542E-15 CLM tag used for the baseline comparisons: clm4_5_4_r151 (baseline directory named clm4_5_4_r150) Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All IG compsets (compsets with CISM / glc_mec) - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Likely new climate For each unit of glacier melt, there is an extra unit of liquid runoff, and one unit less of ice runoff. This results in changes in the following CLM history fields: ERRH2O, ERRH2OSNO, QRGWL, QRUNOFF, QRUNOFF_TO_COUPLER, QSNWCPICE, QSNWCPICE_TO_COUPLER, TWS, VOLR, VOLRMCH And the following changes in l2x cpl hist fields: l2x_Flrl_rofdto, l2x_Flrl_rofgwl, l2x_Flrl_rofi There are also changes in the consequent x2r, r2x and x2l cpl hist fields. I did one test run to confirm that the differences in (l2x_Flrl_rofdto + l2x_Flrl_rofgwl) exactly balanced the differences in l2x_Flrl_rofi. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/HydrologyDrainageMod.F90 ========= Unrelated change: just change a comment M components/clm/src/main/clm_instMod.F90 =============================================================== =============================================================== Tag name: clm4_5_4_r151 Originator(s): erik (Erik Kluzek) Date: Fri Nov 13 15:50:58 MST 2015 One-line Summary: Fix broken r150 tag, with bgc_spinup issue in it Purpose of changes ------------------ Backout changes to -bgc_spinup in CLM build-namelist should NOT have been committed. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2245 2245 setup doesn't work, because of issue iwth -bgc_spinup Notes of particular relevance for users None --------------------------------------- Code reviews and testing ------------------------ Code reviewed by: self CLM testing: None! (but this is the same as what was actually tested for clm4_5_4_r150) CLM tag used for the baseline comparisons: clm4_5_4_r150 Answer changes -------------- Changes answers relative to baseline: No Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: >>>>>>> Remove non-working on-accell option to -bgc_spinup to namelist M components/clm/bld/CLMBuildNamelist.pm M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml =============================================================== =============================================================== Tag name: clm4_5_4_r150 Originator(s): erik (Erik Kluzek) Date: Fri Nov 13 13:42:39 MST 2015 One-line Summary: Some needed answer changing fixes for: fire, disallow negative photosynthis, and reduce LAI of DBT in evergreen state Purpose of changes ------------------ Add in changes to fire fluxes for deadstem (both clm4_5 and clm5_0). Don't allow photosynthis to be less than zero. Reduce the Leaf Area Index (LAI) of Deciduous Broadleaf Tree (DBT) when in an evergreen state (warm enough that all of the leaves never all fall off) by taking into account leaf longevity and only allowing the amount of leaf or fine root Carbon storage transfered to be more than the amount of remaining leaf or fine root Carbon (would be zero unless in an evergreen state). Previously, DBT in evergreen state would have unrealistically high LAI, that would then cause them to take up too much water and then kill the trees. So they had an unrealistic cycle for trees that are supposed to be in an evergreen state. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2242, 2243, and 2244 2244 Change to reduce leaf area of DBT in evergreen state (rfisher) 2243 Deadstem not taken into account in fire-fluxes for clm4_5 (koven) 2242 Disallow negative photosynthis (olesen and bonan) Known bugs introduced in this tag (include bugzilla ID): 2245 2245 setup doesn't work, because of issue iwth -bgc_spinup Notes of particular relevance for users None --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): None Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): None Changes made to namelist defaults (e.g., changed parameter values): None Changes to the datasets (e.g., parameter, surface or initial files): None Substantial timing or memory changes: None Code reviews and testing ------------------------ Code reviewed by: self, koven, rfisher, oleson CLM testing: build-namelist tests: yellowstone - yes unit-tests (components/clm/src): yellowstone - yes tools-tests (components/clm/test/tools): yellowstone - no PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - no regular tests (aux_clm40, aux_clm45): yellowstone_intel - yes yellowstone_pgi - yes yellowstone_gnu (clm45 only) - yes hobart_nag - yes CLM tag used for the baseline comparisons: clm4_5_3_r149 Answer changes -------------- Changes answers relative to baseline: Summarize any changes to answers, i.e., - what code configurations: clm4_5 and clm5_0 - what platforms/compilers: all - nature of change: similar climate Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M components/clm/src/biogeochem/CNFireLi2014Mod.F90 --- Deal with deadstem for fire fluxes M components/clm/src/biogeochem/CNFireLi2016Mod.F90 --- Deal with deadstem for fire fluxes M components/clm/src/biogeochem/CNPhenologyMod.F90 ---- In CNStressDecidPhenology take into account leaf longevity in lgsf, lower leaf and froot Carbon storage to transfer, by the leaf and froot Carbon M components/clm/src/biogeophys/PhotosynthesisMod.F90 - Make ag max of zero or calculated value =============================================================== =============================================================== Tag name: clm4_5_3_r149 Originator(s): sacks (Bill Sacks) Date: Sun Nov 8 12:41:59 MST 2015 One-line Summary: Change cold start initialization of snow pack Purpose of changes ------------------ Change cold start initialization of snow pack to lead to fewer instances where cold start conditions negatively impact the simulation for long periods. Main change is to make initial snow pack over glacier and high-latitude vegetation much smaller. Also make istice consistent with istice_mec. New method developed in conjunction with Jan Lenaerts, Leo van Kampenhout, and Dave Lawrence. Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): none Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone - NOT RUN unit-tests (components/clm/src): yellowstone - OK tools-tests (components/clm/test/tools): yellowstone - NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel - OK yellowstone_pgi - OK yellowstone_gnu (clm45 only) - OK hobart_nag - OK CLM tag used for the baseline comparisons: clm4_5_3_r148 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45/CLM5 cold start runs - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; not investigated carefully If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/main/clm_instMod.F90 ========= Add a comment in changelog entry for clm4_5_3_r148 M ChangeLog M components/clm/doc/ChangeLog ========= Minor tweak to template M .ChangeLog_template =============================================================== =============================================================== Tag name: clm4_5_3_r148 Originator(s): sacks (Bill Sacks) Date: Sat Nov 7 06:41:25 MST 2015 One-line Summary: For CLM5, increase max snow depth to 20 m Purpose of changes ------------------ Change CLM5 snow defaults: - h2osno_max = 20m rather than 1m - nlevsno = 12 rather than 5 - glc_snow_persistence_max_days = 0 rather than 7300 (this is just used for modeling glacial inception when coupled to CISM) Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): For CLM5 cases, need to run initial conditions through init_interp, because of the change in nlevsno. See caveats under clm4_5_3_r146 before doing so. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): - h2osno_max = 20m rather than 1m - nlevsno = 12 rather than 5 - glc_snow_persistence_max_days = 0 rather than 7300 (this is just used for modeling glacial inception when coupled to CISM) Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: This change will lead to some memory increase, and some increase in timing, due to having more snow layers. I have not investigated this closely. The tputcomp tests all passed, but these are not very meaningful - in part because there likely weren't any grid cells with substantially more than 5 layers in these test runs. Code reviews and testing ------------------------ Code reviewed by: self CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: NOT RUN tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r147 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM5 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff, likely new climate in some regions If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: Some runs documented here: https://docs.google.com/document/d/1Y5kCwa0GUnQ0hy7vNwvtFOKaC3ikysfLXdcNVqRygok/edit URL for LMWG diagnostics output used to validate new climate: http://www.staff.science.uu.nl/~kampe004/lnd_diag/tr_eraimod_5_cap-tr_eraimod_12_cap/setsIndex.html - note that this uses 10m snow pack rather than 20m Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Change namelist defaults, as listed above M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Add documentation M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml =============================================================== =============================================================== Tag name: clm4_5_3_r147 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Nov 5 21:51:53 MST 2015 One-line Summary: swenson bedrock and coupler runoff bugfixes Purpose of changes ------------------ Science bugfixs from Sean Swenson for bedrock and coupler runoff fields. The bugfix for bedrock corrects an issue where running with debug flags and bedrock on would generate floating point errors or unreproducible behavior when run with a surface data set that contained bedrock data. coupler runoff fields bugfix: splitting of runoff for mosart did not contain all surface and subsurface fields. surface runoff is the sum of qflx_over, qflx_h2osfc_surf, and qflx_irrig. subsurface runoff is the sum of qflx_drain and qflx_perched_drain Add cropColdStart test mods directory missing from r145. Bugs fixed or introduced ------------------------ Bugs fixed (include bugzilla ID): 2241 Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): use_bedrock is still not true by default because we do not have a fully consistent set of surface data sets and initial conditions for testing. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): none Substantial timing or memory changes: unknown, none expected Code reviews and testing ------------------------ Code reviewed by: Swenson, Andre CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40 pass; 45 ok yellowstone_pgi - 40 pass; 45 ok yellowstone_gnu (clm45 only) - 45 ok hobart_nag - ok 'ok' means - all clm45/50 compsets have only the 'baseline compare' answer changing failures. Fields described below. test BFAIL ERP_D_P15x2_Ld3.f10_f10.ICRUCLM50BGCCROP.yellowstone_intel.clm-cropColdStart because baseline is not in r146. CLM tag used for the baseline comparisons: clm4_5_3_r146 Answer changes -------------- Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm45/clm50 - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff All clm45 and clm50 compsets will have answer changes in the clm and coupler history files because of changes to the runoff passed from clm to the coupler. Only the following fields have changed: clm: TWS, VOLRMCH, VOLR cpl: l2x_Flrl_rofgwl, x2r_Flrl_rofsub, x2r_Flrl_rofsur, r2x_Flrr_volrmch, r2x_Flrr_volr, r2x_Forr_rofl, x2l_Flrr_volrmch, x2l_Flrr_volr, l2x_Flrl_rofsub, l2x_Flrl_rofsur, x2r_Flrl_rofgwl If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: Add missing cropColdStart test mods clm/cime_config/testdefs/testmods_dirs/clm/cropColdStart/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/cropColdStart/include_user_mods clm/cime_config/testdefs/testmods_dirs/clm/cropColdStart List all existing files that have been modified, and describe the changes: Bedrock bug fix, initialize and size arrays correctly. clm/src/biogeophys/SoilWaterMovementMod.F90 Remove expected failure from missing test mods: clm/cime_config/testdefs/ExpectedTestFails.xml Coupler runoff bugfix: clm/src/cpl/lnd_import_export.F90 clm/src/main/lnd2atmType.F90 clm/src/main/clm_driver.F90 clm/src/main/lnd2atmMod.F90 =============================================================== =============================================================== Tag name: clm4_5_3_r146 Originator(s): sacks (Bill Sacks) Date: Wed Nov 4 06:48:41 MST 2015 One-line Summary: Implement vertical interpolation in init_interp Purpose of changes ------------------ The major requirements this addresses are: (1) Handle changes in the vertical discretization of the soil column, including changing the number of layers and/or changing the depth of layers. (2) Handle changes in which soil layers are considered bedrock vs. "standard" soil: Information from bedrock layers should not be transmitted to soil layers, and vice versa. (3) Handle changing number of snow layers There was previously-existing code to handle (1). However, while it was mostly right, it was hard-coded to just handle the interpolation from "standard" vertical levels to "more_vertlevels". And there were a number of special cases that it mis-handled (such as urban points, which have a different vertical discretization; also, spvals were not handled correctly). It was also going to be hard to fit requirements #2 and #3 into the old code. So I removed the old code and replaced it with a new, object-oriented design. Because the number of files involved in init_interp has now grown significantly, I have moved this code to a new directory (src/init_interp). As part of the handling of vertical interpolation, I needed to fix a few inconsistencies in the use of spval in some multi-level fields (see below for details). Finally, to aid with testing of this new code, and because it is useful in general, Sean Swenson added nbedrock to the first history file of a run. Bugs fixed or introduced ------------------------ Known bugs introduced in this tag (include bugzilla ID): [If none, remove this line] - 2239: need new initial conditions files with extra metadata on them Notes of particular relevance for users --------------------------------------- Caveats for users (e.g., need to run initial conditions through init_interp): - If you want to use init_interp on an initial conditions file generated prior to this tag, you will need to first get some new metadata onto your initial conditions file: COL_Z, LEVGRND_CLASS, COL_Z_p and LEVGRND_CLASS_p. You have two options for how to do this: (1) Run the model for a short time using this tag, with a configuration that matches the source file, using the source file as finidat (with use_init_interp = .false.), in order to produce a new restart file with the necessary metadata. Then use that new file as the finidat file for init_interp. (2) If that is not possible, then an alternative is to run the model for a short time using this tag, with cold start initial conditions (finidat = " "). Then use a tool like ncks to copy the misssing fields onto the original source finidat file. Then use that patched file as the finidat file for init_interp. CAUTION: A bug in the netcdf library means that adding fields using ncks can result in data corruption of existing fields. I have observed this in the plant_nalloc field, which fortunately does not seem to be needed at all on the restart file. However, if you use this method, you should compare the old and new files to ensure that nothing other than plant_nalloc has been changed. Changes to CLM's user interface (e.g., new/renamed XML or namelist variables): none Changes made to namelist defaults (e.g., changed parameter values): none Changes to the datasets (e.g., parameter, surface or initial files): Some new initial conditions files: same as old, but with metadata needed for running them through init_interp Substantial timing or memory changes: none Code reviews and testing ------------------------ Code reviewed by: self; CLM-CMT plans to do a more detailed review later CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: ok tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r145 Answer changes -------------- Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 / CLM50 runs that use init_interp to interpolate to a different resolution - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; not investigated closely Changes arise from the fact that multi-level points are now interpolated rather than copied from the input to the output. This changes results for urban points, where the vertical level structure of walls and rooves varies from gridcell to gridcell. There are also FILLDIFFs in SOILLIQ and SOILICE in vector (dov2xy = .false.) history files, due to a bug fix in the treatment of spval for these variables. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Detailed list of changes ------------------------ List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Moved to new directory for init_interp code D components/clm/src/main/initInterpMindist.F90 D components/clm/src/main/initInterp.F90 D components/clm/src/main/test/initInterpMindist_test/CMakeLists.txt D components/clm/src/main/test/initInterpMindist_test/test_init_interp_mindist.pf D components/clm/src/main/test/initInterpMindist_test List all files added and what they do: ========= Added new directory for init_interp code, populated with files that used to be in main A components/clm/src/init_interp A components/clm/src/init_interp/initInterpMindist.F90 A components/clm/src/init_interp/CMakeLists.txt A components/clm/src/init_interp/test/initInterpMindist_test/CMakeLists.txt A components/clm/src/init_interp/test/initInterpMindist_test/test_init_interp_mindist.pf A components/clm/src/init_interp/test/initInterpMindist_test A components/clm/src/init_interp/test ========= Main changes needed for vertical interpolation A components/clm/src/init_interp/initInterpMultilevelBase.F90 A components/clm/src/init_interp/initInterpMultilevelCopy.F90 A components/clm/src/init_interp/initInterpMultilevelInterp.F90 A components/clm/src/init_interp/initInterpMultilevelSnow.F90 A components/clm/src/init_interp/initInterpMultilevelSplit.F90 A components/clm/src/init_interp/initInterpMultilevelContainer.F90 A components/clm/src/init_interp/initInterp.F90 A components/clm/src/init_interp/initInterp2dvar.F90.in A components/clm/src/init_interp/initInterpBounds.F90 ========= Wrappers to transpose and pack that handle allocation for you (needed for pgi) A components/clm/src/utils/array_utils.F90.in ========= Put unit tests around the parts of the vertical interpolation that have significant logic A components/clm/src/init_interp/test/CMakeLists.txt A components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_split.pf A components/clm/src/init_interp/test/initInterpMultilevel_test/initInterpMultilevelMock.pf A components/clm/src/init_interp/test/initInterpMultilevel_test/multilevel_interp_factory.F90 A components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_interp.pf A components/clm/src/init_interp/test/initInterpMultilevel_test/test_init_interp_multilevel_snow.pf A components/clm/src/init_interp/test/initInterpMultilevel_test/CMakeLists.txt A components/clm/src/init_interp/test/initInterpMultilevel_test ========= Allow vrtlay tests to use init_interp A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp/README A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay_interp ========= Add a CLM5 crop test that uses init_interp (init_interp is needed for CLM5 cases with non-cold-start initial conditions) A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/README A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonth_interp List all existing files that have been modified, and describe the changes: ========= Added new directory for init_interp code M components/clm/bld/configure M components/clm/src/CMakeLists.txt M components/clm/src/main/test/CMakeLists.txt ========= Added new metadata to restart file that is needed for the vertical interpolation M components/clm/src/main/initVerticalMod.F90 M components/clm/src/main/subgridRestMod.F90 M components/clm/src/main/ColumnType.F90 M components/clm/src/main/column_varcon.F90 M components/clm/src/main/filterMod.F90 M components/clm/src/utils/restUtilMod.F90.in ========= New routines used to add metadata to restart file and to read variables in init_interp M components/clm/src/main/ncdio_pio.F90.in ========= Remove unnecessary use of spval (the vertical interpolation was not handling this use of spval correctly, because it is inconsistent with other uses of spval in the code; it turns out this use of spval wasn't necessary to begin with) M components/clm/src/biogeochem/ch4Mod.F90 M components/clm/src/soilbiogeochem/SoilBiogeochemNitrifDenitrifMod.F90 ========= Fix bug: calculations were being done on spval, leading to values of 1e75 on the restart file (this was causing problems with the new vertical interpolation) M components/clm/src/biogeophys/WaterStateType.F90 ========= Added new necessary metadata to any initial conditions file that is used for init_interp testing. I have NOT taken the time to do this for other out-of-the-box initial conditions, figuring that they'll be replaced soon anyway. (Any file generated with the new tag will have the necassary metadata on it.) M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm ========= Tangentially-related change: add nbedrock to first history file (changes from Sean Swenson) M components/clm/src/main/GridcellType.F90 M components/clm/src/main/histFileMod.F90 ========= Totally unrelated change: add a comment M components/clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 ========= Added another system test of init_interp, and modify snowlayer tests and some vrtlay tests to use init_interp M components/clm/cime_config/testdefs/testlist_clm.xml M components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_3_monthly/user_nl_clm M components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12/user_nl_clm ========= Put unit tests around the parts of the vertical interpolation that have significant logic M components/clm/src/utils/CMakeLists.txt M components/clm/src/main/CMakeLists.txt ========= Remove old failing vrtlay tests from xfail list, add new vrtlay_interp tests to xfail list (fail due to bug 2239); add two new CLM5 crop tests to xfail list (bugs 2240 and 2241) M components/clm/cime_config/testdefs/ExpectedTestFails.xml ========= New ChangeLog template, with sections removed / added / rearranged M .ChangeLog_template =============================================================== =============================================================== Tag name: clm4_5_3_r145 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue Nov 3 14:25:30 MST 2015 One-line Summary: update clm50 defaults, enable clm50 crop compsets Purpose of changes: CLM 5.0 defaults update from David Lawrence and Erik Kluzek. - New namelist defaults for clm50: - use_bedrock = false - lower_boundray_condition = 2 - limit_lake_evap_irrig = true - use_dynroot = false - Remove QICE from default history output in use cases. - Some history output variables are now inactive by default: CONC_CH4_SAT, CONC_CH4_UNSAT, FPI, N tendency due to vertical transport, vertically resolved N. - Some history output is now active by default: O2 decomp depth unsat, SMINN. - Update cime configuration to allow creation of CLM 5.0 crop compsets. - Update clm4.0 for pio2 from Jim Edwards. Test list update from Ben Andre - Add tests: short CLM 5.0 crop test, renamed aux_clm_short test suite to clm_short_45, create clm_short_50 test suite. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): cime issue #238 (https://github.com/CESM-Development/cime/issues/238) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: Update namelist defaults for clm50 compsets - use_bedrock = false - lower_boundray_condition = 2 - limit_lake_evap_irrig = true - use_dynroot = false List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Erik Kluzek, Ben Andre List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: clm/bld/CLMBuildNamelist.pm - add_default() call for use_bedrock clm/bld/namelist_files/namelist_defaults_clm4_5.xml - new clm50 defaults remove QICE from output clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml clm/bld/namelist_files/use_cases/glacierMEC_pd.xml clm/cime_config/testdefs/testlist_clm.xml - new clm50 crop test, clm_short testlist update clm/cime_config/config_compsets.xml - clm50 crop compset clm/cime_config/config_component.xml - clm50 crop compset clm/src_clm40/main/ncdio_pio.F90.in - pio2 update change default histroy status for variables listed above. clm/src/biogeochem/ch4Mod.F90 clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40, pass; 45, ok yellowstone_pgi - 40, pass; 45, ok yellowstone_gnu (clm45 only) - 45, ok hobart_nag - not finished prior to tag. no problems expected based on prior branch testing. 'ok' means: clm45 tests passed, including baseline comparisons. clm50 test functionality passed, baseline comparisons failed as expected because of changes to namelist defaults. nlcomp will fail for clm50 tests. CLM tag used for the baseline comparisons: clm4_5_3_r144 Changes answers relative to baseline: some Summarize any changes to answers, i.e., - what code configurations: all clm50 compsets - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r144 Originator(s): Sean Swenson, Keith Oleson, Bill Sacks Date: Tue Oct 27 09:47:58 MDT 2015 One-line Summary: Better fix for bug that leads to infinite growth in snow depth Purpose of changes: Back out fix for infinite snow growth that was put in in clm4_5_3_r141. Put in place new fix from Sean Swenson and Keith Oleson that fixes the problem more fundamentally, rather than putting in place a band-aid fix that had its own problems. Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Sean Swenson, Keith Oleson, Bill Sacks, Leo van Kampenhout List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: ok tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r143 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL CLM45 & CLM50 - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but likely to be larger than roundoff/same climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_3_r143 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Oct 22 23:39:50 MDT 2015 One-line Summary: flexibleCN and luna bugfixes. Purpose of changes: flexibleCN and LUNA bugfix changes from Bardan Ghimire, Rosie Fisher and Chonggang Xu. Changes to FlexCN to burn off excess C. Corrected the cpool problem not going to 0 by changing the order of processes. Turn off LUNA for crops and put a boundary cut-off on lnc for very high lnc fed to LUNA Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): none http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new namelist carbon_resp_opt List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: unknown, none expected Code reviewed by: Bardan Ghimire, Chonggang Xu, Rosie Fischer, Dave Lawrence List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: new namelist option carbon_resp_opt clm/bld/CLMBuildNamelist.pm clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/namelist_files/namelist_defaults_clm4_5.xml - update defaults for carbon_storage_excess_opt clm/src/main/controlMod.F90 clm/src/main/clm_varctl.F90 flexibleCN clm/src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 - change order of processes clm/src/biogeochem/CNVegCarbonFluxType.F90 - carbon_resp_opt clm/src/biogeochem/CNCStateUpdate1Mod.F90 - carbon_resp_opt clm/src/biogeochem/NutrientCompetitionMethodMod.F90 - update interface clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 - update interface clm/src/biogeochem/CNDriverMod.F90 - update interface LUNA clm/src/biogeophys/LunaMod.F90 - leaf nitrogen cutoff clm/src/biogeophys/PhotosynthesisMod.F90 - turnoff luna for crops whitespace / remove dead code clm/src/biogeochem/CNPhenologyMod.F90 CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pfunit tests pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass, flexibleCN not expected to be bit for bit with baseline yellowstone_pgi - pass yellowstone_gnu (clm45 only) - pass hobart_nag - not finished at time of tag, expected to be ok based on prior branch testing. CLM tag used for the baseline comparisons: clm4_5_3_r142 Changes answers relative to baseline: some Summarize any changes to answers, i.e., - what code configurations: clm50 with flexibleCN and clm50 with LUNA and crop on - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r142 Originator(s): sacks (Bill Sacks) Date: Mon Oct 19 19:57:35 MDT 2015 One-line Summary: Fix bug in initialization of 12-layer snow pack Purpose of changes: Expand the dzmax arrays to have 12 elements, with the last element being 'huge'. This fixes a bug in the initialization of the snow pack when you try to initialize it to need all 12 layers. And, more generally, this helps avoid special-case code that would otherwise be needed to handle the edge case at the bottom layer. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Leo van Kampenhout List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SnowHydrologyMod.F90 CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: ok tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r141 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_3_r141 Originator(s): Keith Oleson, Bill Sacks Date: Mon Oct 19 14:15:06 MDT 2015 One-line Summary: Fix bug that leads to infinite growth in snow depth Purpose of changes: Restore 'qflx_snow_grnd_col(c) > 0.0_r8' condition in check for whether snow pack should be initialized. This was removed in clm4_5_1_r121, because Leo found that, in some rare cases, with this condition in place, the snow pack wasn't being initialized when it should have been. However, we have more recently found that, without this condition, there are some rare cases when snow depth can grow infinitely - and Keith Oleson traced this bug to the removal of this condition. We haven't managed to trace the entire logic here, but for now we feel it's safest to restore this condition. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Keith Oleson, Bill Sacks List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/CanopyHydrologyMod.F90 CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: NOT RUN tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r140 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL CLM45 & CLM50 - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but likely to be larger than roundoff/same climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_3_r140 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Oct 15 23:45:49 MDT 2015 One-line Summary: mosart, direct to ocean runoff Purpose of changes: Coordinated changes to clm, mosart and rtm to handle direct to ocean runoff. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): 2233 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: unknown, none expected Code reviewed by: Sean Swenson List any svn externals directories updated (cime, rtm, cism, etc.): cime, rtm, mosart List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: SVN_EXTERNAL_DIRECTORIES - update externals clm/cime_config/config_compsets.xml - new compset IMCRUCLM50BGC for testing mosart clm/cime_config/testdefs/ExpectedTestFails.xml - known issue in cime causes test failure clm/cime_config/testdefs/testlist_clm.xml - add IM compset test for mosart add coupler fields for direct ocean runoff clm/src_clm40/main/clm_cpl_indices.F90 clm/src_clm40/main/lnd_import_export.F90 clm/src/biogeophys/WaterfluxType.F90 clm/src/biogeophys/HydrologyDrainageMod.F90 clm/src/cpl/clm_cpl_indices.F90 clm/src/cpl/lnd_import_export.F90 clm/src/main/lnd2atmMod.F90 clm/src/main/lnd2atmType.F90 CLM testing: regular build-namelist tests: yellowstone - n/a, not run unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - generate tests not finished at time of tag, but expected to be ok based on earlier testing. Notes: - ok means that all functionality tests pass. bit for bit comparison with baselines only show differences in the following runoff fields when rtm or mosart is running: clm - TWS, VOLR, VOLRMCH coupler - l2x_Flrl_rofgwl, x2l_Flrr_volr, x2l_Flrr_volrmch, r2x_Forr_rofl, r2x_Flrr_volr, r2x_Flrr_volrmch, x2r_Flrl_rofgwl - bfail for new mosart test - new expected failure for ERI_N2_Ld9.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default because of cime issue. CLM tag used for the baseline comparisons: clm4_5_3_r139 Changes answers relative to baseline: some Summarize any changes to answers, i.e., - what code configurations: all with rtm or mosart - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r139 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue Oct 13 10:17:43 MDT 2015 One-line Summary: dynamic roots from Beth Drewniak Purpose of changes: Introduce dynamic roots from Beth Drewniak. Allows roots to grow dynamically, optimizing for water and nutrient uptake. Crop roots start from zero each season. Generally leads to higher root densities near the surface, but also deeper maximum root depth. Overall, water stress is reduced. If use_bedrock = true, roots are confined exist in soil layers only. Root litter is input into the soil according to the dynamic root distribution. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): none http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new variable use_dynroot List any changes to the defaults for the boundary datasets: updated pft parameter file to c151012, includes merge of dynamic roots and flexCN params. NOTE: ED parameter file was not update. Describe any substantial timing or memory changes: unknown, none expected Code reviewed by: Dave Lawrence List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: clm/src/biogeochem/CNRootDynMod.F90 - implement dynamic roots List all existing files that have been modified, and describe the changes: new namelist variable use_dynroots and defaults clm/bld/CLMBuildNamelist.pm clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/namelist_files/namelist_defaults_clm4_5.xml clm/src/biogeochem/CNDriverMod.F90 - call dynamic roots clm/src/main/controlMod.F90 - read use_dynroot namelist var clm/src/main/clm_varctl.F90 - add use_dynroot switch clm/src/main/pftconMod.F90 - new dynamic root variable clm/src/biogeophys/SoilStateType.F90 - add root_depth_patch variable CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pfunit tests pass tools-tests (components/clm/test/tools): yellowstone - n/a, not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a, not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - a couple of manual tests compiled and ran. full suite not finished at time of tag. NOTE: ok = amelist changes expected for clm45/50. New defaults namelist for clm50 turns on dynamic roots, expect all clm50 compsets to be answer changing. CLM tag used for the baseline comparisons: clm4_5_3_r138 Changes answers relative to baseline: some Summarize any changes to answers, i.e., - what code configurations: clm50 with use_dynroot = true (default) - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r138 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Sun Oct 11 13:23:12 MDT 2015 One-line Summary: new accelerated spinup and soil vertical structure Purpose of changes: Two sets of changes are included in this tag: Accelerated spinup New accelerated spinup mode introduced from Dan Ricciuto. Acceleration factors are calculated directly from the base decomposition rates for each pool and CWD is added to the pools that are accelerated. Wood mortality is also accelerated by a factor of 10 through increasing rates of gap mortality and fire. There are now three accelerated spinup modes Spinup_state = 0 ; no accelerated spinup Spinup_state = 1 ; standard accelerated spinup Spinup_state = 2 ; accelerated accelerated spinup Set through env_run.xml (-bgc_spinup on) and via namelist. In tests, new spinup method accelerates spinup by ~ a factor of 2 for a case starting from pre-spunup conditions but shifting to another forcing dataset. Vertical soil structure The vertical soil structure has been cleaned up and a new vertical soil structure has been added. The namelist options more_vertlayers and deep_soilcolumn have been removed and replaced with the more general soil_layerstruct. Four options for soil_layerstruct are included. 10SL_3.5m - original soil structure from CLM4.5 and prior versions 20SL_8.5m - proposed new soil structure for CLM5 23SL_3.5m - replaces more_vertlayers 49SL_10m - replace deep_soilcolumn Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): 2232 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: remove more_vertlayers, deep_soilcolumn are removed, add soil_layerstruct. new default soil for clm50 compsets List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: improved spinup, none expected for regular runs. Code reviewed by: dlawren List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: remove deep_soilcolumn and more_vertlayers namelist, add soil_layerstruct, update spinup_state: clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/namelist_files/namelist_defaults_clm4_5.xml clm/bld/CLMBuildNamelist.pm switch tests to use new namelist variable: clm/cime_config/testdefs/testmods_dirs/clm/vrtlay/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/user_nl_clm clm/cime_config/testdefs/testmods_dirs/clm/deepsoil_bedrock/user_nl_clm add failing tests clm/cime_config/testdefs/ExpectedTestFails.xml switch soil layer structure from old to new namelist variable clm/src/main/clm_varpar.F90 - switch to new namelist variables for soil vertiacal structure clm/src/main/controlMod.F90 - update namelist variables clm/src/main/clm_varctl.F90 - add soil_layerstruct clm/src/main/initVerticalMod.F90 - switch to new namelist variables for soil vertiacal structure clm/src/biogeophys/SoilHydrologyInitTimeConstMod.F90 - switch to soil_layerstruct clm/src/biogeophys/SoilStateInitTimeConstMod.F90 - switch to soil_layerstruct clm/src/biogeophys/IrrigationMod.F90 clm/src/biogeophys/SoilHydrologyMod.F90 - remove comment new spinup mode: clm/src/biogeochem/CNGapMortalityMod.F90 clm/src/biogeochem/CNVegStructUpdateMod.F90 clm/src/biogeochem/CNVegCarbonStateType.F90 clm/src/biogeochem/CNFireLi2016Mod.F90 clm/src/biogeochem/CNVegStateType.F90 clm/src/biogeochem/CNVegNitrogenStateType.F90 clm/src/biogeochem/CNC14DecayMod.F90 clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeConType.F90 clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 CLM testing: build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - n/a - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - ok yellowstone_pgi - ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok all clm50 compsets fail bit for bit with baseline because of new default soil_layer structure. spinup tests fail because of new spinup code. two new expected fails for clm45 compsets, see bugzilla 2232 CLM tag used for the baseline comparisons: clm4_5_3_r137 Changes answers relative to baseline: some Summarize any changes to answers, i.e., - what code configurations: clm50, spinup - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate clm50 compsets use a new soil layer structure. new accelerated spinup If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r137 Originator(s): erik (Erik Kluzek) Date: Fri Oct 9 12:07:35 MDT 2015 One-line Summary: Bring in changes for fire-emissions as well tuning changes to the fire model to be used with clm5_0 Purpose of changes: Move fire_clm4_5_2_r128_tags/fire06_clm4_5_3_r136 branch originated from Francis Vitt and Fang Li to trunk. This does two things: sends fire-emissions from land to coupler (then to ATM), and has an updated (retuned) version of CNFire from Fang Li. The retuned version is tuned for running with CAM, but there are fire parameter changes that allow it to be tuned with CRU forcing. CLM can test fire-emissions with the -fire_emis option to CLM_BLDNML_OPTS. Coupled configurations with CAM-CHEM will turn on fire-emissions when fire_emis_nl namelist added to drv_flds_in coupler namelist file. Requirements for tag: fire-emissions added and tested, clm4_5 identical, clm5_0 uses newly tuned fire-model Test level of tag: regular Bugs fixed (include bugzilla ID): Bug 2231 – Issue with PTCLM with changes in cime4.0.1 Bug 2229 – clm scripts not compatible with cime > cime2.x.y Bug 2220 – Bug in fire model with ED for RH24 and WIND24 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Add ability to use and generate fire-emission namelist: fire_emis_nl Add cnfire_inparm namelist for clm4_5 with fire_method to either li2014qianfrc (clm4_5 default), or li2016crufrc (clm5_0 default) Add -fire_emis option to CLM build-namelist Add shr_fire_emis_factors_file file to use when fire-emissions are sent to coupler List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, fang-li, fvitt List any svn externals directories updated (cime, rtm, cism, etc.): PTCLM to PTCLM2_151009 List all files eliminated: D components/clm/src/biogeochem/CNFireMod.F90 --- clm4_5 fire model was split into polymorphic versions that include clm4_5 and clm5_0 default versions List all files added and what they do: A components/clm/bld/namelist_files/namelist_defaults_fire_emis.xml -- fire emissions defaults to use when -fire_emis CLM_BLDNML_OPTS option is used ------- Add some clm4_0 specific testmods A components/clm/cime_config/testdefs/testmods_dirs/clm/fire_emis/user_nl_clm ------- Set fire-emission history fields, include fire_emis_elevated=.true. A components/clm/cime_config/testdefs/testmods_dirs/clm/fire_emis/include_user_mods - use default A components/clm/cime_config/testdefs/testmods_dirs/clm/fire_emis A components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN/include_user_mods ---- use default A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/shell_commands - turn -fire_emis A components/clm/cime_config/testdefs/testmods_dirs/clm/40default/user_nl_clm ---- Same as default A components/clm/cime_config/testdefs/testmods_dirs/clm/40default/shell_commands - Same as default, but no -fire_emis A components/clm/cime_config/testdefs/testmods_dirs/clm/40default A components/clm/cime_config/testdefs/testmods_dirs/clm/reduceOutput/shell_commands - -no-fire_emis ------- Fire model polymorphic options ------- CNFireFactory chooses which version, CNFireMethod is abstract class, CNFireBase is base class ------- CNFireLi2014 extends the base class for clm4_5 default version CNFireLi2016 extends base class ------- for clm5_0 default version A components/clm/src/biogeochem/CNFireFactoryMod.F90 --- Factory and namelist to figure out which version to implement. A components/clm/src/biogeochem/CNFireMethodMod.F90 ---- Abstract class to define all interfaces A components/clm/src/biogeochem/CNFireBaseMod.F90 ------ Base class, does initialization and handles lightning and pop-dens datasets A components/clm/src/biogeochem/CNFireLi2014Mod.F90 ---- Implements CNFireArea and CNFireFluxes for the clm4_5 default version A components/clm/src/biogeochem/CNFireLi2016Mod.F90 ---- Implements CNFireArea and CNFireFluxes for the clm5_0 default version, tuned with CAM5 (parameter changes for version tuned to CRUNCEP) ------- Fire emission factors A components/clm/src/biogeochem/FireEmisFactorsMod.F90 -- Fire emission factors from new NetCDF dataset A components/clm/src/biogeochem/CNFireEmissionsMod.F90 -- Gathers carbon emissions from fire sources to be sent to CAM-Chem via the coupler List all existing files that have been modified, and describe the changes: ------- Get working with cime4.0.1 M components/clm/test/tools/TCBCFGtools.sh --- change location of configure and soft link Macros.make M components/clm/test/tools/TCBtools.sh ------ change location of configure and soft link Macros.make ------- Update sample namelist (used for tools tests) M components/clm/tools/clm4_5/mksurfdata_map/mksurfdata_map.namelist ------- M components/clm/bld/unit_testers/build-namelist_test.pl ----------- Add fire_emis tests M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Add default fire_method options M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add cnfire_inparm, and fire_emis_nl namelist groups. Adding: fire_method, fire_emis_elevated, fire_emis_factors_file, and fire_emis_specifier M components/clm/bld/CLMBuildNamelist.pm --------------------------- Add cnfire_inparm namelist and handle fire_emis option simialr to megan and drydep options. Add: fire_emis_elevated, fire_emis_factors_file, and fire_emis_specifier ------- M components/clm/cime_config/testdefs/testmods_dirs/clm/crop/include_user_mods - use 40defaults M components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN/user_nl_clm - use defaults, so just set use_flexibleCN M components/clm/cime_config/testdefs/testmods_dirs/clm/pts/include_user_mods -- use 40defaults M components/clm/cime_config/testdefs/testmods_dirs/clm/default/shell_commands - Add -fire_emis M components/clm/cime_config/testdefs/testlist_clm.xml ------- Change clm4_0 tests to use 40default and add fire_emis test M components/clm/cime_config/user_nl_clm ---- Add mention of -fire_emis option ------- M components/clm/src/dyn_subgrid/dynSubgridControlMod.F90 - NFileName to NLFilename, and read_namelist M components/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 -- Pass NLFilename down to dynSubgrid_init M components/clm/src/cpl/clm_cpl_indices.F90 -------------- Add fire-fluxes, and top of fire emissions M components/clm/src/cpl/lnd_import_export.F90 ------------ Set fire-fluxes and top of fire emission M components/clm/src/biogeochem/VOCEmissionMod.F90 -------- Add an additional coefficient multiply (currently just set to 1.0) M components/clm/src/biogeochem/CNDriverMod.F90 ----------- Pass NLFilename and cnfire_method down to CNDriverInit and cnfire_method to CNDriverNoLeaching, and use for CNFireArea and CNFireFluxes calls M components/clm/src/biogeochem/CNVegCarbonStateType.F90 -- Correct fire variable descriptions (fuelc_col, fuelc_crop_col) MM components/clm/src/main/clm_initializeMod.F90 ----------- Add create_cnfire_method, pass NLFilename down to CNDriverInit, ndep_init, and dynSubgrid_init M components/clm/src/main/lnd2atmMod.F90 ------------------ Add use shr_fire_emis_mod to handle fire emissions to send from land to ATM M components/clm/src/main/ndepStreamMod.F90 --------------- NLFilename passed down rather than using controlMod M components/clm/src/main/lnd2atmType.F90 ----------------- Add wild-fire emissions flux and vertical distribution M components/clm/src/main/atm2lndType.F90 ----------------- Add RH30, PREC10, and PREC60 ( 30 day relative humidity, 10 and 60 day precipitation) Fix bug 2220 for ED M components/clm/src/main/clm_instMod.F90 ----------------- Add cnfire and fireemis objects call initialize for fireemis MM components/clm/src/main/controlMod.F90 ------------------ Call CNFireReadNML M components/clm/src/main/clm_driver.F90 ------------------ Replace cnfire_method object for calls to CNFire module, add call to CNFireEmisUpdate CLM testing: regular build-namelist tests: yellowstone yes unit-tests (components/clm/src): yellowstone yes tools-tests (components/clm/test/tools): yellowstone yes PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone yes regular tests (aux_clm40, aux_clm45): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes hobart_nag yes CLM tag used for the baseline comparisons: clm4_5_3_r137 Changes answers relative to baseline: Yes, for clm5_0 Summarize any changes to answers, i.e., - what code configurations: clm5_0 - what platforms/compilers: All - nature of change (fire tuned differently): =============================================================== =============================================================== Tag name: clm4_5_3_r136 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Oct 8 10:15:33 MDT 2015 One-line Summary: coupler field prep for mosart and externals update Purpose of changes: Changes from T. Craig, M. Vertenstein and S. Swenson to enable separation of runoff fields passed from CLM to RTM Update externals. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): 2223 Known bugs (include bugzilla ID): 2229 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: unknown, none expected Code reviewed by: mvertens, swenson List any svn externals directories updated (cime, rtm, cism, etc.): cime, cism, rtm List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: Split river fluxes sent to the coupler into surface, frozen, subsurface and gwl (glacier/wetland/lake). clm/src_clm40/main/lnd_import_export.F90 clm/src_clm40/main/clm_cpl_indices.F90 clm/src_clm40/main/clm_atmlnd.F90 clm/src/main/lnd2atmMod.F90 clm/src/main/atm2lndType.F90 clm/src/main/lnd2atmType.F90 clm/src/cpl/lnd_import_export.F90 clm/src/cpl/clm_cpl_indices.F90 CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pfunit tests pass tools-tests (components/clm/test/tools): yellowstone - mksurfdata_map unit tests pass tools system tests not cime compatible - bug 2229 PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): NOTE: the filenames for some tests exceeded the allowed length and comparison against the original r136 baselines may show bfail for: ERP_D_P15x2_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_intel.clm-limit_river_withdraw ERP_P15x2_Lm36.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrig_o3_reduceOutput yellowstone_intel - 40 ok, 45 ok yellowstone_pgi - 40 ok, 45 ok yellowstone_gnu (clm45 only) - ok hobart_nag - ok CLM tag used for the baseline comparisons: clm4_5_3_r135 Changes answers relative to baseline: yes Summarize any changes to answers, i.e., - what code configurations: * cism update is answer changing for all IG compsets. see cism for details * splitting river fluxes to the coupler is answer changing for all clm45/50 compsets. - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): much larger than roundoff, assumed new climate. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: no runs were performed URL for LMWG diagnostics output used to validate new climate: no diagnostics created. =============================================================== =============================================================== Tag name: clm4_5_3_r135 Originator(s): sacks (Bill Sacks) Date: Tue Oct 6 05:59:45 MDT 2015 One-line Summary: Substantial reduction in memory allocation for prognostic crops Purpose of changes: With the introduction of "global" crops (78 pfts), memory use was increased substantially. The memory use itself was not a problem, but this led to a significant model slow-down, presumably due to a decrease in cache use efficiency. Much of this increase in memory use was unnecessary: CLM only knows about a minority of the new crop functional types (CFTs) - other CFTs are merged into those known CFTs. Yet space was being set aside in the patch and column-level arrays even for the CFTs that are never used. In this tag, I have reduced prognostic crop memory use by only allocating space in patch and column-level arrays for the crops CLM knows about. This leads to a speedup of prognostic crop cases by about 1.5x. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: New initial conditions files for crops, needed because of the change in number of patches and columns. These were generated by running the old initial conditions through initInterp. Describe any substantial timing or memory changes: Speedup of prognostic crop cases by about 1.5x. Code reviewed by: Erik Kluzek, Ben Andre, Mariana Vertenstein List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes, as described above M components/clm/src/main/subgridMod.F90 M components/clm/src/main/pftconMod.F90 M components/clm/src/main/initGridCellsMod.F90 M components/clm/src/main/clm_varpar.F90 ========= Replace old conditional with use of new is_pft_known_to_model, which is a more general/robust way to accomplish the conditional M components/clm/src/biogeochem/CNPhenologyMod.F90 ========= New initial conditions files for crop M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Remove unneeded 'use' statements M components/clm/src/main/restFileMod.F90 M components/clm/src/main/surfrdMod.F90 CLM testing: build-namelist tests: yellowstone: NOT RUN unit-tests (components/clm/src): yellowstone: ok tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r134 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_3_r134 Originator(s): sacks (Bill Sacks) Date: Sun Oct 4 07:25:05 MDT 2015 One-line Summary: Fixes for transient and preindustrial crop cases Purpose of changes: (1) Important bug fix for transient crop: Collapse crop types in the transient run. Previously, when the transient portion started, (a) the collapse of the 78 pfts to the "known" crops was not happening, and (b) the collapse of irrigated pfts in a non-irrigation run was not happening. These two problems are fixed now. (2) Restore lost transient crop test cases (3) Add 1850 crop raw dataset (created using https://github.com/billsacks/clm_misc-make_crop_1850_dataset) (4) Add out-of-the-box surface datasets for 1850 crop, and out-of-the-box landuse timeseries datasets for transient crop Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2010 (create year-1850 surface datasets for crop) - 2149 (tweak smallville transient test case) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: new out-of-the-box defaults (see below) Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Remove testmods directories that are no longer needed because transient crop cases now pick up a surface dataset and landuse timeseries file out-of-the-box D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/README D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/README D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10 ========= Remove testmods directory that is no longer used (should have been removed in clm4_5_2_r127) D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/README D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/user_nl_clm D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/include_user_mods D components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09 List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Collapse crop types in a transient run M components/clm/src/main/surfrdUtilsMod.F90 M components/clm/src/main/surfrdMod.F90 M components/clm/src/dyn_subgrid/dyncropFileMod.F90 ========= Relax tolerance for check of PCT_NAT_PFT between landuse timeseries file and surface dataset. Transient crop runs exceeded the previous strict tolerance. I'm not sure why they need this relaxed tolerance, but given the manipulations that were needed both to create the year-1850 crop raw dataset and to create the crop landuse timeseries file, I'm willing to accept that this is needed. I have made this tolerance about 2 orders of magnitude looser than the differences I observed, to try to prevent the check from being wrongly triggered in the future. The tolerance of 1e-6 still seems sufficient to catch any real problems. M components/clm/src/dyn_subgrid/dynpftFileMod.F90 ========= Add rules to make some 1850 crop surface datasets M components/clm/tools/clm4_5/mksurfdata_map/Makefile.data ========= Add raw dataset for 1850 crop M components/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml ========= Add some surface datasets for 1850 crop, and landuse timeseries files for crop M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Just add a comment M components/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl ========= Restore transient crop tests that were lost shortly after they were added (when the test list was migrated out of scripts into CLM) M components/clm/cime_config/testdefs/testlist_clm.xml ========= Unrelated change: add a note at Ben's request M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/README M components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/README CLM testing: build-namelist tests: yellowstone: ok unit-tests (components/clm/src): yellowstone: ok tools-tests (components/clm/test/tools): yellowstone: NOT RUN PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone: NOT RUN regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r132 Changes answers relative to baseline: YES - just for transient crop cases Summarize any changes to answers, i.e., - what code configurations: transient crop cases - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): not investigated, likely to be new climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A Note that there were no transient crop tests in the test suite prior to this tag, so the answer changes do not affect any existing tests. Note the following unexpected BFAILs: BFAIL ERP_D_P15x2_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_intel.clm-limit_river_withdraw.clm2.h0.nc : baseline compare clm2.h0 (baseline: compare .base file with clm4_5_3_r132 file) (baseline history file does not exist) BFAIL ERP_D_P15x2_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_intel.clm-limit_river_withdraw.clm2.h1.nc : baseline compare clm2.h1 (baseline: compare .base file with clm4_5_3_r132 file) (baseline history file does not exist) BFAIL ERP_P15x2_Lm36.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrig_o3_reduceOutput.clm2.h1.nc : baseline compare clm2.h1 (baseline: compare .base file with clm4_5_3_r132 file) (baseline history file does not exist) =============================================================== =============================================================== Tag name: clm4_5_3_r133 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Oct 2 16:24:53 MDT 2015 One-line Summary: Bugfix for mksurfdata_map - add file missed by clm4_5_2_r125. Purpose of changes: Bugfix for mksurfdata_map. Add new file needed for soil depth. File should have been included in clm4_5_2_r125. Requirements for tag: tools Test level of tag: mksurfdata_map unit tests Bugs fixed (include bugzilla ID): not in bugzilla Known bugs (include bugzilla ID): systems tests for clm tools not cime3 compatible. http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: clm/tools/clm4_5/mksurfdata_map/src/mksoildepthMod.F90 - include soil depth data List all existing files that have been modified, and describe the changes: none CLM testing: tools build-namelist tests: not run unit-tests (components/clm/src): not run tools-tests (components/clm/test/tools): yellowstone - clm4_5/mksurfdata_map/unit_testers - all tests passed PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone not run regular tests (aux_clm40, aux_clm45): not run. no changes to main clm code. CLM tag used for the baseline comparisons: none. Changes answers relative to baseline: no =============================================================== =============================================================== Tag name: clm4_5_3_r132 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Oct 2 10:29:57 MDT 2015 One-line Summary: Modify soil BGC diffusion and cryoturbation code to allow for spatially variable soil thickness. Purpose of changes: Changes from Dave Lawrence to modify soil BGC diffusion and cryoturbation code to allow for spatially variable soil thickness (e.g. if use_bedrock = true); fixed dzsoi_decomp for new vertical grids Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): 2213, 2219 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: unknown none expected Code reviewed by: Dave Lawrence List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: clm/src/main/initVerticalMod.F90 - use zmin_bedrock param, fix dzsoi_decomp clm/src/main/clm_varcon.F90 - make zmin_bedrock a constant parameter clm/src/main/clm_varpar.F90 - write subsurface discretization levels to land log clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 - spatially variable soil thickness clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 - spatially variable soil thickness CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40, pass; 45, ok (see below) yellowstone_pgi - 40, pass; 45, pass yellowstone_gnu (clm45 only) - pass hobart_nag - pass Only test not bit for bit with baseline is: ERS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_intel.clm-deepsoil_bedrock CLM tag used for the baseline comparisons: clm4_5_3_r131 Changes answers relative to baseline: some Tag changes answers for clm50 bgc compsets where use_bedrock = true. No change to default version of model, doesn't change results for use_bedrock = true when in SP mode. Summarize any changes to answers, i.e., - what code configurations: clm50 bgc with use_bedrock = true - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r131 Originator(s): mvertens, sacks Date: Wed Sep 30 05:13:45 MDT 2015 One-line Summary: Bring in cime3 Purpose of changes: Bring in cime3 (de-CESMization of CIME), which requires the addition of a cime_config directory and a bit of other rework. Also, port the pnetcdf problem with dov2xy=false from clm45 to clm4 code base (workaround for bug 1730 - and 2214 and 2222). This was needed because now pnetcdf is on by default in most runs. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2214: CLM4 test fails in PIO while trying to write h1 file - 2222: CLM4 cases die when writing 1-d history files with pnetcdf Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: rework for cime3 Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): cime2.0.18-p1.1 -> cime3.0.7 cism2_1_02 -> cism2_1_05 rtm1_0_52 -> rtm1_0_53 List all files eliminated: ========= Moved into cime_config D components/clm/cimetest/testlist_clm.xml D components/clm/cimetest/ExpectedTestFails.xml D components/clm/cimetest/testmods_dirs/clm/glcMEC_long/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/glcMEC_long/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC_long/README D components/clm/cimetest/testmods_dirs/clm/glcMEC_long D components/clm/cimetest/testmods_dirs/clm/decStart/include_user_mods D components/clm/cimetest/testmods_dirs/clm/decStart/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/decStart D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetMid/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetMid/README D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetMid/include_user_mods D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetMid D components/clm/cimetest/testmods_dirs/clm/irrigOn_reduceOutput/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/irrigOn_reduceOutput/include_user_mods D components/clm/cimetest/testmods_dirs/clm/irrigOn_reduceOutput/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/irrigOn_reduceOutput D components/clm/cimetest/testmods_dirs/clm/default/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/default/shell_commands D components/clm/cimetest/testmods_dirs/clm/default D components/clm/cimetest/testmods_dirs/clm/crop_trans_sville/README D components/clm/cimetest/testmods_dirs/clm/crop_trans_sville/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/crop_trans_sville/include_user_mods D components/clm/cimetest/testmods_dirs/clm/crop_trans_sville D components/clm/cimetest/testmods_dirs/clm/edTest/shell_commands D components/clm/cimetest/testmods_dirs/clm/edTest/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/edTest D components/clm/cimetest/testmods_dirs/clm/reduceOutput/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/reduceOutput D components/clm/cimetest/testmods_dirs/clm/cropMonthOutput/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/cropMonthOutput/include_user_mods D components/clm/cimetest/testmods_dirs/clm/cropMonthOutput D components/clm/cimetest/testmods_dirs/clm/monthly/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/monthly/include_user_mods D components/clm/cimetest/testmods_dirs/clm/monthly/user_nl_cpl D components/clm/cimetest/testmods_dirs/clm/monthly D components/clm/cimetest/testmods_dirs/clm/glcMEC/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/glcMEC/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC/user_nl_cpl D components/clm/cimetest/testmods_dirs/clm/glcMEC D components/clm/cimetest/testmods_dirs/clm/ciso/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/ciso/include_user_mods D components/clm/cimetest/testmods_dirs/clm/ciso D components/clm/cimetest/testmods_dirs/clm/luna/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/luna/include_user_mods D components/clm/cimetest/testmods_dirs/clm/luna D components/clm/cimetest/testmods_dirs/clm/monthly_noinitial/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/monthly_noinitial/include_user_mods D components/clm/cimetest/testmods_dirs/clm/monthly_noinitial D components/clm/cimetest/testmods_dirs/clm/interp_from_f09/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/interp_from_f09/include_user_mods D components/clm/cimetest/testmods_dirs/clm/interp_from_f09/README D components/clm/cimetest/testmods_dirs/clm/interp_from_f09 D components/clm/cimetest/testmods_dirs/clm/glcMEC_increase/user_nl_cism D components/clm/cimetest/testmods_dirs/clm/glcMEC_increase/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC_increase/README D components/clm/cimetest/testmods_dirs/clm/glcMEC_increase D components/clm/cimetest/testmods_dirs/clm/crop/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/crop/include_user_mods D components/clm/cimetest/testmods_dirs/clm/crop D components/clm/cimetest/testmods_dirs/clm/allActive/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/allActive D components/clm/cimetest/testmods_dirs/clm/cn_conly/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/cn_conly/include_user_mods D components/clm/cimetest/testmods_dirs/clm/cn_conly D components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/shell_commands D components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/README D components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way D components/clm/cimetest/testmods_dirs/clm/snowlayers_12/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/snowlayers_12/include_user_mods D components/clm/cimetest/testmods_dirs/clm/snowlayers_12 D components/clm/cimetest/testmods_dirs/clm/ptsROA/include_user_mods D components/clm/cimetest/testmods_dirs/clm/ptsROA/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/ptsROA D components/clm/cimetest/testmods_dirs/clm/NoVSNoNI/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/NoVSNoNI/include_user_mods D components/clm/cimetest/testmods_dirs/clm/NoVSNoNI D components/clm/cimetest/testmods_dirs/clm/SNICARFRC/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/SNICARFRC/include_user_mods D components/clm/cimetest/testmods_dirs/clm/SNICARFRC D components/clm/cimetest/testmods_dirs/clm/flexibleCN/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/flexibleCN D components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags/README D components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/include_user_mods D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/README D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop D components/clm/cimetest/testmods_dirs/clm/crop_trans_f10/README D components/clm/cimetest/testmods_dirs/clm/crop_trans_f10/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/crop_trans_f10/include_user_mods D components/clm/cimetest/testmods_dirs/clm/crop_trans_f10 D components/clm/cimetest/testmods_dirs/clm/o3/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/o3/include_user_mods D components/clm/cimetest/testmods_dirs/clm/o3 D components/clm/cimetest/testmods_dirs/clm/vrtlay/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/vrtlay/include_user_mods D components/clm/cimetest/testmods_dirs/clm/vrtlay D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetEarly/include_user_mods D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetEarly/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetEarly/README D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetEarly D components/clm/cimetest/testmods_dirs/clm/oldhyd/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/oldhyd D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/README D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/include_user_mods D components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop D components/clm/cimetest/testmods_dirs/clm/irrig_o3_reduceOutput/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/irrig_o3_reduceOutput/include_user_mods D components/clm/cimetest/testmods_dirs/clm/irrig_o3_reduceOutput/README D components/clm/cimetest/testmods_dirs/clm/irrig_o3_reduceOutput D components/clm/cimetest/testmods_dirs/clm/glcMEC_decrease/user_nl_cism D components/clm/cimetest/testmods_dirs/clm/glcMEC_decrease/include_user_mods D components/clm/cimetest/testmods_dirs/clm/glcMEC_decrease/README D components/clm/cimetest/testmods_dirs/clm/glcMEC_decrease D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subset/README D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subset/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subset D components/clm/cimetest/testmods_dirs/clm/USUMB/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/USUMB/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/USUMB D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetLate/include_user_mods D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetLate/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetLate/README D components/clm/cimetest/testmods_dirs/clm/tropicAtl_subsetLate D components/clm/cimetest/testmods_dirs/clm/drydepnomegan/include_user_mods D components/clm/cimetest/testmods_dirs/clm/drydepnomegan/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/drydepnomegan D components/clm/cimetest/testmods_dirs/clm/rootlit/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/rootlit/include_user_mods D components/clm/cimetest/testmods_dirs/clm/rootlit D components/clm/cimetest/testmods_dirs/clm/deepsoil_bedrock/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/deepsoil_bedrock/include_user_mods D components/clm/cimetest/testmods_dirs/clm/deepsoil_bedrock D components/clm/cimetest/testmods_dirs/clm/af_bias_v5/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/af_bias_v5/user_nl_datm D components/clm/cimetest/testmods_dirs/clm/af_bias_v5/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/af_bias_v5 D components/clm/cimetest/testmods_dirs/clm/limit_river_withdraw/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/limit_river_withdraw/include_user_mods D components/clm/cimetest/testmods_dirs/clm/limit_river_withdraw D components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly/include_user_mods D components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly D components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/README D components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/user_nl_clm D components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/include_user_mods D components/clm/cimetest/testmods_dirs/clm/interp_f19_crop D components/clm/cimetest/testmods_dirs/clm/ptsRLA/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/ptsRLA/include_user_mods D components/clm/cimetest/testmods_dirs/clm/ptsRLA D components/clm/cimetest/testmods_dirs/clm/pts/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/pts/README D components/clm/cimetest/testmods_dirs/clm/pts/include_user_mods D components/clm/cimetest/testmods_dirs/clm/pts D components/clm/cimetest/testmods_dirs/clm/ptsRLB/xmlchange_cmnds D components/clm/cimetest/testmods_dirs/clm/ptsRLB/include_user_mods D components/clm/cimetest/testmods_dirs/clm/ptsRLB D components/clm/cimetest/testmods_dirs/clm D components/clm/cimetest/testmods_dirs D components/clm/cimetest D components/clm/bld/clm.buildlib D components/clm/bld/user_nl_clm D components/clm/bld/clm.buildnml ========= This is no longer needed - it is generated automatically by the build script D components/clm/src_clm40/main/ncdio_pio.F90 List all files added and what they do: ========= New stuff needed for cime3 A components/clm/cime_config A components/clm/cime_config/config_component.xml A components/clm/cime_config/config_pes.xml A components/clm/cime_config/config_compsets.xml ========= Moved from elsewhere (new organization needed for cime3) A components/clm/cime_config/testdefs/ExpectedTestFails.xml A components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/USUMB A components/clm/cime_config/testdefs/testmods_dirs/clm/rootlit/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/rootlit/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/rootlit A components/clm/cime_config/testdefs/testmods_dirs/clm/drydepnomegan/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/drydepnomegan/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/drydepnomegan A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate/README A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetLate A components/clm/cime_config/testdefs/testmods_dirs/clm/deepsoil_bedrock/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/deepsoil_bedrock/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/deepsoil_bedrock A components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/limit_river_withdraw A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5/user_nl_datm A components/clm/cime_config/testdefs/testmods_dirs/clm/af_bias_v5 A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_3_monthly/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_3_monthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_3_monthly A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/README A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_crop A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLA A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsRLB A components/clm/cime_config/testdefs/testmods_dirs/clm/pts/README A components/clm/cime_config/testdefs/testmods_dirs/clm/pts/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/pts/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/pts A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/decStart A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_long/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_long/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_long/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_long A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid/README A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetMid A components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/irrigOn_reduceOutput A components/clm/cime_config/testdefs/testmods_dirs/clm/default/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/default/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/default A components/clm/cime_config/testdefs/testmods_dirs/clm/reduceOutput/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/reduceOutput A components/clm/cime_config/testdefs/testmods_dirs/clm/edTest/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/edTest/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/edTest A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/README A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_sville A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC/user_nl_cpl A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly/user_nl_cpl A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/cropMonthOutput A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ciso A components/clm/cime_config/testdefs/testmods_dirs/clm/luna/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/luna/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/luna A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly_noinitial/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly_noinitial/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/monthly_noinitial A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/README A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_from_f09 A components/clm/cime_config/testdefs/testmods_dirs/clm/crop/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/crop/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/crop A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase/user_nl_cism A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_increase A components/clm/cime_config/testdefs/testmods_dirs/clm/cn_conly/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/cn_conly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/cn_conly A components/clm/cime_config/testdefs/testmods_dirs/clm/allActive/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/allActive A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/shell_commands A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_spunup_1way A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/snowlayers_12 A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/ptsROA A components/clm/cime_config/testdefs/testmods_dirs/clm/NoVSNoNI/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/NoVSNoNI/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/NoVSNoNI A components/clm/cime_config/testdefs/testmods_dirs/clm/SNICARFRC/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/SNICARFRC/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/SNICARFRC A components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/flexibleCN A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_changeFlags A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/README A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/README A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/crop_trans_f10 A components/clm/cime_config/testdefs/testmods_dirs/clm/o3/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/o3/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/o3 A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/vrtlay A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/xmlchange_cmnds A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/README A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subsetEarly A components/clm/cime_config/testdefs/testmods_dirs/clm/oldhyd/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/oldhyd A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/README A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/interp_f19_noncrop2crop A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease/README A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease/user_nl_cism A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/glcMEC_decrease A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_o3_reduceOutput/README A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_o3_reduceOutput/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_o3_reduceOutput/include_user_mods A components/clm/cime_config/testdefs/testmods_dirs/clm/irrig_o3_reduceOutput A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/README A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset/user_nl_clm A components/clm/cime_config/testdefs/testmods_dirs/clm/tropicAtl_subset A components/clm/cime_config/testdefs/testmods_dirs/clm A components/clm/cime_config/testdefs/testmods_dirs A components/clm/cime_config/testdefs/testlist_clm.xml A components/clm/cime_config/testdefs A components/clm/cime_config/user_nl_clm A components/clm/cime_config/buildlib A components/clm/cime_config/buildnml List all existing files that have been modified, and describe the changes: ========= Minor changes for cime3 M components/clm/bld/CLMBuildNamelist.pm ========= Port bug 1730 workaround to clm40, fixing bugs 2214 & 2222 M components/clm/src_clm40/main/histFileMod.F90 M components/clm/src_clm40/main/ncdio_pio.F90.in CLM testing: build-namelist tests: yellowstone: ok (ran unit_testers and test_build_namelist tests) unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_3_r130 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): There are large differences in the anomaly forcing, due to a bug fix that has been incorporated in the cime3 changes. This changes answers for this test: SMS_Ld1.f09_g16.ICRUCLM45.yellowstone_intel.clm-af_bias_v5 In addition, there are roundoff-level changes in x2l_Flrr_volr, VOLR and TWS in these tests: ERS_E_Ld9.f45_g37.I.yellowstone_intel.clm-default ERI_D_Ld9.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay Also, this BFAILed: BFAIL ERP_P15x2_Lm36.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrig_o3_reduceOutput.clm2.h1.nc : baseline compare clm2.h1 (baseline: compare .base file with clm4_5_3_r130 file) (baseline history file does not exist) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_3_r130 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Mon Sep 28 11:16:54 MDT 2015 One-line Summary: add namelist option to reduce lake evaporation and irrigation when river water storage falls below a threshold value, and add an additional constraint on stress deciduous leaf onset. Purpose of changes: To maintain river storage above a minimum threshold. To add a second condition that must be fulfilled prior to leaf onset occurring for stress deciduous pfts. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): 2213, 2219 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new namelist flags: limit_lake_evap_and_irrig constrain_stress_deciduous_onset List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: unknown, none expected Code reviewed by: Sean Swenson, Dave Lawrence, Bill Sacks (irrigation) List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none clm/cimetest/testmods_dirs/clm/limit_river_withdraw - testmods for new flag List all existing files that have been modified, and describe the changes: clm/src/biogeochem/CNDriverMod.F90 - add atm2lnd_inst to function call for phenology clm/src/biogeochem/CNPhenologyMod.F90 - add atm2lnd_inst, new check for stress deciduous clm/src/biogeochem/CNSharedParamsMod.F90 - new stress deciduous namelist variable clm/src/main/init_hydrology.F90 - note clm/src/main/controlMod.F90 - add limit_lake_evap_and_irrig namelist variable clm/src/main/clm_driver.F90 - add volr_grc to ApplyIrrigation function call clm/src/main/clm_varctl.F90 - add limit_lake_evap_and_irrig namelist variable clm/src/biogeophys/IrrigationMod.F90 - check for limit irrigation clm/src/biogeophys/LakeFluxesMod.F90 - check for limit lake evap clm/src/biogeophys/test/Irrigation_test/test_irrigation_singlepatch.pf - new unit test clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 - new unit test clm/src/unit_test_stubs/utils/spmdMod_stub.F90 - add stub mpi type clm/bld/namelist_files/namelist_defaults_clm4_5.xml - new namelist variables clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/CLMBuildNamelist.pm clm/cimetest/testlist_clm.xml - new tests CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - pass yellowstone_pgi - pass yellowstone_gnu (clm45 only) pass hobart_nag - run, ok CLM tag used for the baseline comparisons: clm4_5_2_r128/clm4_5_3_r129 Note: new test for limit_river_withdraw got dropped from the testlist during a merge/manage_testlist update. This wasn't noticed until after baselines were generated. This test was added to the baselines manually. Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all clm50 compsets - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): When active, these changes will be climate changing. Lake evaporation and irrigation limitation will change the surface energy balance by reducing latent heat and increasing sensible heat when river storage is below a threshold value. Constraining leaf onset with a secondary constraint will modify vegetation dynamics, e.g. LAI will differ in magnitude and seasonality, which will change the surface energy balance by altering evapotranspiration. Note: all of the aux_clm tests for clm50 compsets triggered were bit for bit with the baseline. The stress deciduous flag is on by default for clm50, and changes should have been observed. Sean Swenson verified that the flag is active in 30 day runs, so it is assumed that the shorter three and nine day runs were not enough to trigger the new code. A new 30 day test was added. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: none URL for LMWG diagnostics output used to validate new climate: none =============================================================== =============================================================== Tag name: clm4_5_3_r129 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri Sep 25 15:30:39 MDT 2015 One-line Summary: bump tag revision number because clm4_5_2_r128 was climate changing. Purpose of changes: Bump tag revision number only. This is the same code base as clm4_5_2_r128. No additional testing was done. Further review of the answer changes in clm4_5_2_r128 indicated that they were a climate changing bugfix, and the patch level version number should have been bumped. S. Swenson and D. Lawrence performed a code review of the changes and decided the original calculation of heat capacity for soil layers > nlevsoi (i.e. bedrock layers) was incorrect due to the heat capacity being scaled by (1-watsat). This scaling, which takes into account the presence of soil voids (the porosity), is not appropriate for the bedrock layers, which are assumed to be rock rather than soil. The new code simply applies the bedrock value for heat capacity without the scaling, which leads to an increase in heat content for bedrock layers. The code change therefore results in non-b4b changes to the simulation. =============================================================== =============================================================== Tag name: clm4_5_2_r128 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Sep 24 10:34:39 MDT 2015 One-line Summary: science bugfixes for root fraction and bedrock thermal properties. not bit for bit with baseline. Purpose of changes: 1) To explicitly set rooting fraction to zero in baresoil patches. 2) To place definition of bedrock heat capacity in a location consistent with other bedrock parameters (e.g. thermal conductivity). Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): 2213, 2219 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none expected Code reviewed by: Sean Swenson, Dave Lawrence List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M clm/src/biogeophys/SoilTemperatureMod.F90 - move heat capacity from SoilTemperature M clm/src/biogeophys/LakeTemperatureMod.F90 - include heat capacity of bedrock M clm/src/biogeophys/SoilStateInitTimeConstMod.F90 - move heat capacity to SoilTemperature M clm/src/main/clm_varcon.F90 - move vol heat copacity from SoilStateInitTimeConstMod M clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 - set root fraction to zero CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - not run regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40 ok, 45 expected not bfb with baseline yellowstone_pgi - 40 ok, 45 expected not bfb with baseline yellowstone_gnu (clm45 only) - 40 ok, 45 expected not bfb with baseline hobart_nag - 45 compiles, bug 2213 CLM tag used for the baseline comparisons: clm4_5_2_r127 Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all clm45 and clm50 compsets - what platforms/compilers: all/all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? Monthly history files were compared from a short CLM Swenson case run. No further testing was performed If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: No runs were performed URL for LMWG diagnostics output used to validate new climate: No diagnostics were generated =============================================================== =============================================================== Tag name: clm4_5_2_r127 Originator(s): sacks (Bill Sacks) Date: Fri Sep 18 10:53:12 MDT 2015 One-line Summary: Fix interpolation of initial conditions for glc_mec and crop <-> non-crop Purpose of changes: Fixes the following bugs in initInterp: (1) interpolation of one glc_mec case to another glc_mec case (due to bug 2126); (2) interpolation of a non-crop case to a crop case (problems with generic crop type); (3) interpolation of a crop case to a non-crop case (problems with generic crop type). Note: If you want to use initInterp with an old glc_mec initial conditions file (generated before this tag), you will need to either: (1) run the same case with this tag or later, without use_init_interp, for at least one time step to generate a new (fixed) initial conditions file; or (2) overwrite the pfts1d_topoglc variable on your original initial conditions file so that it has the correct size. (1) is easier if it is feasible for you to do so. Some related changes to initInterp: - Changed the "user interface" for interpolating initial conditions: There is no longer a finidat_interp_source namelist option; instead, you set finidat to point to the initial conditions file to interpolate, and set the use_init_interp namelist option to .true. - Changed the default behavior so that it does NOT fill missing points (e.g., landunits that are in the new case but not on the input file) with the bare ground patch / vegetated landunit. Instead, it dies with an error message if it can't find a reasonable template column. If you want to it to use the nearest bare ground / vegetated point in this case, then rerun with init_interp_fill_missing_with_natveg set to .true. - Added system and unit tests of initInterp. System tests include both interpolation of a file onto an identical configuration, and ensuring they are bit-for-bit (LII test), and tests that involve interpolation from one resolution or configuration to another. For the latter, I have done some very quick checks, but have NOT carefully checked the current operation to ensure it is correct. Some (mostly) unrelated changes: (1) Write more metadata on column type to 1-d (vector) history files (2) Updated cime; this included, among other things: - fix RESUBMIT bug - no longer set FORCE_COLDSTART for glc_mec runs - addition of LII test Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 1447 (interpinic needs to be extended to work for glc_mec) - 1703 (need better testing of interpinic) - 1725 (interpinic doesn't work right for crop) - 2126 (Error in writing pfts1d_topoglc to restart file) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: - Removed finidat_interp_source namelist option - Added use_init_interp and init_interp_fill_missing_with_natveg namelist option List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): cime: cime2.0.17 -> cime2.0.18-p1.1 (see summary above) List all files eliminated: none List all files added and what they do: ========= Separate testable parts of initInterp into their own module; fix initInterp bugs as noted above A components/clm/src/main/initInterpMindist.F90 ========= Add system tests that exercise initInterp A components/clm/cimetest/testmods_dirs/clm/interp_from_f09/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/interp_from_f09/include_user_mods A components/clm/cimetest/testmods_dirs/clm/interp_from_f09/README A components/clm/cimetest/testmods_dirs/clm/interp_from_f09 A components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/shell_commands A components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/README A components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way/include_user_mods A components/clm/cimetest/testmods_dirs/clm/glcMEC_spunup_1way A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/include_user_mods A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop/README A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/include_user_mods A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop/README A components/clm/cimetest/testmods_dirs/clm/interp_f19_noncrop2crop A components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/README A components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/interp_f19_crop/include_user_mods A components/clm/cimetest/testmods_dirs/clm/interp_f19_crop ========= Add unit tests that exercise initInterp: set_mindist A components/clm/src/main/test/initInterpMindist_test/CMakeLists.txt A components/clm/src/main/test/initInterpMindist_test/test_init_interp_mindist.pf A components/clm/src/main/test/initInterpMindist_test List all existing files that have been modified, and describe the changes: ========= Fix incorrect dimension for pfts1d_topoglc M components/clm/src/main/subgridRestMod.F90 ========= Remove finidat_interp_source namelist option; add use_init_interp and init_interp_fill_missing_with_natveg options M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm M components/clm/src/main/controlMod.F90 M components/clm/src/main/clm_varctl.F90 ========= Refactor to split into two modules; parse new namelist option; change line endings from DOS to UNIX (changes all lines!) M components/clm/src/main/initInterp.F90 ========= Write column type metadata to 1-d (vector) history files M components/clm/src/main/column_varcon.F90 M components/clm/src/main/histFileMod.F90 M components/clm/src/main/restFileMod.F90 ========= Add optional msg argument to check_dim M components/clm/src/main/ncdio_pio.F90.in ========= Add system tests that exercise initInterp M components/clm/cimetest/testlist_clm.xml ========= Add unit tests that exercise initInterp: set_mindist M components/clm/src/main/CMakeLists.txt M components/clm/src/main/test/CMakeLists.txt M components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90.in CLM testing: build-namelist tests: yellowstone: ok This test fails; it looks like it started failing in r125: 131/547 < FAIL> unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok Note: There was one BFAIL: BFAIL ERP_P15x2_Lm36.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrig_o3_reduceOutput.clm2.h1.nc : baseline compare clm2.h1 (baseline: compare .base file with clm4_5_2_r126 file) (baseline history file does not exist) CLM tag used for the baseline comparisons: clm4_5_2_r126 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_2_r126 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Thu Sep 17 14:53:49 MDT 2015 One-line Summary: New decomp_depth_efolding value for clm5.0, set as a namelist parameter instead of netcdf params file. Purpose of changes: Set decomp_depth_efolding as a namelist parameter instead of netcdf params file. New default value for CLM 5.0. Improved error message from build namelist when turning on irrigation. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): 2190 Known bugs (include bugzilla ID): 2213 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: add namelist group bgc_shared and namelist parameter decomp_depth_efolding. New default value for clm 5.0. List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: clm/bld/namelist_files/namelist_defaults_clm4_5.xml - new defaults for decomp_depth_efolding clm/bld/namelist_files/namelist_definition_clm4_5.xml - put decomp_depth_efolding in new namelist group clm/bld/CLMBuildNamelist.pm - logic for new namelist variable, improve error message clm/src/unit_test_stubs/utils/spmdMod_stub.F90 - add mpi_real8 so unit tests will compile clm/src/biogeochem/CNSharedParamsMod.F90 - add new method to set parameters from namelist. clm/src/main/readParamsMod.F90 - pass namelist filename into CNSharedParamsMod::CNParamsReadShared CLM testing: regular build-namelist tests: yellowstone - unit tests pass unit-tests (components/clm/src): yellowstone - passed tools-tests (components/clm/test/tools): yellowstone - n/a PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a regular tests (aux_clm40, aux_clm45): yellowstone_intel : 40 - ok; 45 - ok; 50 - ok, clm50 bgc answer changing yellowstone_pgi : 40 - ok; 45 - ok; 50 - ok, clm50 bgc answer changing yellowstone_gnu (clm45 only) : 45 - ok; 50 - ok, clm50 bgc answer changing hobart_nag : 45 - ok; 50 - ok, clm50 bgc answer changing CLM tag used for the baseline comparisons: clm4_5_2_r125 Changes answers relative to baseline: yes If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: clm50 bgc compsets - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): climate changing for clm 5.0 If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) Parameter change based on experiments described in Koven, C.D., D.M. Lawrence, and W.J. Riley, 2015: Permafrost-carbon feedback: Sensitivity to deep soil decomposability and nitrogen cycle. PNAS, DOI:10.1073/pnas.1415123112. Lawrence, D.M., C.D. Koven, S.C. Swenson, W.J. Riley, and A.G. Slater, 2015: Permafrost thaw and resulting soil moisture changes regulate projected high-latitude CO2 and CH4 emission. ERL, 10, doi:10.1088/1748-9326/10/9/094011. URL for LMWG diagnostics output used to validate new climate: =============================================================== =============================================================== Tag name: clm4_5_2_r125 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue Sep 15 22:08:09 MDT 2015 One-line Summary: clm 5.0 hydrology Purpose of changes: clm 5.0 hydrology Dry surface layer soil evaporative resistance from Sean Swenson Richards equation with adaptive time stepping from Martyn Clark Zero-flux lower boundary for soil moisture Deep soil column (10 meters) Variable depth to bedrock from Sean Swenson Canopy interception and maximum leaf wetted fraction change Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): unknown Known bugs (include bugzilla ID): 2213 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: Added new namelist variables for clm 5.0 hydrology. See namelist definition file for details. use_bedrock, deep_soilcolumn, soilwater_movement_method, upeer_boundary_condition, lower_boundary_condition, soil_resis_method, rooting_profile_method, interception_fraction, maximum_leaf_wetted_fraction. List any changes to the defaults for the boundary datasets: unknown Describe any substantial timing or memory changes: unknown Code reviewed by: Sean Swenson List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: unknown List all files added and what they do: clm/cimetest/testmods_dirs/clm/deepsoil_bedrock - new test for clm50 hydrology options. clm/src/biogeophys/SoilWaterRetentionCurveVanGenuchten1980Mod.F90 - adds placeholder for vanGenuchten soil water retention curve List all existing files that have been modified, and describe the changes: unknown clm/cimetest/testlist_clm.xml - new test clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 clm/tools/clm4_5/mksurfdata_map/src/Srcfiles clm/tools/clm4_5/mksurfdata_map/src/mkvarctl.F90 clm/tools/clm4_5/mksurfdata_map/src/mkfileMod.F90 clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl clm/bld/namelist_files/namelist_defaults_clm4_5.xml - new namelist options clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml clm/bld/namelist_files/namelist_definition_clm4_5.xml clm/bld/CLMBuildNamelist.pm clm/src/cpl/lnd_import_export.F90 clm/src/biogeochem/CNVegStateType.F90 clm/src/main/clm_varpar.F90 clm/src/main/ColumnType.F90 clm/src/main/init_hydrology.F90 clm/src/main/controlMod.F90 clm/src/main/clm_varctl.F90 clm/src/main/pftconMod.F90 clm/src/main/initVerticalMod.F90 clm/src/main/clm_instMod.F90 clm/src/biogeophys/WaterStateType.F90 clm/src/biogeophys/SoilWaterRetentionCurveFactoryMod.F90 clm/src/biogeophys/SoilWaterRetentionCurveClappHornberg1978Mod.F90 clm/src/biogeophys/SoilWaterRetentionCurveMod.F90 clm/src/biogeophys/IrrigationMod.F90 clm/src/biogeophys/SurfaceResistanceMod.F90 clm/src/biogeophys/HydrologyDrainageMod.F90 clm/src/biogeophys/RootBiophysMod.F90 clm/src/biogeophys/BalanceCheckMod.F90 clm/src/biogeophys/SoilWaterMovementMod.F90 clm/src/biogeophys/SoilMoistStressMod.F90 clm/src/biogeophys/CanopyHydrologyMod.F90 clm/src/biogeophys/CanopyTemperatureMod.F90 clm/src/biogeophys/SoilHydrologyMod.F90 clm/src/biogeophys/HydrologyNoDrainageMod.F90 clm/src/biogeophys/SoilStateType.F90 clm/src/biogeophys/SoilHydrologyType.F90 clm/src/biogeophys/BareGroundFluxesMod.F90 clm/src/biogeophys/CanopyFluxesMod.F90 clm/src/biogeophys/SoilStateInitTimeConstMod.F90 clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 CLM testing: build-namelist tests: yellowstone - CLMBuildNamelist unit tests pass unit-tests (components/clm/src): yellowstone - pass tools-tests (components/clm/test/tools): yellowstone - not run PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone - n/a regular tests (aux_clm40, aux_clm45): NOTE: for all aux_clm45 machine/compiler combinations - clm50 compsets not bit for bit with baseline (new hydrology model is used). yellowstone_intel : aux_clm40 - pass; aux_clm45 - clm 45 compsets pass, clm50 compsets not bit for bit yellowstone_pgi : aux_clm40 - pass; aux_clm45 - clm 45 compsets pass, clm50 compsets not bit for bit yellowstone_gnu (clm45 only) : aux_clm45 - clm 45 compsets ok, clm50 compsets not bit for bit hobart_nag : aux_clm45 - clm 45 compsets ok, clm50 compsets not bit for bit. bug 2213 still applies for clm45 compsets. CLM tag used for the baseline comparisons: clm4_5_2_r124 Changes answers relative to baseline: clm40 and clm45 compsets are bit for bit. clm50 compsets are answer changing. If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all clm50 compsets - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff, unknown climate. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: unknown URL for LMWG diagnostics output used to validate new climate: unknown =============================================================== =============================================================== Tag name: clm4_5_2_r124 Originator(s): sacks (Bill Sacks) Date: Sun Sep 13 06:33:20 MDT 2015 One-line Summary: Wind-dependent snow density Purpose of changes: Add an option to make snow density dependent on wind speed. There are two components to this: (1) adding a wind-related factor to the fresh snow density, and (2) adding a wind-related factor for the compaction of the existing snow pack (applying just to the top layer). This allows for more realistic (higher) fresh snow densities in windy conditions. Other than myself, contributors to these changes are: Jan Lenaerts, Leo van Kampenhout and Drew Slater. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new option: 'wind_dependent_snow_density' List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Renamed D components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology.pf List all files added and what they do: ========= Tests of some of the new code A components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_windCompactionFactor.pf ========= Renamed A components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology_initSnowLayers.pf List all existing files that have been modified, and describe the changes: ========= Add wind-dependent snow density parameterizations and control flag M components/clm/src/main/controlMod.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Add tests for some of the new code M components/clm/src/biogeophys/test/SnowHydrology_test/CMakeLists.txt CLM testing: build-namelist tests: yellowstone: ok unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok CLM tag used for the baseline comparisons: clm4_5_2_r123 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all CLM50 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; magnitude not investigated in detail If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_2_r123 Originator(s): sacks (Bill Sacks) Date: Thu Sep 10 10:25:55 MDT 2015 One-line Summary: Fix some problems with pgi and nag compilers Purpose of changes: (1) Fix some problems with the nag compiler: fix bug 2208, but open new bug 2213 (2) Work around some problems that appear with pgi 15.7 (appear to be compiler bugs), which Sean Santos tracked down: The key problem is that PGI is exporting some names that should have been aliased. So for instance: use foo_mod, only: foo_bar => bar In this case, if you have a "use" without an "only" of foo_mod, you get both foo_bar and bar imported into the namespace. Note: I have not been able to test these changes with pgi 15.7, because there is a PIO build problem with pgi 15.7 (3) No longer kill the run if we're using ozone with pgi: The pgi compiler bug that showed up with the ozone code has been fixed as of pgi 15.1, which is now our default. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 2208: nag reports a problem with associate statement to un-allocated LUNA memory in photosynthesisMod (but see new bug 2213) 2094: ozone code doesn't work with the PGI compiler Known bugs (include bugzilla ID): 2213: CLM45 / CLM50 runs without BGC die with NAG compiler, due to associate to unallocated memory http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Fix for pgi 15.7. CLM40 changes from Sean Santos; I ported these changes to CLM45 M components/clm/src_clm40/main/lnd_comp_mct.F90 M components/clm/src_clm40/main/spmdGathScatMod.F90 M components/clm/src/cpl/lnd_comp_mct.F90 M components/clm/src/utils/spmdGathScatMod.F90 ========= Fix for nag from Ben Andre M components/clm/src/biogeophys/PhotosynthesisMod.F90 ========= No longer abort if running with pgi M components/clm/src/biogeophys/OzoneMod.F90 ========= Remove ozone pgi test, because it runs out of wallclock time (but I believe it would now pass: a shorter version now passes) M components/clm/cimetest/testlist_clm.xml ========= Update xFails: Add hobart_nag xFails, remove no-longer-relevant or no-longer-failing tests (many have been passing for a while now) M components/clm/cimetest/ExpectedTestFails.xml CLM testing: unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: ok However, all runs without BGC still die: see bug 2213 and ExpectedTestFails list CLM tag used for the baseline comparisons: clm4_5_2_r122 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_2_r122 Originator(s): sacks (Bill Sacks) Date: Wed Sep 9 05:57:15 MDT 2015 One-line Summary: Update to latest cime Purpose of changes: Update cime to the version likely to be in cesm1_4_beta07 Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: none List any svn externals directories updated (cime, rtm, cism, etc.): cime2.0.7 -> cime2.0.17 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Change instructions: no longer need dot in your path M components/clm/src/README.unit_testing ========= Remove CME and NCK expected failures: these pass now, presumably due to fixes in cime (although since the failures were intermittent, it's hard to say definitively that the problem is fixed) M components/clm/cimetest/ExpectedTestFails.xml - CFAIL CME_Ld5.f10_f10.ICN.yellowstone_intel - CFAIL CME_Ly4.f10_f10.I1850CLM45BGC.yellowstone_intel.clm-monthly - FAIL NCK_Ld1.f10_f10.ICRUCLM45.yellowstone_intel.clm-default - RUN NCK_Ld1.f10_f10.ICRUCLM45.yellowstone_intel.clm-default CLM testing: unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: clm40 ok, CLM45 NOT RUN Note that there is still a problem with clm45 nag CLM tag used for the baseline comparisons: clm4_5_2_r121 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_2_r121 Originator(s): sacks (Bill Sacks) Date: Fri Sep 4 10:55:54 MDT 2015 One-line Summary: Rework snow capping Purpose of changes: Main change is a complete rework of snow capping. Previously, when the snow pack was in a "capped" state (i.e., h2osno >= h2osno_max), new snow and rain were prevented from falling on the snow pack: it was as if someone put a roof over the snow. (However, there were workarounds in place that caused the aerosol concentration and snow grain size in the top snow layer to continue to evolve as if new snow was still falling.) This was problematic in terms of the long-term evolution of the snow pack, since in effect the new, less dense snow was being removed due to capping. So, at least in some regions, the snow pack became much more dense than it should be. The new snow capping scheme allows snow and rain to fall on the top as they would in the uncapped state. But if h2osno exceeds h2osno_max, then mass is removed from the *bottom* snow layer. This removes the need for the old workarounds, and is more physically realistic: This represents the process of converting snow to glacial ice, which in reality occurs at the bottom (most dense) portion of the snow pack. Also, a mostly-unrelated change: rearrange the multi-layer snow history fields, so that index 1 is always the top of the snow pack. These changes are from Leo van Kampenhout (l.vankampenhout@uu.nl) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Rework handling of snow capping. This is now done pretty much entirely in SnowHydrologyMod, in a completely different way than before. M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/LakeFluxesMod.F90 M components/clm/src/biogeophys/SoilFluxesMod.F90 M components/clm/src/biogeophys/WaterStateType.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/main/clm_driver.F90 ========= Remove some old workarounds that were put in place to get reasonable answers with the old snow capping method M components/clm/src/biogeophys/AerosolMod.F90 M components/clm/src/biogeophys/SnowSnicarMod.F90 ========= Rearrange multi-layer snow history fields M components/clm/src/main/histFileMod.F90 M components/clm/doc/UsersGuide/custom.xml ========= Trivial changes (removing associate lines, changing comments, etc.) M components/clm/src/biogeophys/HydrologyDrainageMod.F90 M components/clm/src/biogeophys/SoilHydrologyMod.F90 M components/clm/src/biogeophys/CanopyTemperatureMod.F90 M components/clm/src/biogeophys/UrbanFluxesMod.F90 M components/clm/src/biogeochem/ch4Mod.F90 CLM testing: unit-tests (components/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok hobart_nag: NOT RUN hobart_nag tests still all fail, as they did in clm4_5_1_r120 CLM tag used for the baseline comparisons: clm4_5_1_r120 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All clm4.5 and clm5.0 configurations - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Definitely larger than roundoff. Whether this is the same or new climate has not been investigated carefully. However, at least in some regions with persistent snow cover, this change has a significant effect on the evolution of the snow pack, as described above. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r120 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Sat Aug 29 22:58:57 MDT 2015 One-line Summary: CLM 5 nitrogen models Flexible CN and LUNA Purpose of changes: CLM 5 nitrogen models Flexible CN (Bardan Ghimire, LBNL) and LUNA (Chonggang Xu, LANL). The LUNA model predicts photosynthetic capacities as measured by Vc, max25 and Jmax25 under different environmental conditions (see Ali et al 2015). Requirements for tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ - 2208 https://github.com/CESM-Development/cime/issues - 115, 116 Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: addition of use_luna and use_flexibleCN. use_flexibleCN adds additional namelist options in the clm_nitrogen group. See xml definitions file for details. List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self, Bardan Ghimire, Chonggang Xu List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: src/biogeochem/NutrientCompetitionFlexibleCNMod.F90 - flexibleCN. cimetest/testmods_dirs/clm/flexibleCN - flexible cn regression test src/biogeophys/LunaMod.F90 - luna model cimetest/testmods_dirs/clm/luna - luna regression case List all existing files that have been modified, and describe the changes: new namelist controls for flexibleCN and luna, clm_nitrogen namelist group: bld/namelist_files/namelist_definition_clm4_5.xml bld/namelist_files/namelist_defaults_clm4_5.xml bld/CLMBuildNamelist.pm src/main/clm_varctl.F90 src/main/controlMod.F90 flexibleCN src/biogeochem/CNVegCarbonStateType.F90 - Michaelis-Menten Nitrogen uptake src/biogeochem/CNVegNitrogenStateType.F90 - Michaelis-Menten Nitrogen uptake src/biogeochem/CNGRespMod.F90 - excess carbon storage src/biogeochem/CNGapMortalityMod.F90 - excess carbon storage src/biogeochem/NutrientCompetitionFactoryMod.F90 - add flexible cn option src/biogeochem/NutrientCompetitionMethodMod.F90 - modify interface to accomidate flexiblecn src/biogeochem/CNPhenologyMod.F90 - floating cn evergreen phenology src/biogeochem/CNDriverMod.F90 - update function call args src/main/pftconMod.F90 - flexible cn pft variables src/main/histFileMod.F90 - nlev canopy src/main/clm_driver.F90 - update function call args src/main/clm_instMod.F90 - update function call args src/biogeophys/WaterfluxType.F90 - additional water flux vars src/biogeophys/SoilWaterMovementMod.F90 - soil water work around LUNA src/main/clm_varcon.F90 - new constant for luna src/biogeophys/PhotosynthesisMod.F90 - luna use of vcmax25 and jmax25 src/biogeophys/CanopyFluxesMod.F90 - luna calculation of vcmax25 and jmax25 src/main/atm2lndType.F90 - state data needed for luna src/biogeophys/FrictionVelocityMod.F90 - luna variables src/biogeophys/WaterStateType.F90 - luna variables src/biogeophys/TemperatureType.F90 - luna variables src/biogeophys/SolarAbsorbedType.F90 - luna variables src/biogeophys/QSatMod.F90 - saturated vapor pressure density src/biogeophys/SoilHydrologyType.F90 - luna var src/biogeophys/CanopyStateType.F90 - update vcmax and jmax for luna cimetest/ExpectedTestFails.xml - update for cime bugs 115 and 116 cimetest/testlist_clm.xml - update test list for aux_clm_short, new luna and flexibleCN tests CLM testing: regular, build-namelist build-namelist tests: yellowstone - unit tests : pass, other pass unit-tests (components/clm/src): yellowstone - ok tools-tests (components/clm/test/tools): n/a PTCLM testing (components/clm/tools/shared/PTCLM/test): n/a regular tests (aux_clm40, aux_clm45): yellowstone_intel - 40 ok, 45 ok yellowstone_pgi - 40 ok, 45 ok yellowstone_gnu (clm45 only) ok hobart_nag - not run, tests hang, see bug 2208 Testing notes: * new namelist group clm_nitrogen causes all nlcomp tests to fail * introduces new tests for flexibleCN and luna that do not have baselines in clm4_5_1_r119. * two new expected fails believed to be related to cime issues 115 and 116. * removes the existing aux_clm_short tests and replaces them with a new set of SMS, ERS and ERP tests that are replicated for yellowstone gnu, intel and pgi. CLM tag used for the baseline comparisons: clm4_5_1_r119 Changes answers relative to baseline: none =============================================================== =============================================================== Tag name: clm4_5_1_r119 Originator(s): erik (Erik Kluzek) Date: Wed Aug 26 22:29:10 MDT 2015 One-line Summary: Bring hobart/nag bug fixes to trunk, and fix a few bugs Purpose of changes: Bring hobart/nag bug fixes to trunk. Fix ncl6.3.0 bug for getregional script. Fix use_c13 bug. Update RTM to handle regional direction files. Make sure _r8 constants in ED have a decimal point, so the NAG compiler will treat them as double-precision rather than as integer*2. Move testing from goldbach to hobart. For hobart_nag make all of the tests on just one node (24 processors). Requirements for tag: compile run with hobart/nag (fix bugs 2205 and 2199) move testing from goldbach to hobart Test level of tag: regular Bugs fixed (include bugzilla ID): 2206 (PTCLM stopped working with pft number in surface dataset filenames for mksurfdata.pl) 2205 (Problems with some constants in ED for NAG compiler) 2199 (crayftn compiler issue with continuation in middle of string) 2180 (ncl6.3.0 bug for getregional script) 2174 (use_c13 bug, unformatted write caused model to die) 2156 (Update RTM to handle regional direction files) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, klindsay List any svn externals directories updated (cime, rtm, cism, etc.): PTCLM, cime and rtm cime to cime2.0.07 rtm to rtm1_0_52 PTCLM to PTCLM2_150826 List all files eliminated: Move goldbach to hobart D components/clm/test/tools/tests_posttag_goldbach_nompi List all files added and what they do: Move goldbach to hobart A components/clm/test/tools/tests_posttag_hobart_nompi List all existing files that have been modified, and describe the changes: ------------ Move goldbach to hobart, remove PGI option for hobart M components/clm/test/tools/test_driver.sh M components/clm/bld/unit_testers/build-namelist_test.pl -- Fix ED tests so megan off M components/clm/cimetest/testlist_clm.xml --- Move goldbach tests to hobart Make 2-node hobart_nag tests on a single node M components/clm/tools/shared/mkmapgrids/mkscripgrid.ncl -- Fix so will work with ncl6.3.0, bug 2180 M components/clm/src/README.unit_testing --- add some notes about unit-testing ------------ Bug 2205, some _r8 constants in ED don't have a decimal point ------------ and the NAG compiler then treats them as integer*2. M components/clm/src/ED/main/EDCLMLinkMod.F90 M components/clm/src/ED/main/EDRestVectorMod.F90 M components/clm/src/ED/main/EDInitMod.F90 M components/clm/src/ED/fire/SFMainMod.F90 M components/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 M components/clm/src/ED/biogeophys/EDBtranMod.F90 M components/clm/src/ED/biogeochem/EDGrowthFunctionsMod.F90 M components/clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 M components/clm/src/ED/biogeochem/EDPhysiologyMod.F90 M components/clm/src/ED/biogeochem/EDPatchDynamicsMod.F90 ------------ Bug 2199, write to iulog was unformatted, which caused the model ------------ to die after it had already done formatted writes. M components/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 M components/clm/src/biogeochem/CNVegCarbonStateType.F90 CLM testing: build-namelist tests: yellowstone yes unit-tests (components/clm/src): yellowstone yes tools testing (components/clm/test/tools): yellowstone yes PTCLM testing (components/clm/tools/shared/PTCLM/test): yellowstone yes regular tests (aux_clm40, aux_clm45, aux_clm_short): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu (clm45 only) yes hobart_nag yes hobart_pgi yes hobart_intel yes CLM tag used for the baseline comparisons: clm4_5_1_r118 Changes answers relative to baseline: no =============================================================== =============================================================== Tag name: clm4_5_1_r118 Originator(s): sacks (Bill Sacks) Date: Wed Aug 5 16:22:33 MDT 2015 One-line Summary: Minor rework of glc coupling fields Purpose of changes: This makes CLM compatible with recent CIME changes. (1) Use renamed coupler field, in both clm40 and clm45 (2) In clm45 code, rework clm_cpl_indices to use glc_elevclass_mod (simpler and more robust than the earlier code) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): cime1.1.11 -> cime2.0.0 cism2_0_09 -> cism2_1_02 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Main changes M components/clm/src/cpl/clm_cpl_indices.F90 M components/clm/src/cpl/lnd_import_export.F90 M components/clm/src_clm40/main/clm_cpl_indices.F90 M components/clm/src_clm40/main/lnd_import_export.F90 ========= Document new unit testing method needed for yellowstone, due to cime update M components/clm/src/README.unit_testing ========= Rework test mods due to a fundamental change in how the forced decrease / increase in glc area works M components/clm/cimetest/testmods_dirs/clm/glcMEC_decrease/user_nl_cism M components/clm/cimetest/testmods_dirs/clm/glcMEC_increase/user_nl_cism ========= New failures, which seem to be attributable to the cime update, unrelated to my changes. However, the NCK and CME test failures seem dependent on the order in which tests are run, so these problems are hard to reproduce. Running them as single tests leads to PASSes. M components/clm/cimetest/ExpectedTestFails.xml + CFAIL CME_Ld5.f10_f10.ICN.yellowstone_intel + FAIL NCK_Ld1.f10_f10.ICRUCLM45.yellowstone_intel.clm-default + RUN ERS_D_Mmpi-serial_Ld5.1x1_brazil.ICLM45ED.yellowstone_gnu.clm-edTest CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok - but see caveat below! As noted above, there are three new test failures. The ERS_D ICLM45ED failure seems to be a legitimate bug in CLM. The other two (CFAIL CME_Ld5.f10_f10.ICN.yellowstone_intel and FAIL NCK_Ld1.f10_f10.ICRUCLM45.yellowstone_intel.clm-default) seem to be intermittent failures, likely due to a bug in the test system or elsewhere in cime. These sometimes pass and sometimes fail. They always seem to pass when run as single tests, but sometimes fail when run as part of a test suite. It's not clear if the new cime is to blame directly, or if these are arising now simply because tests are being run in a different order. golbach-nag does not run out-of-the-box with this tag. However, it should run out-of-the-box if you merge in the next commit in cime master (4b52ec73086a4290323dddfde6087a6d6d12ab96). I did my changes with that commit merged in, but this hadn't come to master in time for me to include it in this CLM tag. CLM tag used for the baseline comparisons: clm4_5_1_r117 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Configurations with CISM (IG), both CLM4 and CLM45 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Likely larger than roundoff/same climate (but not investigated closely) These changes are due to a complete rework of the coupling between CISM and CLM, manifested as major changes in the CIME and CISM externals. (The changes in CLM are not directly responsible for the answer changes.) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r117 Originator(s): sacks (Bill Sacks) Date: Tue Jul 28 06:01:04 MDT 2015 One-line Summary: Repartition rain vs. snow from atmosphere Purpose of changes: Add an option to repartition rain vs. snow from atmosphere based on near-surface temperature. This repartitioning uses a ramp-based partitioning that is also used in datm: we ignore the rain vs. snow partitioning sent from the atmosphere, and generate our own rain vs. snow partitioning. A sensible heat flux is generated to conserve energy with this repartitioning. The motivation for this is two-fold: (1) There are biases in CAM which cause rain to be generated in cold conditions. This is particularly a problem for glacier surface mass balance in Greenland. Andrew Gettelman has suggested putting in place this workaround in CLM until CAM can find a robust fix. (2) With the downscaling to glacier elevation classes, it is useful to have a different rain/snow partitioning in each elevation class. This repartitioning is on by default in CLM5, off by default in CLM4.5. If / when the CAM bias is fixed, we could potentially change this code so that it just does the repartitioning over the do_smb filter, similarly to the other downscaling in atm2lndMod. (Rather than doing this correction everywhere - which we do now in order to correct the rain vs. snow partitioning bias in CAM.) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: Removed glcmec_downscale_rain_snow_convert option, added repartition_rian_snow option List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Sean Swenson reviewed the calculation of the sensible heat flux correction List any svn externals directories updated (cime, rtm, cism, etc.): cime1.1.10 -> cime1.1.11 This creates a new shared routine for partitioning rain vs. snow, now shared between datm and CLM. List all files eliminated: List all files added and what they do: ========= Add unit tests for repartitioning of rain vs snow, and supporting utility code A components/clm/src/main/test/atm2lnd_test/test_sens_heat_from_precip_conversion.pf A components/clm/src/main/test/atm2lnd_test/CMakeLists.txt A components/clm/src/main/test/atm2lnd_test/test_partition_precip.pf A components/clm/src/main/test/atm2lnd_test A components/clm/src/unit_test_shr/unittestArrayMod.F90 A components/clm/src/unit_test_shr/test/unittestArray_test/CMakeLists.txt A components/clm/src/unit_test_shr/test/unittestArray_test/test_unittestArray.pf A components/clm/src/unit_test_shr/test/unittestArray_test List all existing files that have been modified, and describe the changes: ========= Repartition rain vs snow from atmosphere, and add a sensible heat flux correction for energy conservation M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/src/main/clm_driver.F90 M components/clm/src/main/clm_varctl.F90 M components/clm/src/main/controlMod.F90 M components/clm/src/main/atm2lndType.F90 M components/clm/src/main/lnd2atmType.F90 M components/clm/src/main/atm2lndMod.F90 M components/clm/src/main/lnd2atmMod.F90 ========= Remove glcmec_rain_snow_threshold M components/clm/src/main/clm_varcon.F90 ========= Remove glcmec_downscale_rain_snow_convert option, add repartition_rain_snow option M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Add an IG CLM5 test M components/clm/cimetest/testlist_clm.xml ========= Remove glcmec_downscale_rain_snow_convert setting (which no longer exists) M components/clm/cimetest/testmods_dirs/clm/glcMEC_changeFlags/user_nl_clm ========= Add unit tests for repartitioning of rain vs snow, and supporting utility code M components/clm/src/main/CMakeLists.txt M components/clm/src/main/test/CMakeLists.txt M components/clm/src/biogeophys/CMakeLists.txt M components/clm/src/unit_test_shr/test/CMakeLists.txt M components/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 M components/clm/src/unit_test_shr/CMakeLists.txt CLM testing: build-namelist tests: yellowstone: ok (changes namelists, as expected) unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r116 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM5 cases - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Likely new climate, but not investigated closely Answer changes are due to new rain vs. snow partitioning, which is on by default in CLM5. Also changes answers for ERP_D_Ld5.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC_changeFlags (expected, since it no longer downscales precip). If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r116 Originator(s): sacks (Bill Sacks) Date: Wed Jul 22 06:39:28 EDT 2015 One-line Summary: Rename some history fields Purpose of changes: (1) Make QSNOMELT point to qflx_snomelt, as it does in CLM4.0, rather than qflx_snow_drain (previously qflx_snow_melt) (2) Turn on QSNOFRZ by default (parallels QSNOMELT) (3) For the 3 history fields that have FOO and FOO_NODYNLNDUSE versions: Rename FOO to FOO_TO_COUPLER and FOO_NODYNLNDUSE to FOO. This is at Sean Swenson's suggestion: He points out that the version without the dyn landuse adjustment (and, soon, the sensible heat adjustment from rain/snow conversion) is the one most people will be interested in, so should be the one without the suffix. (4) Tweak test lists: (a) Move prealpha & prebeta goldbach tests to hobart (b) Move an aux_clm45 pgi test to intel: With recent versions of cime (starting with cime1.1.0), threading tests with pgi take a very long time. So this test took 3 hours with pgi, vs 20 min with intel. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2178 (QSNOMELT incorrect in clm4.5) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= See summary above M components/clm/src/main/lnd2atmType.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/EnergyFluxType.F90 M components/clm/cimetest/testlist_clm.xml CLM testing: unit-tests (models/lnd/clm/src): yellowstone: NOT RUN mac: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r115 Changes answers relative to baseline: YES Just changes a few diagnostic fields, for CLM4.5 and CLM5: - QSNOMELT: changed to qflx_snomelt rather than qflx_snow_drain - QRUNOFF: differs for cases with transient landcover - FSH: differs for cases with CISM - QSNWCPICE: differs for cases with CISM =============================================================== =============================================================== Tag name: clm4_5_1_r115 Originator(s): sacks (Bill Sacks) Date: Wed Jul 15 05:26:37 MDT 2015 One-line Summary: Remove redundant code, rename a variable Purpose of changes: (1) Remove some redundant code in SnowHydrologyMod, related to 'void'. This was supposed to be removed a long time ago. (Apparently the issue this was trying to fix was fixed in a different, more robust way.) (2) Remove redundant, unused copy of accumulMod in utils/ (newer copy is in main/) (3) Rename qflx_snow_melt to qflx_snow_drain, to avoid confusion with the existing qflx_snomelt. (4) Clarify documentation of snowdp_col Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: ========= Redundant and unused (copy in main/ is used) D components/clm/src/utils/accumulMod.F90 List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Rename qflx_snow_melt to qflx_snow_drain M components/clm/src/biogeophys/CanopyHydrologyMod.F90 M components/clm/src/biogeophys/LakeHydrologyMod.F90 M components/clm/src/biogeophys/BalanceCheckMod.F90 M components/clm/src/biogeophys/WaterfluxType.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 - also remove redundant lines of code related to 'void' (see above) M components/clm/src/biogeophys/LakeTemperatureMod.F90 ========= Clarify documentation for snowdp_col M components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M components/clm/src/biogeophys/WaterStateType.F90 ========= Remove SMS_Lm25.f10_f10.ICLM45BGCCROP.goldbach_nag.clm-reduceOutput (runs out of wall-clock time, and we have sufficient test coverage of that configuration) M components/clm/cimetest/testlist_clm.xml CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r114 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r114 Originator(s): sacks (Bill Sacks) Date: Fri Jul 10 19:34:57 MDT 2015 One-line Summary: Update cime external, remove genf90-generated files Purpose of changes: Main purpose is to update the cime external to the version in cesm1_4_beta05. This also required updating the unit test build to use genf90 during the build rather than relying on already-generated files. Making this change led to some genf90'd files being regenerated in-source during the unit test build, which would lead these files to be updated every time we make a tag. To avoid this annoyance, I have removed the genf90'd files from the repository: These are not needed any more in either the unit test or system builds, and it simplifies things to remove them. Then I added an svn:ignore property to ignore files generated by genf90 during the unit test build. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): cime1.0.7 -> cime1.1.10 Among other things, this brings in Jay's big batch system refactor List all files eliminated: ========= Remove genf90-generated files, and some scripts that were used to create them (these are now created as part of the unit test or system build) D components/clm/src/dyn_subgrid/dynVarMod.F90 D components/clm/src/dyn_subgrid/dynVarTimeInterpMod.F90 D components/clm/src/dyn_subgrid/dynVarTimeUninterpMod.F90 D components/clm/src/dyn_subgrid/do_genf90 D components/clm/src/unit_test_stubs/utils/do_genf90 D components/clm/src/unit_test_stubs/utils/restUtilMod_stub.F90 D components/clm/src/unit_test_stubs/main/ncdio_var.F90 D components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90 D components/clm/src/unit_test_stubs/main/do_genf90 D components/clm/src/utils/restUtilMod.F90 D components/clm/src/main/ncdio_pio.F90 List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Generate files with genf90 rather than using pre-generated files M components/clm/src/CMakeLists.txt ========= Document new, simpler method for building and running the unit tests (thanks largely to new default options in run_tests.py) M components/clm/src/README.unit_testing ========= Add svn:ignore property to ignore files generated by genf90 during the unit test build M components/clm/src/dyn_subgrid M components/clm/src/unit_test_stubs/utils M components/clm/src/unit_test_stubs/main CLM testing: build-namelist tests: yellowstone: ok unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r112 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL - what platforms/compilers: intel - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff, according to cime documentation If bitwise differences were observed, how did you show they were no worse than roundoff? N/A: Trusting Sean Santos's cime documentation. These changes were likely due to this change in cime: commit 0d7eab6bd112565ba9eb6eb82b74127ae5a5f390 Author: Sean Patrick Santos Date: Fri May 15 12:35:31 2015 -0600 Use our native gamma/erf implementations on Intel =============================================================== =============================================================== Tag name: clm4_5_1_r113 Originator(s): sacks (Bill Sacks) Date: Thu Jul 9 10:01:13 MDT 2015 One-line Summary: Support backwards compatibility of restart variable names Purpose of changes: Previously, if a restart variable was renamed, backwards compatibility was implemented in an ad-hoc manner. A key point is taht none of these ad-hoc solutions allowed backwards compatibility when running initInterp. (initInterp would just skip any variable if it could not find an exact match on the input [template] file.) This tag provides a standard mechanism for putting in place backwards compatibility when renaming a restart variable. This backwards compatibility carries over to initInterp, by communicating the necessary metadata through a new attribute on the restart file: 'varnames_on_old_files'. In order to use this new mechanism, give a colon-delimited list of variable names in the varname argument to restartvar. For example, if a restart variable FOO has been renamed to BAR, then specify varname='BAR:FOO'. Note that this list is searched in order, and the first item should be the current restart variable name. Also, applied this new mechanism to the recently-added LIQCAN variable. Previously, backwards compatibility of this variable was handled in an ad-hoc manner, which did not work when running initInterp. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Erik List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Add module for higher-level netcdf utilities. Currently contains routine for finding a variable on a netcdf file from a list of possible variables. Also add unit tests for this routine. A components/clm/src/main/ncdio_utils.F90 A components/clm/src/main/test/ncdio_utils_test/test_ncdio_utils.pf A components/clm/src/main/test/ncdio_utils_test/CMakeLists.txt A components/clm/src/main/test/ncdio_utils_test List all existing files that have been modified, and describe the changes: ========= Allow multiple possible names in reading restart files and in reading the 'input' file in initInterp M components/clm/src/utils/restUtilMod.F90.in M components/clm/src/utils/restUtilMod.F90 M components/clm/src/main/initInterp.F90 ========= Apply new mechanism to recently-added LIQCAN restart field. Also remove redundant setting of snocan_patch to 0 if it isn't found on the restart file - not needed since initCold is always called. M components/clm/src/biogeophys/WaterStateType.F90 ========= Changes to support unit testing of ncdio_utils M components/clm/src/main/CMakeLists.txt M components/clm/src/main/test/CMakeLists.txt M components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90.in M components/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90 ========= Fix path to genf90 for new cime organization M components/clm/src/unit_test_stubs/main/do_genf90 M components/clm/src/unit_test_stubs/main/ncdio_var.F90 CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r112 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r112 Originator(s): oleson (Keith Oleson,UCAR/TSS,303-497-1332) Date: Wed Jul 1 10:14:11 MDT 2015 One-line Summary: Justin Perket snow on vegetation Purpose of changes: Incorporate Justin Perket's snow on vegetation changes Requirements for tag: Test level of tag: regular, build-namelist, unit_tests Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: Add snowveg_flag item List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Keith Oleson, Justin Perket, Erik Kluzek List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/README.unit_testing Add note that instructions are for bash shell M components/clm/src/biogeophys/CanopyFluxesMod.F90 snow/liq on vegetation M components/clm/src/biogeophys/WaterStateType.F90 history/restart handling for snow/liq on vegetation M components/clm/src/biogeophys/BalanceCheckMod.F90 line spaces only M components/clm/src/biogeophys/WaterfluxType.F90 history handling for snow on vegetation M components/clm/src/biogeophys/CanopyHydrologyMod.F90 snow/liq on vegetation and snowveg_flag handling M components/clm/src/biogeophys/SurfaceAlbedoMod.F90 snow on vegetation optical properties M components/clm/src/main/controlMod.F90 line spaces only M components/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml snowveg_flag handling M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml snowveg_flag handling M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml snowveg_flag handling M components/clm/bld/CLMBuildNamelist.pm snowveg_flag handling CLM testing: build-namelist tests: yellowstone: ok All CLM45 and CLM50 tests have namelist differences; this is expected due to addition of new namelist item unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r111 Changes answers relative to baseline: Yes, for CLM50 If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM50 - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate. See Justin Perket (perketj@umich.edu) If bitwise differences were observed, how did you show they were no worse than roundoff? NA If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: NA URL for LMWG diagnostics output used to validate new climate: NA =============================================================== =============================================================== Tag name: clm4_5_1_r111 Originator(s): sacks (Bill Sacks) Date: Fri Jun 12 20:19:25 MDT 2015 One-line Summary: Remove temporary hack to get bfb results in InitSnowLayers Purpose of changes: In order to get bit-for-bit results in clm4_5_1_r110 (relative to r109), we put in place a temporary hack in InitSnowLayers that set dz based on the old equations rather than the new, more general ones - thus avoiding roundoff-level changes. This looked like: if (abs(dz(c,0)-3.59_r8) < eps) then ! TODO remove col%dz(c, 0) = snow_depth(c)-col%dz(c,-4)-col%dz(c,-3)-col%dz(c,-2)-col%dz(c,-1) if (abs(dz(c,0)-3.59_r8) > eps) & call endrun(msg=errmsg(__FILE__, __LINE__)) endif This tag removes that temporary hack. In this way, we have separated the answer-changing from non-answer-changing parts of the r110 refactor. Note that the above code confirms that the differences are no larger than roundoff (eps was 1e-9 in the above case, but some tests showed that it could have been much smaller - e.g., ~ 1e-15). Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M components/clm/src/biogeophys/SnowHydrologyMod.F90 CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r110 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM4.5 / CLM5 cold start - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? See above code sample, which confirms that the changes were no greater than roundoff-level. =============================================================== =============================================================== Tag name: clm4_5_1_r110 Originator(s): sacks (Bill Sacks) Date: Fri Jun 12 15:30:11 MDT 2015 One-line Summary: Add flexibility to have more snow layers Purpose of changes: Generalize snow code so that it no longer assumes 5 snow layers. Instead, make the number of snow layers (and the maximum SWE in the snow pack) a runtime parameter, allowing 3 - 12 snow layers. Most changes were made by Leo van Kampenhout (l.vankampenhout@uu.nl). Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2182 (possible threading issue with optimized pgi builds) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: - new namelist parameters: nlevsno, h2osno_max List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: ========= Add tests with different number of snow layers A components/clm/cimetest/testmods_dirs/clm/snowlayers_12/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/snowlayers_12/include_user_mods A components/clm/cimetest/testmods_dirs/clm/snowlayers_12 A components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly/user_nl_clm A components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly/include_user_mods A components/clm/cimetest/testmods_dirs/clm/snowlayers_3_monthly ========= Add unit tests for snow pack initialization (note: these were added mainly to facilitate debugging InitSnowLayers; since this routine is only used in cold-start, these are not critical unit tests, and can be removed if the maintenance cost proves too high) A components/clm/src/biogeophys/test/SnowHydrology_test/CMakeLists.txt A components/clm/src/biogeophys/test/SnowHydrology_test/README A components/clm/src/biogeophys/test/SnowHydrology_test/test_SnowHydrology.pf A components/clm/src/biogeophys/test/SnowHydrology_test List all existing files that have been modified, and describe the changes: ========= Major rework to remove assumption of 5 snow layers - instead allow runtime-setable number of snow layers, between 3 and 12. Also, clean up white space throughout file, and add mode/indentation emacs line. M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Minor changes to remove assumption of 5 snow layers M components/clm/src/biogeophys/SurfaceRadiationMod.F90 M components/clm/src/biogeophys/SoilTemperatureMod.F90 - also: remove unused variables, fix array argument declarations to conform to conventions M components/clm/src/main/initVerticalMod.F90 - also: clean up some white space ========= Add namelist variables to control number of snow layers and maximum SWE M components/clm/src/main/clm_varcon.F90 M components/clm/src/main/clm_varpar.F90 M components/clm/src/main/controlMod.F90 M components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M components/clm/bld/namelist_files/namelist_definition_clm4_5.xml M components/clm/bld/CLMBuildNamelist.pm ========= Minor changes needed for unit testing of SnowHydrologyMod M components/clm/src/biogeophys/CMakeLists.txt M components/clm/src/biogeophys/test/CMakeLists.txt M components/clm/src/biogeophys/SnowSnicarMod.F90 M components/clm/src/biogeophys/SnowHydrologyMod.F90 M components/clm/src/biogeophys/AerosolMod.F90 M components/clm/src/main/CMakeLists.txt M components/clm/src/unit_test_stubs/main/histFileMod_stub.F90 ========= Add tests with different number of snow layers M components/clm/cimetest/testlist_clm.xml ========= Unrelated change: remove unused variables in associate statements M components/clm/src/biogeophys/SoilFluxesMod.F90 ========= Remove a test that now passes (hooray for weird compiler bugs!) M components/clm/cimetest/ExpectedTestFails.xml CLM testing: build-namelist tests: yellowstone: ok All CLM45 and CLM50 tests have namelist differences; this is expected due to addition of 2 new namelist items. unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r109 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r109 Originator(s): sacks (Bill Sacks) Date: Sat Jun 6 06:12:02 MDT 2015 One-line Summary: Fix bug in DivideSnowLayers Purpose of changes: Fix bug in DivideSnowLayers. Leo van Kampenhout (l.vankampenhout@uu.nl) discovered the bug and determined how to fix it. He found this bug in the course of refactoring this routine to introduce loops; without this bug fix, answers differed with his new logic that removes duplication. Specifically: Logic using many IF-statements is employed to see whether or not a layer may be subdivided, depending on the layer thickness. Currently, the test for subdividing the BOTTOM layer are only reachable when the layer above it was also too thick. As it turns out, this is faulty as a situation can arise where the bottom layers grows even though the layer above it was not divided, i.e. dumped mass to it. The current understanding is that this happens through meltwater percolation (liquid h2o is translated to thickness as well). Note that the indentation has not been appropriately corrected, this is because this fix is only temporary (less cluttered logic will be implemented next). Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 2183 (incorrect logic for sub-dividing bottom snow layer in DivideSnowLayers) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Mark Flanner List any svn externals directories updated (cime, rtm, cism, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Bug fix, described above M components/clm/src/biogeophys/SnowHydrologyMod.F90 ========= Remove failures from here, now that we're using the file in cimetest M components/clm/bld/unit_testers/xFail/expectedClmTestFails.xml ========= Change a failure type from FAIL to RUN (presumably due to new test reporting) M components/clm/cimetest/ExpectedTestFails.xml CLM testing: regular tests (aux_clm40, aux_clm45): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r108 Changes answers relative to baseline: YES Summarize any changes to answers, i.e., - what code configurations: All clm4.5 and clm5 - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff. While not investigated carefully, Leo showed that the impacts are relatively small, so this is believed NOT to be climate-changing. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r108 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Fri May 29 15:14:26 MDT 2015 One-line Summary: Crop changes from Sam Levis Purpose of changes: Crop model changes from Sam Levis. Increases the number of crops to 64, with 78 total pfts. Requires new parameters file, surface dataset, and land use timeseries files. Test level of tag: regular Bugs fixed (include bugzilla ID): 2146, 2155 Known bugs (include bugzilla ID): 2180, 2182 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: n/a Describe any changes made to the namelist: n/a List any changes to the defaults for the boundary datasets: Regenerate surface data sets and land use timeseries to increase the number of pfts and crops with data from Levis. New raw datasets: rawdata/pftlanduse.3minx3min.simyr2000.c110913/mksrf_78pft_landuse_rc2000_c130927.nc rawdata/pftlandusedyn.0.5x0.5.simyr1850-2005.c090630/mksrf_78pft_landuse_rc2000_c150130.nc Describe any substantial timing or memory changes: Increased memory usage for all crop runs. Exact consequences requires further study. Code reviewed by: andre, levis List any svn externals directories updated (cime, rtm, cism, etc.): n/a List all files eliminated: n/a List all files added and what they do: components/clm/cimetest/ExpectedTestFails.xml - new expected fails file for upcoming cime xfail integration components/clm/tools/clm4_5/mksurfdata_map/Makefile.data - automate generating all surface data sets List all existing files that have been modified, and describe the changes: clm/bld/CLMBuildNamelist.pm - increase max pft, add info to error message, fix quoted empty string processing (bug 2146) clm/bld/namelist_files/namelist_defaults_clm4_5.xml - point to new datasets clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - update rawdata sets clm/bld/namelist_files/namelist_definition_clm4_5.xml - update sim year range to avoid special cases for testing clm/bld/test_build_namelist/t/test_do_harvest.pm - update test (bug 2146) clm/bld/test_build_namelist/t/test_do_transient_crops.pm - update test (bug 2146) clm/bld/test_build_namelist/t/test_do_transient_pfts.pm - update test (bug 2146) clm/cimetest/testmods_dirs/clm/crop_trans_f10/user_nl_clm - point to new datafiles clm/cimetest/testmods_dirs/clm/crop_trans_sville/user_nl_clm - point to new datafiles clm/src/biogeophys/WaterStateType.F90 - workaround for pgi compiler bug clm/src/biogeochem/CNNDynamicsMod.F90 - new crop model clm/src/biogeochem/CNPhenologyMod.F90 clm/src/biogeochem/CNVegStructUpdateMod.F90 clm/src/biogeochem/CropType.F90 clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 clm/src/biogeochem/VOCEmissionMod.F90 clm/src/biogeophys/CanopyFluxesMod.F90 clm/src/biogeophys/PhotosynthesisMod.F90 clm/src/main/PatchType.F90 clm/src/main/clm_varpar.F90 clm/src/main/pftconMod.F90 clm/src/main/subgridRestMod.F90 clm/src/main/surfrdMod.F90 clm/tools/clm4_5/mksurfdata_map/README clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl - move file writes into functions. write to __dataset_name__.namelist clm/tools/clm4_5/mksurfdata_map/src/mklaiMod.F90 - update for new crops clm/tools/clm4_5/mksurfdata_map/src/mkpftConstantsMod.F90 clm/tools/clm4_5/mksurfdata_map/src/mkpftMod.F90 - update for new crops, fix bug 2155. clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 - work around for an issue causing abort during urban dataset generation clm/tools/clm4_5/mksurfdata_map/src/mkurbanparMod.F90 CLM testing: regular build-namelist tests: yellowstone unit tests - ok unit-tests (components/clm/src): yellowstone_intel - ok regular tests (aux_clm40, aux_clm45): - yellowstone aux clm40 intel - ok - yellowstone aux clm40 pgi - ok - yellowstone aux clm45 intel - ok - all namelist fail - new datasets and parameters - crop - new crop model - baseline failures expected - ed - new parameters file - baseline failures expected, ok'd by rfisher - ERP_D_Ld5.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC_changeFlags - baseline compare fails, roundoff level, unknown cause, ok'd by sacks - ERP_E_Ld5.f19_g16.I1850CRUCLM45CN.yellowstone_intel.clm-default - baseline compare fails, only occurs with esmf - ok'd by mvertens - yellowstone aux clm45 gnu - ok - all namelist fail - new datasets and parameters - crop - new crop model - baseline failures expected - yellowstone aux clm45 pgi - - all namelist fail - new datasets and parameters - crop - new crop model - baseline failures expected - ed - new parameters file - baseline failures expected, ok'd by rfisher - ERI_D_Ld9.f19_g16.I1850CLM45CN.yellowstone_pgi - roundoff in cpl baseline - ERI_D_Ld9.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-drydepnomegan - roundoff in baseline - ERP_P15x2_Lm13.f10_f10.IHISTCLM45BGC.yellowstone_pgi.clm-monthly - new xfail, #2182, dies at runtime in optimized, threaded pgi builds only. - yellowstone mksurfdata_map unit_testers - ok - yellowstone tools tests - ok, known issues with PTCLMmkdata (bug 2180) CLM tag used for the baseline comparisons: clm4_5_1_r106 (bit for bit with clm4_5_1_r107) Changes answers relative to baseline: yes Summarize any changes to answers, i.e., - what code configurations: crop, ed - what platforms/compilers: all - nature of change : answer changes updated crop model. approved by levis. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - simulations with all pfts everywhere were run with merged code. Levis compared current runs with his archived runs. =============================================================== =============================================================== Tag name: clm4_5_1_r107 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Tue May 19 10:05:49 MDT 2015 One-line Summary: Update externals to use github version of cime1.0.7. Purpose of changes: Switch the cime external from using svn to github. Update cime to cime1.0.7. Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (cime, rtm, cism, etc.): cime 1.0.7 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: N/A CLM testing: regular build-namelist tests: unit-tests: ok system-tests: not run unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45): yellowstone_intel 40 - ok; 45 - ok yellowstone_pgi 40 - ok; 45 - ok yellowstone_gnu (clm45 only) 45 - ok CLM tag used for the baseline comparisons: clm4_5_1_r106 Changes answers relative to baseline: none =============================================================== =============================================================== Tag name: clm4_5_1_r106 Originator(s): erik/fvitt Date: Thu May 14 13:22:51 MDT 2015 One-line Summary: Fix CO2 forcing for MEGAN Purpose of changes: Bring in changes from Francis Vitt, and Louisa Emmons to correct CO2 forcing for MEGAN and dry-deposition. Previously, the fixed value of CO2 was being used rather than using the CO2 forcing sent in from the atmosphere model. Also fix some issues with clm4_0 code where some urban diagnostic fields have a different fill-value pattern on restart from startup. Fill-value is now only set over non-land, and areas without urban, are set to zero. +M models/lnd/clm/src_clm40/biogeochem/VOCEmissionMod.F90 + - use time-dependent atmospheric CO2 concentrations rather than + the CCSM_CO2_PPMV constant value + +M models/lnd/clm/src_clm40/biogeochem/DryDepVelocity.F90 + - science updates and bug fixes provided by Maria Val Martin + +M models/lnd/clm/src/biogeochem/VOCEmissionMod.F90 + - use time-dependent atmospheric CO2 concentrations rather than + the CCSM_CO2_PPMV constant value + - use 10-day average of LAI rather than 1-day average + +M models/lnd/clm/src/biogeochem/DryDepVelocity.F90 + - science updates and bug fixes provided by Maria Val Martin + +M models/lnd/clm/src/biogeophys/CanopyStateType.F90 + - get 10-day average of LAI rather than 1-day average Requirements for tag: Fix 2177 and some 2165 clm40 tests Test level of tag: regular Bugs fixed (include bugzilla ID): 2177 (MEGAN improperly uses constant CO2 rather than time varying) 2176 (ED doesn't work with MEGAN -- partial just turn MEGAN off when ED on) 2165 (some clm40 tests have history files differ on restart in urban fillvalue) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Turn MEGAN off when ED on Have build-namelist make sure MEGAN is off when ED is on List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self,fvitt,simone,emmons List any svn externals directories updated (cime, rtm, cism, etc.): cime, cism cime up to cime0_3_21 (bring in optional orbital calculation) cism up to cism2_0_09 (just bring the branch to the trunk) List all files eliminated: None List all files added and what they do: Turn MEGAN off for ED tests A components/clm/cimetest/testmods_dirs/clm/edTest/shell_commands List all existing files that have been modified, and describe the changes: M components/clm/src_clm40/biogeophys/UrbanInitMod.F90 -- Initialize to zero over land M components/clm/src_clm40/main/clm_initializeMod.F90 --- Call urbanInit before reading restart files M components/clm/src_clm40/main/accFldsMod.F90 ---------- Change running mean from 1 day to 10 days M components/clm/src_clm40/main/clmtypeInitMod.F90 ------ Initilize to spval M components/clm/src_clm40/biogeochem/VOCEmissionMod.F90 + - use time-dependent atmospheric CO2 concentrations rather than + the CCSM_CO2_PPMV constant value + - use 10-day average of LAI rather than 1-day average M components/clm/src_clm40/biogeochem/DryDepVelocity.F90 + - science updates and bug fixes provided by Maria Val Martin M components/clm/src/biogeochem/VOCEmissionMod.F90 + - use time-dependent atmospheric CO2 concentrations rather than + the CCSM_CO2_PPMV constant value + - use 10-day average of LAI rather than 1-day average M components/clm/src/biogeochem/DryDepVelocity.F90 + - science updates and bug fixes provided by Maria Val Martin M components/clm/src/biogeophys/CanopyStateType.F90 + - get 10-day average of LAI rather than 1-day average M README_cime --- Update documentation M components/clm/bld/CLMBuildNamelist.pm -- Check that MEGAN off when ED on M components/clm/bld/unit_testers/build-namelist_test.pl - Add check for MEGAN off when ED on M components/clm/bld/clm.buildnml --- only copy drv_flds_in over if it was actually created. M components/clm/tools/README ---- Have documentation point to gen_domain under cime/tools/mapping CLM testing: regular build-namelist tests: yellowstone yes unit-tests (models/lnd/clm/src): yellowstone yes regular tests (aux_clm40, aux_clm45): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes CLM tag used for the baseline comparisons: clm4_5_1_r105 Changes answers relative to baseline: Yes -- MEGAN diagnostic fields only! as well as dry-deposition because of science update Summarize any changes to answers, i.e., - what code configurations: All with MEGAN on - what platforms/compilers: All - nature of change: Diagnostic fields change VOC emissions change =============================================================== =============================================================== Tag name: clm4_5_1_r105 Originator(s): erik (Erik) Date: Thu Apr 16 13:23:19 MDT 2015 One-line Summary: Move test lists to beneath active components, change build scripts from cshell to perl, move to new cime directory structure Purpose of changes: * Move CESM test lists from under scripts to under active components. Now clm and rtm have their own CESM test lists under their "cimetest" directory. * Change build scripts from cshell to perl cshell is buggy with arbitrary problems with line lengths and number of arguments cshell doesn't allow long scripts to be broken up into subroutine -- perl does :::::::::: Get unit-testing working with directory structure change M src/CMakeLists.txt M src/README.unit_testing :::::::::: Get tools testing working with directory structure change M test/tools/TBLtools.sh M test/tools/TSMCFGtools.sh M test/tools/TSMscript_tools.sh M test/tools/TCBCFGtools.sh M test/tools/TCBscripttools.sh M test/tools/TSMncl_tools.sh M test/tools/TBLCFGtools.sh M test/tools/TSMtools.sh M test/tools/TBLscript_tools.sh M test/tools/TCBtools.sh M test/tools/test_driver.sh :::::::::: Get tools working with directory structure change M tools/clm4_0/mksurfdata_map/mksurfdata.pl M tools/clm4_5/mksurfdata_map/mksurfdata.pl M tools/shared/ncl_scripts/getco2_historical.ncl :::::::::: Updates to build M bld/CLMBuildNamelist.pm M bld/configure M bld/queryDefaultNamelist.pl M bld/listDefaultNamelist.pl M bld/unit_testers/xFail/wrapClmTests.pl M bld/unit_testers/xFail/expectedFail.pm M bld/test_build_namelist/test_build_namelist.pl M bld/namelist_files/checkmapfiles.ncl M bld/namelist_files/namelist_definition.xsl :::::::::: Updates to documentation with new directory structure M doc/README M doc/UsersGuide/co2_streams.txt M doc/Quickstart.userdatasets M doc/Quickstart.GUIDE M doc/KnownLimitations M tools/README M tools/README.filecopies M tools/clm4_0/interpinic/README M tools/clm4_0/mksurfdata_map/README M tools/clm4_5/refactorTools/associate/README M tools/clm4_5/refactorTools/clmType/README M tools/clm4_5/mksurfdata_map/README M tools/shared/mkmapgrids/README M tools/shared/mkmapdata/README M tools/shared/mkprocdata_map/README M tools/shared/ncl_scripts/README M tools/README.testing M bld/README CLM testing: build-namelist tests: yellowstone yes unit-tests (models/lnd/clm/src): yellowstone yes regular tests (aux_clm40, aux_clm45, and aux_clm_short): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes CLM tag used for the baseline comparisons: clm4_5_1_r104 Changes answers relative to baseline: =============================================================== =============================================================== Tag name: clm4_5_1_r104 Originator(s): erik (Erik) Date: Tue Jan 27 11:22:48 MST 2015 One-line Summary: Update externals to latest cesm beta tag + bring in shared build for clm4_5/clm5_0 for testing Purpose of changes: * Update externals to cesm1_3_beta15+ shared clm4_5/clm5_0 library build for testing. * Fix BG1850CN @ f09 by changing fglcmask (Bill Sacks) * Update more prealpha/prebeta tests to test with clm4_5 * Create datasets for clm4_5 at ne16 and ne120 resolution (for ne120 create rcp8.5 and rcp4.5 transient datasets) M models/lnd/clm/src/cpl/lnd_comp_mct.F90 ----------- Add only for lnd_import_export use statement ------------ Change so sample subsetting uses the high resolution datasets M models/lnd/clm/tools/shared/ncl_scripts/README.getregional M models/lnd/clm/tools/shared/ncl_scripts/sample_inlist M models/lnd/clm/tools/shared/ncl_scripts/sample_outlist CLM testing: build-namelist tests: yellowstone yes unit-tests (models/lnd/clm/src): yellowstone yes tools testing: yellowstone yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu (clm45 only) yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes CLM tag used for the baseline comparisons: clm4_5_1_r103 Changes answers relative to baseline: YES! Summarize any changes to answers, i.e., - what code configurations: all - what platforms/compilers: time-change all, roundoff-intel - nature of change: roundoff Small change in driver changes time-stamps on history files by roundoff (drvseq5_1_05). Normal cprnc comparison then does NOT compare fields and calls files different. Changes in the intel build (on yellowstone) change answers to roundoff for intel on yellowstone (Machines update between Machines_141125 and Machines_150106a causes answers to change) =============================================================== =============================================================== Tag name: clm4_5_1_r103 Originator(s): sacks (Bill Sacks) Date: Thu Jan 1 06:15:57 MST 2015 One-line Summary: enable transient crops Purpose of changes: (1) Allow transient crops! Note that carbon and nitrogen conservation still is not done, but this at least allows crop areas to evolve in time. (2) Add control flags for which pieces of the transient dynamics should be done: transient natural PFTs, transient crops, and/or harvest. (3) Reworked both source code and unit tests to be able to use the true CLM time manager in unit tests rather than a stub version. Also added functionality to time_info_type to be able to take the date from the end of the current time step or the beginning of the time step. This flexibility was needed because: (a) for crops, with an annual update, I wanted the update time to be consistent with the glacier update time: the first time step after crossing the year boundary (so take time from the start of the time step); (b) for transient PFTs and harvest, for consistency with what was being done before, we need to take the time from the end of the time step. (4) Make CNBalanceCheck more modular and object-oriented. Also, bypass the balance check for newly-active columns, which is needed to avoid balance check errors with transient crops. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: New control flags that control which aspects of transient subgrid dynamics (and harvest) are turned on/off. This lets you turn on/off transient natural PFTs, transient crops, and/or harvest independently. List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: some pieces reviewed by Erik scripts: append_nl_value_n03_scripts4_141201 -> append_nl_value_n07_scripts4_141201 esmf_wrf_timemgr: esmf_wrf_timemgr_141028 -> esmf_wrf_timemgr_141217 List all files eliminated: ========= No longer use stub time manager - use true time manager instead D models/lnd/clm/src/unit_test_stubs/utils/clm_time_manager_stub.F90 List all files added and what they do: ========= Add transient crops A models/lnd/clm/src/dyn_subgrid/dyncropFileMod.F90 ========= Reads and stores namelist items controlling transient dynamics. This allows turning off select pieces of the transient subgrid behavior. A models/lnd/clm/src/dyn_subgrid/dynSubgridControlMod.F90 ========= Add unit test utilities that wrap the clm time manager A models/lnd/clm/src/unit_test_shr/unittestTimeManagerMod.F90 ========= Start adding unit tests for the clm time manager A models/lnd/clm/src/utils/test/clm_time_manager_test/test_clm_time_manager.pf A models/lnd/clm/src/utils/test/clm_time_manager_test/CMakeLists.txt A models/lnd/clm/src/utils/test/clm_time_manager_test A models/lnd/clm/src/utils/test/CMakeLists.txt A models/lnd/clm/src/utils/test ========= Test logic for new control flags A models/lnd/clm/bld/test_build_namelist/t/test_do_transient_pfts.pm A models/lnd/clm/bld/test_build_namelist/t/test_do_harvest.pm A models/lnd/clm/bld/test_build_namelist/t/test_do_transient_crops.pm List all existing files that have been modified, and describe the changes: ========= Reworked both source code and unit tests to be able to use the true CLM time manager in unit tests rather than a stub version. Also added functionality to time_info_type to be able to take the date from the end of the current time step or the beginning of the time step. Note that some unit test builds now need to link against the esmf_wrf_timemgr library, if they use the time manager either directly or indirectly. M models/lnd/clm/src/utils/clm_time_manager.F90 M models/lnd/clm/src/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/dyn_subgrid/dynVarTimeInterpMod.F90.in M models/lnd/clm/src/dyn_subgrid/dynVarTimeInterpMod.F90 M models/lnd/clm/src/dyn_subgrid/dynFileMod.F90 M models/lnd/clm/src/unit_test_stubs/dyn_subgrid/dynFileMod_stub.F90 M models/lnd/clm/src/unit_test_stubs/utils/CMakeLists.txt M models/lnd/clm/src/CMakeLists.txt M models/lnd/clm/src/unit_test_shr/CMakeLists.txt M models/lnd/clm/src/dyn_subgrid/test/dynTimeInfo_test/test_dynTimeInfo.pf M models/lnd/clm/src/dyn_subgrid/test/dynTimeInfo_test/CMakeLists.txt M models/lnd/clm/src/dyn_subgrid/test/dynVar_test/test_dynVarShared.F90 M models/lnd/clm/src/dyn_subgrid/test/dynVar_test/test_dynVarTimeInterp.pf M models/lnd/clm/src/dyn_subgrid/test/dynVar_test/test_dynVarTimeUninterp.pf M models/lnd/clm/src/dyn_subgrid/test/dynVar_test/CMakeLists.txt M models/lnd/clm/src/biogeophys/test/Daylength_test/CMakeLists.txt M models/lnd/clm/src/biogeophys/test/Irrigation_test/CMakeLists.txt M models/lnd/clm/src/utils/CMakeLists.txt ========= Changes related to new control flags, as well as the rework of the time_info%set_current_year interface M models/lnd/clm/src/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/dyn_subgrid/dynHarvestMod.F90 ========= Changes related to new control flags and addition of transient crops M models/lnd/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Moved flanduse_timeseries and other control flags into dynSubgridControlMod M models/lnd/clm/src/main/clm_varctl.F90 M models/lnd/clm/src/main/controlMod.F90 M models/lnd/clm/src/main/restFileMod.F90 M models/lnd/clm/src/main/subgridRestMod.F90 M models/lnd/clm/src/main/clm_varpar.F90 M models/lnd/clm/src/biogeochem/CNDriverMod.F90 M models/lnd/clm/src/biogeochem/CNFireMod.F90 ========= Make CNBalanceCheck more modular and object-oriented M models/lnd/clm/src/biogeochem/CNBalanceCheckMod.F90 - also, bypass balance checks for newly-active columns (needed to avoid balance check errors for newly-active crop columns) M models/lnd/clm/src/biogeochem/CNVegNitrogenStateType.F90 M models/lnd/clm/src/biogeochem/CNVegCarbonStateType.F90 M models/lnd/clm/src/main/clm_instMod.F90 M models/lnd/clm/src/main/clm_driver.F90 - also make alt_calc operate over inactive as well as active points ========= Added new control flags M models/lnd/clm/bld/CLMBuildNamelist.pm M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml ========= Make test files (more) consistent with actual files. This may not have been necessary. M models/lnd/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M models/lnd/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml ========= Minor changes (e.g., changes to comments and other small changes) M models/lnd/clm/src/unit_test_stubs/utils/spmdMod_stub.F90 M models/lnd/clm/src/dyn_subgrid/do_genf90 M models/lnd/clm/src/dyn_subgrid/dynVarTimeUninterpMod.F90 M models/lnd/clm/src/dyn_subgrid/dynVarMod.F90 M models/lnd/clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 M models/lnd/clm/src/biogeophys/ActiveLayerMod.F90 CLM testing: build-namelist tests: yellowstone: ok Note that there are differences from baseline due to new control flags unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r102 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r102 Originator(s): sacks (Bill Sacks) Date: Sat Dec 27 06:52:20 MST 2014 One-line Summary: make new input datasets to support transient crops Purpose of changes: While the main purpose of this tag was to make new input datasets to support transient crops, it also includes a number of reworks of mksurfdata_map, which were either central or tangential to this overall goal. Specifically: (1) Update mksurfdata_map to be able to generate datasets with transient crops. Currently the logic uses the non-prognostic-crop raw data for the transient time series: It takes the area of the generic crop from that timeseries to specify the transient PCT_CROP area, and sets the PCT_CFT areas based on the year-2000 areas. (2) Rewrite mksurfdata_map code that normalizes pct_pft to account for special landunits. The code to handle urban was very confusing, and I hope this new code is at least astep towards being less confusing. Note that this introduces roundoff-level differences. (3) Introduce new mksurfdata_map utility routines: ncd_def_spatial_var and ncd_put_time_slice. These encapsulate behavior that used to be duplicated in the code. (4) Add mksurfdata_map unit tests using the new pfunit-based unit testing framework. However, I have NOT done a full migration of the mksurfdata_map unit tests. Thus, there are still some tests that use the old unit testing framework that I put in place (which leveraged the test stuff that Erik set up for csm_share a while ago). These tests can be migrated to pfunit incrementally: as someone touches code that is under test using the old framework, they could move the relevant tests into the new pfunit-based framework. (5) Create new input datasets, based on the above changes: - new flanduse_timeseries files created because I have changed the information on these files - new surface datasets created because I introduced roundoff-level changes in the surface datasets, and so regenerated all surface datasets now in order to save someone a headache later. - new initial conditions files created so that out-of-the-box initial conditions will be compatible with the surface datasets, according to various consistency checks. See the following files in inputdata for documentation of how the new datasets were created: lnd/clm2/surfdata_map/README_c141219 lnd/clm2/initdata_map/README_c141226 Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: - new surface datasets - new flanduse_timeseries files - new initial conditions files List any changes to the defaults for the boundary datasets: - new surface datasets - new flanduse_timeseries files - new initial conditions files Describe any substantial timing or memory changes: none Code reviewed by: Most changes sent to Erik for review, although I can't remember how much he actually reviewed. List any svn externals directories updated (csm_share, mct, etc.): scripts: append_nl_value_n02_scripts4_141201 -> append_nl_value_n03_scripts4_141201 - point tropicAtl_subset tests to new file List all files eliminated: ========= Move tests elsewhere D models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkpftMod.F90 List all files added and what they do: ========= New object-oriented class for storing and operating on pct_pft data. Encapsulating a bunch of behavior in here allowed me to simplify other code. Before this, I was keeping track of two separate representations of pct_pft: First it was stored as % of grid cell, then it was later converted into % of landunit together with the landunit's % of grid cell. This was starting to get hard to manage, because certain operations could only be done on the first representation, and other operations could only be done on the second representation – and at some point in the processing pipeline, the conversion happened and the first representation was no longer usable. Now there is a single representation, and the class allows any desired operation to be performed on that single representation. This adds some complexity within the class, but removes complexity from the rest of the code, particularly mksurfdat.F90. A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkpctPftTypeMod.F90 ========= This new module contains routines that operate on both instances of pct_pft_type (pctnatpft and pctcft) at once. Thus, this contains higher-level logic than was appropriate for mkpctPftTypeMod. Yet, I wanted these routines in a separate module from mkpftMod as an aid to testing, since mkpftMod has a bunch of dependencies. A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkpftUtilsMod.F90 ========= Moved constants from other places into here, partly to centralize them, and partly to remove problems with circular dependencies A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkpftConstantsMod.F90 ========= Add automated test of making a transient crop surface dataset A models/lnd/clm/test/tools/nl_files/mksrfdt_10x15_crp_1850-2000 ========= Add input file for creating a transient smallville dataset for testing transient crops A models/lnd/clm/tools/clm4_5/mksurfdata_map/single_point_dynpft_files/README.landuse_timeseries_smallvilleIA_hist_simyr1850-1855 A models/lnd/clm/tools/clm4_5/mksurfdata_map/single_point_dynpft_files/landuse_timeseries_smallvilleIA_hist_simyr1850-1855.txt A models/lnd/clm/tools/clm4_5/mksurfdata_map/single_point_dynpft_files/README A models/lnd/clm/tools/clm4_5/mksurfdata_map/single_point_dynpft_files ========= Add pfunit-based unit tests for mksurfdata_map A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/CMakeLists.txt A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpftUtils_test/test_adjust_total_veg_area.pf A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpftUtils_test/CMakeLists.txt A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpftUtils_test/test_convert_from_p2g.pf A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpftUtils_test A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpctPftType_test/CMakeLists.txt A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpctPftType_test/test_mkpctPftType.pf A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/mkpctPftType_test A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test/CMakeLists.txt A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/test A models/lnd/clm/tools/clm4_5/mksurfdata_map/src/README.unit_testing List all existing files that have been modified, and describe the changes: ========= Changes that take advantage of some of the other refactoring described here, plus add logic to allow input dataset to not contain crops even when generating a transient dataset for crops; also add PCT_CROP and PCT_CFT on the landuse_timeseries output file M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkpftMod.F90 ========= Changes that take advantage of some of the other refactoring described here, especially making use of the new pct_pft_type methods. Plus: (a) save pctcft from the initial input file so it can be used when generating landuse_timeseries, (b) remove unwanted landunit percents from the landuse_timeseries file, (c) add transient PCT_CROP and PCT_CFT, (d) complete rewrite of the code that normalizes pct_pft to account for special landunits: the code to handle urban was very confusing, and I hope this new code is at least a step towards being less confusing [this change introduces roundoff-level differences], (e) remove some error-checking code that is now embedded in the pct_pft_type routines M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 ========= Updated for new files M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/Srcfiles ========= Added two new routines: ncd_def_spatial_var and ncd_put_time_slice. These encapsulate behavior that used to be duplicated in the code. Also, moved convert_latlon from mkutilsMod to here, since it is really related to netcdf stuff. M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkncdio.F90 ========= Simplify this module significantly by using the new ncd_def_spatial_var. Also change what fields are present on the transient landuse file: remove some no-longer-desired fields (% of special landunits, etc.). M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkfileMod.F90 ========= Use new ncd_def_spatial_var M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mksoilMod.F90 ========= Delete routines that have been moved to a more appropriate place M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkutilsMod.F90 ========= Remove tests from old test framework for code that I have deleted or migrated to my new modules; and fix some minor errors that appeared when runnng the old unit tests with gfortran. M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkchecksMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkutilsMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkgridmapMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/Srcfiles M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mksurfdata_map.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkdomainMod.F90 ========= Trivial changes (change 'use' statements to reflect migrated code, add comments, etc.) M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mklaiMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkdomainMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkurbanparMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkgridmapMod.F90 ========= Add logic needed for creating transient crop datasets. The main differences are (a) for crop, we create a year-2000 surface dataset together with the transient dataset (rather than a year-1850 surface dataset), and (b) we always use the non-crop transient raw data, even when creating a transient crop dataset. M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl ========= New fsurdat, flanduse_timeseries and finidat files M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Add automated test of making a transient crop surface dataset M models/lnd/clm/test/tools/input_tests_master M models/lnd/clm/test/tools/tests_pretag_yellowstone_nompi CLM testing: build-namelist tests: yellowstone: ok tools tests: yellowstone: ok Note that there were diffs in baseline comparisons for mksurfdata_map tests (and the PTCLM test, which uses mksurfdata_map). See below for details regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r101 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: most clm4_5 runs - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? Note that there were no source code changes in this tag, so the only differences come from differences in datasets (fsurdat, flanduse_timeseries, finidat). I confirmed that differences in fsurdat and flanduse_timeseries are generally roundoff-level. There are greater than roundoff-level diffs in PCT_NAT_PFT at a small number of points, but all of these points have PCT_NATVEG = 0 (this is due to a fix in how PCT_NAT_PFT is determined for points with 0% vegetated landunit, and > 0% urban); this would only affect dynamic landunit runs. Other than that, max normalized RMS diffs are 2e-8, and most are considerbly smaller. finidat files were created as one-offs to ensure that the only differences are in the subgrid weights, arising from these surface dataset differences. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r101 Originator(s): sacks (Bill Sacks) Date: Tue Dec 9 06:27:39 MST 2014 One-line Summary: rework cold start initialization for transient runs Purpose of changes: (1) Do not adjust subgrid weights (or set harvest variables) in cold start initialization. Instead, wait to do this until the first run step. The motivation for this is (a) this is consistent with what is done for glacier (for which prognostic weights aren't available until the run phase), and (b) it simplifies what needs to be done in initialization, particularly for transient crops (which are coming soon). (2) Do not run the biogeophys & biogeochem dyn subgrid conservation code in the first step of a cold start run. This affects the current operation of glacier, and is important in conjunction with (1): this avoids doing a large adjustment of physics or BGC caused by a fictitious change in area in the first time step after cold start. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: mvertens reviewed conceptual changes to clm_time_manager and DaylengthMod; other changes only reviewed by self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Main changes, as documented above M models/lnd/clm/src/main/clm_initializeMod.F90 M models/lnd/clm/src/main/clm_varctl.F90 M models/lnd/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M models/lnd/clm/src/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/dyn_subgrid/dynConsBiogeochemMod.F90 M models/lnd/clm/src/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 ========= Unrelated change: Fix a possible threading bug in DaylengthMod (although this would rarely cause problems: I think this would only cause a problem if you started / restarted exactly on the solstice) M models/lnd/clm/src/utils/clm_time_manager.F90 M models/lnd/clm/src/biogeophys/DaylengthMod.F90 CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r100 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM4.5 cold start runs with glacier and/or transient PFTs - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Changes answers only in initialization. This shows up as changes in the fields set to the coupler in initialization, and/or changes in the initial history file. In offline runs (I compsets), this change does not affect the simulation beyond initialization, but it is expected to change the evolution of the system in coupled runs. However, again note that this only affects cold start runs wth glacier and/or transient PFTs, which would not be typical for production runs. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r100 Originator(s): sacks (Bill Sacks); most changes from Jim Edwards Date: Wed Dec 3 06:21:13 MST 2014 One-line Summary: update pio calls to pio2 API Purpose of changes: Update pio calls to the pio2 API, so that the transition to pio2 will be seamless. Most changes were from Jim Edwards. There are also some other minor changes that are unrelated to this main change, as noted below. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: sacks List any svn externals directories updated (csm_share, mct, etc.): pio: pio1_8_13 -> pio1_9_5 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Main changes needed for pio2 API M models/lnd/clm/src_clm40/main/ncdio_pio.F90 M models/lnd/clm/src_clm40/main/ncdio_pio.F90.in M models/lnd/clm/src/main/ncdio_pio.F90 M models/lnd/clm/src/main/ncdio_pio.F90.in ========= Change 2-d array to 1-d. Jim says he thinks this was also needed for pio2 support. M models/lnd/clm/src_clm40/main/histFileMod.F90 M models/lnd/clm/src/main/histFileMod.F90 ========= Unrelated change: change len to len_trim. Jim says this was needed to fix a problem on some machine. M models/lnd/clm/src_clm40/biogeochem/MEGANFactorsMod.F90 M models/lnd/clm/src/biogeochem/MEGANFactorsMod.F90 CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok CLM tag used for the baseline comparisons: clm4_5_1_r099 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r099 Originator(s): sacks (Bill Sacks) Date: Tue Dec 2 15:05:09 MST 2014 One-line Summary: add ozone stress code from Danica Lombardozzi Purpose of changes: (1) Implement ozone stress. The scientific implementation was done by Danica Lombardozzi. The software reimplementation was done by Bill Sacks. (2) Fix some misc. bugs, including a restart bug that was introduced in r097. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2091: some restarts not bit-for-bit starting in clm4_5_1_r097 - 2029: Memory leak in GetGlobalValuesMod Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ - 2094: ozone code doesn't work with the PGI compiler Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new namelist option, use_ozone List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: ozone changes reviewed by Danica Lombardozzi (she reviewed both the code and the changes seen due to ozone in a short simulation) List any svn externals directories updated (csm_share, mct, etc.): scripts: append_nl_value_n02_scripts4_141129 -> append_nl_value_n02_scripts4_141201 List all files eliminated: List all files added and what they do: ========= Implement ozone stress. Uses polymorphism to handle ozone-on vs. ozone-off. A models/lnd/clm/src/biogeophys/OzoneFactoryMod.F90 A models/lnd/clm/src/biogeophys/OzoneOffMod.F90 A models/lnd/clm/src/biogeophys/OzoneBaseMod.F90 A models/lnd/clm/src/biogeophys/OzoneMod.F90 List all existing files that have been modified, and describe the changes: ========= Calculate and apply ozone stress M models/lnd/clm/src/biogeophys/PhotosynthesisMod.F90 M models/lnd/clm/src/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/main/clm_instMod.F90 - also fix restart bug (bug 2091) M models/lnd/clm/src/main/clm_driver.F90 ========= Add namelist flag to turn ozone on M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/src/main/clm_varctl.F90 M models/lnd/clm/src/main/controlMod.F90 ========= Fix memory leak (bug 2029) M models/lnd/clm/src/main/GetGlobalValuesMod.F90 M models/lnd/clm/src_clm40/main/GetGlobalValuesMod.F90 M models/lnd/clm/src/ED/main/EDRestVectorMod.F90 ========= Workarounds for compiler bugs M models/lnd/clm/src/biogeochem/CNDVType.F90 M models/lnd/clm/src/biogeochem/CNDriverMod.F90 ========= Improve documentation comments for compiler bug workarounds M models/lnd/clm/src/biogeophys/IrrigationMod.F90 M models/lnd/clm/src/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/dyn_subgrid/dynpftFileMod.F90 ========= Add allocation of Points, matching behavior of true routine M models/lnd/clm/src/unit_test_stubs/csm_share/mct_mod_stub.F90 M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml Remove these two entries that now pass: restarts not bit-for-bit error on hist comparison Add this new failure: compiler bug in PGI's handling of polymorphism CLM testing: Note: testing was done on ozone_polymorphism_n09_clm4_5_1_r098, which was before I put in place the abort if you're trying to run ozone with pgi. After that, I ran two tests with pgi (one with ozone and one without), and one test with intel (with ozone) in order to make sure that the abort check was put in properly. After all testing was complete, I reverted accidental whitespace changes in clm_initializeMod.F90 and restFileMod.F90 - I did not run any additional testing after reverting those whitespace changes. unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok Also, ran these two additional tests, with comparisons to baselines - these are tests that I have replaced with new tests: ERS_Ly5.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrigOn_reduceOutput PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_pgi.clm-irrigOn_reduceOutput CLM tag used for the baseline comparisons: clm4_5_1_r098 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_1_r098 Originator(s): sacks (Bill Sacks) Date: Sat Nov 29 06:18:59 MST 2014 One-line Summary: update externals to cesm1_3_beta14 or beyond Purpose of changes: Update most externals to cesm1_3_beta14 or beyond. The one exception is mct, for which I had trouble accessing the tag at the location used in beta14, so I am sticking with the previous mct tag. Some notable changes: (1) update in intel compiler on yellowstone to intel15 (2) robust fix for number of datm streams, using Sean Santos's dynamic vector (3) testmods reworked to use recursive testmods (4) unit_testing, CMake & Machines updated so that unit tests now work on yellowstone Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/refactor_koven_tags/refactor_koven_n02_scripts4_141023 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/append_nl_value_tags/append_nl_value_n02_scripts4_141129 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_141017a +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_141125 -scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_140715 +scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_141122 -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_18 -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_141022 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_141121 -models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_140529 +models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_141028 -models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_12/pio +models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_13/pio -tools/load_balancing_tool https://svn-ccsm-models.cgd.ucar.edu/tools/load_balancing_tool/trunk_tags/load_balancing_tool_140818/ +tools/load_balancing_tool https://svn-ccsm-models.cgd.ucar.edu/tools/load_balancing_tool/trunk_tags/load_balancing_tool_141008 -tools/pyReshaper https://subversion.ucar.edu/asap/pyReshaper/tags/v0.9.1/ +tools/pyReshaper https://proxy.subversion.ucar.edu/pubasap/pyReshaper/tags/v0.9.1 -tools/unit_testing https://svn-ccsm-models.cgd.ucar.edu/unit_testing/trunk_tags/unit_testing_0_08 +tools/unit_testing https://svn-ccsm-models.cgd.ucar.edu/unit_testing/trunk_tags/unit_testing_0_12 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Workaround for pgi internal compiler error M models/lnd/clm/src/main/clm_driver.F90 ========= Rework README, mainly to remove the need for using '--clean' M models/lnd/clm/src/README.unit_testing ========= Move an xfail from goldbach to yellowstone; add xfail for ERS_Ly5 test (bug 2091) M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml ========= Update what machine-comiler combos we test M .ChangeLog_template CLM testing: unit-tests (models/lnd/clm/src): yellowstone: ok regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu (clm45 only): ok goldbach_nag: ok Other than xFails, note that the following failed: *** presumably this failed in the previous tag, so baselines didn't exist BFAIL SMS_D.1x1_mexicocityMEX.I.goldbach_nag.compare_hist.clm4_5_1_r097 *** ozone tests that won't work until an upcoming tag that brings ozone in SFAIL ERS_D.f10_f10.I1850CLM45.goldbach_nag.clm-o3.GC.1128-0838.45.n SFAIL PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_pgi.clm-irrig_o3_reduceOutput.GC.1128-0838.45.p SFAIL ERS_Ly5.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrig_o3_reduceOutput.GC.1128-0838.45.i Also, note that the following test failed: FAIL ERH_D.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-drydepnomegan This appears to be a scripts problem. Since Mariana wants to do away with ERH tests anyway, I just replaced this with: PASS ERI_D.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-drydepnomegan Also ran the following two tests, which have been replaced with (currently-failing) o3 tests: ERS_Ly5.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrigOn_reduceOutput PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_pgi.clm-irrigOn_reduceOutput The PET test passed, but the ERS test failed (see bug 2091) CLM tag used for the baseline comparisons: clm4_5_1_r097 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: ALL - what platforms/compilers: yellowstone-intel - nature of change (roundoff; larger than roundoff/same climate; new climate): NOT INVESTIGATED These diffs are presumably due to the yellowstone-intel compiler upgrade to v15. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r097 Originator(s): mvertens mvertens (Mariana Vertenstein,UCAR/CSEG,303-497-1349) Date: Mon Nov 24 11:06:30 MST 2014 One-line Summary: major refactorization to introduce new soilbiogeochem data types and routines that are independent of either ED or CN datatypes Purpose of changes: Major refactorization to introduce new soilbiogeochem data types and permit ED and CN vegetation to be independent of each other AND both work with either the same soilbiogeochem or in the future potentially different soilbiogeochem modules Requirements for tag: None Test level of tag: regular Bugs fixed (include bugzilla ID): None Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: bld/configure modified to accomodate new directory structure - introduction of soilbiogeochem/ directory Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: Since use_cn and use_ed are now mutually exclusive, CN memory is not longer allocated when ED is run and the memory reduction (for f19_g16) seems to be about 50%. Code reviewed by: mvertens, muszala, sacks List any svn externals directories updated (csm_share, mct, etc.): scripts branch is used https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/refactor_koven_tags/refactor_koven_n01_scripts4_141023 List all files eliminated, added and modified: (see below) List all files added and what they do: (see below) List all existing files that have been modified, and describe the changes: (see below) These are all grouped together here - since splitting the items up simply did not make sense in this case --------------------------------------------------- New module where all instances are now declared (moved from clm_initializeMod) - all calls to instance restarts are here as well - so restFileMod is greatly simplified --------------------------------------------------- A models/lnd/clm/src/main/clm_instMod.F90 --------------------------------------------------- New soilbiogeochem/ directory introduced (new modules and data types) --------------------------------------------------- A models/lnd/clm/src/soilbiogeochem --------------------------------------------------- CN state and flux types split into: SoilBiogeoChem[Carbon|Nitrogen][State|Flux]Type and SoilBiogoechemStateType CNVeg[Carbon|Nitrogen][State|Flux]Type and CNVegStateType --------------------------------------------------- D models/lnd/clm/src/biogeochem/CNStateType.F90 D models/lnd/clm/src/biogeochem/CNCarbonFluxType.F90 D models/lnd/clm/src/biogeochem/CNCarbonStateType.F90 D models/lnd/clm/src/biogeochem/CNNitrogenFluxType.F90 D models/lnd/clm/src/biogeochem/CNNitrogenStateType.F90 A models/lnd/clm/src/biogeochem/CNVegStateType.F90 A models/lnd/clm/src/biogeochem/CNVegCarbonFluxType.F90 A models/lnd/clm/src/biogeochem/CNVegCarbonStateType.F90 A models/lnd/clm/src/biogeochem/CNVegNitrogenStateType.F90 A models/lnd/clm/src/biogeochem/CNVegNitrogenFluxType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemCarbonFluxType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemCarbonStateType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemStateType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNitrogenFluxType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNitrogenStateType.F90 --------------------------------------------------- New modules in soilbiogeochem/ directory that are independent of CNVeg or ED types --------------------------------------------------- D models/lnd/clm/src/biogeochem/CNDecompCascadeConType.F90 D models/lnd/clm/src/biogeochem/CNNitrifDenitrifMod.F90 D models/lnd/clm/src/biogeochem/CNVerticalProfileMod.F90 D models/lnd/clm/src/biogeochem/CNDecompMod.F90 D models/lnd/clm/src/biogeochem/CNAllocationMod.F90 D models/lnd/clm/src/biogeochem/CNDecompCascadeBGCMod.F90 D models/lnd/clm/src/biogeochem/CNDecompCascadeCNMod.F90 D models/lnd/clm/src/biogeochem/CNSoilLittVertTranspMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeConType.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNitrifDenitrifMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNStateUpdate1Mod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeBGCMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNLeachingMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemVerticalProfileMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemCompetitionMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemDecompCascadeCNMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemPrecisionControlMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemNitrogenUptakeMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemPotentialMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemLittVertTranspMod.F90 A models/lnd/clm/src/soilbiogeochem/SoilBiogeochemDecompMod.F90 --------------------------------------------------- Moved CNEcosystemDynMod to CNDRiverMod --------------------------------------------------- D models/lnd/clm/src/biogeochem/CNEcosystemDynMod.F90 A models/lnd/clm/src/biogeochem/CNDriverMod.F90 --------------------------------------------------- Changes to modules in biogeochem/ directory to now use new datatypes (see above) --------------------------------------------------- A models/lnd/clm/src/biogeochem/C14BompbSpikeMod.F90 M models/lnd/clm/src/biogeochem/CNCStateUpdate2Mod.F90 M models/lnd/clm/src/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/biogeochem/CNGRespMod.F90 M models/lnd/clm/src/biogeochem/CNNStateUpdate1Mod.F90 M models/lnd/clm/src/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/biogeochem/CNNStateUpdate3Mod.F90 M models/lnd/clm/src/biogeochem/CNFireMod.F90 M models/lnd/clm/src/biogeochem/CNDVDriverMod.F90 M models/lnd/clm/src/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/biogeochem/MEGANFactorsMod.F90 M models/lnd/clm/src/biogeochem/SatellitePhenologyMod.F90 M models/lnd/clm/src/biogeochem/CNWoodProductsMod.F90 M models/lnd/clm/src/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/biogeochem/ch4Mod.F90 M models/lnd/clm/src/biogeochem/DUSTMod.F90 M models/lnd/clm/src/biogeochem/CNDVLightMod.F90 M models/lnd/clm/src/biogeochem/NutrientCompetitionMethodMod.F90 M models/lnd/clm/src/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/biogeochem/ch4varcon.F90 M models/lnd/clm/src/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/biogeochem/CNCStateUpdate3Mod.F90 M models/lnd/clm/src/biogeochem/CNSharedParamsMod.F90 M models/lnd/clm/src/biogeochem/CNDVType.F90 M models/lnd/clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 M models/lnd/clm/src/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/biogeochem/CNNStateUpdate2Mod.F90 M models/lnd/clm/src/biogeochem/CropType.F90 M models/lnd/clm/src/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/biogeochem/CNVegStructUpdateMod.F90 M models/lnd/clm/src/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/biogeochem/CNC14DecayMod.F90 --------------------------------------------------- Moved frictionvel_type to FrictionVelocityMod --------------------------------------------------- D models/lnd/clm/src/biogeophys/FrictionVelocityType.F90 --------------------------------------------------- Moved aerosol_type to AerosolMod --------------------------------------------------- D models/lnd/clm/src/biogeophys/AerosolType.F90 --------------------------------------------------- Moved photosyns_type to PhotosynthesisMod --------------------------------------------------- D models/lnd/clm/src/biogeophys/PhotosynthesisType.F90 --------------------------------------------------- Moved soilstate cold start initialization to a new module --------------------------------------------------- A models/lnd/clm/src/biogeophys/SoilStateInitTimeConstMod.F90 --------------------------------------------------- Moved soilhydrology time constant initialization to a new module --------------------------------------------------- A models/lnd/clm/src/biogeophys/SoilHydrologyInitTimeConstMod.F90 --------------------------------------------------- ED Refactorization1 (1) EDPhenologyMod changed to EDPhenologyType (2) EDBioType moved as a module type (ed_clm_type) in EDCLMLINKMod.F90 (3) EDVecPatchType no longer needed (for now is_veg, is_bareground and wt_ed are in PatchType.F90) --------------------------------------------------- D models/lnd/clm/src/ED/main/EDBioType.F90 M models/lnd/clm/src/ED/main/EDCLMLinkMod.F90 D models/lnd/clm/src/ED/main/EDVecPatchType.F90 D models/lnd/clm/src/ED/biogeophys/EDPhenologyMod.F90 A models/lnd/clm/src/ED/biogeochem/EDPhenologyType.F90 A models/lnd/clm/src/ED/biogeochem/EDSharedParamsMod.F90 --------------------------------------------------- ED Refactorization2 (1) Modified EDTypesMod.F90 Removed gridcell_edstate_type (array of pointers) and instance gridcelledstate - now have the following ED types and instance defined in clm_instMod.F90 and passed down in clm_initialize and clm_driver (top level) type(ed_site_type), allocatable, target :: ed_allsites_inst(:) type(ed_phenology_type) :: ed_phenology_inst type(ed_clm_type) :: ed_clm_inst so now have ed_allsites_inst which is an array of sites (at this point allocated at the gridcell level - but that could easily be modified to be at some other level like the column level (2) In EDTypesMod.F90 added method map_clmpatch_to_edpatch that maps a CLM vector patch to an ED linked-list patch - there is still a one to one correspondence between an ED patch and a CLM vector patch. The call looks like the following currentPatch => map_clmpatch_to_edpatch(ed_allsites_inst(g), p) (3) In EDTypesMod.F90 added a ED Patch type-bound method set_root_fraction that computes the root fraction for an ED patch (4) In EDTypes.F90 eliminated the following components of userdata type (site) , pointer :: firstsite_pnt => null() ! pointer to the first site in the system type (cohort), pointer :: storesmallcohort => null() ! storage of the smallest cohort for insertion routine type (cohort), pointer :: storebigcohort => null() ! storage of the largest cohort for insertion routine These are no longer needed since the above pointers are now local variables in EDCohortDynamics and EDPatchDynamics --------------------------------------------------- M models/lnd/clm/src/ED/main/EDVecCohortType.F90 M models/lnd/clm/src/ED/main/EDRestVectorMod.F90 M models/lnd/clm/src/ED/main/EDInitMod.F90 M models/lnd/clm/src/ED/main/EDMainMod.F90 M models/lnd/clm/src/ED/main/EDTypesMod.F90 M models/lnd/clm/src/ED/fire/SFMainMod.F90 M models/lnd/clm/src/ED/biogeochem/EDGrowthFunctionsMod.F90 M models/lnd/clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 M models/lnd/clm/src/ED/biogeochem/EDPhysiologyMod.F90 M models/lnd/clm/src/ED/biogeochem/EDPatchDynamicsMod.F90 M models/lnd/clm/src/ED/biogeochem/EDCanopyStructureMod.F90 M models/lnd/clm/src/ED/biogeophys/EDAccumulateFluxesMod.F90 M models/lnd/clm/src/ED/biogeophys/EDSurfaceAlbedoMod.F90 M models/lnd/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 M models/lnd/clm/src/ED/biogeophys/EDBtranMod.F90 --------------------------------------------------- The following changes are implemented below (and in the above routines) (1) _vars% changed to _inst% (2) pft% changed to patch% (3) merged ecophyscon and pftvarcon into single derived type pftcon (in pftconMod) "use EcophysConType, only : ecophyscon" changed to "use pftconMod, only : pftcon" (4) module save statements removed in majority of routines --------------------------------------------------- D models/lnd/clm/src/main/EcophysConType.F90 D models/lnd/clm/src/main/pftvarcon.F90 A models/lnd/clm/src/main/pftconMod.F90 M models/lnd/clm/src/main/initInterp.F90 M models/lnd/clm/src/main/clm_varpar.F90 M models/lnd/clm/src/main/landunit_varcon.F90 M models/lnd/clm/src/main/accumulMod.F90 M models/lnd/clm/src/main/subgridWeightsMod.F90 M models/lnd/clm/src/main/decompInitMod.F90 M models/lnd/clm/src/main/clm_initializeMod.F90 M models/lnd/clm/src/main/subgridRestMod.F90 M models/lnd/clm/src/main/ColumnType.F90 M models/lnd/clm/src/main/subgridMod.F90 M models/lnd/clm/src/main/PatchType.F90 M models/lnd/clm/src/main/ndepStreamMod.F90 M models/lnd/clm/src/main/lnd2atmType.F90 M models/lnd/clm/src/main/atm2lndType.F90 M models/lnd/clm/src/main/histFileMod.F90 M models/lnd/clm/src/main/restFileMod.F90 M models/lnd/clm/src/main/clm_varsur.F90 M models/lnd/clm/src/main/LandunitType.F90 M models/lnd/clm/src/main/GetGlobalValuesMod.F90 M models/lnd/clm/src/main/controlMod.F90 M models/lnd/clm/src/main/initSubgridMod.F90 M models/lnd/clm/src/main/filterMod.F90 M models/lnd/clm/src/main/lnd2glcMod.F90 M models/lnd/clm/src/main/clm_driver.F90 M models/lnd/clm/src/main/clm_varctl.F90 M models/lnd/clm/src/main/subgridAveMod.F90 M models/lnd/clm/src/main/initGridCellsMod.F90 M models/lnd/clm/src/main/atm2lndMod.F90 M models/lnd/clm/src/main/lnd2atmMod.F90 M models/lnd/clm/src/main/ncdio_pio.F90 M models/lnd/clm/src/main/surfrdMod.F90 M models/lnd/clm/src/main/decompMod.F90 M models/lnd/clm/src/main/reweightMod.F90 M models/lnd/clm/src/main/readParamsMod.F90 M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/biogeophys/WaterfluxType.F90 M models/lnd/clm/src/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/biogeophys/SnowSnicarMod.F90 M models/lnd/clm/src/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/biogeophys/LakeTemperatureMod.F90 M models/lnd/clm/src/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/biogeophys/SoilFluxesMod.F90 M models/lnd/clm/src/biogeophys/TemperatureType.F90 M models/lnd/clm/src/biogeophys/HumanIndexMod.F90 M models/lnd/clm/src/biogeophys/PhotosynthesisMod.F90 M models/lnd/clm/src/biogeophys/LakeFluxesMod.F90 M models/lnd/clm/src/biogeophys/AerosolMod.F90 M models/lnd/clm/src/biogeophys/ActiveLayerMod.F90 M models/lnd/clm/src/biogeophys/SurfaceResistanceMod.F90 M models/lnd/clm/src/biogeophys/SoilStateType.F90 M models/lnd/clm/src/biogeophys/SoilHydrologyType.F90 M models/lnd/clm/src/biogeophys/HydrologyDrainageMod.F90 M models/lnd/clm/src/biogeophys/UrbanAlbedoMod.F90 M models/lnd/clm/src/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/biogeophys/RootBiophysMod.F90 M models/lnd/clm/src/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/biogeophys/SoilWaterMovementMod.F90 M models/lnd/clm/src/biogeophys/SoilMoistStressMod.F90 M models/lnd/clm/src/biogeophys/UrbBuildTempOleson2015Mod.F90 M models/lnd/clm/src/biogeophys/CanopyHydrologyMod.F90 M models/lnd/clm/src/biogeophys/EnergyFluxType.F90 M models/lnd/clm/src/biogeophys/CanopyStateType.F90 M models/lnd/clm/src/biogeophys/UrbanFluxesMod.F90 M models/lnd/clm/src/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/biogeophys/UrbanRadiationMod.F90 M models/lnd/clm/src/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/biogeophys/IrrigationMod.F90 M models/lnd/clm/src/biogeophys/CanopyTemperatureMod.F90 M models/lnd/clm/src/biogeophys/HydrologyNoDrainageMod.F90 M models/lnd/clm/src/biogeophys/LakeHydrologyMod.F90 M models/lnd/clm/src/biogeophys/UrbanParamsType.F90 M models/lnd/clm/src/dyn_subgrid/dynLandunitAreaMod.F90 M models/lnd/clm/src/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 M models/lnd/clm/src/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/dyn_subgrid/dynFileMod.F90 M models/lnd/clm/src/dyn_subgrid/dynConsBiogeochemMod.F90 M models/lnd/clm/src/dyn_subgrid/dynPriorWeightsMod.F90 M models/lnd/clm/src/dyn_subgrid/dynVarMod.F90 M models/lnd/clm/src/dyn_subgrid/dynEDMod.F90 M models/lnd/clm/src/dyn_subgrid/dynVarTimeInterpMod.F90 M models/lnd/clm/src/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/dyn_subgrid/dynCNDVMod.F90 M models/lnd/clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 M models/lnd/clm/src/dyn_subgrid/dynVarMod.F90.in M models/lnd/clm/src/dyn_subgrid/dynVarTimeInterpMod.F90.in M models/lnd/clm/src/dyn_subgrid/dynInitColumnsMod.F90 M models/lnd/clm/src/cpl/lnd_comp_esmf.F90 M models/lnd/clm/src/cpl/lnd_import_export.F90 M models/lnd/clm/src/cpl/lnd_comp_mct.F90 M models/lnd/clm/src/utils/accumulMod.F90 M models/lnd/clm/src/utils/domainMod.F90 --------------------------------------------------- Changes for Unit testing --------------------------------------------------- R models/lnd/clm/src/ED/main/CMakeLists.txt M models/lnd/clm/src/unit_test_shr/unittestSubgridMod.F90 M models/lnd/clm/src/CMakeLists.txt --------------------------------------------------- Configuration changes for new soilbiogeochem/ --------------------------------------------------- M models/lnd/clm/bld/configure CLM testing: unit-tests (models/lnd/clm/src): yellowstone - okay regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - okay yellowstone_pgi - okay yellowstone_gnu - okay goldbach_nag - okay goldbach_intel (moved these to yellowstone_intel for future tests) CLM tag used for the baseline comparisons: clm4_5_1_r096 Changes answers relative to baseline: NO (bit-for-bit) =============================================================== =============================================================== Tag name: clm4_5_1_r096 Originator(s): erik (Erik) Date: Wed Nov 19 02:11:09 MST 2014 One-line Summary: Several answer changing bug-fixes: snow grain size, lake hydrology, default settings, organic soil Purpose of changes: Bring in several bug-fixes most of which change answers. Snow grain size bug that Mark Flanner discovered under snow layer combination. Lake hydrology fix from Zack Subin that would rarely cause the code to abort. Snow depth fix from Sean Swenson. Use Priglent inversion as recommended by Charlie Koven. Correct population density for 2000 conditions from 1850 to peroperly be 2000. Modify all Carbon on spinup from Dave Lawrenece. Add option to square or not square the organic fraction (default is to square for clm4_5 and to NOT for clm5_0). Bug with pervious road that Keith Oleson found. Simplify an if for urban to consistently use a double precision constant. Point to the new CLMNCEP_V5 dataset. For clm4_0 rcp6 and rcp8.5 pftdyn datasets are updated for after 2005. Some fixes that don't change answers. Get the Prigilent inversion and usephfact options working again. Fix a bug in interp_source option that Sean Swenson found. Split out test datasets for getregional script since the datasets all have to be at the same resolution as the domain file. Also read filelist rather than use env variables. Requirements for tag: Fix the bugs below Test level of tag: regular Bugs fixed (include bugzilla ID): 1934 -- snow grain size (both clm4_0 and clm4_5). Just fix on clm4_5 side. 1717 -- lake hydrology fix (clm4_5 only) 1941 -- snowdp fix from Sean (both clm4_0 and clm4_5) (fix in clm4_5_1_r087) 1759 -- ngwh for clm4_0 datasets (apply cesm1_2_x_n10_clm4_5_10) 1772 -- use Priglent inversion 1838 -- pop dens is 1850 for 2000 compsets 1774 -- modify all Carbon on spinup 1765 -- remove duplicate setting of bd and tkdry 1764 -- Bug with pervious road 2066 -- getregional_datasets.pl bug for long lists of files 2067 -- get Prigilent inversion and usephfact options working. 2081 -- point to new CLMNCEP_V5 version 2061 -- make constant consistently double precision rather than have an if around it 2089 -- bug in interp_source that Sean Swenson found Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Add organic_frac_squared=.false. logical as a clm5_0 default feature The old behavior organic_frac_squared=.true. is on as before for clm4_5. List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, dlawren, swensosc, mflanner, subin List any svn externals directories updated (csm_share, mct, etc.): datm update datm to datm8_141113 update CRUNCEP_V5 dataset version used List all files eliminated: None List all files added and what they do: =========== Split out getregional lists (all files in list MUST be at same res as domain file) A models/lnd/clm/test/tools/nl_files/getregional_05popd A models/lnd/clm/test/tools/nl_files/getregional_T62 A models/lnd/clm/test/tools/nl_files/getregional_ndep A models/lnd/clm/tools/shared/ncl_scripts/sample_inlist_0.5popd A models/lnd/clm/tools/shared/ncl_scripts/sample_inlist_ndep A models/lnd/clm/tools/shared/ncl_scripts/sample_outlist_0.5popd A models/lnd/clm/tools/shared/ncl_scripts/sample_outlist_ndep A models/lnd/clm/tools/shared/ncl_scripts/sample_inlist_T62 A models/lnd/clm/tools/shared/ncl_scripts/sample_outlist_T62 List all existing files that have been modified, and describe the changes: =========== Change getregional tests M models/lnd/clm/test/tools/input_tests_master M models/lnd/clm/test/tools/tests_posttag_nompi_regression =========== Bring in M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.pl M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.ncl M models/lnd/clm/tools/shared/ncl_scripts/sample_inlist M models/lnd/clm/tools/shared/ncl_scripts/sample_outlist M models/lnd/clm/bld/unit_testers/build-namelist_test.pl ------------ correct number of tests M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml ---- Update rcp6 and rcp8.5 pftdyn datasets M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ---- Add organic_frac_squared, set fin_use_fsat=.false. by default M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml -- Add organic_frac_squared M models/lnd/clm/bld/namelist_files/use_cases/2000_control.xml ------ Correct year for popd from 1850 to 2000 M models/lnd/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml - Correct year for popd from 1850 to 2000 M models/lnd/clm/bld/CLMBuildNamelist.pm - Add organic_frac_squared and change setup_logic_more_vertlayers to setup_logic_soilstate M models/lnd/clm/src/utils/CMakeLists.txt - Add namelist util to source list M models/lnd/clm/src/biogeochem/ch4Mod.F90 ------------- Pass fsurdat to initCold M models/lnd/clm/src/biogeochem/CNCarbonStateType.F90 -- Use nlevdecomp_full in place of nlevdecomp M models/lnd/clm/src/main/initInterp.F90 --------------- Change use of rbufsli to rbufslo M models/lnd/clm/src/main/clm_initializeMod.F90 -------- Pass nlfilename into soilstate_vars init M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 ---- Add col%itype(c) == icol_road_perv to an if condition M models/lnd/clm/src/biogeophys/SoilTemperatureMod.F90 - Use 1.0_r8 constant always, rather than integer 1 for clm4_5 M models/lnd/clm/src/biogeophys/SnowHydrologyMod.F90 --- Use Mass-weighted combination of radius for combo M models/lnd/clm/src/biogeophys/SoilStateType.F90 ------ Add organic_frac_squared logical and namelist read for it. Add two if's that determine if organic_frac should be squared or not. M models/lnd/clm/src/biogeophys/LakeHydrologyMod.F90 --- Break apart if-condition for snl==-1 CLM testing: regular build-namelist tests: yellowstone yes unit-tests (models/lnd/clm/src): yellowstone yes (although still fails) regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes short tests (aux_clm_short) (generally these are NOT used when making a tag): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes tools testing: (when tools modified or scripts updated (for PTCLM)) yellowstone interactive yes PTCLM (models/lnd/clm/tools/shared/PTCLM/test) yellowstone yes CLM tag used for the baseline comparisons: Changes answers relative to baseline: Yes! Summarize any changes to answers, i.e., - what code configurations: clm4_5 and clm5_0 - what platforms/compilers: all - nature of change (similar climate, except new clm5_0 feature) clm4_0 for rcp6 and rcp8.5 changes answers by using the new good wood harvest datasets for after 2005. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) /home/erik/noorg_clm451r092_I1850CRUCLM45BGC -- clm4_5 default version /home/erik/clm451r092_I1850CRUCLM45BGC -------- clm4_5 with organic_frac_squared=.false. (clm5_0 default version) =============================================================== =============================================================== Tag name: clm4_5_1_r095 Originator(s): andre (Benjamin Andre,UCAR/CSEG,303-497-1391) Date: Mon Nov 10 17:54:18 MST 2014 One-line Summary: refactoring N comp by Jinyun Tang (LBL) and transpiration sink isolation by Gautam Bisht (LBL) Purpose of changes: Bring in two refactorings: Jinyun Tang (LBL) - isolation of the routines to do soil nutrient competition dynamics into a module, and allow for different implementations through runtime polymorphism. Gautam Bisht (LBL) - new function to make transpiration sink distribution independent of subsurface flow physics Test level of tag: regular Bugs fixed (include bugzilla ID): 2039 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: andre, cmt List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none Nutrient Competition List all files added and what they do: A clm/src/biogeochem/NutrientCompetitionFactoryMod.F90 - factory module to select soil nutrient competition method A clm/src/biogeochem/NutrientCompetitionMethodMod.F90 - abstract base class for soil nutrient competition dynamics A clm/src/biogeochem/NutrientCompetitionCLM45defaultMod.F90 - existing clm45 implementation of soil nutrient competition dynamics List all existing files that have been modified, and describe the changes: M clm/src/biogeochem/CNDecompMod.F90 - add nutrient competition method to function parameters M clm/src/biogeochem/CropType.F90 - rename UpdateAccVars() to work around pgi compiler error, remove dependency on temperature_type M clm/src/biogeochem/CNAllocationMod.F90 - move code into clm45 default nutrient competition module M clm/src/biogeochem/CNEcosystemDynMod.F90 - add nutrient competition method to function parameters MM clm/src/main/clm_initializeMod.F90 - add nutrient competition method to function parameters M clm/src/main/clm_driver.F90 - add nutrient competition method to function parameters, call to renamed CropUpdateAccVars M clm/src/main/readParamsMod.F90 - add nutrient competition method to function parameters Transpiration Sink: List all existing files that have been modified, and describe the changes: M clm/src/biogeophys/SoilWaterMovementMod.F90 - move transpiration sink into separate function so it is independent of physics. CLM testing: build-namelist tests: n/a unit-tests (models/lnd/clm/src): no yellowstone regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 4.0 ok; 4.5 ok yellowstone_pgi - 4.0 ok; 4.5 ok yellowstone_gnu - n/a; 4.5 ok goldbach_nag - 4.0 ok; 4.5 ok goldbach_intel - 4.0 ok; 4.5 ok short tests (aux_clm_short) - no tools testing: (when tools modified or scripts updated (for PTCLM)) - n/a CLM tag used for the baseline comparisons: clm4_5_1_r094 Changes answers relative to baseline: no =============================================================== =============================================================== Tag name: clm4_5_1_r094 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Nov 7 13:43:38 MST 2014 One-line Summary: misc. glacier-related updates Purpose of changes: (1) Add dlnd, satm and srof externals, so that TG compsets (CISM-only, forced by dlnd) can be run from a CLM tag. This will facilitate CISM development and testing. (2) Remove CLM's dependence on the CISM grid. Previously, CLM used the CISM grid to determine which fglcmask file to use. But the differences between the fglcmask files were inconsequential (all of them included the full area of Greenland, which is what was important). I have created a new set of fglcmask files that are independent of the CISM grid, and point CLM to these new files. This will make it easier to add new CISM grids in the future, because no changes will be needed in CLM for this purpose. However, note that the use of these new files means that the number of virtual landunits & columns changes for glcmec runs. (3) In subgridAveMod, fix c2l routines: change pft%wtlunit to col%wtlunit (bugz 2077) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 2077 (c2l references pft instead of col) - 2085 (listDefaultNamelist is broken) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: change fglcmask List any changes to the defaults for the boundary datasets: change fglcmask Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): ========= Added new externals for the sake of running TG compsets (CISM-only) from a CLM tag +models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 +models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm +models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/srof ========= Other externals updates -tools/unit_testing https://svn-ccsm-models.cgd.ucar.edu/unit_testing/trunk_tags/unit_testing_0_07 +tools/unit_testing https://svn-ccsm-models.cgd.ucar.edu/unit_testing/trunk_tags/unit_testing_0_08 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: ========= Remove dependence on CISM grid M models/lnd/clm/bld/listDefaultNamelist.pl - also fix bug 2085 M models/lnd/clm/bld/unit_testers/build-namelist_test.pl M models/lnd/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M models/lnd/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M models/lnd/clm/bld/clm.buildnml.csh M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_overall.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults.xsl M models/lnd/clm/bld/user_nl_clm M models/lnd/clm/bld/CLMBuildNamelist.pm M models/lnd/clm/doc/UsersGuide/custom.xml M models/lnd/clm/src_clm40/main/controlMod.F90 M models/lnd/clm/src_clm40/main/clm_varctl.F90 M models/lnd/clm/src/main/controlMod.F90 M models/lnd/clm/src/main/clm_varctl.F90 ========= Change pft%wtlunit to col%wtlunit in c2l routines (which currently aren't called from anywhere in the code) (bugz 2077) M models/lnd/clm/src/main/subgridAveMod.F90 ========= fix numbers of build-namelist unit test failures, due to removal of a test M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: ok expected failure in 23, due to change in fglcmask unit-tests (models/lnd/clm/src): yellowstone: still broken, due to internal compiler error roo2 (mac laptop): ok See notes in clm4_5_1_r090. Point (2) has been fixed, but point (1) remains. regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu: ok goldbach_nag: ok goldbach_intel: ok CLM tag used for the baseline comparisons: clm4_5_1_r093 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: IG compsets - both CLM40 and CLM45 - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): only changes diagnostic cpl hist fields - no change in the simulation The changes arise from the new glcmask files, which exclude a few points from the glcmask that used to be included. These points are all outside of Greenland, so they are not important for coupling to CISM. However, it means that a few virtual columns have been removed. This, in turn, changes the values of some l2x topo, tsrf and qice fields sent to the coupler. But this does NOT feed back on the simulation in any way. Some tests also exhibit diffs in the CLM diagnostic fields PCT_GLC_MEC and QICE_FORC. Again, these are due to changes in where we have virtual columns, and do not affect the simulation. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r093 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Nov 7 13:17:47 MST 2014 One-line Summary: change cold-start snow initialization, update cism external Purpose of changes: (1) Change cold-start snow initialization logic. The original logic did different snow initialization depending on whether we are inside or outside the glcmask. That's a problem in that answers change depending on the glcmask. The new logic instead uses a latitude threshold for determining where to initialize a non-zero snow pack. Note that this will change answers for all cold-start cases, including non-glcmec cases. (2) Update CISM to version 2. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self; Dave Lawrence agreed with the change to snow initialization List any svn externals directories updated (csm_share, mct, etc.): -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140916 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism2_0_02 List all files eliminated: List all files added and what they do: List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/main/clm_initializeMod.F90 CLM testing: regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu: ok goldbach_nag: ok goldbach_intel: ok CLM tag used for the baseline comparisons: clm4_5_1_r092 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: (1) any CLM4.5 case with cold start initialization, due to change in snow initialization (2) any case that includes CISM, due to answer changes in the CISM external I carefully checked the yellowstone-intel clm4.5 tests to ensure that: (a) FAILed compare_hist all had finidat = ' ' (b) PASSed compare_hist either had non-blank finidat OR were single-point - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated, but expected to be larger than roundoff/same climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r092 Originator(s): muszala (Stefan Muszala) Date: Tue Nov 4 06:10:16 MST 2014 One-line Summary: bug fixes from santos that address valgrind problems. update rtm external Purpose of changes: Addresses issues found with Valgrind by Santos. Update RTM. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: santos, self List any svn externals directories updated (csm_share, mct, etc.): -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_39 +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_40 List all files eliminated: N/A List all files added and what they do: A + models/lnd/clm/src/main/dtypes.h List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml -- update failures M SVN_EXTERNAL_DIRECTORIES -- rtm update to 40 M models/lnd/clm/src_clm40/main/ncdio_pio.F90 M models/lnd/clm/src_clm40/main/ncdio_pio.F90.in M models/lnd/clm/src/main/ncdio_pio.F90 M models/lnd/clm/src/main/ncdio_pio.F90.in -- example changes: - status = pio_inq_vardimid(ncid, vardesc , dids) + status = pio_inq_vardimid(ncid, vardesc , dids(1:ndims)) M models/lnd/clm/src_clm40/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/biogeophys/FrictionVelocityMod.F90 -- change 1.0_8 to 1.0_r8 CLM testing: build-namelist tests: N/A unit-tests (models/lnd/clm/src): N/A regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel 40- OK 45- OK yellowstone_pgi 40- OK 45- OK yellowstone_gnu 40- N/A 45- OK goldbach_nag 40- OK 45- OK goldbach_intel 40- OK 45- OK tools testing: (when tools modified or scripts updated (for PTCLM)) N/A CLM tag used for the baseline comparisons: clm4_5_1_r091 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_1_r091 Originator(s): muszala (Stefan Muszala) Date: Mon Oct 27 09:48:56 MDT 2014 One-line Summary: update externals. fix bug so CLM runs with Intel 14x. Purpose of changes: Update externals. Fix bug in VOCEmissionMod.F90 that prevented CLM from running with Intel 14x on yellowstone. Bring in workaround for bug 1730 from Sacks. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ - see CLM test fail list Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, clm developers, particularly Bill Sacks. List any svn externals directories updated (csm_share, mct, etc.): -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_141009 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_141023 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_141001 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_141017a -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_141003 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_141022 -models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_140416 +models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_140925 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: -- reflect changes in new testlists M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml -- Sacks' workaround for bug 1730 M models/lnd/clm/src/main/histFileMod.F90 M models/lnd/clm/src/main/ncdio_pio.F90 M models/lnd/clm/src/main/ncdio_pio.F90.in -- remove duplicate assignment of 0_r8 to meg_out(imeg)%flux_out M models/lnd/clm/src/biogeochem/VOCEmissionMod.F90 -- update externals M SVN_EXTERNAL_DIRECTORIES CLM testing: Please view the CLM expected fail list for new test failures. They are matched to bugzilla bug ids. regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 40 - OK 45 - OK yellowstone_pgi - 40 - OK 45 - OK yellowstone_gnu - 40 - N/A 45 - OK goldbach_nag - 40 - OK 45 - OK goldbach_intel - 40 - OK 45 - OK CLM tag used for the baseline comparisons: clm4_5_1_r090 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_1_r090 Originator(s): sacks (Bill Sacks) Date: Thu Oct 16 06:39:52 MDT 2014 One-line Summary: modularize irrigation; do some unit test rework Purpose of changes: (1) Pull irrigation code out of CanopyFluxes and CanopyHydrology, into its own module (2) Pull out the locally-created filters from CanopyFluxes and BareGroundFluxes into filterMod, in order to support pulling irrigation out of CanopyFluxes. This will also be needed to support pulling other hydrology stuff out of CanopyFluxes. (3) Add unit tests for irrigation (4) Rework some irrigation infrastruture, and add some more unit test utility routines Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 2063 (HumanIndexMod fails to compile with gfortran) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: muszala List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: ========= Renamed to unit_test_stubs D models/lnd/clm/src/unit_test_mocks/util_share/ncdio_var.F90 D models/lnd/clm/src/unit_test_mocks/util_share/restUtilMod_mock.F90 D models/lnd/clm/src/unit_test_mocks/util_share/GetGlobalValuesMod_mock.F90 D models/lnd/clm/src/unit_test_mocks/util_share/ncdio_var.F90.in D models/lnd/clm/src/unit_test_mocks/util_share/restUtilMod_mock.F90.in D models/lnd/clm/src/unit_test_mocks/util_share/clm_time_manager_mock.F90 D models/lnd/clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90 D models/lnd/clm/src/unit_test_mocks/util_share/spmdMod_mock.F90 D models/lnd/clm/src/unit_test_mocks/util_share/do_genf90 D models/lnd/clm/src/unit_test_mocks/util_share/CMakeLists.txt D models/lnd/clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90.in D models/lnd/clm/src/unit_test_mocks/util_share D models/lnd/clm/src/unit_test_mocks/csm_share/shr_mpi_mod_mock.F90 D models/lnd/clm/src/unit_test_mocks/csm_share/mct_mod_mock.F90 D models/lnd/clm/src/unit_test_mocks/csm_share/CMakeLists.txt D models/lnd/clm/src/unit_test_mocks/csm_share D models/lnd/clm/src/unit_test_mocks/main/CMakeLists.txt D models/lnd/clm/src/unit_test_mocks/main/histFileMod_mock.F90 D models/lnd/clm/src/unit_test_mocks/main D models/lnd/clm/src/unit_test_mocks/dyn_subgrid/dynFileMod_mock.F90 D models/lnd/clm/src/unit_test_mocks/dyn_subgrid/CMakeLists.txt D models/lnd/clm/src/unit_test_mocks/dyn_subgrid D models/lnd/clm/src/unit_test_mocks/CMakeLists.txt D models/lnd/clm/src/unit_test_mocks ========= Remove unnecessary files D models/lnd/clm/src/ED/CMakeLists.txt D models/lnd/clm/src/ED/biogeophys/CMakeLists.txt List all files added and what they do: ========= Pull out irrigation code into its own module A models/lnd/clm/src/biogeophys/IrrigationMod.F90 ========= Add some unit test utility code (and some tests for the utility code) A models/lnd/clm/src/unit_test_shr/unittestFilterBuilderMod.F90 A models/lnd/clm/src/unit_test_shr/unittestSimpleSubgridSetupsMod.F90 A models/lnd/clm/src/unit_test_shr/test/unittestFilterBuilder_test/test_filterBuilder.pf A models/lnd/clm/src/unit_test_shr/test/unittestFilterBuilder_test/CMakeLists.txt A models/lnd/clm/src/unit_test_shr/test/unittestFilterBuilder_test A models/lnd/clm/src/unit_test_shr/test/CMakeLists.txt A models/lnd/clm/src/unit_test_shr/test ========= Renamed from unit_test_mocks to unit_test_stubs; also renamed individual files from mock to stub (or 'fake' for ncdio_pio, because it does more than a stub); also, moved some stubs to match the current organization of the main source tree A models/lnd/clm/src/unit_test_stubs/utils/restUtilMod_stub.F90 A models/lnd/clm/src/unit_test_stubs/utils/do_genf90 A models/lnd/clm/src/unit_test_stubs/utils/restUtilMod_stub.F90.in A models/lnd/clm/src/unit_test_stubs/utils/CMakeLists.txt A models/lnd/clm/src/unit_test_stubs/utils/clm_time_manager_stub.F90 A models/lnd/clm/src/unit_test_stubs/utils/spmdMod_stub.F90 A models/lnd/clm/src/unit_test_stubs/utils A models/lnd/clm/src/unit_test_stubs/csm_share/shr_mpi_mod_stub.F90 A models/lnd/clm/src/unit_test_stubs/csm_share/mct_mod_stub.F90 A models/lnd/clm/src/unit_test_stubs/csm_share/CMakeLists.txt A models/lnd/clm/src/unit_test_stubs/csm_share A models/lnd/clm/src/unit_test_stubs/main/histFileMod_stub.F90 A models/lnd/clm/src/unit_test_stubs/main/ncdio_var.F90 A models/lnd/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90 A models/lnd/clm/src/unit_test_stubs/main/ncdio_var.F90.in A models/lnd/clm/src/unit_test_stubs/main/GetGlobalValuesMod_stub.F90 A models/lnd/clm/src/unit_test_stubs/main/do_genf90 A models/lnd/clm/src/unit_test_stubs/main/ncdio_pio_fake.F90.in A models/lnd/clm/src/unit_test_stubs/main/CMakeLists.txt A models/lnd/clm/src/unit_test_stubs/main A models/lnd/clm/src/unit_test_stubs/dyn_subgrid/dynFileMod_stub.F90 A models/lnd/clm/src/unit_test_stubs/dyn_subgrid/CMakeLists.txt A models/lnd/clm/src/unit_test_stubs/dyn_subgrid A models/lnd/clm/src/unit_test_stubs/CMakeLists.txt A models/lnd/clm/src/unit_test_stubs ========= Add dependencies (direct & indirect) of IrrigationMod A models/lnd/clm/src/biogeochem/CMakeLists.txt ========= Add unit tests for irrigation (see README file for some design notes) A models/lnd/clm/src/biogeophys/test/Irrigation_test/test_irrigation_deficit.pf A models/lnd/clm/src/biogeophys/test/Irrigation_test/test_irrigation_multipatch.pf A models/lnd/clm/src/biogeophys/test/Irrigation_test/IrrigationWrapperMod.F90 A models/lnd/clm/src/biogeophys/test/Irrigation_test/test_irrigation_singlepatch.pf A models/lnd/clm/src/biogeophys/test/Irrigation_test/CMakeLists.txt A models/lnd/clm/src/biogeophys/test/Irrigation_test/README A models/lnd/clm/src/biogeophys/test/Irrigation_test List all existing files that have been modified, and describe the changes: ========= Pull irrigation out of CanopyFluxes into its own routine, and also pull out the filters that used to be created locally in CanopyFluxes and BareGroundFluxes M models/lnd/clm/src/main/clm_driver.F90 ========= Pull out filters that used to be created locally in CanopyFluxes and BareGroundFluxes, so that they can be reused in irrigation and elsewhere M models/lnd/clm/src/main/filterMod.F90 ========= Add calls to irrigation init & restart routines MM models/lnd/clm/src/main/clm_initializeMod.F90 M models/lnd/clm/src/main/restFileMod.F90 ========= Irrigation computations are now done in the irrigation module; also, the relevant filter is now created outside CanopyFluxes MM models/lnd/clm/src/biogeophys/CanopyFluxesMod.F90 ========= Irrigation computations are now done in the irrigation module M models/lnd/clm/src/biogeophys/CanopyHydrologyMod.F90 M models/lnd/clm/src/biogeophys/LakeHydrologyMod.F90 ========= Irrigation variables are now defined in the irrigation module M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/biogeophys/WaterfluxType.F90 M models/lnd/clm/src/biogeophys/HydrologyDrainageMod.F90 ========= Filter is now created outside BareGroundFluxes; also, moved some bare ground initialization from CanopyFluxes (needed because the filters are no longer created locally, so CanopyFluxes does not know what points it would need to set for bare ground) MM models/lnd/clm/src/biogeophys/BareGroundFluxesMod.F90 ========= Remove some dependencies to aid unit testing (this includes combining two branches of a conditional (allowlakeprod) that were doing the same thing M models/lnd/clm/src/biogeophys/SoilStateType.F90 ========= Add dependencies (direct & indirect) of IrrigationMod for unit testing M models/lnd/clm/src/utils/CMakeLists.txt M models/lnd/clm/src/main/CMakeLists.txt M models/lnd/clm/src/ED/main/CMakeLists.txt M models/lnd/clm/src/CMakeLists.txt M models/lnd/clm/src/biogeophys/CMakeLists.txt M models/lnd/clm/src/biogeophys/test/CMakeLists.txt ========= Add new unit test utilities M models/lnd/clm/src/unit_test_shr/CMakeLists.txt ========= Make newly-added subgrid units active by default M models/lnd/clm/src/unit_test_shr/unittestSubgridMod.F90 ========= Unrelated fix for gfortran MM models/lnd/clm/src/biogeophys/HumanIndexMod.F90 ========= Change whitespace M models/lnd/clm/src/dyn_subgrid/CMakeLists.txt CLM testing: build-namelist tests: yellowstone: not run unit-tests (models/lnd/clm/src): yellowstone: fail due to two issues: (1) The unit tests currently won't build on yellowstone due to an ICE that will probably be fixed when we remove dependencies of SoilStateType. (2) In addition, even once that's fixed, the yellowstone unit tests either need (a) a bump in the unit testing external (unit_testing_0_08) and Machines external (Machines_141007) (I didn't do that for my tag because it pulls in a bump in the intel compiler version to 14 rather than 13.1), or (b) the following diffs: Index: tools/unit_testing/python/machine_setup.py =================================================================== --- tools/unit_testing/python/machine_setup.py (revision 64421) +++ tools/unit_testing/python/machine_setup.py (working copy) @@ -52,7 +52,7 @@ mod.load("ncarenv/1.0") mod.load("ncarbinlibs/1.0") if compiler == "intel": - mod.load("intel/13.1.2") + mod.load("intel/14.0.2") elif compiler == "pgi": mod.load("pgi/13.9") mod.load("ncarcompilers/1.0") Index: scripts/ccsm_utils/Machines/config_compilers.xml =================================================================== --- scripts/ccsm_utils/Machines/config_compilers.xml (revision 64421) +++ scripts/ccsm_utils/Machines/config_compilers.xml (working copy) @@ -547,7 +547,7 @@ -xHost -xHost $(TRILINOS_PATH) - /glade/u/home/santos/pFUnit/pFUnit_Intel + /glade/u/home/sacks/pFUnit/pFUnit3.0.1_Intel14.0.2_Serial However, I have run the unit tests on my mac, with gfortran, and they all pass regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu: NOT RUN goldbach_nag: ok goldbach_intel: ok CLM tag used for the baseline comparisons: clm4_5_1_r089 Changes answers relative to baseline: YES (but only because of bug 1998) If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: irrigation - what platforms/compilers: all - nature of change: larger than roundoff, not investigated closely The changes are entirely due to the btran bug (bug 1998): The old flow was: - compute btran - calculate irrigation (depends on btran) - hack btran for soybeans whereas the new flow is: - call CanopyFluxes: computes btran and hacks btran for soybeans - calculate irrigation (depends on btran) I have confirmed that answers are bit-for-bit for both irrigation tests (for both cpl and clm hist files), when I introduce the following diffs in both the trunk and the branch: Index: src/biogeophys/CanopyFluxesMod.F90 =================================================================== --- src/biogeophys/CanopyFluxesMod.F90 (revision 64406) +++ src/biogeophys/CanopyFluxesMod.F90 (working copy) @@ -842,9 +842,6 @@ btran(p) = min(1._r8, btran(p) * 3.33_r8) end if end if - if (pft%itype(p) == nsoybean .or. pft%itype(p) == nsoybeanirrig) then - btran(p) = min(1._r8, btran(p) * 1.25_r8) - end if end do if ( use_ed ) then @@ -894,9 +891,6 @@ btran(p) = min(1._r8, btran(p) * 3.33_r8) end if end if - if (pft%itype(p) == nsoybean .or. pft%itype(p) == nsoybeanirrig) then - btran(p) = min(1._r8, btran(p) * 1.25_r8) - end if end do call Photosynthesis (bounds, fn, filterp, & =============================================================== =============================================================== Tag name: clm4_5_1_r089 Originator(s): erik (Erik) Date: Mon Oct 13 13:46:43 MDT 2014 One-line Summary: Bring new urban building temperature to trunk as a clm5.0 feature as well as human-stress index calculations Purpose of changes: New prognostic internal building air temperature methodology for CLM5.0. Retain the older simpler method for CLM4.5. The namelist toggle to switch between them is: building_temp_method. By default for clm4_5 physics the older method is used and for clm5_0 the newer one is used. Also add in a package of human-stress index calculations. Again this is by default on for clm5_0 and off for clm4_5. The new building air temperature methodology, solves the system of equations for internal: air, roof, floor, and wall (shade and sunlit) Temperatures. It uses the LAPACK subroutine DGESV to solve the system. It also figures out the energy flux needed to either cool the building air temperature to a maximum allowed temperature or to heat it to the minimum allowed temperature. Add in the new load balancing tool and the PyReshaper tool (changes multi-variable monthly history files into single-variable time-series files). Remove the now unneeded clm4_5 interpinic as well as the mkmapgrids FORTRAN program. Requirements for tag: Bring in new clm5_0 building temperature as option, Fix scripts bugs, bug 2053/2032 Test level of tag: regular Bugs fixed (include bugzilla ID): 2053 Problems with dry-deposition for fully coupled cases with CLM4.5 in cesm1_3_beta13 2032 rtm.buildnml.csh kills cesm_setup if GET_REFCASE is FALSE 1685 Drydeposition potentially using "rs" variable before it's defined (over water) (was fixed but came back) Fix bugs: 2024, 2035, 2037 in scripts SBN and namelistcompare issues Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: new namelists and namelist items add: clmu_inparm and clm_humanindex_inparm namelists move: urban_hac and urban_traffic to clmu_inparm add: calc_human_stress_indices to clm_humanindex_inparm building_temp_method to clmu_inparm List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: clm5_0 memory use is greater than clm4_5 due to human_stress_indices calculations and output on history (I've seen it make up to a 20% difference) Code reviewed by: self, oleson, sacks, mvertens, andre List any svn externals directories updated (csm_share, mct, etc.): scripts, Machines, rtm, cism, csm_share Update to cesm1_3_alpha13c externals. Also add in load_balancing_tool and pyReshaper scripts to scripts4_141009 Machines to Machines_141001 rtm to rtm1_0_39 cism to cism1_140916 csm_share to share3_141003 load_balancing_tool to load_balancing_tool_140818 pyReshaper to v0.9.1 List all files eliminated: --------- Remove clm4_5 interpinic, online interpinic supersedes it D models/lnd/clm/tools/clm4_5/interpinic D models/lnd/clm/tools/clm4_5/interpinic/* D models/lnd/clm/tools/clm4_5/interpinic/src/* --------- Remove mkmapgrids program D models/lnd/clm/tools/shared/mkmapgrids/src D models/lnd/clm/tools/shared/mkmapgrids/src/* D models/lnd/clm/tools/shared/mkmapgrids/mkmapgrids.namelist D models/lnd/clm/tools/shared/mkmapgrids/mkmapgrids.csh --------- shr_sys_mod was mocked only because of it's use of shr_mpi_mod --------- mock shr_mpi_mod instead and standard shr_sys_mod can be used D models/lnd/clm/src/unit_test_mocks/csm_share/shr_sys_mod_mock.F90 List all files added and what they do: A models/lnd/clm/src/biogeophys/HumanIndexMod.F90 -- New module to calculate a bunch of human stress index values. A models/lnd/clm/src/biogeophys/UrbBuildTempOleson2015Mod.F90 --- New module for calculating the prognostic internal building air temperature. --------- New simple unit tester for humanstress indices module A models/lnd/clm/src/biogeophys/test/HumanStress_test/test_humanstress.pf A models/lnd/clm/src/biogeophys/test/HumanStress_test/CMakeLists.txt A models/lnd/clm/src/biogeophys/test/HumanStress_test A models/lnd/clm/src/unit_test_mocks/csm_share/shr_mpi_mod_mock.F90 -- shell for most shr_mpi_ calls that do nothing (so assumes MPI is NOT being done) shr_mpi_abort does a stop List all existing files that have been modified, and describe the changes: --------- remove the mkgriddata and clm4_5 interpinic tools from testing M models/lnd/clm/test/tools/input_tests_master M models/lnd/clm/test/tools/tests_posttag_yong M models/lnd/clm/test/tools/tests_posttag_nompi_regression M models/lnd/clm/test/tools/tests_pretag_yellowstone_nompi --------- remove documentation of mkgriddata and clm4_5 interpinic tools --------- but add documentation on ncl script M models/lnd/clm/tools/shared/mkmapgrids/README M models/lnd/clm/tools/README --------- Add new namelists: clmu_inparm and clm_humanindex_inparm M models/lnd/clm/bld/configure ---- use same configuration for clm4_5 AND clm5_0 M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Set calc_human_stress_indices, and building_temp_method by clm4_5/clm5_0 M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Move urban_hac and urban_traffic to new clmu_inparm namelist and add building_temp_method and add calc_human_stress_indices to clm_humanindex_inparm namelist M models/lnd/clm/bld/CLMBuildNamelist.pm -------- Handle new namelists: clmu_inparm and clm_humanindex_inparm --------- Get unit tests working again, and add a simple humanindex test M models/lnd/clm/src/utils/CMakeLists.txt M models/lnd/clm/src/ED/main/CMakeLists.txt M models/lnd/clm/src/ED/biogeophys/CMakeLists.txt M models/lnd/clm/src/README.unit_testing M models/lnd/clm/src/biogeophys/CMakeLists.txt M models/lnd/clm/src/biogeophys/test/CMakeLists.txt M models/lnd/clm/src/unit_test_mocks/util_share/spmdMod_mock.F90 - set mpicom M models/lnd/clm/src/unit_test_mocks/csm_share/CMakeLists.txt M models/lnd/clm/src/CMakeLists.txt --------- Fix so can work with drydeposition namelist and without megan namelist M models/lnd/clm/src/biogeochem/VOCEmissionMod.F90 - Don't initialize if megan namelist is turned off (bug 2053) M models/lnd/clm/src/biogeochem/DryDepVelocity.F90 - Make sure rs is set before used (bug 1685) --------- Handle new building temperature options, add new constants MM models/lnd/clm/src/main/clm_varcon.F90 -------- Bunch of new constants for urban-building (should be moved to modules that use them). Also pass is_simple_buildtemp into init method MM models/lnd/clm/src/main/clm_initializeMod.F90 - Pass building temp type down to relevent init methods add initialization for humanindex_vars, initialize drydepvel_vars (bug 2053) M models/lnd/clm/src/main/restFileMod.F90 ------- Pass building temp type logicals down to energyflux_vars and temperature_vars restart methods M models/lnd/clm/src/main/LandunitType.F90 ------ Add documentation, correct error in documentation MM models/lnd/clm/src/main/controlMod.F90 -------- Move urban namelist items to UrbanReadNML, add HumanIndexReadNML M models/lnd/clm/src/main/clm_driver.F90 -------- Pass humanindex_vars down as needed --------- Change for new shr_cal_mod names M models/lnd/clm/src/ED/biogeophys/EDPhenologyMod.F90 - Use full name of month "january" instead of "jan" --------- Add new building temperature module and add capability to do old clm4_5 --------- method as well as new method, also do human_stress_indices calculations MM models/lnd/clm/src/biogeophys/SoilTemperatureMod.F90 --- Add BuildingHAC for simple building temp method (should move to it's own module) add if statements for building_temp_method type, call BuildingTemperature when prognostic method used M models/lnd/clm/src/biogeophys/SoilFluxesMod.F90 -------- Change name of eflx_building_heat to eflx_building_heat_errsoi M models/lnd/clm/src/biogeophys/TemperatureType.F90 ------ Add building temperature variables (should move to urbBuildTemp module), pass building_temp method logical down for initialization, add documentation headers hist, cold, and restart init depends on building temp method logical M models/lnd/clm/src/biogeophys/LakeFluxesMod.F90 -------- calc_human_stress_indices (should move to method in humanIndexMod) MM models/lnd/clm/src/biogeophys/BareGroundFluxesMod.F90 -- calc_human_stress_indices (should move to method in humanIndexMod) MM models/lnd/clm/src/biogeophys/CanopyFluxesMod.F90 ------ calc_human_stress_indices (should move to method in humanIndexMod) M models/lnd/clm/src/biogeophys/EnergyFluxType.F90 ---- change name of eflx_building_heat_col to eflx_building_heat_errsoi_col, add some new building temperature flux terms add documentation, alloc, hist, restart and cold initialization depends on building temperature method type (should move to urban building module) M models/lnd/clm/src/biogeophys/UrbanFluxesMod.F90 ---- Add private functions: wasteheat, simple_wasteheatfromac, calc_simple_internal_building_temp (should move to building_temp modules). calc_human_stress_indices (should move to method in humanIndexMod) M models/lnd/clm/src/biogeophys/UrbanParamsType.F90 --- Add methods: UrbanReadNML, IsSimpleBuildTemp, IsProgBuildTemp, add clmu_inparm namelist, and move urban_* items there and add building_temp_method to it. CLM testing: build-namelist tests: yellowstone yes unit-tests (models/lnd/clm/src): yellowstone yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes short tests (aux_clm_short): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes tools testing: yellowstone interactive yes PTCLM (models/lnd/clm/tools/shared/PTCLM/test) yellowstone yes CLM tag used for the baseline comparisons: clm4_5_1_r088 Changes answers relative to baseline: No for CLM40 and CLM45 But, answers DO change for CLM50 (except scripts tag update changes history files for IG and irrigation compsets/tests) =============================================================== =============================================================== Tag name: clm4_5_1_r088 Originator(s): muszala (Stefan Muszala) Date: Wed Oct 1 09:24:43 MDT 2014 One-line Summary: Pull out ED deps. in TemperatureTypeMod, can now compile with pgi 14.7 Purpose of changes: Pull out the dependency on EDBioType in TemperatureType.F90. The ED variables related to phenology now reside in EDPhenologyMod.F90. This refactor also had the effect of getting past a PGI 14.7 ICE which looks like it was due to the use of EDbio_vars in TemperatureType.F90. When I pulled out lines 1227 and 1226 of biogeophys/TemperatureType.F90 (in clm4_5_1_r087) and passed the two EDbio_vars variables through the argument list the ICE went away. This tag breaks ED restart tests. We went ahead with the tag because we had to fix a more general problem with the CESM and CAM builds and PGI 14.7. the ED v0.1.0 branch does not have these modifications and may be used as an alternative. A new clm tag will shortly follow that addresses any remaining problems. Requirements for tag: None Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): Temporary addition of a cal parameter type in a branch tag. Will be merged into csm_share trunk shortly. -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/branch_tags/share_ece_tags/share_ece_01_140723 List all files eliminated: N/A List all files added and what they do: ! new home for ED phenology variables and type-bound procedures that ! allow for accumulation of buffers A + models/lnd/clm/src/ED/biogeophys/EDPhenologyMod.F90 ! put some CMakeLists.txt in place for ED unittests A + models/lnd/clm/src/ED/main/CMakeLists.txt A + models/lnd/clm/src/ED/biogeophys/CMakeLists.txt A + models/lnd/clm/src/ED/CMakeLists.txt List all existing files that have been modified, and describe the changes: ! pull out deps. on EDBioType M models/lnd/clm/src/biogeophys/TemperatureType.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/main/clm_initializeMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/main/restFileMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/main/clm_driver.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/biogeochem/EDPhysiologyMod.F90 ! pull out ED_GDD_patch and phen_cd_status_patch M models/lnd/clm/src/ED/main/EDBioType.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/biogeochem/EDPhysiologyMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/main/EDMainMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/main/EDCLMLinkMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/main/EDRestVectorMod.F90 ! add code for new class instance, EDphenology_inst M models/lnd/clm/src/ED/main/EDInitMod.F90 ! for ED unit tests M models/lnd/clm/src/CMakeLists.txt ! update CNED failures M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: ERS CNED tests are failing in this tag. It is expected. expectedClmTestFails.xml is updated to reflect this. build-namelist tests: N/A yellowstone regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel 40-OK, 45-OK yellowstone_pgi 40-OK, 45-OK goldbach_nag 40-OK, 45-OK goldbach_intel 40-OK, 45-OK CLM tag used for the baseline comparisons: clm4_5_1_r087 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_1_r087 Originator(s): erik (Erik) Date: Tue Sep 30 12:07:10 MDT 2014 One-line Summary: Fix two balance check errors, and turn abort for balance check back on to appropriate levels Purpose of changes: Fix two balance check errors that were causing problems for simulations. Also some of the balance check aborts were turned off in clm4_5_1_r082, so turn them back on again. Tighten water balance error from 1.e-4 to 1.e-5. Tighten LW, surface-flux and solar radiation balance errors from 1.e-3 to 1.e-5 and add warning for 1.e-7. Turn surface-flux balance and soil balance check errors abort back on. Soil balance tightened to 1.e-4 (from 1.e-3) with warnings shown at 1.e-6. Also bring in an update to PTCLM, and allow tools tester to be submitted to geyser or caldera. Just as an aside as something that was already done. Requirements for tag: Fix bugs: 2026 and 1941 Test level of tag: regular Bugs fixed (include bugzilla ID): 2026 Soil balance error 1941 snowdp balance error Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, swensosc List any svn externals directories updated (csm_share, mct, etc.): Update PTCLMmkdata version Update PTCLM to PTCLM2_140816 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/test/tools/test_driver.sh ------ Allow to run in caldera and geyser as well M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 ---- Add aborts back as well as warnings and tighten some error conditions and warnings. M models/lnd/clm/src/biogeophys/SoilTemperatureMod.F90 - Remove calculation and use of the heat capacity of frozen h2osfc layer but use the heat capacity of the liquid layer as balance check doesn't know about the frozen, and the discrepency causes balance check errors. M models/lnd/clm/src/biogeophys/SnowHydrologyMod.F90 --- snow includes dew. CLM testing: build-namelist tests: yellowstone YES regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel YES yellowstone_pgi YES yellowstone_gnu (optional) YES goldbach_nag YES goldbach_intel YES CLM tag used for the baseline comparisons: clm4_5_1_r086 Changes answers relative to baseline: YES! Summarize any changes to answers, i.e., - what code configurations: CLM4.5 - what platforms/compilers: ALL - nature of change: same climate (some shorter simulations are still exact) Simulations that Keith ran to test the snowdp change are: /glade/p/work/oleson/urb2dev_n00_clm4_5_52/scripts/urb2dev_n03_clm4_5_64_V2DomLam_I20TRCRUCLM45 /glade/p/work/oleson/urb2dev_n00_clm4_5_52/scripts/urb2dev_n03_clm4_5_64_V2DomLam_IRCP85CRUCLM45 URL for LMWG diagnostics output used to validate new climate: For soil balance error... http://www.cgd.ucar.edu/staff/swensosc/public/diagnostics/ColdtestTRENDYspinupf091850CRU-ColdtestTRENDYspinupf091850CRU_control/setsIndex.html =============================================================== =============================================================== Tag name: clm4_5_1_r086 Originator(s): muszala (Stefan Muszala) Date: Thu Sep 25 09:04:08 MDT 2014 One-line Summary: critical ED modifications from r fisher, fix bug 2043 Purpose of changes: add modifications to ED, particularly for cold deciduous. add fix for bug 2043. Consider these ED baselines as fixed (ie. unless you are modifying ED science, these should now be BFB). Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): 2043 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, r fisher List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/main/lnd2atmMod.F90 -- change intent of waterstate_vars to inout (fixes bug 2043) M models/lnd/clm/src/main/clm_driver.F90 -- add EDbio_vars to edmodel actual argument list M models/lnd/clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 -- change leaves_off_switch and laimemory handling M models/lnd/clm/src/ED/biogeochem/EDPhysiologyMod.F90 -- rework cold deciduous and threshold code. add fragmentation_scaler routine M models/lnd/clm/src/ED/main/EDMainMod.F90 -- change argument lists to include EDbio_vars for ecosystem_dynamics and phenology M models/lnd/clm/src/ED/biogeophys/EDSurfaceAlbedoMod.F90 -- overhaul ED norman radiation code M models/lnd/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 -- tweak calculation of jmax25top and tpu25top M models/lnd/clm/src/biogeophys/SurfaceAlbedoMod.F90 -- clean up two comments CLM testing: ED compsets change values. For bug 2043. Confirmed that a gnu compile on yellowstone gets passed the intent problem. build-namelist tests: N/A regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 40 OK - 45 OK yellowstone_pgi - 40 OK - 45 OK goldbach_nag - 40 OK - 45 OK goldbach_intel - 40 OK - 45 OK CLM tag used for the baseline comparisons: clm4_5_1_r085 Changes answers relative to baseline: Only for ED compsets =============================================================== =============================================================== Tag name: clm4_5_1_r085 Originator(s): sacks (Bill Sacks) Date: Fri Sep 19 10:22:30 MDT 2014 One-line Summary: replace conditionals with polymorphism for soil water retention curve Purpose of changes: The main motivation for this tag was the need to introduce a soil_suction_inverse routine, which will be used for irrigation. It is important that soil_suction_inverse remains consistent with soil_suction for every soil water retention curve method. In talking with Ben Andre and Erik, we felt the best way to ensure this consistency was to have a separate, small module for each soil retention curve method. We felt the best way to implement this was via polymorphism. Polymorphism is arguably overkill in this simple case, but we thought it would be good to convert it to polymorphism partly as an example that we and others can follow in more complex cases where it will provide greater benefit. To add a new soil retention curve method: (1) Create a module similar to SoilWaterRetentionCurveClappHornberg1978Mod.F90 (2) Modify the select case statement in SoilWaterRetentionCurveFactoryMod.F90 so that it is able to create an instance of your new type Note that this refactor also combines the soil_suction and soil_hk parameterization options into a single option. Dave Lawrence and Rosie Fisher felt that was preferable, and Jinyun Tang was okay with this. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: andre, muszala, Jinyun Tang List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: ========= Replaced with the 3 new files noted below D models/lnd/clm/src/biogeophys/SoiWatRetCurveParMod.F90 List all files added and what they do: ========= Replacement for SoiWatRetCurveParMod, implemented using polymorphism. Note that I have also added a soil_suction_inverse routine, which is not yet used or tested. I'll be using (and testing) this in an upcoming tag, where I refactor the irrigation code to use this. A models/lnd/clm/src/biogeophys/SoilWaterRetentionCurveMod.F90 A models/lnd/clm/src/biogeophys/SoilWaterRetentionCurveFactoryMod.F90 A models/lnd/clm/src/biogeophys/SoilWaterRetentionCurveClappHornberg1978Mod.F90 List all existing files that have been modified, and describe the changes: ========= Minor changes to accommodate the refactored code M models/lnd/clm/src/main/clm_initializeMod.F90 M models/lnd/clm/src/main/init_hydrology.F90 M models/lnd/clm/src/main/clm_driver.F90 M models/lnd/clm/src/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/biogeophys/SoilWaterMovementMod.F90 M models/lnd/clm/src/biogeophys/SoilMoistStressMod.F90 M models/lnd/clm/src/biogeophys/HydrologyNoDrainageMod.F90 CLM testing: regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu: NOT RUN goldbach_nag: ok goldbach_intel: ok CLM tag used for the baseline comparisons: clm4_5_1_r084 Changes answers relative to baseline: NO - bfb =============================================================== =============================================================== Tag name: clm4_5_1_r084 Originator(s): sacks (Bill Sacks) Date: Thu Sep 18 14:39:44 MDT 2014 One-line Summary: make glc_dyn_runoff_routing spatially-varying, based on input from glc Purpose of changes: Dave Lawrence, Bill Lipscomb and Jeremy Fyke have pointed out that glc_dyn_runoff_routing needs to be spatially-varying: Even when we're coupling to CISM, we should continue to use the old scheme in regions that don't have an active icesheet model underneath (which currently includes Antarctica and all of the world's smaller glaciers - i.e., everything except Greenland). Furthermore, we have introduced a new ability into CISM to run in diagnostic mode, without sending calving/runoff fluxes to the coupler. In this case, too, CLM should revert to using the old scheme (glc_dyn_runoff_routing = .false.). To accomplish both of these things, I have introduced a new coupling field, through which GLC tells CLM which areas have an icesheet that is "active" in the sense of sending fluxes to the coupler. In this CLM tag, CLM sets a spatially-varying version of glc_dyn_runoff_routing based on this new coupler field - replacing the old, namelist-settable version of this flag. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: - removed glc_dyn_runoff_routing List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): drv: drvseq5_0_15 -> drvseq5_0_17 - fix for some multi-instance runs - add icemask_coupled_fluxes field cism: cism1_140602 -> cism1_140914 - Add zero_gcm_fluxes option; send icemask_coupled_fluxes field to coupler scripts: scripts4_140916b -> scripts4_140916c - Rename CLM_UPDATE_GLC_AREAS to GLC_TWO_WAY_COUPLING List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Rename CLM_UPDATE_GLC_AREAS xml variable to GLC_TWO_WAY_COUPLING. Remove glc_dyn_runoff_routing namelist variable (this is now a spatially-varying field, tied more tightly to CISM). M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/CLMBuildNamelist.pm M models/lnd/clm/bld/unit_testers/build-namelist_test.pl M models/lnd/clm/bld/user_nl_clm M models/lnd/clm/bld/env_run.xml ========= Receive icemask_coupled_fluxes from CISM M models/lnd/clm/src/cpl/lnd_import_export.F90 M models/lnd/clm/src/cpl/clm_cpl_indices.F90 ========= Set spatially-varying glc_dyn_runoff_routing field based on icemask_coupled_fluxes, and use this in place of the old scalar glc_dyn_runoff_routing flag M models/lnd/clm/src/main/glc2lndMod.F90 M models/lnd/clm/src/main/clm_driver.F90 M models/lnd/clm/src/main/clm_varctl.F90 M models/lnd/clm/src/main/controlMod.F90 M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/biogeophys/HydrologyDrainageMod.F90 M models/lnd/clm/src/dyn_subgrid/dynSubgridDriverMod.F90 ========= Add write statement to workaround a pgi compiler problem M models/lnd/clm/src/main/restFileMod.F90 CLM testing: build-namelist tests: yellowstone: ok (baseline comparisons fail for clm45 & clm50, as expected) regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok yellowstone_gnu: NOT RUN goldbach_nag: ok goldbach_intel: ok Note: Most testing was run on glc_runoff_routing_n06_clm4_5_1_r083. After that tag, I added the following write statement in restFileMod, as a workaround for a PGI compiler bug: write(iulog,*) 'about to call aerosol_vars%restart: ', ubound(waterstate_vars%h2osoi_ice_col) After that addition, I just reran a subset of tests: 6 yellowstone-intel tests, 10 yellowstone-pgi tests (including the 2 that had failed due to the compiler bug), and 5 goldbach-nag tests. NOTE: Unit test build is currently failing due to a change in r082. Stefan is working on a fix. CLM tag used for the baseline comparisons: clm4_5_1_r083 Changes answers relative to baseline: YES Summarize any changes to answers, i.e., - what code configurations: CLM45 with CISM (i.e., IG) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): (1) Large changes in runoff from glaciers in IG compsets, due to setting glc_dyn_runoff_routing to .false. outside of Greenland. (2) Roundoff-level changes in icemask for some resolutions and compilers, due to changes in the cism external. If bitwise differences were observed, how did you show they were no worse than roundoff? For icemask changes: examined cprnc RMS errors. For other changes, diffs are greater than roundoff. If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_1_r083 Originator(s): muszala (Stefan Muszala) Date: Wed Sep 17 09:21:31 MDT 2014 One-line Summary: only update scripts and run new baselines. this due to an error in yellowstone pgi test naming (clm_aux45 changed to aux_clm45) Purpose of changes: Update scripts due to an error in a previous scripts tag in which I named pgi tests as clm_aux45 instead of aux_clm45. These were for tests moved from goldbach to yellowstone. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts4_140910 -> scripts4_140916b List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: N/A - update test list failures M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): WJS: note: Stef told me that he actually ran all the yellowstone tests, and all were okay yellowstone_intel - 40 OK yellowstone_pgi goldbach_nag - 40 OK, 45 OK goldbach_intel - 40 OK, 45 OK Note 1: Due to the fact that in older baselines component_gen_comp was failing due to a scripts error (now fixed) some older baselines don't have the clm history files. As one example: BFAIL ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.compare_hist.clm4_5_1_r082 - rerun. compare against clm4_5_1_r081, then it passes. PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.memleak PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.compare_hist.clm4_5_1_r082_oldPgi PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.memcomp.clm4_5_1_r082_oldPgi PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.tputcomp.clm4_5_1_r082_oldPgi PASS ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput.C.140917-082253.nlcomp For these tests, I made sure that clm2 history files were in the clm4_5_1_r083 baseline dirs. They should pass going forward. ERI_D.f10_f10.ICRUCLM50BGC.goldbach_intel.clm-reduceOutput PEM.f10_f10.ICLM45BGCCROP.goldbach_intel.clm-crop SSP.f19_g16.I1850CLM45BGC.yellowstone_pgi.clm-default Note 2: For these N2 tests, there was a scripts problem with st_archiver in the previous tag. These should pass next time. BFAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.compare_hist.clm4_5_1_r082 BFAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.compare_hist.clm4_5_1_r082 BFAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.clm2.h0.compare_hist.clm4_5_1_r082 BFAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.clm2.h1.compare_hist.clm4_5_1_r082 CLM tag used for the baseline comparisons: clm4_5_1_r082 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_1_r082 Originator(s): muszala (Stefan Muszala) Date: Thu Sep 11 14:07:58 MDT 2014 One-line Summary: Merge in a number of ED changes to address science bugs and infrastructure (particularly restarts) Purpose of changes: Merge in ED changes. Most of these have to do with science changes from rfisher. There is also a refactor and added functionality for ED restarts. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): Bug 2041, 2042 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, rfisher List any svn externals directories updated (csm_share, mct, etc.): scripts4_140814a -> scripts4_140910 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml -- add failing N2 tests (due to scripts, see below). M models/lnd/clm/bld/build-namelist -- documentation at top of file should reference CLMBuildNamelist.pm M models/lnd/clm/src/main/clm_driver.F90 -- add call to SurfaceAlbedo for use_ed logical branch, add EDBioVars as -- argument to temperature_vars%UpdateAccVars and add canopystate_vars as -- argument to BalanceCheck (for ED) M models/lnd/clm/src/main/decompMod.F90 -- added openMP output...remove mods after getting ED working with openMP M models/lnd/clm/src/ED/biogeochem/EDCanopyStructureMod.F90 -- change handling of CWD_AG and CWD_BG M models/lnd/clm/src/ED/biogeochem/EDGrowthFunctionsMod.F90 -- added tree_sai function M models/lnd/clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 -- signifcant reworking of entire module M models/lnd/clm/src/ED/biogeochem/EDPhysiologyMod.F90 -- minor code clean up M models/lnd/clm/src/ED/biogeochem/EDPatchDynamicsMod.F90 -- change an error check to: if (( areatot - area ) > 0._r8 ) then M models/lnd/clm/src/ED/main/EDBioType.F90 -- add infrastrucutre (define, allocate, etc...) M models/lnd/clm/src/ED/main/EDMainMod.F90 -- major update for updating canopy biomass pools M models/lnd/clm/src/ED/main/EDCLMLinkMod.F90 -- modify calls for history file output and error checking M models/lnd/clm/src/ED/main/EDRestVectorMod.F90 -- add resp_clm as restart variable and use SHR_ASSERT instead of call assert. major refactor -- of createPatchCohortStructure to handle arbitrary number of cohorts and patches M models/lnd/clm/src/ED/main/EDInitMod.F90 -- add logical to deal with different values of assignemnt dc%laimemory M models/lnd/clm/src/ED/main/EDTypesMod.F90 -- add cohort_type and change paramters: numCohortsPerPatch, cohorts_per_gcell and fire_threshold M models/lnd/clm/src/ED/fire/SFMainMod.F90 -- clean up write statemnts and a bug fix: change tau_b(dg_sf) -> tau_b(c) M models/lnd/clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 -- change tpu25top(FT) = 0.06_r8 * jmax25top(FT) to tpu25top(FT) = 0.167_r8 * jmax25top(FT), some cleanup M models/lnd/clm/src/biogeophys/SnowHydrologyMod.F90 -- clean up a use statement M models/lnd/clm/src/biogeophys/BareGroundFluxesMod.F90 -- remove whitespaces after a statement M models/lnd/clm/src/biogeophys/SurfaceAlbedoMod.F90 -- add a use_ed block to prevent some unassigned pointer errors M models/lnd/clm/src/biogeophys/BalanceCheckMod.F90 -- add canopystate_vars as argument for elai and esai for more verbose error -- reporting added by rfisher M models/lnd/clm/src/biogeophys/TemperatureType.F90 -- add arg. for UpdateAccVars for new calculations, split out use_ed and use_crop M models/lnd/clm/src/biogeophys/SurfaceRadiationMod.F90 -- add associate statement to tlai for error reporting M UpDateChangeLog.pl -- fix tiny typo M SVN_EXTERNAL_DIRECTORIES -- for scripts update CLM testing: build-namelist tests: yellowstone - N/A, no namelist changes made in this tag regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 45 OK , 40 OK Changes due to update of scripts from scripts4_140814a -> scripts4_140910 (this change came with scripts4_140828) FAIL namelist compare: user_nl_clm differs These should pass next time. NEW: flanduse_timeseries = '$DIN_LOC_ROOT/lnd/clm2/surfdata_map/surfdata.pftdyn_1x1_tropicAtl_TEST_simyr1939-1943_c140108.nc' BASELINE: flanduse_timeseries = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/surfdata.pftdyn_1x1_tropicAtl_TEST_simyr1939-1943_c140108.nc' FAIL SMS_Ly3.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetLate.GC.newPgi_45_intel.nlcomp FAIL SMS_Ly5.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetMid.GC.newPgi_45_intel.nlcomp FAIL SMS_Ly8.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetEarly.GC.newPgi_45_intel.nlcomp These fail due to a bug in scripts4_140905c. Alice is aware of this and will provide a fix for a future CLM tag (but 2041): FAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel FAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default Failure that looks like a hardware problem and that Erik is looking into (bug 2042): RUN ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.GC.newPgi_45_intel yellowstone_pgi - 45 OK , 40 OK New PGI tests brought over from goldbach. These are expected as there are no baselines for this on yellowstone BFAIL ERI_D.f10_f10.ICLM45BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI_D.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI_D.f19_g16.ICLM45BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI_D.f19_g16.ICLM45.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI_D.f19_g16.ICRUCLM50BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI.f10_f10.ICLM45BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI.f19_g16.ICLM45BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERI.f19_g16.ICRUCLM50BGC.yellowstone_pgi.clm-reduceOutput.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.I20TRCLM45BGC.yellowstone_pgi.clm-decStart.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.I20TRCLM45BGC.yellowstone_pgi.clm-decStart.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.I20TRCLM45BGC.yellowstone_pgi.clm-decStart.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_pgi.clm-ciso.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_pgi.clm-ciso.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_pgi.clm-ciso.compare_hist.clm4_5_1_r081 BFAIL ERS.f10_f10.I1850CLM45BGC.yellowstone_pgi.clm-default.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL ERS.f10_f10.I1850CLM45BGC.yellowstone_pgi.clm-default.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL ERS.f10_f10.I1850CLM45BGC.yellowstone_pgi.clm-default.compare_hist.clm4_5_1_r081 BFAIL ERS_Lm3.1x1_smallvilleIA.ICLM45BGCCROP.yellowstone_pgi.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL ERS_Lm3.1x1_smallvilleIA.ICLM45BGCCROP.yellowstone_pgi.compare_hist.clm4_5_1_r081 BFAIL SMS_D.1x1_vancouverCAN.ICLM45.yellowstone_pgi.clm-default.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL SMS_D.1x1_vancouverCAN.ICLM45.yellowstone_pgi.clm-default.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL SMS_D.1x1_vancouverCAN.ICLM45.yellowstone_pgi.clm-default.compare_hist.clm4_5_1_r081 BFAIL SMS.f45_f45.ICLM45.yellowstone_pgi.clm-ptsRLB.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL SMS.f45_f45.ICLM45.yellowstone_pgi.clm-ptsRLB.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL SMS.f45_f45.ICLM45.yellowstone_pgi.clm-ptsRLB.compare_hist.clm4_5_1_r081 BFAIL SMS.f45_f45.ICLM45.yellowstone_pgi.clm-ptsROA.compare_hist.clm4_5_1_r081 BFAIL SMS_Ld5.f19_g16.IRCP45CLM45BGC.yellowstone_pgi.clm-decStart.clm2.h0.compare_hist.clm4_5_1_r081 BFAIL SMS_Ld5.f19_g16.IRCP45CLM45BGC.yellowstone_pgi.clm-decStart.clm2.h1.compare_hist.clm4_5_1_r081 BFAIL SMS_Ld5.f19_g16.IRCP45CLM45BGC.yellowstone_pgi.clm-decStart.compare_hist.clm4_5_1_r081 goldbach_nag - 45 OK , 40 OK Baseline missing from previous tag: ERI_D.f19_g16.ICLM45BGC.goldbach_nag.clm-reduceOutput- file /fs/cgd/csm/ccsm_baselines/clm4_5_1_r081/ERI_D.f19_g16.ICLM45BGC.goldbach_nag.clm-reduceOutput/cpl.hi.nc does not exis SMS.f09_g16.ICRUCLM45.goldbach_nag.clm-af_bias_v5 file /fs/cgd/csm/ccsm_baselines/clm4_5_1_r081/SMS.f09_g16.ICRUCLM45.goldbach_nag.clm-af_bias_v5/cpl.hi.nc does not exist SMS_Ly1.f19_g16.ICLM45BGCCROP.goldbach_nag.clm-reduceOutput file /fs/cgd/csm/ccsm_baselines/clm4_5_1_r081/SMS_Ly1.f19_g16.ICLM45BGCCROP.goldbach_nag.clm-reduceOutput/cpl.hi.nc does no SMS_Ly1.f19_g16.ICLM45BGCCROP.goldbach_nag.clm-reduceOutput file /fs/cgd/csm/ccsm_baselines/clm4_5_1_r081/SMS_Ly1.f19_g16.ICLM45BGCCROP.goldbach_nag.clm-reduceOutput/cpl.hi.nc does no t exist goldbach_intel - 45 OK , 40 OK goldbach_pgi (These have been moved to yellowstone due to PGI 14.1 throwing and ICE on goldbach) CLM tag used for the baseline comparisons: clm4_5_1_r081 Changes answers relative to baseline: Only for ED compsets =============================================================== =============================================================== Tag name: clm4_5_1_r081 Originator(s): mvertens (Mariana Vertenstein) Date: Sun Aug 24 19:39:50 MDT 2014 One-line Summary: major infrastructure changes and directory reorganization under src Purpose of changes: Overview of previous code design ========================================== - data structures arranged by subgrid type (pps, cps, lps, grc) - all functional categorization lumped in that one subgrid type which led to - centralization rather than modularization of all data - definition and instantiation in ONE big module (clmtype.F90) - allocation and initialization in ONE big module (clmtypeInitMod.F90) - history variables all in ONE big module (histFldsMod.F90) - restart variables in effectively TWO big modules (biogeophysicsRestMod.F90 and CNRestMod.F90) - time constant initialization in ONE complex module (initTimeConst.F90) - time varying cold start initialization in ONE module (initColdMod.F90) - accumulation variables in ONE module (accumulMod.F90) Overview of new code design ========================================== data structures arranged by scientific functional categories (e.g. temperature_type, waterstate_type, energyflux_type, etc) - a given data structure now contains ALL subgrid levels are in the data structure - and variables in the data structure are now appended with a unique suffix to indicate their subgrid levels (new suffixes: _patch, _col, _lun, _grc) - this does NOT effect the science code base, ONLY the associate statements - there are separate module for each data type definition (e.g. TemperatureType.F90, WaterstateType.F90, EnergyFluxType.F90, etc) - each data type has associated methods for - Allocation: variables now initialized as NaNs upon allocation - Cold Start Initialization: cold start initialization of variables is now ALWAYS done and overwritten if finidat is read in as spun up dataset (also now have on line interpolation of initial conditions as part of this refactor as well) - History initialization of variables All history fields now initialized as spval - Restart initialization of variables - Accumulation Initialization initialization and accumulation update of variables - Instantiation of datatypes is now separate from their declaration (for now in clm_initialize.F90 - will be moved in the future) Centralized routines that no longer exist: ========================================== Data types : clmtype.F90, clmtypeInitMod.F90 Initialization : initTimeConst.F90, initCold.F90 History : histFldsMod.F90 Accumulation : accumulMod.F90 Restart : biogeophysRestMod.F90, CNRestMod.F90 Biogeochemistry: CNSetValue.F90 New Type modules that now replace clmtype.F90 ========================================== main/atm2lndType.F90 main/lnd2atmType.F90 main/ColumnType.F90 main/EcophysConType.F90 main/GridcellType.F90 main/LandunitType.F90 main/PatchType.F90 biogeochem/CNCarbonFluxType.F90 biogeochem/CNCarbonStateType.F90 biogeochem/CNDecompCascadeConType.F90 biogeochem/CNDVType.F90 biogeochem/CNNitrogenFluxType.F90 biogeochem/CNNitrogenStateType.F90 biogeochem/CNStateType.F90 biogeochem/CropType.F90 biogeophys/AerosolType.F90 biogeophys/CanopyStateType.F90 biogeophys/EnergyFluxType.F90 biogeophys/FrictionVelocityType.F90 biogeophys/LakeStateType.F90 biogeophys/PhotosynthesisType.F90 biogeophys/SoilHydrologyType.F90 biogeophys/SoilStateType.F90 biogeophys/SolarAbsorbedType.F90 biogeophys/SurfaceAlbedoType.F90 biogeophys/TemperatureType.F90 biogeophys/UrbanParamsType.F90 biogeophys/WaterfluxType.F90 biogeophys/WaterStateType.F90 ED/main/EDBioType.F90 ED/main/EDEcophysConType.F90 ED/main/EDVecCohortType.F90 ED/main/EDVecPatchType.F90 Instantiation of Types 2) Public Types: - the following are public types that can BE PASSED AS ARGUMENTS - the type instances FOR NOW are clm_initialized and then used by the driver - this will be generalized in the future type(ch4_type) :: ch4_vars type(carbonstate_type) :: carbonstate_vars type(carbonstate_type) :: c13_carbonstate_vars type(carbonstate_type) :: c14_carbonstate_vars type(carbonflux_type) :: carbonflux_vars type(carbonflux_type) :: c13_carbonflux_vars type(carbonflux_type) :: c14_carbonflux_vars type(nitrogenstate_type) :: nitrogenstate_vars type(nitrogenflux_type) :: nitrogenflux_vars type(dgvs_type) :: dgvs_vars type(crop_type) :: crop_vars type(cnstate_type) :: cnstate_vars type(dust_type) :: dust_vars type(vocemis_type) :: vocemis_vars type(drydepvel_type) :: drydepvel_vars type(aerosol_type) :: aerosol_vars type(canopystate_type) :: canopystate_vars type(energyflux_type) :: energyflux_vars type(frictionvel_type) :: frictionvel_vars type(lakestate_type) :: lakestate_vars type(photosyns_type) :: photosyns_vars type(soilstate_type) :: soilstate_vars type(soilhydrology_type) :: soilhydrology_vars type(solarabs_type) :: solarabs_vars type(surfalb_type) :: surfalb_vars type(surfrad_type) :: surfrad_vars type(temperature_type) :: temperature_vars type(urbanparams_type) :: urbanparams_vars type(waterflux_type) :: waterflux_vars type(waterstate_type) :: waterstate_vars type(atm2lnd_type) :: atm2lnd_vars type(glc2lnd_type) :: glc2lnd_vars type(lnd2atm_type) :: lnd2atm_vars type(lnd2glc_type) :: lnd2glc_vars type(glc_diagnostics_type) :: glc_diagnostics_vars type(EDbio_type) :: EDbio_vars - private Types (now som modules have their own PRIVATE types) DUSTMod.F90 : type(dust_type) VOCEmissionMod.F90: type(vocemis_type) ch4Mod.F90 : type(ch4_type) API Changes: ========================================== Original APIs: clmtype was in effect a global common block and all routines had use statements into it difficult to track any intent or flow through system difficult to set up functional unit testing (.e.g. CanopyFluxesMod.F90, etc) Refactorized APIs: all new datatype instances are passed as arguments science code is effectively the same since only the associate statements have been modified New Directory Structure under clm/ ========================================== bld/ doc/ src/biogeochem/ src/biogeophys/ src/cpl/ src/dyn_subgrid/ src/ED/ src/ED/biogeochem src/ED/biogeophys src/ED/fire src/ED/main src/main/ src/unit_test_mocks/ src/unit_test_shr/ src/utils/ src_clm4_0/ test/ tools/ Advantages of refactorization: ========================================== - Lets compiler enforce intent attributes - Makes functional unit testing easier since module drivers can be constructed with relevant mock data more easily - Makes more sense scientifically since now easier to extend code logic as to where you want to introduce new variables - Easier to maintain code since code flow is easier to follow and to modify - Easy to move variables around from one data type to another since now know everything that is logically connected to that variable that needs to be moved Offers new modularity for trading in and out new formulations of targeted functionality Requirements for tag: N.A. Test level of tag: regular Bugs fixed (include bugzilla ID): N.A. Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: directory restructuring has been reflected in changes to configure in setting up the Filepath Describe any changes made to the namelist: clm_hydrology1_inparm changed to clm_canopyhydrology_inparm List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: myself, Bill Sacks List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: D clm/src/clm4_5 D clm/src/clm4_5/biogeochem D clm/src/clm4_5/biogeochem/CNCStateUpdate2Mod.F90 D clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 D clm/src/clm4_5/biogeochem/CNRestMod.F90 D clm/src/clm4_5/biogeochem/CropRestMod.F90 D clm/src/clm4_5/biogeochem/CNGRespMod.F90 D clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 D clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 D clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 D clm/src/clm4_5/biogeochem/CNNStateUpdate3Mod.F90 D clm/src/clm4_5/biogeochem/CNFireMod.F90 D clm/src/clm4_5/biogeochem/CNMRespMod.F90 D clm/src/clm4_5/biogeochem/MEGANFactorsMod.F90 D clm/src/clm4_5/biogeochem/SatellitePhenologyMod.F90 D clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 D clm/src/clm4_5/biogeochem/ch4RestMod.F90 D clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 D clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 D clm/src/clm4_5/biogeochem/CNWoodProductsMod.F90 D clm/src/clm4_5/biogeochem/CNSummaryMod.F90 D clm/src/clm4_5/biogeochem/ch4Mod.F90 D clm/src/clm4_5/biogeochem/DUSTMod.F90 D clm/src/clm4_5/biogeochem/CNDVLightMod.F90 D clm/src/clm4_5/biogeochem/CNInitMod.F90 D clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 D clm/src/clm4_5/biogeochem/ch4varcon.F90 D clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 D clm/src/clm4_5/biogeochem/CNDecompMod.F90 D clm/src/clm4_5/biogeochem/CNCStateUpdate3Mod.F90 D clm/src/clm4_5/biogeochem/CNDVMod.F90 D clm/src/clm4_5/biogeochem/ED D clm/src/clm4_5/biogeochem/ED/EDCanopyStructureMod.F90 D clm/src/clm4_5/biogeochem/ED/EDSetValuesMod.F90 D clm/src/clm4_5/biogeochem/ED/EDGrowthFunctionsMod.F90 D clm/src/clm4_5/biogeochem/ED/EDPhysiologyMod.F90 D clm/src/clm4_5/biogeochem/ED/EDCohortDynamicsMod.F90 D clm/src/clm4_5/biogeochem/ED/EDPatchDynamicsMod.F90 D clm/src/clm4_5/biogeochem/CNSharedParamsMod.F90 D clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 D clm/src/clm4_5/biogeochem/ch4InitMod.F90 D clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 D clm/src/clm4_5/biogeochem/CNNStateUpdate2Mod.F90 D clm/src/clm4_5/biogeochem/DryDepVelocity.F90 D clm/src/clm4_5/biogeochem/CNAllocationMod.F90 D clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 D clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 D clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 D clm/src/clm4_5/biogeochem/CNSetValueMod.F90 D clm/src/clm4_5/biogeochem/CNVegStructUpdateMod.F90 D clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 D clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 D clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 D clm/src/clm4_5/biogeochem/CNDVInitMod.F90 D clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 D clm/src/clm4_5/main D clm/src/clm4_5/main/clm_varcon.F90 D clm/src/clm4_5/main/initInterp.F90 D clm/src/clm4_5/main/clm_varpar.F90 D clm/src/clm4_5/main/landunit_varcon.F90 D clm/src/clm4_5/main/initTimeConstMod.F90 D clm/src/clm4_5/main/subgridWeightsMod.F90 D clm/src/clm4_5/main/decompInitMod.F90 D clm/src/clm4_5/main/clm_initializeMod.F90 D clm/src/clm4_5/main/subgridRestMod.F90 D clm/src/clm4_5/main/clm_glclnd.F90 D clm/src/clm4_5/main/paramUtilMod.F90 D clm/src/clm4_5/main/accFldsMod.F90 D clm/src/clm4_5/main/subgridMod.F90 D clm/src/clm4_5/main/clmtypeInitMod.F90 D clm/src/clm4_5/main/ndepStreamMod.F90 D clm/src/clm4_5/main/init_hydrology.F90 D clm/src/clm4_5/main/initColdMod.F90 D clm/src/clm4_5/main/column_varcon.F90 D clm/src/clm4_5/main/histFileMod.F90 D clm/src/clm4_5/main/pft2colMod.F90 D clm/src/clm4_5/main/clm_atmlnd.F90 D clm/src/clm4_5/main/findHistFields.pl D clm/src/clm4_5/main/clm_varsur.F90 D clm/src/clm4_5/main/restFileMod.F90 D clm/src/clm4_5/main/CMakeLists.txt D clm/src/clm4_5/main/controlMod.F90 D clm/src/clm4_5/main/spitfireSF D clm/src/clm4_5/main/spitfireSF/SFParamsMod.F90 D clm/src/clm4_5/main/spitfireSF/SFMainMod.F90 D clm/src/clm4_5/main/test D clm/src/clm4_5/main/test/subgridWeights_test D clm/src/clm4_5/main/test/subgridWeights_test/test_subgridWeights.pf D clm/src/clm4_5/main/test/subgridWeights_test/CMakeLists.txt D clm/src/clm4_5/main/test/clm_glclnd_test D clm/src/clm4_5/main/test/clm_glclnd_test/test_clm_glclnd.pf D clm/src/clm4_5/main/test/clm_glclnd_test/CMakeLists.txt D clm/src/clm4_5/main/test/CMakeLists.txt D clm/src/clm4_5/main/initSubgridMod.F90 D clm/src/clm4_5/main/filterMod.F90 D clm/src/clm4_5/main/clm_varctl.F90 D clm/src/clm4_5/main/clm_driver.F90 D clm/src/clm4_5/main/surfrdUtilsMod.F90 D clm/src/clm4_5/main/ED D clm/src/clm4_5/main/ED/EDInitTimeConst.F90 D clm/src/clm4_5/main/ED/EDCLMLinkMod.F90 D clm/src/clm4_5/main/ED/EDClmType.F90 D clm/src/clm4_5/main/ED/EDRestVectorMod.F90 D clm/src/clm4_5/main/ED/EDHistFldsMod.F90 D clm/src/clm4_5/main/ED/EDClmTypeInitMod.F90 D clm/src/clm4_5/main/ED/EDPftvarcon.F90 D clm/src/clm4_5/main/ED/EDParamsMod.F90 D clm/src/clm4_5/main/ED/EDInitMod.F90 D clm/src/clm4_5/main/ED/EDTypesMod.F90 D clm/src/clm4_5/main/ED/EDMainMod.F90 D clm/src/clm4_5/main/subgridAveMod.F90 D clm/src/clm4_5/main/initGridCellsMod.F90 D clm/src/clm4_5/main/initSoilParVICMod.F90 D clm/src/clm4_5/main/pftvarcon.F90 D clm/src/clm4_5/main/surfrdMod.F90 D clm/src/clm4_5/main/decompMod.F90 D clm/src/clm4_5/main/FuncPedotransferMod.F90 D clm/src/clm4_5/main/clmtype.F90 D clm/src/clm4_5/main/reweightMod.F90 D clm/src/clm4_5/main/readParamsMod.F90 D clm/src/clm4_5/main/histFldsMod.F90 D clm/src/clm4_5/biogeophys D clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 D clm/src/clm4_5/biogeophys/SLakeInitMod.F90 D clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 D clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 D clm/src/clm4_5/biogeophys/UrbanInputMod.F90 D clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 D clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 D clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 D clm/src/clm4_5/biogeophys/TridiagonalMod.F90 D clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 D clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 D clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 D clm/src/clm4_5/biogeophys/UrbanMod.F90 D clm/src/clm4_5/biogeophys/QSatMod.F90 D clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 D clm/src/clm4_5/biogeophys/SurfaceResistanceMod.F90 D clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 D clm/src/clm4_5/biogeophys/CLMVICMapMod.F90 D clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 D clm/src/clm4_5/biogeophys/SNICARMod.F90 D clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 D clm/src/clm4_5/biogeophys/CMakeLists.txt D clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 D clm/src/clm4_5/biogeophys/RootBiophysMod.F90 D clm/src/clm4_5/biogeophys/test D clm/src/clm4_5/biogeophys/test/CMakeLists.txt D clm/src/clm4_5/biogeophys/test/Daylength_test D clm/src/clm4_5/biogeophys/test/Daylength_test/test_daylength.pf D clm/src/clm4_5/biogeophys/test/Daylength_test/CMakeLists.txt D clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 D clm/src/clm4_5/biogeophys/SoilWaterMovementMod.F90 D clm/src/clm4_5/biogeophys/H2OSfcMod.F90 D clm/src/clm4_5/biogeophys/SoilMoistStressMod.F90 D clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 D clm/src/clm4_5/biogeophys/SoiWatRetCurveParMod.F90 D clm/src/clm4_5/biogeophys/ED D clm/src/clm4_5/biogeophys/ED/EDAccumulateFluxesMod.F90 D clm/src/clm4_5/biogeophys/ED/EDSurfaceAlbedoMod.F90 D clm/src/clm4_5/biogeophys/ED/EDPhotosynthesisMod.F90 D clm/src/clm4_5/biogeophys/ED/EDBtranMod.F90 D clm/src/clm4_5/biogeophys/FracWetMod.F90 D clm/src/clm4_5/biogeophys/UrbanInitMod.F90 D clm/src/clm4_5/biogeophys/SLakeRestMod.F90 D clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 D clm/src/clm4_5/biogeophys/SLakeCon.F90 D clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 D clm/src/clm4_5/biogeophys/BandDiagonalMod.F90 D clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 D clm/src/clm4_5/biogeophys/DaylengthMod.F90 D clm/src/clm4_5/dyn_subgrid D clm/src/clm4_5/dyn_subgrid/test D clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test D clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf D clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights.pf D clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/CMakeLists.txt D clm/src/clm4_5/dyn_subgrid/test/dynVar_test D clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarShared.F90 D clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarTimeInterp.pf D clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarTimeUninterp.pf D clm/src/clm4_5/dyn_subgrid/test/dynVar_test/CMakeLists.txt D clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test D clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test/test_dynTimeInfo.pf D clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test/CMakeLists.txt D clm/src/clm4_5/dyn_subgrid/test/CMakeLists.txt D clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test D clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf D clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test/CMakeLists.txt D clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 D clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 D clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 D clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 D clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 D clm/src/clm4_5/dyn_subgrid/dynConsBiogeochemMod.F90 D clm/src/clm4_5/dyn_subgrid/dynPriorWeightsMod.F90 D clm/src/clm4_5/dyn_subgrid/dynVarMod.F90 D clm/src/clm4_5/dyn_subgrid/dynEDMod.F90 D clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90 D clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 D clm/src/clm4_5/dyn_subgrid/dynCNDVMod.F90 D clm/src/clm4_5/dyn_subgrid/dynConsBiogeophysMod.F90 D clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90 D clm/src/clm4_5/dyn_subgrid/dynVarMod.F90.in D clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90.in D clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90.in D clm/src/clm4_5/dyn_subgrid/do_genf90 D clm/src/clm4_5/dyn_subgrid/dynInitColumnsMod.F90 D clm/src/clm4_5/dyn_subgrid/CMakeLists.txt D clm/src/util_share D clm/src/util_share/organicFileMod.F90 D clm/src/util_share/spmdGathScatMod.F90 D clm/src/util_share/clm_time_manager.F90 D clm/src/util_share/clm_nlUtilsMod.F90 D clm/src/util_share/clm_varorb.F90 D clm/src/util_share/abortutils.F90 D clm/src/util_share/accumulMod.F90 D clm/src/util_share/getdatetime.F90 D clm/src/util_share/fileutils.F90 D clm/src/util_share/dtypes.h D clm/src/util_share/ncdio_pio.F90 D clm/src/util_share/SimpleMathMod.F90 D clm/src/util_share/spmdMod.F90 D clm/src/util_share/domainMod.F90 D clm/src/util_share/ncdio_pio.F90.in D clm/src/util_share/restUtilMod.F90 D clm/src/util_share/quadraticMod.F90 D clm/src/util_share/restUtilMod.F90.in D clm/src/util_share/CMakeLists.txt D clm/src/util_share/GetGlobalValuesMod.F90 D clm/src/clm4_0 D clm/src/clm4_0/biogeochem D clm/src/clm4_0/biogeochem/CNCStateUpdate2Mod.F90 D clm/src/clm4_0/biogeochem/CNC13StateUpdate2Mod.F90 D clm/src/clm4_0/biogeochem/CNGapMortalityMod.F90 D clm/src/clm4_0/biogeochem/CropRestMod.F90 D clm/src/clm4_0/biogeochem/CNGRespMod.F90 D clm/src/clm4_0/biogeochem/CNNStateUpdate1Mod.F90 D clm/src/clm4_0/biogeochem/CNBalanceCheckMod.F90 D clm/src/clm4_0/biogeochem/CNNStateUpdate3Mod.F90 D clm/src/clm4_0/biogeochem/CNFireMod.F90 D clm/src/clm4_0/biogeochem/CNMRespMod.F90 D clm/src/clm4_0/biogeochem/MEGANFactorsMod.F90 D clm/src/clm4_0/biogeochem/CNPrecisionControlMod.F90 D clm/src/clm4_0/biogeochem/CNWoodProductsMod.F90 D clm/src/clm4_0/biogeochem/CNSummaryMod.F90 D clm/src/clm4_0/biogeochem/DUSTMod.F90 D clm/src/clm4_0/biogeochem/CNDVLightMod.F90 D clm/src/clm4_0/biogeochem/CNPhenologyMod.F90 D clm/src/clm4_0/biogeochem/CNCStateUpdate1Mod.F90 D clm/src/clm4_0/biogeochem/CNDecompMod.F90 D clm/src/clm4_0/biogeochem/STATICEcosysDynMod.F90 D clm/src/clm4_0/biogeochem/CNCStateUpdate3Mod.F90 D clm/src/clm4_0/biogeochem/CNDVMod.F90 D clm/src/clm4_0/biogeochem/CNC13StateUpdate1Mod.F90 D clm/src/clm4_0/biogeochem/CNrestMod.F90 D clm/src/clm4_0/biogeochem/CNC13StateUpdate3Mod.F90 D clm/src/clm4_0/biogeochem/VOCEmissionMod.F90 D clm/src/clm4_0/biogeochem/CNDVEcosystemDynIniMod.F90 D clm/src/clm4_0/biogeochem/CNAnnualUpdateMod.F90 D clm/src/clm4_0/biogeochem/CNNStateUpdate2Mod.F90 D clm/src/clm4_0/biogeochem/C13SummaryMod.F90 D clm/src/clm4_0/biogeochem/DryDepVelocity.F90 D clm/src/clm4_0/biogeochem/CNC13FluxMod.F90 D clm/src/clm4_0/biogeochem/CNAllocationMod.F90 D clm/src/clm4_0/biogeochem/CNNDynamicsMod.F90 D clm/src/clm4_0/biogeochem/CNEcosystemDynMod.F90 D clm/src/clm4_0/biogeochem/CNSetValueMod.F90 D clm/src/clm4_0/biogeochem/CNVegStructUpdateMod.F90 D clm/src/clm4_0/biogeochem/CNDVEstablishmentMod.F90 D clm/src/clm4_0/main D clm/src/clm4_0/main/clm_varcon.F90 D clm/src/clm4_0/main/clm_varpar.F90 D clm/src/clm4_0/main/CNiniTimeVar.F90 D clm/src/clm4_0/main/dynlandMod.F90 D clm/src/clm4_0/main/decompInitMod.F90 D clm/src/clm4_0/main/clm_initializeMod.F90 D clm/src/clm4_0/main/subgridRestMod.F90 D clm/src/clm4_0/main/clm_glclnd.F90 D clm/src/clm4_0/main/accFldsMod.F90 D clm/src/clm4_0/main/subgridMod.F90 D clm/src/clm4_0/main/clmtypeInitMod.F90 D clm/src/clm4_0/main/ndepStreamMod.F90 D clm/src/clm4_0/main/pftdynMod.F90 D clm/src/clm4_0/main/iniTimeConst.F90 D clm/src/clm4_0/main/histFileMod.F90 D clm/src/clm4_0/main/pft2colMod.F90 D clm/src/clm4_0/main/clm_atmlnd.F90 D clm/src/clm4_0/main/findHistFields.pl D clm/src/clm4_0/main/restFileMod.F90 D clm/src/clm4_0/main/clm_varsur.F90 D clm/src/clm4_0/main/controlMod.F90 D clm/src/clm4_0/main/initSurfAlbMod.F90 D clm/src/clm4_0/main/filterMod.F90 D clm/src/clm4_0/main/clm_driver.F90 D clm/src/clm4_0/main/clm_varctl.F90 D clm/src/clm4_0/main/subgridAveMod.F90 D clm/src/clm4_0/main/initGridCellsMod.F90 D clm/src/clm4_0/main/CNiniSpecial.F90 D clm/src/clm4_0/main/pftvarcon.F90 D clm/src/clm4_0/main/surfrdMod.F90 D clm/src/clm4_0/main/decompMod.F90 D clm/src/clm4_0/main/clmtype.F90 D clm/src/clm4_0/main/histFldsMod.F90 D clm/src/clm4_0/main/mkarbinitMod.F90 D clm/src/clm4_0/biogeophys D clm/src/clm4_0/biogeophys/BalanceCheckMod.F90 D clm/src/clm4_0/biogeophys/SurfaceRadiationMod.F90 D clm/src/clm4_0/biogeophys/SoilTemperatureMod.F90 D clm/src/clm4_0/biogeophys/SnowHydrologyMod.F90 D clm/src/clm4_0/biogeophys/UrbanInputMod.F90 D clm/src/clm4_0/biogeophys/Biogeophysics1Mod.F90 D clm/src/clm4_0/biogeophys/Biogeophysics2Mod.F90 D clm/src/clm4_0/biogeophys/FracWetMod.F90 D clm/src/clm4_0/biogeophys/UrbanInitMod.F90 D clm/src/clm4_0/biogeophys/FrictionVelocityMod.F90 D clm/src/clm4_0/biogeophys/TridiagonalMod.F90 D clm/src/clm4_0/biogeophys/SurfaceAlbedoMod.F90 D clm/src/clm4_0/biogeophys/Hydrology1Mod.F90 D clm/src/clm4_0/biogeophys/Hydrology2Mod.F90 D clm/src/clm4_0/biogeophys/BiogeophysicsLakeMod.F90 D clm/src/clm4_0/biogeophys/BiogeophysRestMod.F90 D clm/src/clm4_0/biogeophys/UrbanMod.F90 D clm/src/clm4_0/biogeophys/SoilHydrologyMod.F90 D clm/src/clm4_0/biogeophys/QSatMod.F90 D clm/src/clm4_0/biogeophys/clm_driverInitMod.F90 D clm/src/clm4_0/biogeophys/HydrologyLakeMod.F90 D clm/src/clm4_0/biogeophys/BareGroundFluxesMod.F90 D clm/src/clm4_0/biogeophys/SNICARMod.F90 D clm/src/clm4_0/biogeophys/CanopyFluxesMod.F90 List all files added and what they do: A clm/src_clm40 A clm/src_clm40/biogeochem A clm/src_clm40/biogeochem/CNCStateUpdate2Mod.F90 A clm/src_clm40/biogeochem/CNC13StateUpdate2Mod.F90 A clm/src_clm40/biogeochem/CNGRespMod.F90 A clm/src_clm40/biogeochem/CNBalanceCheckMod.F90 A clm/src_clm40/biogeochem/CNNStateUpdate3Mod.F90 A clm/src_clm40/biogeochem/CNSummaryMod.F90 A clm/src_clm40/biogeochem/CNPhenologyMod.F90 A clm/src_clm40/biogeochem/STATICEcosysDynMod.F90 A clm/src_clm40/biogeochem/CNCStateUpdate1Mod.F90 A clm/src_clm40/biogeochem/CNC13StateUpdate1Mod.F90 A clm/src_clm40/biogeochem/CNrestMod.F90 A clm/src_clm40/biogeochem/VOCEmissionMod.F90 A clm/src_clm40/biogeochem/CNAnnualUpdateMod.F90 A clm/src_clm40/biogeochem/CNNStateUpdate2Mod.F90 A clm/src_clm40/biogeochem/C13SummaryMod.F90 A clm/src_clm40/biogeochem/CNAllocationMod.F90 A clm/src_clm40/biogeochem/DryDepVelocity.F90 A clm/src_clm40/biogeochem/CNNDynamicsMod.F90 A clm/src_clm40/biogeochem/CNSetValueMod.F90 A clm/src_clm40/biogeochem/CNGapMortalityMod.F90 A clm/src_clm40/biogeochem/CropRestMod.F90 A clm/src_clm40/biogeochem/CNNStateUpdate1Mod.F90 A clm/src_clm40/biogeochem/CNFireMod.F90 A clm/src_clm40/biogeochem/CNMRespMod.F90 A clm/src_clm40/biogeochem/MEGANFactorsMod.F90 A clm/src_clm40/biogeochem/CNWoodProductsMod.F90 A clm/src_clm40/biogeochem/CNPrecisionControlMod.F90 A clm/src_clm40/biogeochem/DUSTMod.F90 A clm/src_clm40/biogeochem/CNDVLightMod.F90 A clm/src_clm40/biogeochem/CNDecompMod.F90 A clm/src_clm40/biogeochem/CNDVMod.F90 A clm/src_clm40/biogeochem/CNCStateUpdate3Mod.F90 A clm/src_clm40/biogeochem/CNC13StateUpdate3Mod.F90 A clm/src_clm40/biogeochem/CNDVEcosystemDynIniMod.F90 A clm/src_clm40/biogeochem/CNC13FluxMod.F90 A clm/src_clm40/biogeochem/CNEcosystemDynMod.F90 A clm/src_clm40/biogeochem/CNVegStructUpdateMod.F90 A clm/src_clm40/biogeochem/CNDVEstablishmentMod.F90 A clm/src_clm40/main A clm/src_clm40/main/spmdGathScatMod.F90 A clm/src_clm40/main/organicFileMod.F90 A clm/src_clm40/main/clm_varcon.F90 A clm/src_clm40/main/clm_varpar.F90 A clm/src_clm40/main/CNiniTimeVar.F90 A clm/src_clm40/main/abortutils.F90 A clm/src_clm40/main/accumulMod.F90 A clm/src_clm40/main/decompInitMod.F90 A clm/src_clm40/main/clm_glclnd.F90 A clm/src_clm40/main/accFldsMod.F90 A clm/src_clm40/main/subgridMod.F90 A clm/src_clm40/main/pftdynMod.F90 A clm/src_clm40/main/pft2colMod.F90 A clm/src_clm40/main/clm_atmlnd.F90 A clm/src_clm40/main/quadraticMod.F90 A clm/src_clm40/main/GetGlobalValuesMod.F90 A clm/src_clm40/main/clm_time_manager.F90 A clm/src_clm40/main/filterMod.F90 A clm/src_clm40/main/clm_varctl.F90 A clm/src_clm40/main/subgridAveMod.F90 A clm/src_clm40/main/dtypes.h A clm/src_clm40/main/CNiniSpecial.F90 A clm/src_clm40/main/surfrdMod.F90 A clm/src_clm40/main/domainMod.F90 A clm/src_clm40/main/lnd_import_export.F90 A clm/src_clm40/main/restUtilMod.F90 A clm/src_clm40/main/clmtype.F90 A clm/src_clm40/main/mkarbinitMod.F90 A clm/src_clm40/main/restUtilMod.F90.in A clm/src_clm40/main/dynlandMod.F90 A clm/src_clm40/main/getdatetime.F90 A clm/src_clm40/main/clm_initializeMod.F90 A clm/src_clm40/main/subgridRestMod.F90 A clm/src_clm40/main/fileutils.F90 A clm/src_clm40/main/clmtypeInitMod.F90 A clm/src_clm40/main/ndepStreamMod.F90 A clm/src_clm40/main/SimpleMathMod.F90 A clm/src_clm40/main/iniTimeConst.F90 A clm/src_clm40/main/lnd_comp_esmf.F90 A clm/src_clm40/main/histFileMod.F90 A clm/src_clm40/main/clm_cpl_indices.F90 A clm/src_clm40/main/findHistFields.pl A clm/src_clm40/main/restFileMod.F90 A clm/src_clm40/main/clm_varsur.F90 A clm/src_clm40/main/controlMod.F90 A clm/src_clm40/main/CMakeLists.txt A clm/src_clm40/main/initSurfAlbMod.F90 A clm/src_clm40/main/clm_nlUtilsMod.F90 A clm/src_clm40/main/clm_driver.F90 A clm/src_clm40/main/clm_varorb.F90 A clm/src_clm40/main/initGridCellsMod.F90 A clm/src_clm40/main/lnd_comp_mct.F90 A clm/src_clm40/main/pftvarcon.F90 A clm/src_clm40/main/ncdio_pio.F90 A clm/src_clm40/main/spmdMod.F90 A clm/src_clm40/main/decompMod.F90 A clm/src_clm40/main/ncdio_pio.F90.in A clm/src_clm40/main/histFldsMod.F90 A clm/src_clm40/biogeophys A clm/src_clm40/biogeophys/BalanceCheckMod.F90 A clm/src_clm40/biogeophys/SoilTemperatureMod.F90 A clm/src_clm40/biogeophys/UrbanInputMod.F90 A clm/src_clm40/biogeophys/SnowHydrologyMod.F90 A clm/src_clm40/biogeophys/Biogeophysics1Mod.F90 A clm/src_clm40/biogeophys/FrictionVelocityMod.F90 A clm/src_clm40/biogeophys/TridiagonalMod.F90 A clm/src_clm40/biogeophys/Hydrology1Mod.F90 A clm/src_clm40/biogeophys/BiogeophysRestMod.F90 A clm/src_clm40/biogeophys/UrbanMod.F90 A clm/src_clm40/biogeophys/QSatMod.F90 A clm/src_clm40/biogeophys/clm_driverInitMod.F90 A clm/src_clm40/biogeophys/HydrologyLakeMod.F90 A clm/src_clm40/biogeophys/BareGroundFluxesMod.F90 A clm/src_clm40/biogeophys/SNICARMod.F90 A clm/src_clm40/biogeophys/CanopyFluxesMod.F90 A clm/src_clm40/biogeophys/SurfaceRadiationMod.F90 A clm/src_clm40/biogeophys/Biogeophysics2Mod.F90 A clm/src_clm40/biogeophys/UrbanInitMod.F90 A clm/src_clm40/biogeophys/FracWetMod.F90 A clm/src_clm40/biogeophys/SurfaceAlbedoMod.F90 A clm/src_clm40/biogeophys/Hydrology2Mod.F90 A clm/src_clm40/biogeophys/BiogeophysicsLakeMod.F90 A clm/src_clm40/biogeophys/SoilHydrologyMod.F90 A clm/src/main A clm/src/main/organicFileMod.F90 A clm/src/main/clm_varcon.F90 A clm/src/main/initInterp.F90 A clm/src/main/landunit_varcon.F90 A clm/src/main/clm_varpar.F90 A clm/src/main/abortutils.F90 A clm/src/main/accumulMod.F90 A clm/src/main/subgridWeightsMod.F90 A clm/src/main/decompInitMod.F90 A clm/src/main/subgridMod.F90 A clm/src/main/atm2lndType.F90 A clm/src/main/lnd2atmType.F90 A clm/src/main/column_varcon.F90 A clm/src/main/EcophysConType.F90 A clm/src/main/GetGlobalValuesMod.F90 A clm/src/main/initSubgridMod.F90 A clm/src/main/lnd2glcMod.F90 A clm/src/main/glc2lndMod.F90 A clm/src/main/filterMod.F90 A clm/src/main/surfrdUtilsMod.F90 A clm/src/main/clm_varctl.F90 A clm/src/main/subgridAveMod.F90 A clm/src/main/initVerticalMod.F90 A clm/src/main/glcDiagnosticsMod.F90 A clm/src/main/lnd2atmMod.F90 A clm/src/main/atm2lndMod.F90 A clm/src/main/surfrdMod.F90 A clm/src/main/FuncPedotransferMod.F90 A clm/src/main/readParamsMod.F90 A clm/src/main/clm_initializeMod.F90 A clm/src/main/subgridRestMod.F90 A clm/src/main/paramUtilMod.F90 A clm/src/main/ColumnType.F90 A clm/src/main/PatchType.F90 A clm/src/main/ndepStreamMod.F90 A clm/src/main/init_hydrology.F90 A clm/src/main/histFileMod.F90 A clm/src/main/findHistFields.pl A clm/src/main/restFileMod.F90 A clm/src/main/clm_varsur.F90 A clm/src/main/controlMod.F90 A clm/src/main/LandunitType.F90 A clm/src/main/CMakeLists.txt A clm/src/main/test A clm/src/main/test/subgridWeights_test A clm/src/main/test/subgridWeights_test/test_subgridWeights.pf A clm/src/main/test/subgridWeights_test/CMakeLists.txt A clm/src/main/test/clm_glclnd_test A clm/src/main/test/clm_glclnd_test/test_clm_glclnd.pf A clm/src/main/test/clm_glclnd_test/CMakeLists.txt A clm/src/main/test/CMakeLists.txt A clm/src/main/clm_driver.F90 A clm/src/main/GridcellType.F90 A clm/src/main/initGridCellsMod.F90 A clm/src/main/pftvarcon.F90 A clm/src/main/ncdio_pio.F90 A clm/src/main/decompMod.F90 A clm/src/main/ncdio_pio.F90.in A clm/src/main/reweightMod.F90 A clm/src/ED A clm/src/ED/biogeochem A clm/src/ED/biogeochem/EDCanopyStructureMod.F90 A clm/src/ED/biogeochem/EDGrowthFunctionsMod.F90 A clm/src/ED/biogeochem/EDPhysiologyMod.F90 A clm/src/ED/biogeochem/EDCohortDynamicsMod.F90 A clm/src/ED/biogeochem/EDPatchDynamicsMod.F90 A clm/src/ED/main A clm/src/ED/main/EDBioType.F90 A clm/src/ED/main/EDEcophysConType.F90 A clm/src/ED/main/EDParamsMod.F90 A clm/src/ED/main/EDMainMod.F90 A clm/src/ED/main/EDCLMLinkMod.F90 A clm/src/ED/main/EDVecCohortType.F90 A clm/src/ED/main/EDVecPatchType.F90 A clm/src/ED/main/EDRestVectorMod.F90 A clm/src/ED/main/EDPftvarcon.F90 A clm/src/ED/main/EDInitMod.F90 A clm/src/ED/main/EDTypesMod.F90 A clm/src/ED/fire A clm/src/ED/fire/SFParamsMod.F90 A clm/src/ED/fire/SFMainMod.F90 A clm/src/ED/biogeophys A clm/src/ED/biogeophys/EDAccumulateFluxesMod.F90 A clm/src/ED/biogeophys/EDSurfaceAlbedoMod.F90 A clm/src/ED/biogeophys/EDPhotosynthesisMod.F90 A clm/src/ED/biogeophys/EDBtranMod.F90 A clm/src/Notes M clm/src/unit_test_shr/unittestSubgridMod.F90 A clm/src/utils A clm/src/utils/spmdGathScatMod.F90 A clm/src/utils/clm_time_manager.F90 A clm/src/utils/clm_nlUtilsMod.F90 A clm/src/utils/clm_varorb.F90 A clm/src/utils/accumulMod.F90 A clm/src/utils/getdatetime.F90 A clm/src/utils/fileutils.F90 A clm/src/utils/dtypes.h A clm/src/utils/spmdMod.F90 A clm/src/utils/SimpleMathMod.F90 A clm/src/utils/domainMod.F90 A clm/src/utils/restUtilMod.F90 A clm/src/utils/quadraticMod.F90 A clm/src/utils/CMakeLists.txt A clm/src/utils/restUtilMod.F90.in A clm/src/biogeochem A clm/src/biogeochem/CNCStateUpdate2Mod.F90 A clm/src/biogeochem/CNDecompCascadeConType.F90 A clm/src/biogeochem/CNNitrifDenitrifMod.F90 A clm/src/biogeochem/CNGRespMod.F90 A clm/src/biogeochem/CNBalanceCheckMod.F90 A clm/src/biogeochem/CNNStateUpdate3Mod.F90 A clm/src/biogeochem/CNDVDriverMod.F90 A clm/src/biogeochem/SatellitePhenologyMod.F90 A clm/src/biogeochem/CNPhenologyMod.F90 A clm/src/biogeochem/CNCarbonFluxType.F90 A clm/src/biogeochem/CNCarbonStateType.F90 A clm/src/biogeochem/CNCStateUpdate1Mod.F90 A clm/src/biogeochem/VOCEmissionMod.F90 A clm/src/biogeochem/CNAnnualUpdateMod.F90 A clm/src/biogeochem/CNNStateUpdate2Mod.F90 A clm/src/biogeochem/CropType.F90 A clm/src/biogeochem/CNAllocationMod.F90 A clm/src/biogeochem/CNNDynamicsMod.F90 A clm/src/biogeochem/DryDepVelocity.F90 A clm/src/biogeochem/CNDecompCascadeBGCMod.F90 A clm/src/biogeochem/CNSoilLittVertTranspMod.F90 A clm/src/biogeochem/CNDecompCascadeCNMod.F90 A clm/src/biogeochem/CNC14DecayMod.F90 A clm/src/biogeochem/CNGapMortalityMod.F90 A clm/src/biogeochem/CNNStateUpdate1Mod.F90 A clm/src/biogeochem/CNFireMod.F90 A clm/src/biogeochem/CNNitrogenFluxType.F90 A clm/src/biogeochem/CNMRespMod.F90 A clm/src/biogeochem/MEGANFactorsMod.F90 A clm/src/biogeochem/CNVerticalProfileMod.F90 A clm/src/biogeochem/CNCIsoFluxMod.F90 A clm/src/biogeochem/CNWoodProductsMod.F90 A clm/src/biogeochem/CNPrecisionControlMod.F90 A clm/src/biogeochem/ch4Mod.F90 A clm/src/biogeochem/DUSTMod.F90 A clm/src/biogeochem/CNDVLightMod.F90 A clm/src/biogeochem/ch4varcon.F90 A clm/src/biogeochem/CNDecompMod.F90 A clm/src/biogeochem/CNCStateUpdate3Mod.F90 A clm/src/biogeochem/CNSharedParamsMod.F90 A clm/src/biogeochem/CNDVType.F90 A clm/src/biogeochem/CNStateType.F90 A clm/src/biogeochem/CNEcosystemDynMod.F90 A clm/src/biogeochem/CNNitrogenStateType.F90 A clm/src/biogeochem/CNVegStructUpdateMod.F90 A clm/src/biogeochem/CNDVEstablishmentMod.F90 A clm/src/biogeophys A clm/src/biogeophys/SnowSnicarMod.F90 A clm/src/biogeophys/SnowHydrologyMod.F90 A clm/src/biogeophys/TridiagonalMod.F90 A clm/src/biogeophys/FrictionVelocityType.F90 A clm/src/biogeophys/LakeFluxesMod.F90 A clm/src/biogeophys/PhotosynthesisMod.F90 A clm/src/biogeophys/AerosolType.F90 A clm/src/biogeophys/ActiveLayerMod.F90 A clm/src/biogeophys/QSatMod.F90 A clm/src/biogeophys/SoilHydrologyType.F90 A clm/src/biogeophys/HydrologyDrainageMod.F90 A clm/src/biogeophys/LakeStateType.F90 A clm/src/biogeophys/BareGroundFluxesMod.F90 A clm/src/biogeophys/SolarAbsorbedType.F90 A clm/src/biogeophys/CanopyHydrologyMod.F90 A clm/src/biogeophys/UrbanFluxesMod.F90 A clm/src/biogeophys/SurfaceAlbedoMod.F90 A clm/src/biogeophys/UrbanRadiationMod.F90 A clm/src/biogeophys/PhotosynthesisType.F90 A clm/src/biogeophys/CanopyTemperatureMod.F90 A clm/src/biogeophys/HydrologyNoDrainageMod.F90 A clm/src/biogeophys/DaylengthMod.F90 A clm/src/biogeophys/WaterfluxType.F90 A clm/src/biogeophys/BalanceCheckMod.F90 A clm/src/biogeophys/SoilTemperatureMod.F90 A clm/src/biogeophys/WaterStateType.F90 A clm/src/biogeophys/LakeTemperatureMod.F90 A clm/src/biogeophys/FrictionVelocityMod.F90 A clm/src/biogeophys/SoilFluxesMod.F90 A clm/src/biogeophys/TemperatureType.F90 A clm/src/biogeophys/SurfaceAlbedoType.F90 A clm/src/biogeophys/AerosolMod.F90 A clm/src/biogeophys/SoilStateType.F90 A clm/src/biogeophys/SurfaceResistanceMod.F90 A clm/src/biogeophys/UrbanAlbedoMod.F90 A clm/src/biogeophys/CanopyFluxesMod.F90 A clm/src/biogeophys/CMakeLists.txt A clm/src/biogeophys/RootBiophysMod.F90 A clm/src/biogeophys/test A clm/src/biogeophys/test/CMakeLists.txt A clm/src/biogeophys/test/Daylength_test A clm/src/biogeophys/test/Daylength_test/test_daylength.pf A clm/src/biogeophys/test/Daylength_test/CMakeLists.txt A clm/src/biogeophys/SurfaceRadiationMod.F90 A clm/src/biogeophys/SoilWaterMovementMod.F90 A clm/src/biogeophys/SoilMoistStressMod.F90 A clm/src/biogeophys/SoiWatRetCurveParMod.F90 A clm/src/biogeophys/EnergyFluxType.F90 A clm/src/biogeophys/CanopyStateType.F90 A clm/src/biogeophys/BandDiagonalMod.F90 A clm/src/biogeophys/SoilHydrologyMod.F90 A clm/src/biogeophys/LakeCon.F90 A clm/src/biogeophys/LakeHydrologyMod.F90 A clm/src/biogeophys/UrbanParamsType.F90 A clm/src/dyn_subgrid A clm/src/dyn_subgrid/dynLandunitAreaMod.F90 A clm/src/dyn_subgrid/dynTimeInfoMod.F90 A clm/src/dyn_subgrid/dynSubgridDriverMod.F90 A clm/src/dyn_subgrid/dynFileMod.F90 A clm/src/dyn_subgrid/dynConsBiogeochemMod.F90 A clm/src/dyn_subgrid/dynEDMod.F90 A clm/src/dyn_subgrid/dynVarMod.F90 A clm/src/dyn_subgrid/dynVarTimeUninterpMod.F90 A clm/src/dyn_subgrid/dynVarMod.F90.in A clm/src/dyn_subgrid/dynVarTimeInterpMod.F90.in A clm/src/dyn_subgrid/dynVarTimeUninterpMod.F90.in A clm/src/dyn_subgrid/do_genf90 A clm/src/dyn_subgrid/CMakeLists.txt A clm/src/dyn_subgrid/test A clm/src/dyn_subgrid/test/dynLandunitArea_test A clm/src/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf A clm/src/dyn_subgrid/test/dynLandunitArea_test/CMakeLists.txt A clm/src/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights.pf A clm/src/dyn_subgrid/test/dynVar_test A clm/src/dyn_subgrid/test/dynVar_test/test_dynVarShared.F90 A clm/src/dyn_subgrid/test/dynVar_test/test_dynVarTimeInterp.pf A clm/src/dyn_subgrid/test/dynVar_test/test_dynVarTimeUninterp.pf A clm/src/dyn_subgrid/test/dynVar_test/CMakeLists.txt A clm/src/dyn_subgrid/test/dynTimeInfo_test A clm/src/dyn_subgrid/test/dynTimeInfo_test/test_dynTimeInfo.pf A clm/src/dyn_subgrid/test/dynTimeInfo_test/CMakeLists.txt A clm/src/dyn_subgrid/test/CMakeLists.txt A clm/src/dyn_subgrid/test/dynInitColumns_test A clm/src/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf A clm/src/dyn_subgrid/test/dynInitColumns_test/CMakeLists.txt A clm/src/dyn_subgrid/dynHarvestMod.F90 A clm/src/dyn_subgrid/dynPriorWeightsMod.F90 A clm/src/dyn_subgrid/dynpftFileMod.F90 A clm/src/dyn_subgrid/dynVarTimeInterpMod.F90 A clm/src/dyn_subgrid/dynCNDVMod.F90 A clm/src/dyn_subgrid/dynConsBiogeophysMod.F90 A clm/src/dyn_subgrid/dynInitColumnsMod.F90 A clm/src/unit_test_mocks/util_share/restUtilMod_mock.F90 A clm/src/unit_test_mocks/util_share/restUtilMod_mock.F90.in List all existing files that have been modified, and describe the changes: ALL files have been modified - see the general description for an overview of what was done - the following files have not had their directories changed - so the summary is below M clm/bld/configure - needed to account for change in filepath M clm/bld/CLMBuildNamelist.pm M clm/bld/namelist_files/namelist_definition_clm4_5.xml - see namelist changes mentioned above M clm/src/unit_test_mocks/util_share/clm_time_manager_mock.F90 M clm/src/unit_test_mocks/util_share/ncdio_var.F90 M clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90 M clm/src/unit_test_mocks/util_share/do_genf90 M clm/src/unit_test_mocks/util_share/CMakeLists.txt M clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90.in M clm/src/unit_test_mocks/main/histFileMod_mock.F90 M clm/src/CMakeLists.txt - unit test changes needed to account for introduction of new data types and directory structure M clm/src/cpl/lnd_comp_esmf.F90 M clm/src/cpl/lnd_import_export.F90 M clm/src/cpl/lnd_comp_mct.F90 - coupling interface changes needed to account for introduction of new data types CLM testing: build-namelist tests: yellowstone yes goldbach yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes NOTE for goldbach_nag - four ED compare_hist tests fail with small bit for bit differences. In fact - looking more closely, these tests are also run for pgi and intel on goldbach - and values for LITTER_IN and LITTER_OUT are 0. for those compilers but non-zero for nag. With the refactoring code - those fields are again 0. for intel and pgi - but totally different and non-zero for nag. Apparently, this is known problem that will get resolved in when new ED changes are brount in ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest short tests (aux_clm_short) (generally these are NOT used when making a tag): N/A tools testing: N/A CLM tag used for the baseline comparisons: clm4_5_1_r080 Changes answers relative to baseline: No - BFB =============================================================== =============================================================== Tag name: clm4_5_1_r080 Originator(s): erik (Erik) Date: Sat Aug 16 15:01:35 MDT 2014 One-line Summary: Update externals to CESM trunk versions, allow eighth degree as a valid resolution Purpose of changes: Update all the externals to the very latest CESM trunk versions (based off of current cesm1_3_alpha13a). Requirements for tag: Get working with trunk externals Test level of tag: regular Bugs fixed (include bugzilla ID): 2012 Problem with mksurfdata_map for eighth degree grid... Scripts issues as follows: 2024 nlcompareonly option to create_test not working correctly for reporting 2019 ERH tests don't save the base env_run.xml, so have trouble when resubmitted... 2018 Failed tests in cesm1_3_beta11 needed for CLM 2005 Remove untested named compsets and grids 1999 T85_g16 has inconsistent land domain and surface datasets Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: Archiving updated in support of time series generation Running test suite now builds some shared libraries built only once Describe any changes made to the namelist: Resolve env and xml vars used in user_nl_* Default for drv_in profile_timer changed from 4 to 1 Some PE layours change: f10_f10 change from 15x1 to 30x1 List any changes to the defaults for the boundary datasets: Add 0.125x0.125 mapping files Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): Update to cesm1_3_alpha13a externals scripts to scripts4_140814a scripts/doc to doc_140519 Machines to Machines_140811 CMake to CMake_Fortran_utils_140715 drv to drvseq5_0_15 cism to cism1_140602 timing to timing_140416 pio to pio1_8_12 cprnc to cprnc_140625 mapping to mapping_140702b (note: gen_domain changes answers) unit_testing to unit_testing_0_07 PTCLM to PTCLM2_140816 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkgridmapMod.F90 -- decrease tolerance M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.ncl - Loop over variables and then cat the files together at the end. This makes the process possible for high resolution and speeds up lower resolution sub-setting as well. M models/lnd/clm/tools/clm4_0/mksurfdata_map/src/mkgridmapMod.F90 -- decrease tolerance M models/lnd/clm/tools/shared/mkmapdata/README --------------------- Fix/update documentation M models/lnd/clm/bld/unit_testers/build-namelist_test.pl ----------- Correct test count M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Add 0.125x0.125 mapping files M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Add 0.125x0.125 as a valid resolution M models/lnd/clm/bld/namelist_files/createMapEntry.pl --- Correct path, get working again CLM testing: build-namelist tests: yellowstone yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes yellowstone_gnu yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes short tests (aux_clm_short) (generally these are NOT used when making a tag): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes tools testing: yellowstone interactive yes CLM tag used for the baseline comparisons: clm4_5_1_r079 Changes answers relative to baseline: Yes! (PE layouts that change) Summarize any changes to answers, i.e., - what code configurations: non single-point configurations, where PE layout changes (f10_f10) - what platforms/compilers: all - nature of change (roundoff) x2l_Flrr_volr changes to roundoff =============================================================== =============================================================== Tag name: clm4_5_1_r079 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Thu Jul 31 17:09:57 MDT 2014 One-line Summary: G. Bisht (LBL) soil temperature refactor; machines update for goldbach-intel Purpose of changes: Refactor soil temperature module to break the construction of the linear system LHS matrix and RHS vector into small physics based routines. Update machines external to fix compiling with goldbach-intel. Requirements for tag: regular Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: gbisht, self, muszala List any svn externals directories updated (csm_share, mct, etc.): Machines List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 - break creation of linear system into small physics based routines. CLM testing: regular build-namelist tests: N/A regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 4.0 ok, 4.5 ok yellowstone_pgi - 4.0 ok, 4.5 ok goldbach_nag - 4.0 ok, 4.5 ok (see note below) goldbach_intel - 4.0 ok, 4.5 ok goldbach_pgi - 4.0 ok, 4.5 ok NOTE for goldbach_nag - four ED compare_hist tests fail with small non bit for bit differences. This is the same issue described in clm4_5_1_r078 tag notes. ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest tools testing: N/A CLM tag used for the baseline comparisons: clm4_5_1_r078 Changes answers relative to baseline: No, bit for bit =============================================================== =============================================================== Tag name: clm4_5_1_r078 Originator(s): muszala (Stefan Muszala) Date: Wed Jul 23 20:42:00 MDT 2014 One-line Summary: Add lai stream capability and the ability to run with V5 cruncep data. Code written by swenson, modified and tested by muszala. Purpose of changes: Add lai stream capability with use_lai_streams namelist variable. Also add a datm_mode option s.t. we can use cruncep V5 data if a user wants. Move anomaly focing code out of CLM and into DATM. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: add setup_logic_lai_streams which controls use_lai_streams namelist variable List any changes to the defaults for the boundary datasets: added option to use V5 cruncep data sets. V4 is default. Describe any substantial timing or memory changes: N/A Code reviewed by: kluzek, swenson, self List any svn externals directories updated (csm_share, mct, etc.): -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/addclm50_tags/addclm50_n06_ED_scripts_015_140305_rev +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/addclm50_tags/addclm50_n09_ED_scripts_015_140305_rev -models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_140312 +models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_140723 -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140418 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: ! add setup_logic_lai_streams function M models/lnd/clm/bld/CLMBuildNamelist.pm ! add entries for: stream_year_first_lai, stream_year_last_lai, model_year_align_lai, stream_fldfilename_lai, lai_mapalgo M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml ! add default values for items added in namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ! add use_lai_streams code. lai_init, lai_interp M models/lnd/clm/src/clm4_5/biogeochem/SatellitePhenologyMod.F90 ! some comment clean up. add use_lai_streams logical M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 ! add use_lai_streams namelist handling and mpi_bcast call M models/lnd/clm/src/clm4_5/main/controlMod.F90 ! remove snomaly forcing streams since they are now in the datm M models/lnd/clm/src/cpl/clm_cpl_indices.F90 ! remove anomaly forcing code since it is now in datm M models/lnd/clm/src/cpl/lnd_import_export.F90 CLM testing: build-namelist tests: yellowstone - failed 20 tests of 537. This is expected due to the addition of the use_lai_streams namelist variable. regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): -- nlcomp tests for 45 will fail -- yellowstone_intel - 40 OK - 45 OK yellowstone_pgi - 40 OK - 45 OK goldbach_nag - 40 OK - 45 OK goldbach_intel - 40 OK - 45 OK goldbach_pgi - 40 OK - 45 OK Both NAG tests on goldbach for ED compsets failed BFB. All other compilers and machines pass. Error probably related to a non-BFB error that I (spm) see with varying numbers of time-steps on an ERS test and am currently trying to fix. FAIL ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest.GC.again_45_nag.compare_hist.clm4_5_1_r077_redo FAIL ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_nag.clm-edTest.GC.again_45_nag.clm2.h0.compare_hist.clm4_5_1_r077_redo FAIL SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest.GC.again_45_nag.compare_hist.clm4_5_1_r077_redo FAIL SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest.GC.again_45_nag.clm2.h0.compare_hist.clm4_5_1_r077_redo CLM tag used for the baseline comparisons: clm4_5_1_r077 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_1_r077 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Thu Jul 10 21:55:11 MDT 2014 One-line Summary: Refactor from Jinyun Tang (LBL) to make hydrology more modular and eventually allow runtime selection of different physics implementations. Purpose of changes: Refactor a number of routines in clm45 hydrology to move duplicate code into reusable routines, make the code more modular for eventual unit testing and run time selection of different physics. Requirements for tag: bit for bit, regular testing Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self, clm-cmt List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: models/lnd/clm/src/util_share/SimpleMathMod.F90 - reuseable array functions models/lnd/clm/src/clm4_5/main/init_hydrology.F90 - initialize different hydrology submodules. models/lnd/clm/src/clm4_5/main/FuncPedotransferMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/SurfaceResistanceMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/RootBiophysMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/SoilWaterMovementMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/SoilMoistStressMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/SoiWatRetCurveParMod.F90 - modularize List all existing files that have been modified, and describe the changes: models/lnd/clm/bld/query-xFail - check python version and provide an error message of it is too old. models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 - remove unused min/max variables that conflict with intrinsics with gfortran. models/lnd/clm/src/clm4_5/main/initTimeConstMod.F90 - modularize models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 - allocate new variable, fix bounds on porosity models/lnd/clm/src/clm4_5/main/controlMod.F90 - initialize new hydrology modules models/lnd/clm/src/clm4_5/main/clmtype.F90 - add new variable models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 - modularize models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 - modularize CLM testing: regular build-namelist tests: yellowstone - n/a regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - OK clm40, OK clm45 yellowstone_pgi - OK clm40, OK clm45 goldbach_nag - OK clm40, OK clm45 goldbach_intel - OK clm40, OK clm45 goldbach_pgi - OK clm40, OK clm45 short tests (aux_clm_short) (generally these are NOT used when making a tag): yellowstone_intel - n/a yellowstone_pgi - n/a goldbach_nag - n/a tools testing: yellowstone interactive - n/a goldbach interactive - n/a CLM tag used for the baseline comparisons: clm4_5_1_r076 Changes answers relative to baseline: No, bit for bit Note: SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTest is not bit for bit. This is the same test Stef had problems with in clm4_5_75. He has looked at it and given the ok make the tag as is. =============================================================== =============================================================== Tag name: clm4_5_1_r076 Originator(s): erik (Erik) Date: Mon Jul 7 14:24:07 MDT 2014 Orig Date: Wed Jun 25 13:49:49 MDT 2014 (Date of what was tagged as clm4_6_0, before we changed naming convention) One-line Summary: Answer changes for fire code from Fang Li Purpose of changes: Several changes to CN Fire model. Some fixes for non-transient, as well as limiting of fire for high tropical forest coverage. Change some units from per time-step to per second. Change Lightning input dataset from just cloud to ground to total lightning. Some fire parameters were also changed and re-tuned for Qian forcing. Some more documentation on fire fields was added. When -ed_mode is sent to CLM build-namelist, a particular ED params dataset is used over the default. Make a simple change that allows ED to run when CN is off. Add a 1850 and transient 20thC and rcp=8.5 datasets for 1x1_brazil. Requirements for tag: Fix bugs 1805 and 1719, lower fire amount in amazon Test level of tag: regular Bugs fixed (include bugzilla ID): 1996 -- change cloud to ground lightning dataset to total lightning 1995 -- change units from per time-step to per second 1805 -- fire fix for non-transient 1719 -- remove double counting of baf in fire area 1992 -- allow ED to run when use_cn=.false. 1988 -- Add ED params dataset. 1991 -- transient datasets for 1x1_brazil Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: specific rparams file for ED add 1850 and 20thC, rcp8.5 datasets for 1x1_brazil use lightning file that is total lightning not just cloud-to-ground Describe any substantial timing or memory changes: The test SMS.f19_g16.IRCP45CN.yellowstone_pgi showed a memory increase Code reviewed by: self, lifang List any svn externals directories updated (csm_share, mct, etc.): scripts scripts to addclm50_n06_ED_scripts_015_140305_rev List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/build-namelist_test.pl --------- Add some ED tests M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - Different params file for ED, and add 1850 and 20thC, rcp8.5 datasets for 1x1_brazil and use lightning file that is total lightning not just cloud-to-ground M models/lnd/clm/bld/CLMBuildNamelist.pm ------------------------- Pass use_ed when getting paramfile M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 ------------- Revisions from Fang Li (2014), change parameters, add documentation, tropical forests will only burn if > 60% coverage, change some fields units to per second rather than per time-step, M models/lnd/clm/src/clm4_5/biogeochem/CNInitMod.F90 ------------- change units for nfire, and farea_burned M models/lnd/clm/src/clm4_5/main/clmtype.F90 --------------------- Change units for nfire, lfc, lfc2, baf_crop, baf_peatf, fbac, fbac1, farea_burned M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 ----------------- Change units for: LFC2, NFIRE, FAREA_BURNED, BAF_CROP, BAF_PEATF M models/lnd/clm/src/clm4_5/biogeophys/ED/EDPhotosynthesisMod.F90 Allow to work when use_cn is .false., use c3psn+1 in finding index for dr array. CLM testing: build-namelist tests: yes NOTE: 191 of the 537 compare tests fail, compared to clm4_5_75, because of changes in the namelist. yellowstone yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes CLM tag used for the baseline comparisons: clm4_5_75 Changes answers relative to baseline: If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CN and BGC with CLM4.5 physics CLM4.0 for all modes, and CLM4.5 with SP or ED should be identical - what platforms/compilers: All - nature of change: new climate Fang Li, ran simulations with Qian forcing on yellowstone and tuned fire parameters to that forcing. However, her simulations had a minor bug in the conversion of total lightning to just cloud-to-ground (latitude in degree's was used for a cosine, rather than latitude in radians -- see bug 1996). =============================================================== =============================================================== Tag name: clm4_5_75 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Fri May 30 11:18:35 MDT 2014 One-line Summary: update externals to rtm1_0_38 and esmf_wrf_timemgr_140529 Purpose of changes: update externals to rtm1_0_38 and esmf_wrf_timemgr_140529. These modifications are based on valgrind errors that orginated in src/riverroute. Tested in clm4_5_72 to make sure everything was still BFB (at least w.r.t. CLM testing). Retested (results below) against clm4_5_73. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID):N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_37 +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_38 -models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_130213 +models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_140529 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: N/A CLM testing: in addition to other clm tests I updated the rtm and esmf externals in cesm1_3_alpha09c and ran two B cases. Note: There is one ED test (SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_nag.clm-edTestGb) that failed for me in my testing but passed for Bill with the same checkout. We will keep an eye on this, but it's not super-critical at the moment. Differences are very small in cpl. voc fields (largest RMS difference is 1e-13). >>more TestStatus PASS ERS_PT.T31_g37.B1850CN.yellowstone_gnu PASS ERS_PT.T31_g37.B1850CN.yellowstone_gnu.memleak >>more TestStatus PASS ERS.ne30_g16.B1850C5CN.yellowstone_intel PASS ERS.ne30_g16.B1850C5CN.yellowstone_intel.memleak build-namelist tests: N/A regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel - 40 OK - 45 OK - component gen comp - OK yellowstone_pgi - 40 OK - 45 OK - component gen comp - OK goldbach_nag - 40 OK - 45 OK goldbach_intel - 40 OK - 45 OK goldbach_pgi - 40 OK - 45 OK CLM tag used for the baseline comparisons: clm4_5_73 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_74 Originator(s): sacks (sacks) Date: Wed May 28 16:05:36 MDT 2014 One-line Summary: misc. bfb changes - see detailed summary below Purpose of changes: (1) Rename fpftdyn to flanduse_timeseries, and make related changes to names throughout the code. This rename is in preparation for an upcoming tag where this file will take on more general uses (e.g., transient crop areas). (2) Decrease thresholds for water, snow and energy balance checks (these were too permissive) (3) Move stuff out of clm_varcon into landunit_varcon (for constants specific to CLM's landunits) and column_varcon (for constants specific to CLM's columns) - analogous to the existing pftvarcon (4) Move some routines out of initGridCellsMod into a new initSubgridMod (5) Make time_info a public member of dyn_file_type, which allows removing a bunch of delegation methods. And rename some things in time_info_type for clarity. (6) Rework metadata for the description of landunit, column and pft types on the history and restart files, to centralize these descriptions to the appropriate place in the code. (7) Add general-purpose functionality for setting up subgrid structure for unit tests (8) Move unit tests into source tree, rather than being in test/unit_testers. Now the top-level script is in models/lnd/clm/src. (9) Fix baseline comparisons for PTCLM tests Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): - 1928 (create landunit_varcon.F90 and column_varcon.F90 from parts of clm_varcon.F90) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: fpftdyn renamed to flanduse_timeseries List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts: addclm50_n03_ED_scripts_015_140305_rev -> addclm50_n04_ED_scripts_015_140305_rev - Rename CLM's fpftdyn to flanduse_timeseries in tests; update perl5lib tools/unit_testing: unit_testing_0_04 -> unit_testing_0_05 - the major change here is allowing rebuilds with intel without needing to specify --clean models/lnd/clm/tools/PTCLM: PTCLM2_140423 -> PTCLM2_140521 - rename fpftdyn -> flanduse_timeseries, and other related renames List all files eliminated: ========= Renamed D models/lnd/clm/tools/clm4_5/mksurfdata_map/pftdyn_hist_simyr1850-2005.txt D models/lnd/clm/tools/clm4_0/mksurfdata_map/pftdyn_hist_simyr1850-2005.txt ========= Move unit tests into source tree D models/lnd/clm/test/unit_testers/clm4_5/biogeophys/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test/test_daylength.pf D models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test D models/lnd/clm/test/unit_testers/clm4_5/biogeophys D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/test_update_landunit_weights.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarShared.F90 D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarTimeInterp.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarTimeUninterp.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/test_dynTimeInfo.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test/test_init_columns.pf D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test D models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/GetGlobalValuesMod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90.in D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/clm_time_manager_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_pio_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/spmdMod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/do_genf90 D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_pio_mock.F90.in D models/lnd/clm/test/unit_testers/clm4_5/mock/util_share D models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/shr_sys_mod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/mct_mod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share D models/lnd/clm/test/unit_testers/clm4_5/mock/main/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/mock/main/histFileMod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/main D models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/dynFileMod_mock.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid D models/lnd/clm/test/unit_testers/clm4_5/mock/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/mock D models/lnd/clm/test/unit_testers/clm4_5/CMakeLists.txt D models/lnd/clm/test/unit_testers/clm4_5/README D models/lnd/clm/test/unit_testers/clm4_5 D models/lnd/clm/test/unit_testers List all files added and what they do: ========= Renamed A models/lnd/clm/tools/clm4_5/mksurfdata_map/landuse_timeseries_hist_simyr1850-2005.txt A models/lnd/clm/tools/clm4_0/mksurfdata_map/landuse_timeseries_hist_simyr1850-2005.txt ========= Move stuff out of clm_varcon into landunit_varcon (for constants specific to CLM's landunits) and column_varcon (for constants specific to CLM's columns) - analogous to the existing pftvarcon. A models/lnd/clm/src/clm4_5/main/landunit_varcon.F90 A models/lnd/clm/src/clm4_5/main/column_varcon.F90 ========= Move some routines out of initGridCellsMod - these are lower-level routines that can also be used by unit test code. So initGridCellsMod contains higher-level stuff that is specific to how the subgrid structure is set up in a production run; and initSubgridMod contains lower-level stuff that doesn't know or care how things are actually set up, conceptually. A models/lnd/clm/src/clm4_5/main/initSubgridMod.F90 ========= Add general-purpose functionality for setting up subgrid structure for unit tests A models/lnd/clm/src/unit_test_shr/unittestSubgridMod.F90 A models/lnd/clm/src/unit_test_shr/CMakeLists.txt A models/lnd/clm/src/unit_test_shr ========= Move unit tests into source tree; also modify some unit tests to take advantage of the new unittestSubgridMod; also add tests of subgridWeightsMod and clm_glclnd A models/lnd/clm/src/clm4_5/main/test/subgridWeights_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/main/test/subgridWeights_test/test_subgridWeights.pf A models/lnd/clm/src/clm4_5/main/test/subgridWeights_test A models/lnd/clm/src/clm4_5/main/test/CMakeLists.txt A models/lnd/clm/src/clm4_5/main/test/clm_glclnd_test/test_clm_glclnd.pf A models/lnd/clm/src/clm4_5/main/test/clm_glclnd_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/main/test/clm_glclnd_test A models/lnd/clm/src/clm4_5/main/test A models/lnd/clm/src/clm4_5/biogeophys/test/CMakeLists.txt A models/lnd/clm/src/clm4_5/biogeophys/test/Daylength_test/test_daylength.pf A models/lnd/clm/src/clm4_5/biogeophys/test/Daylength_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/biogeophys/test/Daylength_test A models/lnd/clm/src/clm4_5/biogeophys/test A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test/test_update_landunit_weights.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynLandunitArea_test A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarShared.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarTimeInterp.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynVar_test/test_dynVarTimeUninterp.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynVar_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynVar_test A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test/test_dynTimeInfo.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynTimeInfo_test A models/lnd/clm/src/clm4_5/dyn_subgrid/test/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test/test_init_columns.pf A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid/test/dynInitColumns_test A models/lnd/clm/src/clm4_5/dyn_subgrid/test A models/lnd/clm/src/README.unit_testing A models/lnd/clm/src/unit_test_mocks/util_share/ncdio_var.F90 A models/lnd/clm/src/unit_test_mocks/util_share/GetGlobalValuesMod_mock.F90 A models/lnd/clm/src/unit_test_mocks/util_share/ncdio_var.F90.in A models/lnd/clm/src/unit_test_mocks/util_share/clm_time_manager_mock.F90 - also add a routine to this mock, needed because of refactor of dyn_file_type / dyn_time_inof A models/lnd/clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90 A models/lnd/clm/src/unit_test_mocks/util_share/spmdMod_mock.F90 A models/lnd/clm/src/unit_test_mocks/util_share/do_genf90 A models/lnd/clm/src/unit_test_mocks/util_share/CMakeLists.txt A models/lnd/clm/src/unit_test_mocks/util_share/ncdio_pio_mock.F90.in A models/lnd/clm/src/unit_test_mocks/util_share A models/lnd/clm/src/unit_test_mocks/csm_share/shr_sys_mod_mock.F90 A models/lnd/clm/src/unit_test_mocks/csm_share/mct_mod_mock.F90 A models/lnd/clm/src/unit_test_mocks/csm_share/CMakeLists.txt A models/lnd/clm/src/unit_test_mocks/csm_share A models/lnd/clm/src/unit_test_mocks/main/CMakeLists.txt A models/lnd/clm/src/unit_test_mocks/main/histFileMod_mock.F90 A models/lnd/clm/src/unit_test_mocks/main A models/lnd/clm/src/unit_test_mocks/dyn_subgrid/dynFileMod_mock.F90 A models/lnd/clm/src/unit_test_mocks/dyn_subgrid/CMakeLists.txt A models/lnd/clm/src/unit_test_mocks/dyn_subgrid A models/lnd/clm/src/unit_test_mocks/CMakeLists.txt A models/lnd/clm/src/unit_test_mocks A models/lnd/clm/src/CMakeLists.txt List all existing files that have been modified, and describe the changes: ========= Renamed fpftdyn -> flanduse_timeseries, and other related changes to variable names (NOTE: Some source files are listed both here and elsewhere in the ChangeLog entry) M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata_map.namelist M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl M models/lnd/clm/tools/clm4_5/mksurfdata_map/README M models/lnd/clm/tools/shared/ncl_scripts/sample_inlist M models/lnd/clm/tools/shared/ncl_scripts/sample_outlist M models/lnd/clm/tools/clm4_0/mksurfdata_map/src/clm_varctl.F90 M models/lnd/clm/tools/clm4_0/mksurfdata_map/mksurfdata_map.namelist M models/lnd/clm/tools/clm4_0/mksurfdata_map/mksurfdata.pl M models/lnd/clm/tools/clm4_0/mksurfdata_map/README M models/lnd/clm/bld/CLMBuildNamelist.pm M models/lnd/clm/bld/unit_testers/build-namelist_test.pl M models/lnd/clm/bld/test_build_namelist/t/input/namelist_defaults_clm4_5_test.xml M models/lnd/clm/bld/test_build_namelist/t/input/namelist_definition_clm4_5_test.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_usr_files.xml M models/lnd/clm/doc/UsersGuide/trouble_shooting.xml M models/lnd/clm/doc/UsersGuide/single_point.xml M models/lnd/clm/doc/UsersGuide/tools.xml M models/lnd/clm/doc/UsersGuide/adding_files.xml M models/lnd/clm/doc/UsersGuide/custom.xml M models/lnd/clm/doc/UsersGuide/ptclm.xml M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/surfrdUtilsMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_0/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_0/main/pftdynMod.F90 M models/lnd/clm/src/clm4_0/main/controlMod.F90 M models/lnd/clm/src/clm4_0/main/clm_varctl.F90 M models/lnd/clm/src/clm4_0/main/clm_driver.F90 M models/lnd/clm/src/clm4_0/biogeophys/BiogeophysRestMod.F90 ========= Updated PTCLM external to rename fpftdyn -> flanduse_timeseries M models/lnd/clm/tools/SVN_EXTERNAL_DIRECTORIES ========= Decrease threshold for water & snow balance checks by 3 orders of magnitude; decrease threshold for energy balance checks by 2 orders of magnitude M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 ========= Separate clm_varcon into clm_varcon, column_varcon and landunit_varcon M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 M models/lnd/clm/src/clm4_5/main/CMakeLists.txt ========= Move some routines out of initGridCellsMod, into a new initSubgridMod (see detailed notes above) M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 ========= Make time_info a public member of dyn_file_type. This allows us to remove all methods from dyn_file_type (which were just delegating responsibility to time_info_type). Also rename some methods and variables in time_info_type. M models/lnd/clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90.in - also change intent(in) to intent(inout), fixing a gfortran problem M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90.in M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90.in M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90 ========= Rework metadata for the description of landunit, column and pft types on the history and restart files. Point is to centralize the definition of these different types as much as possible (rather than, e.g., having restFileMod know about the translation between landunit indices and names). For the history file, I am removing the metadata from the PCT_LANDUNIT long name, instead putting it in global metadata, as is done for the restart file. M models/lnd/clm/src/clm4_5/main/subgridWeightsMod.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 ========= Change 'use' statements based on my split of clm_varcon into clm_varcon, landunit_varcon and column_varcon M models/lnd/clm/src/clm4_5/biogeochem/CNWoodProductsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNInitMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4InitMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/main/initTimeConstMod.F90 M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/subgridMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/initColdMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/ED/EDCLMLinkMod.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeochemMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynEDMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynCNDVMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeophysMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynInitColumnsMod.F90 ========= Fix baseline comparisons for PTCLM tests M models/lnd/clm/test/tools/TSMscript_tools.sh M models/lnd/clm/test/tools/TBLscript_tools.sh ========= Just changes in whitespace M models/lnd/clm/src/clm4_5/dyn_subgrid/CMakeLists.txt M models/lnd/clm/src/clm4_5/biogeophys/CMakeLists.txt CLM testing: Most testing done on dynlu_crops_n01_addclm50bld_n06_clm4_5_72; PTCLM and tools testing done on dynlu_crops_n03_addclm50bld_n06_clm4_5_72 Note that the branch was up-to-date with addclm50bld_n06_clm4_5_72; this is identical to clm4_5_73 except for a fix to the build-namelist tests (see below) build-namelist tests: yellowstone: ok compared against addclm50bld_n06_clm4_5_72 (essentially clm4_5_73) expected diffs for transient cases The following tests also failed when comparing the baseline against itself (NOTE: this is apparently fixed in clm4_5_73): 466/497 < FAIL> 467/497 < FAIL> 496/497 < FAIL> 497/497 < FAIL> regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel: ok yellowstone_pgi: ok goldbach_nag: ok goldbach_intel: ok goldbach_pgi: ok Most comparisons were done against clm4_5_72. These comparisons failed due to a problem with component_gen_comp and the SSP test; manual comparisons show these to be identical to Erik's tests (for clm4_5_73): FAIL SSP.f19_g16.I1850CLM45BGC.yellowstone_pgi.clm-default.GC.0520-2021.45.p.clm2.h0.compare_hist.clm4_5_72 FAIL SSP.f19_g16.I1850CLM45BGC.yellowstone_pgi.clm-default.GC.0520-2021.45.p.clm2.h1.compare_hist.clm4_5_72 For new tests added in Erik's upcoming tag (clm4_5_73), I did manual comparisons against Erik's baselines (cpl & clm hist for the yellowstone tests, just cpl for goldbach tests) - all PASS. Note that CLM hist files were NOT compared for any goldbach tests, because there were no CLM hist file baselines for clm4_5_72. tools testing: yellowstone interactive: ok Compared against addclm50bld_n06_clm4_5_72 (essentially clm4_5_73) Failures in the following baseline comparisons, due to changed name of output file (surfdata.pftdyn -> landuse.timeseries). Manual comparisons showed the output files to be identical in all cases: 010 bl754 TBLtools.sh clm4_0 mksurfdata_map tools__s namelist ...................................\c rc=7 FAIL 012 blg54 TBLtools.sh clm4_5 mksurfdata_map tools__s namelist ...................................\c rc=7 FAIL 018 bl974 TBLscript_tools.sh clm4_0 mksurfdata_map mksurfdata.pl mksrfdt_1x1_brazil_1850-2000^tools_\c rc=7 FAIL 030 bli74 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_1x1_brazil_1850-2000^tools_\c rc=7 FAIL 040 blfg4 TBLscript_tools.sh shared PTCLM PTCLMmkdata PTCLM_USUMB_Global_clm4_5^buildtools ......\c rc=7 FAIL Other than that, all tests & baseline comparisons passed CLM tag used for the baseline comparisons: clm4_5_72, except where noted above Changes answers relative to baseline: NO - bfb =============================================================== =============================================================== Tag name: clm4_5_73 Originator(s): erik (Erik Kluzek,UCAR/TSS,303-497-1326) Date: Wed May 28 15:33:10 MDT 2014 One-line Summary: Add the stub ability for clm5_0 physics to CLM build system Purpose of changes: Add phys=clm5_0 as an option to the build. Currently, nothing is different in the code, this just adds the capability to do clm5_0 as a seperate configuration. The one thing that is different between clm4_5 and clm5_0 is the setting of urban_hac. Fix several issues needed for CAM: problem in DryDeposition (reoccurance of bug 1883, that was fixed and then unfixed in clm4_5_48), fix for internal compiler errors. CAM has been using the branch version of this since: cam5_3_29. Work on updates for PTCLM. Add some new sites for Rosie and Jinyun. Correct the call to mkmapdata.sh. Have CLM1PT forcing directory to use DIN_LOC_ROOT_CLMFORC so you can point it to a location seperate from DIN_LOC_ROOT. Add a new support script to PTCLM to submit a list of sites to batch: PTCLMsublist. Also allow release_tags in version find. Get buildtools to work on edison/hopper. Get tools to work on hopper and edison, and update mapping to use ESMF6.3.0. Fix various bugs: internal compiler error on janus, trigger an error if user_datm.streams.txt file is readonly (rather than hang). Use DIN_LOC_ROOT_CLMFORC for CLM1PT. Requirements for tag: Fix bug 1883 and 1985 for Cheryl and CAM, fix PTCLM, add clm5_0, tools on hopper/edison Test level of tag: regular Bugs fixed (include bugzilla ID): 1985 Internal compiler error on yellowstone with CLM in CAM standalone build 1965 Internal compiler error on janus with CLM on janus 1938 Upgrade mkmapdata to ESMF6.3.0 1937 Using a read-only user_datm.streams.txt file causes cesm_setup to hang 1936 CLM1PT forcing directory needs to use DIN_LOC_ROOT_CLMFORC 1935 Changes needed to get tools to build on hopper... 1933 Correct call to mkmapdata.sh in PTCLM 1925 Add more sites to PTCLM 1904 check for LSF_PJL_TYPE in regridbatch.sh doesn't work correctly 1883 uninitialized variable in DryDepVelocity.F90 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: Add clm5_0 as a new supported physics type Describe any changes made to the namelist: Set urban_hac according to physics clm5_0=ON_WASTEHEAT, and clm4_5=ON List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self (clm_phys perl object reviewed by team: bandre, muszala, sacks) List any svn externals directories updated (csm_share, mct, etc.): scripts, datm scripts to addclm50_n03_ED_scripts_015_140305_rev datm to datm8_140312 csm_share to share3_140418 Use trunk version rather than branch tools/mapping to mapping_131217a PTCLM to PTCLM2_140423 List all files eliminated: D models/lnd/clm/bld/unit_testers/env_run.xml -- This file is now built dynamically when the tester is run. List all files added and what they do: A models/lnd/clm/bld/query-xFail --- Add a script from Ben Andre to read and report on expected fails. A models/lnd/clm/bld/env_run.xml --- envxml_dir option is now required, so this provides a env_*.xml file that can be read by default, when build-namelist is called outside of CESM for testing. A models/lnd/clm/bld/config_files/clm_phys_vers.pm - Enter physics version as a string i.e.: clm4_0 and then have the ability to interpret it as different types so you can do logical operations on physics versions List all existing files that have been modified, and describe the changes: M models/lnd/clm/tools/shared/mkmapdata/regridbatch.sh ------- Add ability to run on hopper M models/lnd/clm/tools/shared/mkmapdata/mkmapdata.sh --------- Add ability to run on hopper/edison remove jaguarpf, and upgrade to ESMF6.3.0 M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.ncl - New version of NCL requires load before "begin" statement M models/lnd/clm/bld/configure ------------- Allow phys=clm5_0 and add in new clm_phys_vers object M models/lnd/clm/bld/README ---------------- Update info on files M models/lnd/clm/bld/CLMBuildNamelist.pm --- Put list of required options at top of help, and make envxml_dir a required option. Add in use of clm_phys_vers object. Make sure use_ed_spitfire is only on if use_ed is on. If CLM_UPDATE_GLC_AREAS=TRUE and phys=clm4_0 trigger an error M models/lnd/clm/bld/config_files/config_definition_clm4_5.xml - Add all three physics version options M models/lnd/clm/bld/config_files/config_definition_clm4_0.xml - Add info. about clm4_5/clm5_0 options M models/lnd/clm/bld/unit_testers/build-namelist_test.pl - Add use of clm_phys_vers, and create env_run.xml on the fly. Also add new tests for glacier update areas, and ED: usespitfireButNOTED, useEDclm40, useEDContradict2 useEDContradict, clm40andUpdateGlc, clm40andUpdateGlc, UpdateGlcContradict, UpdateGlcNoGLCMe, and tests for clm5_0 M models/lnd/clm/bld/test_build_namelist/t/test_vichydro.pm ----- Needs to use clm_phys_vers object M models/lnd/clm/bld/test_build_namelist/test_build_namelist.pl - Needs to use clm_phys_vers object M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - urban_hac depends on clm4_5/clm5_0 -------------- Add phys=clm5_0 for all use-cases that test on phys M models/lnd/clm/bld/namelist_files/use_cases/2000_control.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml ------------- Fix bugs 1883 (rs over lake) and 1983 (CAM internal compiler error) and 1965 ------------- (janus internal compiler error) M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 ----- Set rs over lake M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 -------- Add use only for ncd_pio M models/lnd/clm/src/clm4_5/main/restFileMod.F90 -------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 ---------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 ------ Add use only for ncd_pio M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 -- Add use only for ncd_pio M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 ----- Add use only for clmtype and dynVarTimeUninterpMod fixes internal compiler error on janus (1965) M models/lnd/clm/src/clm4_0/biogeochem/DryDepVelocity.F90 ----- Set rs over lake M models/lnd/clm/src/clm4_0/main/clm_initializeMod.F90 -------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/main/subgridRestMod.F90 ----------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/main/pftdynMod.F90 ---------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/main/iniTimeConst.F90 ------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/main/restFileMod.F90 -------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/main/surfrdMod.F90 ---------------- Add use only for ncd_pio M models/lnd/clm/src/clm4_0/biogeophys/UrbanInputMod.F90 ------ Add use only for ncd_pio M models/lnd/clm/src/clm4_0/biogeophys/BiogeophysRestMod.F90 -- Add use only for ncd_pio CLM testing: a) regular b) build_namelist c) tools build-namelist tests: yellowstone yes regular tests (aux_clm40, aux_clm45, with '-model_gen_comp clm2'): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes goldbach_intel yes goldbach_pgi yes tools testing: yellowstone interactive yes CLM tag used for the baseline comparisons: clm4_5_72 Changes answers relative to baseline: No (bit-for-bit) =============================================================== =============================================================== Tag name: clm4_5_72 Originator(s): muszala (Stefan Muszala) Date: Mon May 5 17:47:52 MDT 2014 One-line Summary: Introduce code for Ecosystem Demography (CLM(ED)) Model Purpose of changes: Introduce code for Ecosystem Demography (CLM(ED)) Model - first functional tag. A large chunk of this code was written and re-written by Rosie Fisher. "Introduce code for Ecosystem Demography (CLM(ED)) Model. Adds capability to allow plant functional types to compete for light, to represent recovery from disturbance, and to allow disturbances (i.e. fire) to only afflict some fraction of the canopy, and to represent vegetation at the scale of cohorts of trees. Note that this is a large change and includes: 1. Significant alterations to canopy albedo and surface radiation calculations 2. New photosynthesis scheme, based on existing science but to allow for more complex canopy structure 3. Introduction of a new allocation and growth scheme, (no consistent with that in CLM(CN) 4. Removal, for now, of Nitrogen limitation capabilities 5. Introduction of the SPITFIRE fire model, which interacts with ED via it's representation of size-structured mortality and removal of litter pools. 6. Introduction of a simple seed bank model to allow persistence of vegetation through fire events. 7. For ED compsets there exists a cohort dimension on the restart files. Cold starts and restarts work for the following. The 1x1_brazil is the most heavily tested case both from science and SE standpoint: 1x1_brazil.ICLM45CNED.yellowstone_[intel | pgi] 5x5_amazon.ICLM45CNED.yellowstone_[intel | pgi] 1x1_brazil.ICLM45CNED.goldbach_[nag | intel | pgi ] 5x5_amazon.ICLM45CNED.goldbach_[nag | intel | pgi ] Cold starts work for: f10_f10.ICLM45CNED.yellowstone_[intel | pgi]. f19_g16.ICLM45CNED.yellowstone_[intel | pgi] Code Origins: The ED code in CLM is originally based on code by Moorcroft (www.oeb.harvard.edu/faculty/moorcroft/code_and_data/index.html) and has been heavily modified in regards to both scientific implementation and assumptions. Fom a software engineering perspective, ED was rewritten from C into F2003 and the structure of the code has been significantly altered to fit into the CESM/CLM framework. The SPITFIRE code is based on (http://www.biogeosciences.net/7/1991/2010/bg-7-1991-2010.pdf) and has been significantly altered and extended to fit into the CESM/CLM framework. Other points: 1. removed many unused variables and module uses as reported by nag 2. changing text wrapping on comments so they end at 139 characters 3. changed many text based logical operators (.ne., .lt., .ge.) with their math. equivlanet (/=, > , <=) Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: ED functionality brought in. Does not change existing configurations. For ED, there are now a CNED and BGCED compset. Use CNED until BGC issues are worked out. Describe any changes made to the namelist: ED functionality brought in. Does not change existing configurations. For ED compsets, there are two new namelist variables. They are: use_ed = .true. use_ed_spit_fire = .true. use_ed_spit_fire is set to true by default if use_ed is on. Unless you are running our ED tests, you will have to change your user_nl_clm to something like: paramfile ='/glade/p/cesmdata/cseg/inputdata/lnd/clm2/edParams/CLMPARAMS_ED_011514.nc' finidat = '' hist_mfilt = 365 hist_nhtfrq = -24 hist_empty_htapes = .true. hist_fincl1='NPP','GPP','BTRAN','TOTVEGC','H2OSOI','TLAI','LITTER_IN','LITTER_OUT', 'STORVEGC','FIRE_AREA','SCORCH_HEIGHT','FIRE_INTENSITY','FIRE_TFC_ROS','fire_fuel_mef', 'LITTERC','fire_fuel_bulkd','fire_fuel_sav','FIRE_NESTEROV_INDEX','PFTbiomass', 'PFTleafbiomass','FIRE_ROS','WIND','TFC_ROS','DISPVEGC','AREA_TREES','AREA_PLANT' If on goldbach, use: paramfile ='/fs/cgd/csm/inputdata/lnd/clm2/edParams/CLMPARAMS_ED_011514.nc' List any changes to the defaults for the boundary datasets: N/A. Describe any substantial timing or memory changes: Code reviewed by: Stefan Muszala and Rosie Fisher. Detailed code review by Bill Sacks, Mariana Vertenstein, Ben Andre, and Erik Kluzek. Discussion of code review included Dave Lawrence, Forrest Hoffmann and Ryan Knox. List any svn externals directories updated (csm_share, mct, etc.): Changed externals to a branch_tag that supports ED compsets. We are using branch_tags because trunk scripts does not work with all of the current CLM tests: -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/add_dynlu_tests_tags/add_dynlu_tests_n03_scripts4_140305 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/ED_scripts_tags/ED_scripts_015_140305_rev List all files eliminated: N/A List all files added and what they do: ### SPITFIRE code # new SPITEFIRE directory A + models/lnd/clm/src/clm4_5/main/spitfireSF # main SPITFIRE code A + models/lnd/clm/src/clm4_5/main/spitfireSF/SFParamsMod.F90 # handle SPITFIRE parameters A + models/lnd/clm/src/clm4_5/main/spitfireSF/SFMainMod.F90 # pull out, move to a shared location, place in own module A + models/lnd/clm/src/util_share/quadraticMod.F90 ### new source and directories for ED ## ED code required for biogeophysics # ED directory in biogeophys A + models/lnd/clm/src/clm4_5/biogeophys/ED # Calculates daily carbon flux drivers from hourly calculations. A + models/lnd/clm/src/clm4_5/biogeophys/ED/EDAccumulateFluxesMod.F90 # Calculates absorbed, reflected and transmitted radiation in diffuse and direct streams for # each of the canopy layer x PFT x leaf layer three-dimensional matrix. Uses iterative Norman # radiation transfer scheme. A + models/lnd/clm/src/clm4_5/biogeophys/ED/EDSurfaceAlbedoMod.F90 # Main photosynthesis model. Calculates leaf level fluxes on a canopy layer x PFT x leaf layer # three-dimensional matrix. Sums to canopy to produce overall canopy conductance. Unpacks leaf- # level fluxes into cohort level fluxes. Uses same scientific assumptions as CLM4.5. A + models/lnd/clm/src/clm4_5/biogeophys/ED/EDPhotosynthesisMod.F90 # Generates PFT specific BTRAN vector for each ED patch. Includes option for SPA-like calculations. A + models/lnd/clm/src/clm4_5/biogeophys/ED/EDBtranMod.F90 ## ED code required for biogeochemistry # ED directory in biogeochem A + models/lnd/clm/src/clm4_5/biogeochem/ED # Determines which cohorts are in the upper and lower canopy layers. Sets leaf area index inputs to biogeophysics calculations. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDCanopyStructureMod.F90 # Initializes some ED-specific variables to zero at startup. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDSetValuesMod.F90 # Contains allometric relationships between vegetation properties (height, dbh, LAI, dead biomass, live biomass, crown area) biogeochem/ED/EDPatchDynamicsMod.F90 : Creates patches, fuses similar patches, controls disturbance and generation of area. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDGrowthFunctionsMod.F90 # Creates, fuses, terminates, sorts, counts and copies cohort structures. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDCohortDynamicsMod.F90 # Contains all calculations of derivatives of biomass, litter and seed pools. Also includes phenology model, seed and litter production and decay models, and canopy optimization model. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDPhysiologyMod.F90 # Creates, fuses, terminates, sorts, counts and copies patch structures. A + models/lnd/clm/src/clm4_5/biogeochem/ED/EDPatchDynamicsMod.F90 ## ED core functionality and types that interact with CLM (generally not science) # ED directory in main A + models/lnd/clm/src/clm4_5/main/ED # Transmits required information for CLM (tlai, htop, tile weights). Updates ED-specific history field variables. A + models/lnd/clm/src/clm4_5/main/ED/EDCLMLinkMod.F90 # Initializes ED PFT parameter structure. A + models/lnd/clm/src/clm4_5/main/ED/EDInitTimeConst.F90 # Contains ED-specific variables for CLM A + models/lnd/clm/src/clm4_5/main/ED/EDClmType.F90 # Adds history field variables specific to ED to history file. A + models/lnd/clm/src/clm4_5/main/ED/EDHistFldsMod.F90 # Prints out and reads in ED state vector to/from history files. A + models/lnd/clm/src/clm4_5/main/ED/EDRestVectorMod.F90 # Initializes ED-specific variables for CLM A + models/lnd/clm/src/clm4_5/main/ED/EDClmTypeInitMod.F90 # Allocates ED PFT specific variables. A + models/lnd/clm/src/clm4_5/main/ED/EDPftvarcon.F90 # Initializes ED site, patch and cohort structures, either to restarting or bare ground values. A + models/lnd/clm/src/clm4_5/main/ED/EDInitMod.F90 # Allocates and initializes ED parameters (that are not PFT specific). A + models/lnd/clm/src/clm4_5/main/ED/EDParamsMod.F90 # Main ED model routine. Calls all other daily ED dynamics, integrates variables, checks carbon balance. A + models/lnd/clm/src/clm4_5/main/ED/EDMainMod.F90 # Contains ED type structures (cohort, site, patch) and static values. A + models/lnd/clm/src/clm4_5/main/ED/EDTypesMod.F90 # utility routine to help in reading parameter files A + models/lnd/clm/src/clm4_5/main/paramUtilMod.F90 # transfers weights calculated internally by ED into wtcol. A + models/lnd/clm/src/clm4_5/dyn_subgrid/dynEDMod.F90 List all existing files that have been modified, and describe the changes: ### build modifications # add ED source directories for build M models/lnd/clm/bld/configure # build namelist additions for ED M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml # setup_cmdl_ed_mode addition. sets namelist vars M models/lnd/clm/bld/CLMBuildNamelist.pm ### util_share modifications # add function is_beg_curr_day() M models/lnd/clm/src/util_share/clm_time_manager.F90 # modify get_proc_bounds to include beg, end cohort M models/lnd/clm/src/util_share/accumulMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/util_share/ncdio_pio.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/util_share/ncdio_pio.F90.in ### 4_5 Modifications # change text based logical with math style (.gt. to >, .ne. to /=) # modify get_proc_bounds to include beg, end cohort M models/lnd/clm/src/clm4_5/biogeochem/CNRestMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNInitMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 # move some variables from stack to heap. # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) # change spacing, text wrapping M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/main/initInterp.F90 # change spacing, text wrapping M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 # add call to call EDInitTimeConst # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/main/initTimeConstMod.F90 # add decomposition for cohort dimension M models/lnd/clm/src/clm4_5/main/decompInitMod.F90 # add call for ed_init M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 # add code for cohort dimension M models/lnd/clm/src/clm4_5/main/subgridMod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/main/initColdMod.F90 # change get_proc_global to support cohort dimension M models/lnd/clm/src/clm4_5/main/histFileMod.F90 # add use_ed logical to support cohort dimension M models/lnd/clm/src/clm4_5/main/restFileMod.F90 # broadcast ed namelist variables M models/lnd/clm/src/clm4_5/main/controlMod.F90 # use_ed logical to call edmodel M models/lnd/clm/src/clm4_5/main/clm_driver.F90 # change spacing, text wrapping M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 # clean up unsued variables from nag compiler warnings M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 # add routine set_cohort_decomp # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/main/initSoilParVICMod.F90 # change spacing, text wrapping # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 # add support for cohort dimension M models/lnd/clm/src/clm4_5/main/decompMod.F90 # change spacing, text wrapping M models/lnd/clm/src/clm4_5/main/clmtype.F90 # add call to read ED and SPITFIRE params R + models/lnd/clm/src/clm4_5/main/readParamsMod.F90 # add routine set_cohort_decomp # add use_ed logical(s) M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/SLakeInitMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 # change get_proc_global to support cohort dimension M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 # change get_proc_global to support cohort dimension M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 # change get_proc_global to support cohort dimension M models/lnd/clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 # change get_proc_global to support cohort dimension M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 # change spacing, text wrapping # add use_ed logical(s) # calculate ed root fractionation M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 # add use_ed logical(s) for forc_solai and parsun # use_ed reporting M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 # add use_ed logical(s) for norman_radiation # change spacing M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 # remove unused variable M models/lnd/clm/src/clm4_5/biogeophys/SLakeCon.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 # change text based logical with math style (.gt. to >, .ne. to /=) M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 # add use_ed logical(s) for call dyn_ED M models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 ### 4_0 Modifications # add optional cohort argument for new ED dimension to get_proc_global and # get_proc_bounds_old for 40 backward compatibility M models/lnd/clm/src/clm4_0/main/decompMod.F90 CLM testing: --SNICARFRC - moved ERI_D.T31_g37.ICLM45.goldbach_nag.clm-SNICARFRC to goldbach and nag. This is a BFAIL. --Testing for new ED compsets. All compare hist portions were BFAIL's since this is the first time the tests are being put in place. --ED and yellowstone [ intel | pgi ] PASS ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.yellowstone_[intel | pgi].clm-edTest PASS SMS.f10_f10.ICLM45CNED.yellowstone_[intel | pgi].clm-edTest PASS SMS.f19_g16.ICLM45CNED.yellowstone_[intel | pgi].clm-edTest PASS SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.yellowstone_[intel | pgi].clm-edTest --ED and goldbach [nag | intel | pgi ] PASS ERS_D_Mmpi-serial.1x1_brazil.ICLM45CNED.goldbach_[nag | intel | pgi ].clm-edTestGb PASS SMS_D_Mmpi-serial.5x5_amazon.ICLM45CNED.goldbach_[nag | intel | pgi ].clm-edTestGb --CLM history file comparison: yellowstone [intel | pgi] - OK I ran without the -model_gen_comp option, but ran component_gen_comp and summarize_cprnc_diffs by hand for both yellowstone_intel and yellowstone_pgi. These are both OK --Regular tests (aux_clm testlist) yellowstone_intel - OK yellowstone_pgi - OK goldbach_nag - OK goldbach_intel - OK goldbach_pgi - OK CLM tag used for the baseline comparisons: clm4_5_71 Changes answers relative to baseline: No. Existing compsets do not change. If you run with an *ED* compset, then results will differ, but that is expected. =============================================================== =============================================================== Tag name: clm4_5_71 Originator(s): Bill Sacks & Jeremy Fyke Date: Fri May 2 13:00:10 MDT 2014 One-line Summary: 2-way feedbacks for glacier, veg columns compute glacier SMB, and related changes Purpose of changes: (1) Bring in two-way feedbacks for glacier when coupled to CISM, via dynamic landunits, so that CLM's glacier area remains consistent with CISM's glacier area. Also update CLM's glacier topography to be consistent with CISM. (2) Add an elevation class "0", which provides surface mass balance over the vegetated portion of the grid cell. This is used to achieve glacial inception in CISM. Along with this change, also (a) set the topographic height of non-glacier areas based on bare land topography from CISM, and (b) change the downscaling of atmospheric fields so that they are also downscaled over vegetated columns within CISM's ice mask, to achieve greater consistency between what's happening in the glacier and vegetated portions of CISM's domain. (Note that, because longwave radiation is normalized, downscaling it over the vegetated column also changes answers over glacier columns.) These changes were primarily from Jeremy Fyke. (3) Rework some consistency checks to play nicely with dynamic landunits. (4) Rework unit test build to use libraries for the clm source and csm_share source (5) Misc. other changes, as noted below. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 1969 (incorrect values for QSNWCPICE_NODYNLNDUSE) - 1929 (dynFileMod breaks with gfortran 4.8) - 1832 (logic for weights error check differs between clm4.0 and clm4.5) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: Describe any changes made to the namelist: List any changes to the defaults for the boundary datasets: Describe any substantial timing or memory changes: Code reviewed by: List any svn externals directories updated (csm_share, mct, etc.): ========= Add tests -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140305 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/add_dynlu_tests_tags/add_dynlu_tests_n03_scripts4_140305 ========= Pull in Machines_140318, needed for goldbach -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/arfs_tags/arfs_01_mach140218 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/arfs_tags/arfs_02_mach140218 ========= Changes needed for elevation class 0, etc. -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_07 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_10 -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140416 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140501 List all files eliminated: ========= Point to real decompMod rather than mock D models/lnd/clm/test/unit_testers/clm4_5/mock/main/decompMod_boundsTypeDecl.F90 List all files added and what they do: ========= Most of reweightMod.F90 moved here; also includes the following changes: (1) renames some subroutines (2) adds some diagnostic fields that are written to the history file (3) adds some utility routines such as get_landunit_weight A models/lnd/clm/src/clm4_5/main/subgridWeightsMod.F90 ========= Add code to initialize newly-active columns A models/lnd/clm/src/clm4_5/dyn_subgrid/dynInitColumnsMod.F90 ========= New unit tests A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test/test_init_columns.pf A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynInitColumns_test ========= Need new mocks and new real files now that we use the real decompMod, and also because of endrun calls A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/GetGlobalValuesMod_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/mct_mod_mock.F90 A models/lnd/clm/src/util_share/CMakeLists.txt ========= Need stub histFileMod now that many modules include calls to hist_addfld A models/lnd/clm/test/unit_testers/clm4_5/mock/main/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock/main/histFileMod_mock.F90 List all existing files that have been modified, and describe the changes: ========= Update glacier cover and topographic heights based on values from CISM; rework code to accommodate icemask and elevation class 0 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/cpl/clm_cpl_indices.F90 M models/lnd/clm/src/cpl/lnd_import_export.F90 ========= Change to be consistent with clm4_5 version, adding elevation class 0 and x2s%icemask (neither of which are used in the clm4_0 version) M models/lnd/clm/src/clm4_0/main/clm_glclnd.F90 ========= Add glc_do_dynglacier namelist option, which triggers off of CLM_UPDATE_GLC_AREAS; rename glc_dyntopo to glc_dyn_runoff_routing and make it also trigger off of CLM_UPDATE_GLC_AREAS; add glc_snow_persistence_max_days; add dynpft_consistency_checks and finidat_consistency_checks groups M models/lnd/clm/bld/user_nl_clm M models/lnd/clm/bld/CLMBuildNamelist.pm M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/unit_testers/build-namelist_test.pl M models/lnd/clm/bld/unit_testers/env_run.xml M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 - also use parameter for file name length, so that other modules can ensure consistency of char length ========= Add functions to convert between col%itype and icemec class; also add landunit_names vector M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 ========= Add functions add_landunit, add_column, add_patch (cleans up this code, and will assist with setting up unit tests) and use new functions from clm_varcon M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 ========= Use new functions from initGridCellsMod M models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 ========= Remove old consistency checks for restart file, add new ones (these changes are needed so that consistency checks work right with dynamic landunits, and we're adding some new consistency checks that weren't in place before) M models/lnd/clm/src/util_share/clm_time_manager.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90.in M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 - also read all subgrid weights and glc topo on restart (some of these used to be read in BiogeophysRestMod; we need all of them with dynamic landunits) - also remove redundant mcdate, mcsec - also add icemask restart variable M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 - also remove redundant PFT_WTGCELL, PFT_WTLUNIT, PFT_WTCOL (equivalent variables are already output by subgridRestMod) - also add snow_persistence ========= Remove old consistency checks for pftdyn file, add new ones (these changes are needed so that consistency checks work right with dynamic landunits) M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 - also call init_subgrid_weights_mod, put call to update_clm_s2x in loop over clumps, and move deallocation of topo_glc_mec to later M models/lnd/clm/src/clm4_5/main/clm_varsur.F90 M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 ========= Add new dimensions for multi-level fields - for subgrid weight diagnostics; add a dimension to accommodate fields dimensioned by glc_nec+1 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 ========= Call new routines (update_clm_x2s, set_subgrid_diagnostic_fields, initialize_new_columns) M models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 ========= Make various code operate over veg as well as icemec columns M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 - also put call to update_clm_s2x in a loop over clumps M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 - also compute snow_persistence, and rework some code for clarity ========= Add initialization of icemask & snow_persistence; change initialization of glc_topo and h2osno M models/lnd/clm/src/clm4_5/main/initColdMod.F90 ========= Add l2g_scale_type = natveg M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 ========= Fix QSNWCPICE_NODYNLNDUSE, add l2g_scale_type for QICE & related fields, add SNOW_PERSISTENCE and ICE_MASK, change _FORC fields to include elevation class 0 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 ========= Get rid of associate statement that caused problems with some compilers M models/lnd/clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 ========= Track old col%active values, needed for initializing new columns M models/lnd/clm/src/clm4_5/dyn_subgrid/dynPriorWeightsMod.F90 ========= Moved most functionality to subgridWeightsMod.F90 (now just a small wrapper to some of the stuff in subgridWeightsMod, whose main purpose is to avoid a dependency of subgridWeightsMod on filterMod) M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Changes for unit tests: (1) Update unit tests to use libraries for the clm source and csm_share source (2) New unit test (3) Make unit tests work with latest CLM trunk (4) Point to real decompMod rather than mock M models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90 M models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90.in M models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/spmdMod_mock.F90 M models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/shr_sys_mod_mock.F90 M models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/main/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/CMakeLists.txt M models/lnd/clm/src/clm4_5/main/CMakeLists.txt M models/lnd/clm/src/clm4_5/dyn_subgrid/CMakeLists.txt ========= Add icemask & snow_persistence; remove unused glc_frac, glc_rofi & glc_rofl M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 ========= Minor changes to 'use' statements for the sake of breaking dependencies for unit tests M models/lnd/clm/src/util_share/GetGlobalValuesMod.F90 M models/lnd/clm/src/clm4_5/main/decompMod.F90 ========= Changes to comments only M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/util_share/domainMod.F90 ========= Remove no-longer-failing test, change failType of a test (it was RUN rather than FAIL at least as far back as clm4_5_69) M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: ok regular tests NOTE: Most tests were done on bare_land_smb_n15_clm4_5_70, which did NOT include r59820 (add a comma in histFileMod to fix a syntax error caught by nag). After r59820, reran all goldbach_nag tests, plus one goldbach_pgi and one goldbach_intel. yellowstone_intel: ok yellowstone_pgi: ok goldbach_nag: ok goldbach_intel: ok goldbach_pgi: ok component_gen_comp on yellowstone_intel & yellowstone_pgi: ok CLM tag used for the baseline comparisons: clm4_5_69 (clm4_5_68 for a few tests with missing baselines in clm4_5_69) Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All IG compsets (i.e., GLC compsets) - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): New Climate Changes are: (1) All IG compsets (clm4.0 & clm4.5) change due to new CISM external (2) In addition, IG compsets with CLM4.5 change further due to: (a) 2-way feedbacks (CLM updated to match CISM) (b) downscaling done over vegetated landunits within the icemask (3) Also, the QSNWCPICE_NODYNLNDUSE history diagnostic field changes for ALL CLM4.5 runs, due to fixing bug 1969. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_70 Originator(s): muszala (Stefan Muszala) Date: Fri Apr 18 08:24:44 MDT 2014 One-line Summary: bring in SHR_ASSERT macros Purpose of changes: bring in SHR_ASSERT macros for Santos. Add: #include "shr_assert.h" to source files Remove: use shr_assert_mod , only : shr_assert then replace "call shr_assert" with SHR_ASSERT_ALL when asserting more than one dim - call shr_assert((ubound(carr) == (/bounds%endc/)), errMsg(__FILE__, __LINE__)) + SHR_ASSERT_ALL((ubound(parr) == (/bounds%endp/)), errMsg(__FILE__, __LINE__)) and use SHR_ASSERT when asserting one dimen - call shr_assert(bounds%level == BOUNDS_LEVEL_CLUMP, errMsg(__FILE__, __LINE__)) + SHR_ASSERT(bounds%level == BOUNDS_LEVEL_CLUMP, errMsg(__FILE__, __LINE__)) Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: Self, Santos, Sacks List any svn externals directories updated (csm_share, mct, etc.): -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_140218 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/arfs_tags/arfs_01_mach140218 -scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_140109 +scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_140403 -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_34 +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_37 -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140303 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140416 -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_131231 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/branch_tags/shr_assert_macro_tags/shr_assert_macro_n04_share3_140115 -models/utils/mct https://github.com/quantheory/MCT/tags/compiler_fixes_n03_MCT_2.8.3 +models/utils/mct https://github.com/quantheory/MCT/tags/compiler_fixes_n04_MCT_2.8.3 -models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_9/pio +models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_11/pio List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/DaylengthMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BandDiagonalMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90.in M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90.in M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90.in M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynPriorWeightsMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeophysMod.F90 M SVN_EXTERNAL_DIRECTORIES CLM testing: Tested with two sets of externals because: The old scripts do not allow any PGI tests to run while the new scripts have various issues with ERI tests and selected PGI tests plus nag debug runs (fixed in a more recent machines tag). By running with two sets of externals, I am confident that the source mods for SHR_ASSERT are working correctly. When CLM gets updates in scripts and machines, these will be updated in later tags. I) Those included in this tag II) Those that are a part of cesm1_3_alpha09b (only tested clm45). build-namelist tests: N/A regular tests: for (I) above: yellowstone_intel - OK - component_comp_gen - OK goldbach_nag - OK goldbach_intel - OK regular tests: for (II) above: 1) Yellowstone + Intel : all ERI tests are completing ref1 and ref2 but die a silent death in the base case. Erik looks like you changed Testlists, so the two VIC tests might be expected. >>./cs.status.70Intel.yellowstone | grep -v CLM50 | grep -v PASS | grep -v tputcomp | grep -v ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay | grep -v ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay | grep -v ERS_Ld211_D_P112x1.f10_f10.ICNCROP RUN ERI.f09_g16.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI.f10_f10.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI.f19_g16.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_D.T31_g37.ICLM45.yellowstone_intel.clm-SNICARFRC.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_D.f09_g16.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_D.f10_f10.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_D.f19_g16.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run RUN ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.GC.70Intel -- ref1 and ref2 run, then no output in non-ref run FAIL ERS_D.f10_f10.ICLM45VIC.yellowstone_intel.clm-vrtlay.GC.70Intel.compare_hist.clm4_5_69 69 Comparing hist file with baseline hist file, /glade/scratch/muszala/ERS_D.f10_f10.ICLM45VIC.yellowstone_intel.clm-vrtlay.GC.70Intel/run/ERS_D.f10_f10.ICLM45VIC.yellowstone_intel.clm-vrtlay.GC.70Intel.cpl.hi.0001-01-12-00000 .nc /glade/p/cesmdata/cseg/ccsm_baselines/clm4_5_69/ERS_D.f10_f10.ICLM45VIC.yellowstone_intel.clm-vrtlay/cpl.hi.nc 70 ncdump1 done 71 ncdump2 done 72 comparing split files x[a-z][a-z] 73 xaa 74 6979,6981c6979,6981 < 0.983149585541109, 0.972017300931466, 0.972017300784614, < 0.972017300929172, 0.964088275988772, 0.971373805810303, < 0.977583443108289, 0.983149585551217, 0.983149585568791, --- 75 FAIL 76 hist file comparison is FAIL FAIL SMS.f19_g16.ICLM45VIC.yellowstone_intel.clm-default.GC.70Intel.compare_hist.clm4_5_69 60 /glade/u/spooldir/1397689222.575650.shell: Storing new baseline in /glade/p/cesmdata/cseg/ccsm_baselines/clm4_5_70/SMS.f19_g16.ICLM45VIC.yellowstone_intel.clm-default 61 Comparing hist file with baseline hist file, /glade/scratch/muszala/SMS.f19_g16.ICLM45VIC.yellowstone_intel.clm-default.GC.70Intel/run/SMS.f19_g16.ICLM45VIC.yellowstone_intel.clm- default.GC.70Intel.cpl.hi.0001-01-06-00000.n c /glade/p/cesmdata/cseg/ccsm_baselines/clm4_5_69/SMS.f19_g16.ICLM45VIC.yellowstone_intel.clm-default/cpl.hi.nc 62 ncdump1 done 63 ncdump2 done 64 comparing split files x[a-z][a-z] 65 xad 66 18300c18300 < 0.983410370293909, 0.984052369383093, 0.979227772964994, --- > 0.983410370293909, 0.984052369383093, 0.979228345951215, 18341,18350c18341,18350 67 FAIL 68 hist file comparison is FAIL 2) Goldbach + NAG seems to have passed OK, _D runs have failed as expected. The reporting is messed up...ie., TestStatus.out look OK, but TestStatus does not. >> ./cs.status.70nag.goldbach | grep -v CLM50 | grep -v PASS | grep -v tputcomp | grep -v _D FAIL ERI.f10_f10.ICLM45BGC.goldbach_nag.clm-reduceOutput.GC.70nag --look at /scratch/cluster/muszala/tests/ERI.f10_f10.ICLM45BGC.goldbach_nag.clm-reduceOutput.GC.70nag/TestStatus.out and there is no FAIL FAIL ERI.f19_g16.ICLM45BGC.goldbach_nag.clm-reduceOutput.GC.70nag --/scratch/cluster/muszala/tests/ERI.f19_g16.ICLM45BGC.goldbach_nag.clm-reduceOutput.GC.70nag and there is no FAIL 3) Goldbach + Intel - these look all like passes to me: see /scratch/cluster/muszala/tests/*/TestStatus.out >> ./cs.status.70intel.goldbach | grep -v CLM50 | grep -v PASS | grep -v tputcomp FAIL ERI.f10_f10.ICLM45BGC.goldbach_intel.clm-reduceOutput.GC.70intel -- Shows FAIL but PASSes in TestStatus.out FAIL ERI.f19_g16.ICLM45BGC.goldbach_intel.clm-reduceOutput.GC.70intel -- Shows FAIL but PASSes in TestStatus.out FAIL ERI_D.T31_g37.I1850CLM45.goldbach_intel.clm-reduceOutput.GC.70intel -- Shows FAIL but PASSes in TestStatus.out FAIL ERI_D.f10_f10.ICLM45BGC.goldbach_intel.clm-reduceOutput.GC.70intel -- Shows FAIL but PASSes in TestStatus.out FAIL ERI_D.f19_g16.ICLM45BGC.goldbach_intel.clm-reduceOutput.GC.70intel -- Shows FAIL but PASSes in TestStatus.out 4) Goldbach + PGI - some are indicating FAIL with TestStatus.out shows PASSes, others are straight out FAILs ./cs.status.70pgi.goldbach | grep -v CLM50 | grep -v PASS | grep -v tputcomp | grep -v SMS_Ly1.f19_g16.ICLM45BGCCROP.frankfurt_pgi FAIL ERI.f10_f10.ICLM45BGC.goldbach_pgi.clm-reduceOutput.GC.70pgi -- Shows FAIL but PASSes in TestStatus.out FAIL ERI.f19_g16.ICLM45BGC.goldbach_pgi.clm-reduceOutput.GC.70pgi -- ref1 fail: 22 g005.cgd.ucar.edu - daemon did not report back when launched 23 g006.cgd.ucar.edu - daemon did not report back when launched 24 g009.cgd.ucar.edu - daemon did not report back when launched 25 /usr/mpi/pgi/openmpi-1.4.3-qlc/bin/orted: error while loading shared libraries: libpgc.so: cannot open shared object file: No such file or directory FAIL ERI_D.f10_f10.ICLM45BGC.goldbach_pgi.clm-reduceOutput.GC.70pgi -- Shows FAIL but PASSes in TestStatus.out FAIL ERI_D.f19_g16.ICLM45.goldbach_pgi.clm-reduceOutput.GC.70pgi -- ref1 fail: 1 /usr/mpi/pgi/openmpi-1.4.3-qlc/bin/orted: error while loading shared libraries: libpgc.so: cannot open shared object file: No such file or directory 2 /usr/mpi/pgi/openmpi-1.4.3-qlc/bin/orted: error while loading shared libraries: libpgc.so: cannot open shared object file: No such file or directory 3 /usr/mpi/pgi/openmpi-1.4.3-qlc/bin/orted: error while loading shared libraries: libpgc.so: cannot open shared object file: No such file or directory 4 -------------------------------------------------------------------------- 5 A daemon (pid 29755) died unexpectedly with status 127 while attempting FAIL ERI_D.f19_g16.ICLM45BGC.goldbach_pgi.clm-reduceOutput.GC.70pgi -- ref1 fail: 23 g021.cgd.ucar.edu - daemon did not report back when launched 24 g022.cgd.ucar.edu - daemon did not report back when launched 25 g023.cgd.ucar.edu - daemon did not report back when launched 26 /usr/mpi/pgi/openmpi-1.4.3-qlc/bin/orted: error while loading shared libraries: libpgc.so: cannot open shared object file: No such file or directory FAIL SMS_Ld5.f19_g16.IRCP45CLM45BGC.goldbach_pgi.clm-decStart.GC.70pgi -- ERROR in /var/spool/torque/mom_priv/jobs/19500.goldbach.cgd.ucar.edu.SC: file /fs/cgd/csm/ccsm_baselines/clm4_5_69/SMS_Ld5.f19_g16.IRCP45CLM45BGC.goldbach_pgi.clm-decStart/cpl.hi.nc does not exist CLM tag used for the baseline comparisons: clm4_5_69 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_69 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Tue Mar 18 21:12:34 MDT 2014 One-line Summary: start unit testing build-namelist Purpose of changes: start doing unit testing on construction of the clm namelist. This involved moving the contents of build-namelist into CLMBuildNamelist.pm and bringing in perl infrastructure to supplement Test::More. Initial test suites are implented for several name list variables. Requirements for tag: N/A Test level of tag: regular, build_namelist Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, clm-cmt List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: models/lnd/clm/bld: CLMBuildNamelist.pm - contents of build-namelist test_build_namelist/perl5lib/* - CPAN modules needed for unit testing test_build_namelist/t/input/* - mock input files for build-namelist tests test_build_namelist/t/template_test_XXX.pm - template for new tests test_build_namelist/t/test_*.pm - unit tests test_build_namelist/test_build_namelist.pl - unit test driver test_build_namelist/README List all existing files that have been modified, and describe the changes: models/lnd/clm/bld: build-namelist - moved contents into CLMBuildNamelist.pm, now just a driver calling main function. CLM testing: build-namelist tests: yellowstone - OK new and existing generate tests goldbach - OK new and existing generate tests regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK goldbach_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: N/A short tests (aux_clm_short): yellowstone_intel - OK yellowstone_pgi - OK tools testing: N/A CLM tag used for the baseline comparisons: clm4_5_68 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_68 Originator(s): erik (Erik) Date: Fri Mar 7 16:43:23 MST 2014 One-line Summary: Update scripts to version that turns on transient CO2 streams for transient compsets, and update CISM (changes answers) Purpose of changes: Bring in the scripts version that by default had transient CO2 for any transient compsets. You can still turn it off by setting DATM_CO2_TSERIES=FALSE in env_run.xml. Also bring in the latest CISM version that has answer changes for any IG compsets. It fixes fields sent from CISM to the coupler (fixes an exact restart problem). Requirements for tag: update scripts and CISM, transient and IG compsets have different answers Test level of tag: regular Bugs fixed (include bugzilla ID): 979 (adding CO2 streams) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts to scripts4_140305 cism to cism1_140303 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: None CLM testing: build-namelist tests: yellowstone yes regular tests (aux_clm): yellowstone_intel yes yellowstone_pgi yes goldbach_nag yes goldbach_pgi yes goldbach_intel yes history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel yes yellowstone_pgi yes tools testing: None CLM tag used for the baseline comparisons: clm4_5_67 Changes answers relative to baseline: Yes! Summarize any changes to answers: - what code configurations: transient and IG compsets - what platforms/compilers: all - nature of change: larger than roundoff =============================================================== =============================================================== Tag name: clm4_5_67 Originator(s): mvertens Date: Thu Mar 6 16:53:23 MST 2014 One-line Summary: removed initSurfAlb as part of the initialization Purpose of changes: removed the call to initSurfAlb as well as part of the initialization and also removed the routine from the clm4.5 code base Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): None Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None M bld/namelist_files/namelist_defaults_clm4_5.xml - clmi.ICRUCLM45BGCCROPmp24.0241-01-01.10x15_USGS_simyr2000_c140111.nc had not in fact been created - this effected the PEM test in the goldbach clm45 test suite - the default namelist has not been backed up to the original clmi.ICRUCLM45BGCCROPmp24.0241-01-01.10x15_USGS_simyr2000_c131028.nc for now Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: mvertens List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M src/clm4_5/biogeochem/ch4InitMod.F90 - removal of code block that is no longer needed due to removal of initSurfAlb (this had already been commented out in clm4_5_66) M src/clm4_5/main/initInterp.F90 - minor bug fix the turn off spval by default for nonactive points D src/clm4_5/main/initSurfAlbMod.F90 M src/clm4_5/main/clm_initializeMod.F90 - removal of call initSurfAlb (main purpose of this tag) - removal of code to upgrade old initial data files to have new metadata a new scheme should be put in place with a namelist option to take clm4.5 restart datasets that have been created prior to the introduction of initInterp and introduce the new metadata at run time M src/clm4_5/main/initColdMod.F90 - had to introduce setting values for the following variables in order to remove call to initSurfAlb cps%albgrd_pur(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgri_pur(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgrd_bc(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgri_bc(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgrd_oc(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgri_oc(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgrd_dst(bounds%begc:bounds%endc,:) = 0.2_r8 cps%albgri_dst(bounds%begc:bounds%endc,:) = 0.2_r8 M src/clm4_5/main/clm_driver.F90 - just comments CLM testing: regular tests (aux_clm): OK means only failures were expected yellowstone_intel : OK yellowstone_pgi : OK goldbach_nag : OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi : OK CLM tag used for the baseline comparisons: clm4_5_66 Changes answers relative to baseline: some - for all compsets where finidat is set to blank, then answers will change relative to baseline for all compsets wehre finidat is pointing to a dataset, answers will be bfb compared to baseline =============================================================== =============================================================== Tag name: clm4_5_66 Originator(s): mvertens Date: Mon Mar 3 10:50:24 MST 2014 One-line Summary: refactoring of initialization and introduction of run-time finidat interpolation Purpose of changes: refactoring of initialization and introduction of run-time finidat interpolation Completely rewrote clm_initialize to leverage new initialization scheme In the new scheme, cold start initialization is ALWAYS called and values are overwritten by either an appropriate finidat file OR by calling finidat_interp to interplate finidat to the output resolution/mask. Requirements for tag: Test level of tag: regular, short, tools, build_namelist, doc Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1930 (MEGAN does not work correctly with prognostic crops on) 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: NONE Describe any changes made to the namelist: - ALL clm4.5 finidat files have been updated to contain new metadata that will enable the online interpinic to operate on them. The time stamp on all new files has the date c140111. The files have been created so that they are bit-for-bit compatible with the code base. - The following new namelist variables have been added to the namelist_definition_clm4_5.xml file - finidat_interp_source if non-blank, then interpinic will be called to interpolate finidat_interp_source and create output file specified by finidat_interp_dest. - finidat_interp_dest if finidat_interp_source is set to non-blank, then interpinic will be called to interpolate finidat_interp_source and create output file finidat_interp_dest List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: mvertens, sacks List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: The following file pairs were renamed and subsequently extensively modfified Summaries of the modifications are below: ------- New module initColdMod.F90 contains calls to initialize the cold start for ------- the entire model. The cold start values are then overwritten with either ------- an finidat file or an interpolation file using finidat_interp_source. D models/lnd/clm/src/clm4_5/main/mkarbinitMod.F90 A models/lnd/clm/src/clm4_5/main/initColdMod.F90 ------- Renamed file D models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 A models/lnd/clm/src/clm4_5/biogeochem/SatellitePhenologyMod.F90 ------- Renamed file D models/lnd/clm/src/clm4_5/biogeochem/CNDVEcosystemDynIniMod.F90 A models/lnd/clm/src/clm4_5/biogeochem/CNDVInitMod.F90 ------- Renamed file, removed initch4, merged routines initTimeConst_ch4 and makearbinit_ch4 --------into new routine initColdCH4.Also removed almost all associate statements ------- (but kept the intput/output documentation) and used the explicit clmtype definition. D models/lnd/clm/src/clm4_5/biogeochem/initch4Mod.F90 A models/lnd/clm/src/clm4_5/biogeochem/ch4InitMod.F90 ------- Renamed and combined files ------- Migrated all CN cold start initialization for both soil and ------- special landuntis into new routine initColdCN in new module biogeochem/CNInitMod. D models/lnd/clm/src/clm4_5/main/CNiniSpecial.F90 D models/lnd/clm/src/clm4_5/main/CNiniTimeVar.F90 A models/lnd/clm/src/clm4_5/biogeochem/CNInitMod.F90 ------- Renamed file and merged routines makearbinit and snow_depth2Lake ------- into one new routine initColdSlake. Also removed almost all associate ------- statements (but kept the intput/output documentation) and used the explict ------- clmtype definition. A models/lnd/clm/src/clm4_5/biogeophys/SLakeInitMod.F90 D models/lnd/clm/src/clm4_5/biogeophys/initSlakeMod.F90 ------- Renamed iniTimeConst, removed associate statements but kept ------- the documentation of input/output and also explictly listed ------- full clmtype variables D models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 A models/lnd/clm/src/clm4_5/main/initTimeConstMod.F90 List all files added and what they do: ------- New run-time interpolation of input finidat to target resolution/mask ------- using the new namelist variables finidat_interp_source and finidat_interp_dest A models/lnd/clm/src/clm4_5/main/initInterp.F90 ------- Obtain/write global index space value for target point at given clmlevel A models/lnd/clm/src/util_share/GetGlobalValuesMod.F90 List all existing files that have been modified, and describe the changes: ------- In all files, unless otherwise noted added call to errMsg(__FILE__,__LINE__) ------- and in some cases optional arguments of decomp_index and clmlevel also added ------- See documentation for namelist changes above M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ------ Overloaded endrun subroutine to also print out global index infromation by ------ calling new GetGlobalWrite routine if optional arguments decomp_index and ------ clm_level are passed in M models/lnd/clm/src/util_share/abortutils.F90 ------- Replaced missing value setting of huge(1) with ispval M models/lnd/clm/src/util_share/accumulMod.F90 ------- Replaced endrun with call to shr_sys_abort M models/lnd/clm/src/util_share/domainMod.F90 ------- Added in missing values and special values for variable metadata - this ------- is needed needed by initInterp M models/lnd/clm/src/util_share/restUtilMod.F90 M models/lnd/clm/src/util_share/restUtilMod.F90.in ------- Completely rewrote clm_initialize to leverage new initialization scheme ------- In the new scheme, cold start initialization is ALWAYS called and values ------- are overwritten by either an appropriate finidat file OR by calling ------- finidat_interp to interplate finidat to the output resolution/mask. M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ------- Added new metadata and variables - include global indices for parent subgrid ------- level(s) (i.e. column, landunit and gridcell for pfts) M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 ------- Moved view_factor routine and associated variable from a separate routine ------- in UrbanMod to part of the initTimeConstUrban subroutine in UrbanInitMod M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 ------- Removed vf_xx variables from restart file and also ------- removed do_initsurfalb variable M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 ------- In all files, unless otherwise noted added call to errMsg(__FILE__,__LINE__) ------- and in some cases optional arguments of decomp_index and clmlevel also added M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90.in M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/MEGANFactorsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4RestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4varcon.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSharedParamsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 M models/lnd/clm/src/clm4_5/main/decompInitMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/ndepStreamMod.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/surfrdUtilsMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/initSoilParVICMod.F90 M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/main/decompMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CLMVICMapMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeochemMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynPriorWeightsMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90.in CLM testing: regular tests (aux_clm): NOTE1: that all namelists compares where finidat was not blank will fail - since new finidat files are used that have new metadata - BUT - the results are still bfb yellowstone_intel - OK expected failures: ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay yellowstone_pgi - OK expected failures: ERH_D.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-default ERS.f09_g16.ICLM45VIC.yellowstone_pgi.clm-vrtlay goldbach_nag - OK goldbach_intel - OK goldbach_pgi - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: NOTE1: all .h1 tests fail since there are new meta data fields for cols1d_active, pfts1d_active - and FILLDIFF is different yellowstone_intel OK yellowstone_pgi OK CLM tag used for the baseline comparisons: clm4_5_65 Changes answers relative to baseline: No - bit-for-bit =============================================================== =============================================================== Tag name: clm4_5_65 Originator(s): mvertens (Mariana Vertenstein,UCAR/CSEG,303-497-1349) Date: Tue Feb 25 13:45:38 MST 2014 One-line Summary: Turn off MEGAN vocs when crops is running Purpose of changes: MEGAN does not currently work with prognostic crops. It needs a table of pft-specific values, and that this table has only been created for the 16 "standard" (non-crop) pfts. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1930 (MEGAN does not work correctly with prognostic crops on) 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.):s scripts4_140214a -> scripts4_140220 Machines_140214 -> Machines_140218 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: ======== Add a new logical variable - use_voc - that is true by default but is set to false if prognostic cop is activated ======== M src/clm4_5/main/clm_initializeMod.F90 M src/clm4_5/main/clm_atmlnd.F90 M src/clm4_5/main/controlMod.F90 M src/clm4_5/main/clm_varctl.F90 M src/clm4_5/main/clm_driver.F90 M src/clm4_5/main/histFldsMod.F90 M src/clm4_0/main/clm_varctl.F90 M src/cpl/clm_cpl_indices.F90 CLM testing: build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK expected failures ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay expected non-bfb failures due to VIC/CROP changes ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel SMS_Ly1.f19_g16.ICLM45BGCCROP.yellowstone_intel yellowstone_pgi - OK expected failures ERH_D.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-default ERS.f09_g16.ICLM45VIC.yellowstone_pgi.clm-vrtlay goldbach_nag - OK expected non-bfb failures due to VIC/CROP changes ERS_D.f10_f10.ICLM45BGCCROP.goldbach_nag.clm-allActive ERS_Lm3.1x1_numaIA.ICLM45BGCCROP.goldbach_nag goldbach_intel - OK expected non-bfb failures due to VIC/CROP changes ERS_Ly20.1x1_numaIA.ICLM45BGCDVCROP.goldbach_intel.clm-crop PEM.f10_f10.ICLM45BGCCROP.goldbach_intel.clm-crop goldbach_pgi - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK - yellowstone_pgi - OK CLM tag used for the baseline comparisons: clm4_5_64 Changes answers relative to baseline: No - except for VOC fields when prognostic crop is on (this is a diagnostic only and does not impact the answers) =============================================================== =============================================================== Tag name: clm4_5_64 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Wed Feb 19 09:19:40 MST 2014 One-line Summary: fix and clean ncdio_pio.F90.in. clean clm_time_manager. update externals. Purpose of changes: Note 1: This is the last tag that is tested on frankfurt; new tests are on goldbach. Note 2: Pts. mode is being deprecated for science use as of this tag. Use PTCLM. Pts. mode remains in place in our test system. Note 3: There is an unresolved problem with higher resolutions when dov2xy is .false. and we are using pnetcdf. Please see bug 1730. ncdio_pio.F90.in - fix initialization problem where count and start are sometimes used without being set. clm_time_manager - clean out unused variables update externals to support ED compsets, move pts. mode tests to testmods. Update Machines and pio to address bug 1730. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.):s scripts4_140209 -> scripts4_140214a Machines_140213 -> Machines_140214 pio1_8_8 -> pio1_8_9 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/util_share/clm_time_manager.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90.in CLM testing: build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK Expected due to change in pts. mode: BFAIL SMS.f45_f45.I.yellowstone_intel.clm-ptsRLA.GC.64Intel.compare_hist.clm4_5_63 BFAIL SMS.f45_f45.I.yellowstone_intel.clm-ptsROA.GC.64Intel.compare_hist.clm4_5_63 BFAIL SMS_D_Mmpi-serial.f45_f45.ICLM45.yellowstone_intel.clm-ptsRLA.GC.64Intel.compare_hist.clm4_5_63 BFAIL SMS_Mmpi-serial.f45_f45.ICLM45.yellowstone_intel.clm-ptsRLA.GC.64Intel.compare_hist.clm4_5_63 Expected due to change in pio_buffer_size_limit FAIL ERI.f09_g16.ICLM45BGC.yellowstone_intel.GC.64Intel.nlcomp FAIL ERI_D.f09_g16.ICLM45BGC.yellowstone_intel.GC.64Intel.nlcomp New Failure for VIC but due to dov2xy problem ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay yellowstone_pgi - OK Expected due to change in pts. mode: BFAIL SMS.f45_f45.I.yellowstone_pgi.clm-ptsRLB.GC.64Pgi.compare_hist.clm4_5_63 BFAIL SMS_D_Mmpi-serial.f45_f45.ICLM45.yellowstone_pgi.clm-ptsRLA.GC.64Pgi.compare_hist.clm4_5_63 BFAIL SMS_Mmpi-serial.f45_f45.ICLM45.yellowstone_pgi.clm-ptsRLA.GC.64Pgi.compare_hist.clm4_5_63 Expected due to change in pio_buffer_size_limit FAIL ERI.f09_g16.I1850CRUCLM45BGC.yellowstone_pgi.GC.64Pgi.nlcomp FAIL ERI.f09_g16.ICLM45BGC.yellowstone_pgi.GC.64Pgi.nlcomp FAIL ERI_D.f09_g16.ICLM45BGC.yellowstone_pgi.GC.64Pgi.nlcomp FAIL ERS_D.hcru_hcru.ICRUCLM45BGC.yellowstone_pgi.GC.64Pgi.nlcomp FAIL ERS_D.hcru_hcru.ICRUCN.yellowstone_pgi.GC.64Pgi.nlcomp New Failure for VIC but due to dov2xy problem ERS.f09_g16.ICLM45VIC.yellowstone_pgi.clm-vrtlay frankfurt_nag - OK Expected due to change in pts. mode: BFAIL SMS_D_Mmpi-serial.f45_f45.ICLM45.frankfurt_nag.clm-ptsRLA.GC.64Nag.compare_hist.clm4_5_63 BFAIL SMS_Mmpi-serial.f45_f45.ICLM45.frankfurt_nag.clm-ptsRLA.GC.64Nag.compare_hist.clm4_5_63 BFAIL SMS_Mmpich.f45_f45.ICLM45.frankfurt_nag.clm-ptsRLA.GC.64Nag.compare_hist.clm4_5_63 frankfurt_intel - OK frankfurt_pgi - OK Expected due to change in pts. mode: BFAIL SMS.f45_f45.ICLM45.frankfurt_pgi.clm-ptsRLB.GC.64Pgi.compare_hist.clm4_5_63 BFAIL SMS.f45_f45.ICLM45.frankfurt_pgi.clm-ptsROA.GC.64Pgi.compare_hist.clm4_5_63 history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_63 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_63 Originator(s): sacks (sacks) Date: Fri Feb 14 07:22:37 MST 2014 One-line Summary: add some code needed for dynamic landunits; activate 0-weight veg landunit sometimes Purpose of changes: (1) Add grc%landunit_indices(:,:), so you can find a given l index if you have the g index (this will be needed in a few places for dynamic landunits) (2) Add code to update landunit weights; currently has no effect because landunit areas don't change yet (3) Refactor logic in the is_active_X routines, and add logic to activate a virtual vegetated landunit under some conditions (needed for coupling with CISM, and helpful for dynamic landunits). Specifically, we activate a virtual (0-weight) vegetated landunit for any grid cell that is NOT 100% istice (i.e., standard glacier) (we exclude grid cells that are 100% istice to avoid the performance penalty, because these aren't used for coupling with CISM, and the only way this glacier can retreat is if another landunit, like crop, increases there, which will rarely happen). Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: Performance about 6% worse for clm4.5 IG runs at f09 (i.e., with glcmec), because of the new virtual vegetated columns. Not investigated for f19 or T31, but probably a similar performance hit. There were also a few memcomp failures Code reviewed by: quick review by mvertens List any svn externals directories updated (csm_share, mct, etc.): Machines: Machines_140207a -> Machines_140213 (to fix pgi on yellowstone) List all files eliminated: ========= Remove "IN_PROGRESS" D models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90.IN_PROGRESS List all files added and what they do: ========= Add code to update landunit weights; currently has no effect because landunit areas don't change yet A models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90 ========= Add unit tests for dynLandunitAreaMod A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/test_update_landunit_weights_one_gcell.pf A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynLandunitArea_test/test_update_landunit_weights.pf List all existing files that have been modified, and describe the changes: ========= Add grc%landunit_indices(:,:), so you can find a given l index if you have the g index (this will be needed in a few places for dynamic landunits) M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 - just add a comment M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 ========= Call new code in dynLandunitAreaMod M models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 ========= Refactor logic in the is_active_X routines, and add logic to activate a virtual vegetated landunit under some conditions (needed for coupling with CISM, and helpful for dynamic landunits) M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Add unit tests for dynLandunitAreaMod M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/CMakeLists.txt M models/lnd/clm/src/clm4_5/dyn_subgrid/CMakeLists.txt M models/lnd/clm/src/clm4_5/main/CMakeLists.txt CLM testing: regular tests (aux_clm): NOTE: frankfurt intel & pgi ran on a slightly older version of the branch (dynlu_weight_updates_glacier_n05_clm4_5_62, which did not include some final minor refactoring to reweightMod); frankfurt nag & yellowstone intel/pgi ran on the final version yellowstone_intel: ok yellowstone_pgi: ok frankfurt_intel: ok frankfurt_pgi: ok frankfurt_nag: ok history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: ok yellowstone_pgi: ok A bunch of expected failures in h1 (1-d) hist files, due to the newly-active points. Also: Did a manual test to show that answers are the same for virtual vegetated columns as they would be if the column had non-zero weight (to ensure there are no dependencies on whether a column's weight is non-zero). CLM tag used for the baseline comparisons: clm4_5_62 for yellowstone (but clm4_5_61 for component_gen_comp); clm4_5_61 for frankfurt Changes answers relative to baseline: NO However, note diffs in 1-d hist files due to newly-active points. =============================================================== =============================================================== Tag name: clm4_5_62 Originator(s): erik (Erik) Date: Mon Feb 10 04:16:07 MST 2014 One-line Summary: Get PTCLM working robustly, US-UMB test working, add CO2 streams to datm, add more consistency testing between compsets and user settings Purpose of changes: US-UMB fix in scripts and datm update. Fix so build-namelist will abort if there is an inconsistency with CLM_BLDNML_OPTS and user_nl_clm. Add CO2 streams as a built-in option to datm. Turn CO2 streams on with the DATM_CO2_TSERIES env_run.xml variable. Can be set to: none,20tr,rcp2.6,rcp4.5,rcp6.0,rcp8.5, by default is none. Requirements for tag: datm -- CO2 update, streams improvements Fix build-namelist consistency issues Fix bug 1847 -- end1d in hist for clm4_0 Add envxml_dir check that cndv and fpftdyn aren't on the same time Add PTCLM tests to test_Driver Add PTCLM test system in Make PTCLM more robust Test level of tag: regular, tools, build_namelist Bugs fixed (include bugzilla ID): 1918 -- sort options in build-namelist 1917 -- remove WRF resolutions 1903 -- buildtools fails for PTCLM 1900 -- Remove BUILDHEAT and Qanth from output for CLM testing 1896 -- CLM build-namelist should abort if use_cndv AND fpftdyn are set. 1881 -- Add envxml_casedir option to CLM build-namelist 1879 -- need error triggered when use_crop and CLM_BLDNL_OPTS are not consistent 1847 -- 'histfilemod_mp_hist_restart_ncd_$END1D' is being used without being defined Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: frankfurt switched out for goldbach Describe any changes made to the namelist: CLM build-namelist changed to ensure user changes don't conflict with command-line options List any changes to the defaults for the boundary datasets: remove WRF datasets Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts to scripts4_140209 Machines to Machines_140207a datm to datm8_140114 pio to pio1_8_8 cprnc to cprnc_140203 PTCLM to PTCLM2_140204 List all files eliminated: move frankfurt to goldbach D models/lnd/clm/test/tools/tests_posttag_frankfurt_nompi List all files added and what they do: goldbach, and add PTCLM tools testing A + models/lnd/clm/test/tools/tests_posttag_goldbach_nompi A models/lnd/clm/test/tools/TCBscripttools.sh ------------ Add script to run buildtools for PTCLM A models/lnd/clm/test/tools/config_files/PTCLM__s -------- Config for PTCLM A models/lnd/clm/test/tools/nl_files/PTCLM_USUMB_clm4_0 A models/lnd/clm/test/tools/nl_files/PTCLM_USUMB_clm4_5 A models/lnd/clm/test/tools/nl_files/PTCLM_USUMB_Cycle_clm4_5 A models/lnd/clm/test/tools/nl_files/PTCLM_USUMB_Global_clm4_5 A models/lnd/clm/bld/unit_testers/myuser_nl_clm --- New build-namelist tests A models/lnd/clm/bld/unit_testers/env_run.xm List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/build-namelist ----- Sort options, add -envxml_dir option, check that user hasn't contradicted themself with CLM_BLDNML_OPTS and user_nl_clm, remove options: -noio, -nofire, -snicar_frc, -vsoilc, -exlaklayers, -clm4me use Cwd::abs_path and remove home-grown absolute_path, add some more docmentation and comments, redo some ordering and names M models/lnd/clm/bld/clm.buildnml.csh --- add -envxml_dir so will use env_*.xml files to expand env variables M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml --- Add irrig setting, remove WRF files M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Add irrig, bgc_spinup, and bgc_mode M models/lnd/clm/bld/namelist_files/namelist_defaults_overall.xml -- Remove WRF resolutions: us20, wus12 M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_0.xml - Remove WRF resolutions M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Remove WRF resolutions, add bgc_mode M models/lnd/clm/bld/unit_testers/build-namelist_test.pl ----------- Add a bunch of new tests M models/lnd/clm/bld/user_nl_clm ----- make note of variables that should be done by command-line build-namelist options M models/lnd/clm/src/clm4_0/main/histFileMod.F90 --- Fix bug 1847 ------------ Add PTCLM testing M models/lnd/clm/test/tools/README.testnames M models/lnd/clm/test/tools/test_driver.sh ------ Remove bluefire, lynx, mirage, jaguarpf -- switch frankfurt for goldbach M models/lnd/clm/test/tools/TBLscript_tools.sh M models/lnd/clm/test/tools/TSMscript_tools.sh M models/lnd/clm/test/tools/input_tests_master M models/lnd/clm/test/tools/tests_posttag_nompi_regression M models/lnd/clm/test/tools/tests_posttag_yong M models/lnd/clm/test/tools/tests_pretag_yellowstone_nompi ------------ Don't die if debug and files were not created. M models/lnd/clm/tools/clm4_0/mksurfdata_map/mksurfdata.pl M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl CLM testing: regular, build-namelist, tools build-namelist tests: yellowstone regular tests (aux_clm): yellowstone_intel yellowstone_pgi goldbach_nag edison_intel history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel yellowstone_pgi tools testing: yellowstone interactive goldbach interactive CLM tag used for the baseline comparisons: clm4_5_61 Changes answers relative to baseline: None, bit-for-bit =============================================================== =============================================================== Tag name: clm4_5_61 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Tue Feb 4 09:45:43 MST 2014 One-line Summary: add 3-d snow history fields; continue harvest past end of pftdyn timeseries Purpose of changes: There are two separate sets of changes in this tag; both apply just to CLM4.5: (1) Addition of 3-d snow history fields: These history fields (inactive by default) provide diagnostics for each layer of the snow pack. This involved adding some additional history file infrastructure to handle the variable number of snow pack layers. See the new section in the user's guide (custom.xml) for a description of how these new history fields work, and how to interpret them. (2) Change the harvest logic for transient runs that extend past the end of the pftdyn dataset: Until now, harvest was set to 0 when you passed the end of the pftdyn dataset. With this tag, this behavior is changed, so that for all years past the end of the pftdyn dataset, harvest rates remain fixed at the last year's value. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: Memory use increases slightly, as observed by memcomp failures for a few tests. This is presumably due to new fields in clmtype. Code reviewed by: Erik reviewed changes for the 3-d snow history fields. List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Add 3-d snow history fields M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/doc/UsersGuide/custom.xml ========= Add a new snow diagnostic, sub_surf_abs_SW M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 ========= Continue harvest past end of pftdyn time series M models/lnd/clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/test_dynTimeInfo.pf CLM testing: regular tests (aux_clm): yellowstone_intel: ok yellowstone_pgi: ok frankfurt_intel: ok frankfurt_pgi: ok frankfurt_nag: ok history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: ok yellowstone_pgi: ok Only failures are these expected failures (see notes on answer changes below): FAIL SMS_Ly3.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetLate.compare_hist.clm4_5_60.clm2.h0 FAIL SMS_Ly5.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetMid.compare_hist.clm4_5_60.clm2.h0 FAIL SMS_Ly8.1x1_tropicAtl.I20TRCLM45BGC.yellowstone_intel.clm-tropicAtl_subsetEarly.compare_hist.clm4_5_60.clm2.h0 CLM tag used for the baseline comparisons: clm4_5_60 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 transient runs that continue past the end of the pftdyn dataset - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate: Harvest rates are now continued past the end of the pftdyn dataset (staying fixed at their value from the last year), for the remainder of the simulation. This leads to potentially large answer changes for transient runs that continue past the end of the pftdyn dataset. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_60 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Thu Jan 30 18:27:03 MST 2014 One-line Summary: refactor build-namelist Purpose of changes: : break build-namelist into small unit-testable functions instead of a single massive script. Use output functions to standardize screen output for errors, warnings and messages so that results can be automatically searched by scripts. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, Erik List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/build-namelist - major refactor described above M models/lnd/clm/bld/namelist_files/namelist_defaults_overall.xml - fix incorrect comments M models/lnd/clm/bld/unit_testers/build-namelist_test.pl - minor updates to check new output style CLM testing: build-namelist tests: yellowstone - pass compare and generate with only xfails frankfurt - not tested, CLM-CMT believes tests may be broken. regular tests (aux_clm): yellowstone_intel - ok yellowstone_pgi - ok frankfurt_intel - ok frankfurt_pgi - ok frankfurt_nag - ok history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - ok yellowstone_pgi - ok short tests (aux_clm_short): yellowstone_intel - ok tools testing: N/A CLM tag used for the baseline comparisons: clm4_5_59 Changes answers relative to baseline: none, bit for bit =============================================================== =============================================================== Tag name: clm4_5_59 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Wed Jan 22 15:04:12 MST 2014 One-line Summary: use new get_curr_yearfrac function in clm_time_manager Purpose of changes: Use the new get_curr_yearfrac function in clm_time_manager in place of dyn_time_weights. The reason is that, as Erik pointed out, dyn_time_weights was out of place in dynUtilsMod, and really this functionality belongs in the clm_time_manager module. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: ========= Made obsolete by get_curr_yearfrac in clm_time_manager D models/lnd/clm/src/clm4_5/dyn_subgrid/dynUtilsMod.F90 D models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/dynUtilsMod_mock.F90 List all files added and what they do: ========= Mock out get_curr_yearfrac: return a fixed fraction A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/clm_time_manager_mock.F90 List all existing files that have been modified, and describe the changes: ========= Fix get_curr_yearfrac to be real rather than integer M models/lnd/clm/src/util_share/clm_time_manager.F90 ========= Use get_curr_yearfrac instead of dyn_time_weights M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynCNDVMod.F90 M models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90.in ========= Update unit tests to pull in clm_time_manager (mock) rather than dynUtilsMod (mock) M models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/CMakeLists.txt M models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/CMakeLists.txt CLM testing: regular tests (aux_clm): yellowstone_intel: ok yellowstone_pgi: ok frankfurt_intel: ok frankfurt_pgi: ok frankfurt_nag: ok history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: ok yellowstone_pgi: ok CLM tag used for the baseline comparisons: clm4_5_58 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45, either transient or with DV - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): roundoff If bitwise differences were observed, how did you show they were no worse than roundoff? Ran testing on an intermediate tag, where I computed the time weights in both the old and new ways. I confirmed that the difference in time weights (which is the only change in this tag) is always less than 1e-13. Actually, this difference is always less than 2e-16, double-precision roundoff. Also examined cpl hist diffs for a few select tests. Diffs are generally ~ 1e-6 after 5 days. =============================================================== =============================================================== Tag name: clm4_5_58 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Wed Jan 22 14:11:30 MST 2014 One-line Summary: major refactor of transient pft code, in prep for dynamic landunits Purpose of changes: Major refactor of transient pft code, in prep for dynamic landunits. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): 1899 (harvest rates remain non-zero even after the end of the pftdyn dataset) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Erik; design reviewed by CLM-CMT List any svn externals directories updated (csm_share, mct, etc.): scripts: scripts4_140102 -> scripts4_140114 Machines: Machines_131206b -> Machines_140107 csm_share: share3_131226 -> share3_131231 pio: pio1_8_3 -> pio1_8_6 CMake: New external added List all files eliminated: ========= renamed to dynConsBiogeophysMod.F90 D models/lnd/clm/src/clm4_5/main/dynlandMod.F90 ========= renamed to dynpftFileMod.F90; much of the stuff in here moved to other files in the dyn_subgrid directory D models/lnd/clm/src/clm4_5/main/pftdynMod.F90 ========= renamed D models/lnd/clm/test/unit List all files added and what they do: ========= Rename test/unit to test/unit_testers, and add unit tests for some of the stuff in dyn_subgrid. This also involved adding some mocks - particularly of ncdio_pio. A models/lnd/clm/test/unit_testers/clm4_5/biogeophys/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test/test_daylength.pf A models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/biogeophys/Daylength_test A models/lnd/clm/test/unit_testers/clm4_5/biogeophys A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarShared.F90 A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarTimeInterp.pf A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/test_dynVarTimeUninterp.pf A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynVar_test A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/test_dynTimeInfo.pf A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/dynTimeInfo_test A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/dyn_subgrid A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_var.F90.in A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_pio_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/spmdMod_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/do_genf90 A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share/ncdio_pio_mock.F90.in A models/lnd/clm/test/unit_testers/clm4_5/mock/util_share A models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/shr_sys_mod_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock/csm_share A models/lnd/clm/test/unit_testers/clm4_5/mock/main/decompMod_boundsTypeDecl.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/main/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock/main A models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/dynFileMod_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/dynUtilsMod_mock.F90 A models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock/dyn_subgrid A models/lnd/clm/test/unit_testers/clm4_5/mock/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/mock A models/lnd/clm/test/unit_testers/clm4_5/CMakeLists.txt A models/lnd/clm/test/unit_testers/clm4_5/README A models/lnd/clm/test/unit_testers/clm4_5 A models/lnd/clm/test/unit_testers ========= Make a new directory to hold all of the stuff related to dynamic subgrid weights. Currently this means transient PFTs, but soon it will also mean dynamic landunits. This includes stuff that used to be in pftdynMod and dynlandMod, as well as a bit from clm_driver. I have added a new driver for the dyn_subgrid stuff (dynSubgridDriverMod), and pulled out much of the shared, lower-level functionality into new modules (dynTimeInfoMod, dynFileMod, dynVarMod, dynVarTimeInterpMod, dynVarTimeUninterpMod, dynUtilsMod [which will soon go away]). In addition, I have separated the many routines in pftdynMod into separate modules, each with a single, better-defined function. A models/lnd/clm/src/clm4_5/dyn_subgrid/dynTimeInfoMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynSubgridDriverMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynHarvestMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynFileMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeochemMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynPriorWeightsMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynpftFileMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynCNDVMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynConsBiogeophysMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarMod.F90.in A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeInterpMod.F90.in A models/lnd/clm/src/clm4_5/dyn_subgrid/dynLandunitAreaMod.F90.IN_PROGRESS A models/lnd/clm/src/clm4_5/dyn_subgrid/dynUtilsMod.F90 A models/lnd/clm/src/clm4_5/dyn_subgrid/dynVarTimeUninterpMod.F90.in A models/lnd/clm/src/clm4_5/dyn_subgrid/CMakeLists.txt A models/lnd/clm/src/clm4_5/dyn_subgrid ========= script to generate files from their .in files using genf90 A models/lnd/clm/src/clm4_5/dyn_subgrid/do_genf90 ========= move check_sums_equal_1 to a new module, partly to reduce dependencies of unit tests, and partly because it is cleaner design to have it outside of surfrdMod A models/lnd/clm/src/clm4_5/main/surfrdUtilsMod.F90 List all existing files that have been modified, and describe the changes: ========= Add dyn_subgrid directory M models/lnd/clm/bld/configure ========= Add get_curr_yearfrac function (currently broken, will be fixed in next tag) M models/lnd/clm/src/util_share/clm_time_manager.F90 ========= Change type(file_desc_t) to class(file_desc_t); add 'only' clause to use statements to allow compilation with pgi M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90.in ========= Change 'use statement' for reworked dyn_subgrid code M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 ========= Move surfrd_check_urban and surfrd_check_sums_equal_1 to more appropriate places M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 ========= Move some code into dynSubgridDriverMod M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 - also remove incorrect header comment ========= Add compute_higher_order_weights routine M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Add unit testing support M models/lnd/clm/src/clm4_5/main/CMakeLists.txt ========= Frankfurt-PGI tests now pass! M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: ok regular tests (aux_clm): yellowstone_intel: ok yellowstone_pgi: ok frankfurt_intel: ok frankfurt_pgi: ok frankfurt_nag: ok history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: ok yellowstone_pgi: ok CLM tag used for the baseline comparisons: clm4_5_57 Changes answers relative to baseline: YES (very limited; see below) Can change answers due to the following: (1) Changes answers for harvest when a run starts inside the pftdyn timeseries but extends beyond it, without an intervening restart (see bug 1899) (2) Could theoretically change answers for yellowstone-pgi or hopper-pgi due to machines updates, but no changes showed up in the yellowstone test suite =============================================================== =============================================================== Tag name: clm4_5_57 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Tue Jan 7 14:17:04 MST 2014 One-line Summary: change CNDV water conservation to use the pftdyn method Purpose of changes: For my dynamic landunit work, I was trying to reconcile what's going on with prescribed transient PFTs (pftdyn) vs CNDV. The reason is that I'm trying to set up an overall control flow for dynamic landunits, and you need to be able to run either of these in conjunction with dynamic landunits. In doing this, I noticed that water conservation is handled differently for pftdyn vs CNDV: For pftdyn, water conservation is done as described in section 21.2 of the CLM4.5 tech note: water contents are summed before and after transition, and the difference is put in the runoff term CNDV appears not to use this before & after difference. Instead, it does a correction for canopy water in pftdynMod: pftdyn_wbal. For dynamic landunits, we're planning to use an approach like what is currently done for pftdyn. I think it's going to be messy and confusing to try to maintain the current CNDV approach when it's possible to have CNDV in conjunction with dynamic landunits. Thus, I am changing CNDV to use the pftdyn approach to water conservation, whether or not you are running with dynamic landunits. This will change answers for CNDV/BGCDV cases in CLM4.5, though I expect the effects to be small. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: erik; concept approved by Sam Levis List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Main changes are here; also minor (somewhat related) cleanup: fix some section heading comments, add a timer (ndep_interp) - pulling out some stuff that used to be (inappropriately) in the pftdynwts timer section M models/lnd/clm/src/clm4_5/main/clm_driver.F90 ========= Remove a bunch of now-unneeded code, especially from pftdynMod M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 CLM testing: NOTE: Most testing was done from tag cndv_water_conservation_n01_clm4_5_55 - up-to-date with clm4_5_55, NOT clm4_5_56. I then updated to clm4_5_56 and reran just the three tests that were run for that tag (see its ChangeLog entry, below), with comparison to clm4_5_56. regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK CLM tag used for the baseline comparisons: clm4_5_55 for most tests, clm4_5_56 for three tests (see above note) Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 with DV (CNDV / BGCDV) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated, but expected to be larger than roundoff/same climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_56 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Thu Jan 2 09:06:32 MST 2014 One-line Summary: update scripts external to fix I20TRCLM45BGC compset Purpose of changes: update scripts external to fix I20TRCLM45BGC compset Requirements for tag: fix bug 1869 Test level of tag: limited (see below) Bugs fixed (include bugzilla ID): 1869 (I20TRCLM45BGC compset improperly defined) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts: scripts4_131203 -> scripts4_140102 List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: none CLM testing: ONLY TESTED THE 3 AFFECTED TESTS from the yellowstone & frankfurt aux_clm test suites: PASS PET_P15x2_Lm13.f10_f10.I20TRCLM45BGC.yellowstone_pgi.clm-reduceOutput.GC.140102-060037 PASS ERS_D.f10_f10.I20TRCLM45BGC.frankfurt_pgi.clm-decStart.GC.140102-060448 PASS ERS_Mmpich.f10_f10.I20TRCLM45BGC.frankfurt_nag.clm-decStart.GC.140102-060608 CLM tag used for the baseline comparisons: clm4_5_55 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Only I20TRCLM45BGC compsets - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): New climate for this compset: correctly uses CLM4.5 instead of CLM4.0 code. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_55 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Dec 27 16:21:45 MST 2013 One-line Summary: add hooks to Sean Santos's unit test framework, and begin to add CLM unit tests Purpose of changes: (1) add hooks to Sean Santos's unit test framework (2) begin to add CLM unit tests Note: this tag currently does NOT have the CMake utilities that are needed to run the unit tests. Instead, the instructions show how to point to a version of these in my directory. That's because, as of the time I submitted this tag for testing, the necessary working version of the CMake utilities was not yet tagged. In the near future, another external could be added to pull in these CMake utilities in the CLM directory tree. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self, santos List any svn externals directories updated (csm_share, mct, etc.): tools/unit_testing: added external models/csm_share: share3_131101 -> share3_131226 (to get changes needed for building unit tests) List all files eliminated: None List all files added and what they do: ========= Set up unit test directories & CMakeLists.txt files, and add unit tests for DaylengthMod. Note that the tests themselves are in Daylength_test/test_daylength.pf. See the README file for how to run the tests A models/lnd/clm/test/unit A models/lnd/clm/test/unit/clm4_5 A models/lnd/clm/test/unit/clm4_5/README A models/lnd/clm/test/unit/clm4_5/CMakeLists.txt A models/lnd/clm/test/unit/clm4_5/mock A models/lnd/clm/test/unit/clm4_5/mock/decompMod_boundsTypeDecl.F90 A models/lnd/clm/test/unit/clm4_5/mock/CMakeLists.txt A models/lnd/clm/test/unit/clm4_5/Daylength_test A models/lnd/clm/test/unit/clm4_5/Daylength_test/test_daylength.pf A models/lnd/clm/test/unit/clm4_5/Daylength_test/CMakeLists.txt ========= Add CMakeLists.txt files that are needed to build unit tests A models/lnd/clm/src/clm4_5/main/CMakeLists.txt A models/lnd/clm/src/clm4_5/biogeophys/CMakeLists.txt List all existing files that have been modified, and describe the changes: ========= Remove unneeded 'use' statement, to prevent pulling in more than is necessary for the unit test build M models/lnd/clm/src/clm4_5/main/clmtype.F90 ========= Externals updated - see above M SVN_EXTERNAL_DIRECTORIES CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK Also ran the new unit tests, as per the instructions in models/lnd/clm/test/unit/clm4_5/README - all PASS CLM tag used for the baseline comparisons: clm4_5_54 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_54 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Dec 27 15:55:05 MST 2013 One-line Summary: update externals to cesm1_3_beta06 Purpose of changes: Update externals to cesm1_3_beta06 versions. However, do NOT update RTM, because the latest version of RTM results in failures for ERI _N2 tests. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): none Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: not investigated Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (.../trunk_tags/clm4_5_53) (revision 56268) +++ SVN_EXTERNAL_DIRECTORIES (.../branch_tags/clm_update_externals_cesm1_3_beta06_tags/clm_update_externals_cesm1_3_beta06_n02_clm4_5_53) (revision 56268) @@ -1,25 +1,25 @@ # CESM scripts, machines and driver -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_131126a -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130930b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_02 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_131203 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_131206b +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_07 # Model components: Data atmosphere, and stub components as well as land-ice model -models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_131116 -models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_03/socn -models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_03/sice -models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_03/sglc -models/wav/swav https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_03/swav -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_34 -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_130924 +models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_131201 +models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/socn +models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice +models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc +models/wav/swav https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/swav +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_34 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_131008 # Utilities: csm_share, esmf, timing, MCT, PIO -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_130918 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_131101 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_130213 -models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_130506 +models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_131108 models/utils/mct https://github.com/quantheory/MCT/tags/compiler_fixes_n03_MCT_2.8.3 -models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_7_2/pio +models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_3/pio # Mapping tools: -tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_130529 +tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_131120 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130716 models/lnd/clm/tools/shared/gen_domain https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130716/gen_domain_files List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= NCK tests no longer fail; change fail type of ERS_D.f19_g16.IGRCP26CN.frankfurt_pgi from CFAIL to RUN M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: generate only (no baselines from clm4_5_53) CLM tag used for the baseline comparisons: clm4_5_53 Changes answers relative to baseline: YES, but only for multi-instance If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Multi-instance - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated, but suspected to be new climate Note that NCK tests newly pass (they had been failing), so this answer change for multi-instance tests is expected. These answer changes show up in the following tests: FAIL CME_N2.f10_f10.I1850CLM45BGC.yellowstone_intel.clm-default.GC.131227-063851.compare_hist.clm4_5_53 FAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.GC.131227-063851.compare_hist.clm4_5_53 FAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.GC.131227-063851.compare_hist.clm4_5_53 FAIL CME_N2.f10_f10.I1850CLM45BGC.yellowstone_intel.clm-default.compare_hist.clm4_5_53.clm2.h0 FAIL ERI_N2.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.compare_hist.clm4_5_53.clm2.h0 If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_53 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Thu Dec 19 07:59:56 MST 2013 One-line Summary: refactor restart interfaces Purpose of changes: Refactor restart interfaces. Most work done by mvertens. 1) Add two *.F90.in files that use genf90.pl go generate source. This saves time in dev. and maintenence. If you modify only the *.F90 file, your changes will be lost. Instead modify the *F90.in file, then run genf90.pl on that file. If you have questions, ask a clm developer for help. 2) Restart capability has now been encapsulated in a subroutine call that uses Fortran 2003 interfaces over type and dimension. For example: - if (flag == 'define') then - call ncd_defvar(ncid=ncid, varname='grainc_storage_to_xfer', xtype=ncd_double, & - dim1name='pft',long_name='grain C shift storage to transfer',units='gC/m2/s') - else if (flag == 'read' .or. flag == 'write') then - call ncd_io(varname='grainc_storage_to_xfer', data=pcf%grainc_storage_to_xfer, & - dim1name=namep, ncid=ncid, flag=flag, readvar=readvar) - if (flag=='read' .and. .not. readvar) then - if (is_restart()) call endrun - end if - end if + call restartvar(ncid=ncid, flag=flag, varname='grainc_storage_to_xfer', xtype=ncd_double, & + dim1name='pft', & + long_name='grain C shift storage to transfer', units='gC/m2/s', & + interpinic_flag='interp', readvar=readvar, data=pcf%grainc_storage_to_xfer) Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary data sets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: ! renamed for consistency D models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 List all files added and what they do: A models/lnd/clm/src/util_share/dtypes.h A models/lnd/clm/src/util_share/ncdio_pio.F90.in A models/lnd/clm/src/util_share/restUtilMod.F90 A models/lnd/clm/src/util_share/restUtilMod.F90.in A models/lnd/clm/src/clm4_5/biogeochem/CNRestMod.F90 List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/util_share/clm_time_manager.F90 M models/lnd/clm/src/util_share/accumulMod.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4RestMod.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeRestMod.F90 CLM testing: regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK frankfurt_intel - OK frankfurt_pgi - OK frankfurt_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_52 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_52 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Tue Nov 26 22:07:32 MST 2013 One-line Summary: turn on longwave radiation downscaling for glc_mec by default Purpose of changes: Turn on longwave radiation downscaling for glc_mec by default Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: set glcmec_downscale_longwave to true by default List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts: updated to scripts4_131126a, to get tweaked test list List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Change glcmec_downscale_longwave to true by default M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 ========= Remove build-namelist tests that are no longer xFails M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml ========= Update scripts to scripts4_131126a, to get tweaked test list M SVN_EXTERNAL_DIRECTORIES CLM testing: build-namelist tests: yellowstone: ok frankfurt regular tests (aux_clm): yellowstone_intel: ok yellowstone_pgi: ok frankfurt_intel: ok frankfurt_pgi: ok frankfurt_nag: ok Expected baseline failures: FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.131126-131812.compare_hist.clm4_5_51 FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.131126-131812.nlcomp Two BFAILs due to changed tests; I reran them as their old versions and confirmed that answers changed, as expected: BFAIL ERS_D.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC_changeFlags.GC.131126-131807.compare_hist.clm4_5_51 BFAIL PEM_D.f19_g16.IG1850CLM45.yellowstone_pgi.clm-glcMEC.GC.131126-214346.compare_hist.clm4_5_51 history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: ok yellowstone_pgi: ok CLM tag used for the baseline comparisons: clm4_5_51 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 with glc_mec (IG compsets) - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff; not investigated whether it is same climate or new climate If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_51 Originator(s): sacks (sacks) Date: Tue Nov 26 05:46:29 MST 2013 One-line Summary: rework downscaling of atm fields for glc_mec Purpose of changes: A number of changes related to downscaling atm -> clm fields for glc_mec landunits: (1) Add new options to downscale precip (division into rain/snow) and longwave radiation (developed by Bill Lipscomb). Both are currently off by default. (2) Move downscaling code out of clm_driverInit into clm_atmlnd.F90 - this is a more appropriate module, and is a step towards modularity, because the code to deal with the atmospheric forcing fields lives in the same module as the definition of these atmospheric forcing fields. (3) Ensure that all code uses the downscaled, column-level fields where possible. Previously, some code (which did not operate over glc_mec landunits) used the non-downscaled, gridcell-level version of fields such as forc_t. This was a problem because (a) it was confusing and error-prone, and (b) we will soon be bringing in code to do downscaling over other landunits as well as glc_mec landunits. (4) To support (3), and make it harder for someone to accidentally use the gridcell-level version of a field when they should be using the downscaled, column-level version: Broke clm_a2l into two pieces - one containing fields that aren't downscaled, and one containing fields that are downscaled. For fields that are downscaled, clearly distinguished the non-downscaled versions so they couldn't be used by accident. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: For glc_mec runs, added two new namelist options: glcmec_downscale_rain_snow_convert and glcmec_downscale_longwave. List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: erik; portions reviewed by Bill Lipscomb, Mariana, Stefan List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Significant changes: includes stuff that used to be in clm_driverInitMod; added new downscaling code for precip & lwrad; split atm2lnd type into two types; reworked initialization interfaces; removed unused field rainf, because it currently isn't used and could theoretically become inconsistent with the downscaled rain/snow M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 ========= Removed downscaling code from here M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 ========= Added call to downscale_forcings M models/lnd/clm/src/clm4_5/main/clm_driver.F90 ========= Changed interface to init_atm2lnd: previously, clm_initialize used clm_a2l from clm_atmlnd and passed it to init_atm2lnd_type (also in clm_atmlnd) - there was no reason for this, it was confusing, and to some extent broke modularity M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ========= New parameter M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 ========= Removed some variables M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 ========= New namelist control variables M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/bld/build-namelist M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml ========= Removed unused rainf, added pointer to allow lnd_import_export to remain identical between clm4_0 and clm4_5 code M models/lnd/clm/src/clm4_0/main/clm_atmlnd.F90 ========= Fixed test ids for failing build-namelist tests M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml ========= Minor changes in lots of places to (a) use fields from a2l_downscaled_col instead of clm_a2l, and (b) index those fields by column rather than by gridcell M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/cpl/lnd_comp_esmf.F90 M models/lnd/clm/src/cpl/lnd_import_export.F90 M models/lnd/clm/src/cpl/lnd_comp_mct.F90 CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK Note the following expected nlcomp failures: FAIL ERS_D.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC.GC.131125-104703.nlcomp FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.131125-104751.nlcomp FAIL PEM_D.f19_g16.IG1850CLM45.yellowstone_pgi.GC.131125-104751.nlcomp Also, there were BFAILs for the following, implying that the baselines didn't exist; I'm not too concerned because baseline comparisons passed for similar tests: BFAIL SMS_Ly1_Mmpich.f19_g16.ICLM45BGCCROP.frankfurt_nag.clm-reduceOutput.GC.131125-104832.compare_hist.clm4_5_50 BFAIL ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel.clm-monthly_noinitial.GC.131125-104703.compare_hist.clm4_5_50 history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK As above, baselines were missing for this test: BFAIL2 ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel.clm-monthly_noinitial.compare_hist.clm4_5_50.clm2.h0 (baseline history file does not exist) BFAIL2 ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel.clm-monthly_noinitial.compare_hist.clm4_5_50.clm2.h1 (baseline history file does not exist) CLM tag used for the baseline comparisons: clm4_5_50 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_50 Originator(s): erik (Erik) Date: Sun Nov 24 18:51:11 MST 2013 One-line Summary: Bring in a crop of b4b bugfixes, fix getregional script, start move of PTCLM to PTCLMmkdata tool Purpose of changes: Bring in a crop of bit-for-bit bug-fixes to the trunk for November. Fix the getregional_datasets script and initial move of PTCLM to just be a CLM tool under models/lnd/clm/tools/shared to create single-point datasets. New option to create_newcase "-user_mods_dir" for a directory with user chagnes such as user_nl_* namelist modification files, xmlchange_cmnds file with xmlchanges to make, and SourceMods/src.*/* files. Requirements for tag: bit-for-bit bug-fixes and work on PTCLM Test level of tag: regular Bugs fixed (include bugzilla ID): 1868 (Add user_mods_dir to create_newcase) 1854 (Remove fndepdat no longer used) 1842 (Remove unused variables from gridcell type) (Bill) 1835 (Add write statement to pftdyn so you can see what it is doing) 1828 (Clarify modulo used in irrigation code) (Bill) 1770 (Remove sitespf_pt valid_values list for clm4_0) 1724 (getregional script does NOT work) 1625 (Problem setting finidat in CLM for RUN_TYPE=hybrid/branch) 1543 (large-file format does NOT work in latest clm) 1481 (Provide a more direct way to set a user provided finidat file) 1437 (problems with link_dirtree -- no longer needed) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: SourceMods directory on create_newcase New -user_mods_dir option to create_newcase that will copy SourceMods/src.*/* files to the new case. Also copies user_nl_* files and xmlchange_cmnds Describe any changes made to the namelist: remove outnc_large_files Remove outnc_large_files -- wasn't functional (now always use 64-bit format) Remove fndepdat from namelist_definition/defaults no longer used. List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self, sacks (fixes for 1842 and 1828) List any svn externals directories updated (csm_share, mct, etc.): scripts scripts updated to scripts4_131119 List all files eliminated: None List all files added and what they do: ------ Add externals for tools so PTCLM shows up in tool directory ------ Add same files and a README file for getregional A models/lnd/clm/tools/SVN_EXTERNAL_DIRECTORIES A models/lnd/clm/tools/shared/ncl_scripts/sample_inlist A models/lnd/clm/tools/shared/ncl_scripts/sample_outlist A models/lnd/clm/tools/shared/ncl_scripts/README.getregional List all existing files that have been modified, and describe the changes: ------ Get getregional_datasets script working again. Now operates on ------ lists of files. M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.pl M models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.ncl M models/lnd/clm/test/tools/TSMscript_tools.sh --- Copy sample_*list files M models/lnd/clm/test/tools/nl_files/getregional - change arguments ------ M models/lnd/clm/tools/clm4_0/interpinic/src/interpinic.F90 ------ Remove clm_startfile option and outnc_large_files M models/lnd/clm/bld/config_files/config_definition_clm4_0.xml -- Remove valid_values from sitespf_pt so can be anything. M models/lnd/clm/bld/unit_testers/build-namelist_test.pl -- Remove clm_startfile option and move some clm4_5 configure options to build-namelist M models/lnd/clm/bld/build-namelist --- Remove clm_startfile option and outnc_large_files M models/lnd/clm/bld/clm.buildnml.csh - Remove clm_startfile option set finidat/nrevsn like how is done in CAM ------ Remove fndepdat and outnc_large_files M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0_tools.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_0.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 -------- Remove some gridcell variables not sued M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 ------------- write on what's going on M models/lnd/clm/src/clm4_5/main/controlMod.F90 ------------ Remove outnc_large_files M models/lnd/clm/src/clm4_5/main/clmtype.F90 --------------- Remove a bunch of gridcell variables not needed M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 - Add seconds_since_irrig_start_time temporary to clarify ------ M models/lnd/clm/src/clm4_0/main/pftdynMod.F90 --- add some writes statements M models/lnd/clm/src/clm4_0/main/controlMod.F90 -- remove outnc_large_files CLM testing: build-namelist tests: yellowstone yes frankfurt no NOTE: there were some tests that were passing but in the xFail list from before clm4_5_49 that I marked as working. regular tests (aux_clm): yellowstone_intel yes yellowstone_pgi yes frankfurt_intel yes frankfurt_pgi yes frankfurt_nag yes history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel yes yellowstone_pgi yes tools testing: yellowstone interactive yes CLM tag used for the baseline comparisons: clm4_5_49 Changes answers relative to baseline: no bit-for-bit =============================================================== =============================================================== Tag name: clm4_5_49 Originator(s): muszala (Stefan Muszala) Date: Sat Nov 16 07:51:27 MST 2013 One-line Summary: Swenson anomaly forcing - Part 1 purpose of changes: add additional data streams to modify existing data streams for purposes such as bias correction or specifying future changes relative to baseline data streams, e.g. specifying future atmospheric forcing anomalies when running CLM with data atmosphere. Paired with datm8_131115. For what to set in user_nl_cpl, user_nl_datm, see testing section. This is part 1 of 2. Part 2 will address a general way to handle streams in the DATM that is triggered off of an AF compset. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: swenson, kluzek, self List any svn externals directories updated (csm_share, mct, etc.): update to datm8_131116 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_0/main/clm_atmlnd.F90 M models/lnd/clm/src/cpl/clm_cpl_indices.F90 M models/lnd/clm/src/cpl/lnd_import_export.F90 M SVN_EXTERNAL_DIRECTORIES CLM testing: Testing for features of this tag: f09_g16, ICRUCLM45 anomaly forcing namelists user_nl_cpl: cplflds_custom = 'Sa_prec_af->a2x', 'Sa_prec_af->x2l','Sa_tbot_af->a2x', 'Sa_tbot_af->x2l','Sa_pbot_af->a2x', 'Sa_pbot_af->x2l','Sa_shum_af->a2x', 'Sa_shum_af->x2l','Sa_u_af->a2x', 'Sa_u_af->x2 l','Sa_v_af->a2x', 'Sa_v_af->x2l','Sa_swdn_af->a2x', 'Sa_swdn_af->x2l','Sa_lwdn_af->a2x', 'Sa_lwdn_af->x2l' user_nl_datm: anomaly_forcing = 'Anomaly.Forcing.Precip','Anomaly.Forcing.Temperature', 'Anomaly.Forcing.Pressure','Anomaly.Forcing.Humidity','Anomaly.Forcing.Uwind', 'Anomaly.Forcing.Vwind','Anomaly.Forcing.Shortwave','Anomaly.Forcing.Longwave' bias correction namelists user_nl_cpl: cplflds_custom = 'Sa_precsf->a2x', 'Sa_precsf->x2l' user_nl_datm: bias_correct = 'BC.CRUNCEP.GPCP.Precip' build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK frankfurt_intel - OK frankfurt_pgi - OK frankfurt_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_48 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_48 Originator(s): muszala (Stefan Muszala) Date: Thu Nov 14 08:28:31 MST 2013 One-line Summary: bug fixes for CLM dry deposition and MEGAN VOC emissions Purpose of changes: Bring in bug fixes from fvitt for CLM dry deposition and MEGAN VOC emissions. Any changes to answers are limited to rare circumstances. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: fvitt, self List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_0/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/DryDepVelocity.F90 CLM testing: build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK frankfurt_intel - OK frankfurt_pgi - OK frankfurt_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_47 Changes answers relative to baseline: Answers change for some of the chemistry configurations, but these changes will not be see in the CLM standalone tests. =============================================================== =============================================================== Tag name: clm4_5_47 Originator(s): muszala (Stefan Muszala) Date: Tue Nov 12 09:26:20 MST 2013 One-line Summary: fix Bug 1858 - AGDD now reset annually Purpose of changes: Fix bug 1858. AGDD is now reset annually. Replace -99999_r8 with a parameter in accumulMod.F90 which is used in accFldsMod.F90 Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): 1858 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: slevis, sacks, muszala List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/util_share/accumulMod.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 CLM testing: build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK frankfurt_intel - OK frankfurt_pgi - OK frankfurt_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_46 Changes answers relative to baseline: Generally, No. There may be changes in DV compsets over very long simulation times and at very high resolutions. =============================================================== =============================================================== Tag name: clm4_5_46 Originator(s): sacks (sacks) Date: Fri Nov 8 17:26:02 MST 2013 One-line Summary: remove zeroing out of slope for special landunits Purpose of changes: Previously, there was code to zero out slope for grid cells with 100% special landunits. However, there were a number of problems with this: (1) With dynamic landunits, this is problematic, because a grid cell could start as 100% special landunits, then later become < 100% special landunits (e.g., due to retreating glaciers) (2) Moreover, why should the slope of a special landunit depend on whether the grid cell has 100% special landunits. This seems to be saying that, e.g., the slope of a glacier landunit depends on whether the grid cell is entirely glacier or part glacier and part natural veg. (3) And I guess moreover, why is the slope zeroed out for special landunits in the first place? From talking with Erik, we decided thish code was probably a relic from a time when the surface dataset had some bad values (e.g., over Greenland / Antarctica). This is no longer the case, so this code is no longer needed. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: erik List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK Following are the baseline failures, which are expected (see below): FAIL SMS_D.1x1_mexicocityMEX.ICLM45.frankfurt_intel.clm-default.GC.131107-223431.compare_hist.clm4_5_45 FAIL SMS_D.1x1_vancouverCAN.ICLM45.frankfurt_pgi.clm-default.GC.131107-223435.compare_hist.clm4_5_45 FAIL ERS_Mmpi-serial.1x1_mexicocityMEX.I1PTCLM45.frankfurt_nag.clm-default.GC.131107-223439.compare_hist.clm4_5_45 FAIL SMS_D_Mmpi-serial.1x1_vancouverCAN.I1PTCLM45.frankfurt_nag.clm-default.GC.131107-223439.compare_hist.clm4_5_45 FAIL ERI.f09_g16.ICLM45BGC.yellowstone_intel.GC.131107-223256.compare_hist.clm4_5_45 FAIL ERI_D.f09_g16.ICLM45BGC.yellowstone_intel.GC.131107-223256.compare_hist.clm4_5_45 FAIL ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.GC.131107-223256.compare_hist.clm4_5_45 FAIL ERS_Lm3.1x1_vancouverCAN.I1PTCLM45.yellowstone_intel.clm-default.GC.131107-223256.compare_hist.clm4_5_45 FAIL SMS.1x1_mexicocityMEX.I1PTCLM45.yellowstone_intel.clm-default.GC.131107-223256.compare_hist.clm4_5_45 FAIL ERI.f09_g16.I1850CRUCLM45BGC.yellowstone_pgi.GC.131107-223301.compare_hist.clm4_5_45 FAIL ERI.f09_g16.ICLM45BGC.yellowstone_pgi.GC.131107-223301.compare_hist.clm4_5_45 FAIL ERI_D.f09_g16.ICLM45BGC.yellowstone_pgi.GC.131107-223301.compare_hist.clm4_5_45 history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK Following are the baseline failures, which are expected: FAIL ERS_Lm3.1x1_vancouverCAN.I1PTCLM45.yellowstone_intel.clm-default.compare_hist.clm4_5_45.clm2.h0 FAIL ERS_Lm3.1x1_vancouverCAN.I1PTCLM45.yellowstone_intel.clm-default.compare_hist.clm4_5_45.clm2.h1 FAIL SMS.1x1_mexicocityMEX.I1PTCLM45.yellowstone_intel.clm-default.compare_hist.clm4_5_45.clm2.h0 FAIL SMS.1x1_mexicocityMEX.I1PTCLM45.yellowstone_intel.clm-default.compare_hist.clm4_5_45.clm2.h1 CLM tag used for the baseline comparisons: clm4_5_45 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Any configuration that includes a grid cell that has 100% special landunits, including at least some urban - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but expected to be larger than roundoff/same climate - since this only affects a very small number of grid cells, and (I believe) only the urban pervious road in those grid cells If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_45 Originator(s): sacks (sacks) Date: Fri Nov 8 16:10:22 MST 2013 One-line Summary: refactor daylength calculation, and other minor changes Purpose of changes: (1) Compute daylength in a single place, and compute necessary variables at initialization rather than having them on the restart file (2) Compute daylength-related variables at initialization rather than having them on the restart file, both to clean things up and to fix some daylength bugs at initialization (these bugs were fixed in a kludgey way in clm4_5_44, and now are fixed robustly) (3) Fix daylength calculation at the poles (previously blew up due to roundoff errors) (doesn't change behavior currently, but could change behavior / answers if there were a vegetated landunit at the pole) (4) Fix sminn on restart, so that crop restarts can be bfb (bug 1846) (5) Add all_active namelist variable that makes even 0-weight points active, for testing purposes Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 1846 (crop restarts aren't exact due to sminn field) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: add all_active namelist variable List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: (1) & (2) reviewed by erik, (3) by self List any svn externals directories updated (csm_share, mct, etc.): scripts: update from scripts4_131030 -> scripts4_131107a List all files eliminated: none List all files added and what they do: ========= Compute daylength in a single place A models/lnd/clm/src/clm4_5/biogeophys/DaylengthMod.F90 List all existing files that have been modified, and describe the changes: ========= Refactor daylength calculation to just compute daylength in a single, central place, and compute necessary variables at initialization rather than having them on the restart file M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 - also fix sminn on restart (bug 1846) M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 M models/lnd/clm/src/clm4_5/main/CNiniTimeVar.F90 M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 ========= Add all_active namelist variable that makes even 0-weight points active, for testing purposes M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Add & remove tests from xFail list M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml *** No longer tested (replaced by CME_Ly4) - Runs out of time. CME_Ly4.f10_f10.I1850CLM45BGC.yellowstone_intel.clm-monthly passes *** Now passes - restarts not exact due to bug 1846: crop restarts are not exact due to sminn field + Diffs in cpl log files in rofl, rofi and volr + Diffs in cpl log files in rofl, rofi and volr CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK Note that we get the following failures in compare_hist: FAIL PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.GC.131107-214732.compare_hist.clm4_5_44 FAIL PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrigOn_reduceOutput.GC.131107-214732.compare_hist.clm4_5_44 However, I think that's expected due to the oddities in the clm4_5_44 tag with openmp - see notes in the ChangeLog for clm4_5_44 for details. Note that this one is identical to clm4_5_43, suggesting that clm4_5_45 undoes the problem introduced in clm4_5_44: PASS PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.compare_hist.clm4_5_43.cpl.hi history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK In particular, note that the clm hist comparison passes for the two above tests that had unexpected cpl diffs: PASS PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.compare_hist.clm4_5_44.clm2.h0 PASS PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_intel.clm-irrigOn_reduceOutput.compare_hist.clm4_5_44.clm2.h0 CLM tag used for the baseline comparisons: clm4_5_44 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_44 Originator(s): sacks (sacks) Date: Fri Nov 8 08:19:56 MST 2013 One-line Summary: temporary hack to daylength initialization to provide baselines for the next tag Purpose of changes: The next tag (clm4_5_45) involves a major refactor to the daylength calculation. That refactor is bfb in most respects, but gives differences in the first time step in a few situations. This tag (clm4_5_44) does the minimal changes needed to get the same results as clm4_5_45, in order to have more confidence when testing clm4_5_45. Note that the changes here are a kludge that will be reverted in clm4_5_45. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1850 (incorrect daylength in first timestep of some runs) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK There were a number of expected compare_hist failures, as described below. There was one unexpected compare_hist failure: FAIL PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.GC.145424.compare_hist.clm4_5_43 Diffs are just in voc cpl fields. CLM hist files are identical at the end of the run. An SMS version of this test passes. Interestingly, the kludgey code that I have added for clm4_5_44 isn't even executed in this test... so the only diffs should be in variable declarations and 'use' statements. And clm4_5_45 (which I will tag soon) gives identical cpl hist files to clm4_5_43. So there may just be a compiler fluke in the compilation of this tag with openmp enabled. Because clm4_5_45 will give identical results to clm4_5_43, I'm not worrying about this. history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK A number of expected compare_hist failures, as described below. CLM tag used for the baseline comparisons: clm4_5_43 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 with BGC (and probably CN), involving either (a) initial conditions interpolated from a different resolution, or (b) a change in start date relative to the ref date of an initial file (which shows up in ERI tests) - see bug 1850 - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but almost certainly larger than roundoff/same climate, since this code mod just changes things in the first timestep. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_43 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Wed Nov 6 09:21:25 MST 2013 One-line Summary: allocate memory for most landunits in every grid cell (needed for dynamic landunits) Purpose of changes: Main change is to allocate memory for most landunits in every grid cell, to support dynamic landunits. Note that we ALWAYS do this extra memory allocation, so that the user isn't required to do interpinic between a non-dynamic run and a dynamic landunit run. (If we eventually change the restart file format / processing so that you can add / remove 0-weight points at will, then we could potentially add some logic to only do this extra allocation if we're using dynamic landunits.) Supporting changes are (1) determining which grid cells have enough parameter data to support urban landunits, (2) new initial conditions files, (3) in mksurfdata_map, don't set soil parameters to 0 under glacier, and set urban parameters even if urban cover is 0% Other changes are: (1) only do snow balance check over active columns (2) fix interpinic bug (bug 1839) (3) newer files for testing interpinic NOTE: All CLM4.5 initial conditions will need to be interpinic'ed to be usable in this tag (this has been done for all out-of-the-box initial conditions) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 1839 (interpinic uses wrong column when there are 0-weight columns in the input file) - 1840 (snow balance check is executed over inactive columns) - 1825 (surface datasets need urban parameters even when pcturb is 0 everywhere): partial fix - still waiting on new USUMB dataset Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: new initial conditions for CLM4.5 List any changes to the defaults for the boundary datasets: new initial conditions for CLM4.5 Describe any substantial timing or memory changes: Significant memory increases for all CLM4.5 configurations - memory is now allocated for all natural veg landunits, all crop landunits (if using create_crop_landunit), and most urban landunits. Also, significant performance decrease (~ 10%) associated with the above change, which I believe is mainly due to decreased cache friendliness. Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: ========= Change files used for testing interpinic, in order to use files that are up-to-date with the current version of CLM, for a more accurate test D models/lnd/clm/tools/clm4_5/interpinic/clmi.I2000CLM45BGC.2000-01-01.10x15_simyr2000_c130607.nc List all files added and what they do: ========= Change files used for testing interpinic, in order to use files that are up-to-date with the current version of CLM, for a more accurate test A models/lnd/clm/tools/clm4_5/interpinic/clmi.I2000CLM45BGC.2000-01-01.10x15_simyr2000_c131104.nc List all existing files that have been modified, and describe the changes: ========= Determine which grid cells should have urban landunits M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varsur.F90 M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 ========= Create a natural veg landunit and crop landunit in all grid cells; create an urban landunit in all grid cells for which we have determined that urban is "valid" M models/lnd/clm/src/clm4_5/main/subgridMod.F90 ========= Change a comment M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 ========= Only do snow balance check over active columns (fixes bug 1840) M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 ========= Use new initial conditions files that are consistent with the expanded 1-d memory structures M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Don't set soil parameters to 0 under glacier; set urban parameters even if urban cover is 0% M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mksoilMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 ========= Fix interpinic bug (bug 1839) M models/lnd/clm/tools/clm4_5/interpinic/src/interpinic.F90 ========= Change files used for testing interpinic, in order to use files that are up-to-date with the current version of CLM, for a more accurate test M models/lnd/clm/tools/clm4_5/interpinic/interpinic.runoptions ========= Add two tests to the xFail list M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml + Runs out of time. CME_Ly4.f10_f10.I1850CLM45BGC.yellowstone_intel.clm-monthly passes + Needs new surface dataset CLM testing: NOTE: main tests were done with dynlu_allocate_memory_n11_fix_cndv_time_averages_n01_clm4_5_41; tools tests with dynlu_allocate_memory_n12_clm4_5_42; build-namelist tests with a slightly older tag build-namelist tests: yellowstone: OK. However, the "correct" comparisons are spurious, because I think the build-namelist test is broken regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK tools testing: yellowstone interactive: OK *** Expected failures 027 smiS4 TSMscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............\c rc=6 FAIL 028 bliS4 TBLscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............\c rc=4 FAIL *** Expected baseline failure (uses new input & output file, and there is a non-bfb change in interpinic) 016 blh54 TBLtools.sh clm4_5 interpinic tools__ds runoptions ....................................\c rc=7 FAIL *** Expected diffs in PCT_SAND, PCT_CLAY, SOIL_COLOR 012 blg54 TBLtools.sh clm4_5 mksurfdata_map tools__s namelist ...................................\c rc=7 FAIL 020 bli24 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_T31_crpglc_2000^tools__ds \c rc=7 FAIL 022 bli53 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__o ....\c rc=7 FAIL 024 bli54 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__ds ...\c rc=7 FAIL 026 bli57 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__do ...\c rc=7 FAIL *** Expected diffs in urban fields 030 bli74 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_1x1_brazil_1850-2000^tools_\c rc=7 FAIL 032 bliT4 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_1x1_numaIA_mp24_2000^tools_\c rc=7 FAIL CLM tag used for the baseline comparisons: clm4_5_42 Changes answers relative to baseline: NO - but see note below: NOTE: This tag has the potential to change answers for cases using initial conditions that were interpinic'ed using the out-of-the-box interpinic, because of bugs in interpinic. This applies to CLM4.5 cases @ ne30 and hcru resolutions, as well as CLM4.5 cases using DV @ f09. However, no diffs showed up in the test suite, so it's possible that this isn't a problem. =============================================================== =============================================================== Tag name: clm4_5_42 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Mon Nov 4 09:45:36 MST 2013 One-line Summary: fix bug 1857 for CLM4.5 - CNDV running temperature means are incorrect Purpose of changes: Fix bug 1857 for CLM4.5 (not yet fixed for CLM4.0!). From the bugzilla entry: In this code in CNDVMod: do p = bounds%begp, bounds%endp g = pft%gridcell(p) if (kyr == 2) then ! slevis: add ".and. start_type==arb_ic" here? tmomin20(g) = t_mo_min(p) ! NO, b/c want to be able to start dgvm agdd20(g) = agdd(p) ! w/ clmi file from non-dgvm simulation end if tmomin20(g) = (19._r8 * tmomin20(g) + t_mo_min(p)) / 20._r8 agdd20(g) = (19._r8 * agdd20(g) + agdd(p) ) / 20._r8 end do Notice that this is a loop over p, but it's updating gridcell-level variables. This means that the running temperature means aren't at all what they purport to be. e.g., in a grid cell with the 17 natural PFTs and nothing else, the grid cell-level values will get the 17 pft values averaged in each year, rather than getting a single pft value per year. This means that these temperature variables are closer to a single year's value than to a running mean. The fix here should be simple: just change tmomin20 & agdd20 to pft-level variables. WARNING: USE CAUTION WHEN USING THIS TAG WITH AN OLDER RESTART FILE FROM A CLM4.5 DV CASE (this is not a problem for any out-of-the-box initial conditions files, but could apply if you have your own initial file from a DV run): In this case, the two DV-related variables AGDD20 and TMOMIN20 will be reset to their arbitrary initial conditions. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1857 - partial fix (still open for clm4.5) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEcosystemDynIniMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 ========= Remove SSP compare_hist BFAIL from xFAIL list (Ben fixed this in the last tag) M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK All baseline comparisons pass except the following expected failure: FAIL PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.GC.104252.compare_hist.clm4_5_41 history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK All baseline comparisons pass except the following expected failure: FAIL PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel.clm-reduceOutput.compare_hist.clm4_5_41.clm2.h0 CLM tag used for the baseline comparisons: clm4_5_41 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM4.5 with DV - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): Definitely larger than roundoff, but not investigated as to whether it's same climate or new climate. If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_41 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Wed Oct 30 17:33:16 MDT 2013 One-line Summary: update scripts to convert clm4_5 CPP flags to namelist variables. Purpose of changes: Convert clm4_5 CPP flags in controlMod.F90 into namelist variables, update scripts infrastructure to generate cases with namelist variables for bgc (CN, CNDV, methane, vsoilc_centbgc), crop, extra lake layers, vic, nofire, noio, sitespf_pt, snicarfrc, maxpatch_pft. Requirements for tag: Test level of tag: regular, tools, build_namelist Bugs fixed (include bugzilla ID): 1728 (scripts4_20131030 tag). 1770 (clm4_5 portion). Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: removed clm4_5 CPP flags: NOFIRE, LCH4, NITRIF, VERTSOILC, EXTRALAKELAYERS, VICHYDRO, CENTURY, CN, CNDV, CROP, SNICAR, VANCOUVER, NOIO, MEXICOCITY Describe any changes made to the namelist: added namelist variables: use_nofire, use_lch4, use_nitrif_denitrif, use_vertsoilc, use_extralakelayers, use_vichydro, use_century_decomp, use_cn, use_cndv, use_crop, use_snicar_frc, use_vancouver, use_mexicocity, use_noio All new namelist variables are logicals. List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: Erik Kluzek, Bill Sacks List any svn externals directories updated (csm_share, mct, etc.): scripts4_131030 List all files eliminated: D models/lnd/clm/bld/config_files/config_definition.xml - split into clm4_X variants List all files added and what they do: A models/lnd/clm/bld/config_files/config_definition_clm4_5.xml A models/lnd/clm/bld/config_files/config_definition_clm4_0.xml List all existing files that have been modified, and describe the changes: M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl - set crop=off by default to pick up correct defaults. M models/lnd/clm/bld/configure - completely remove clm4_5 only cpp flags, add physics dependent logic to clm4_0 flags. M models/lnd/clm/bld/queryDefaultNamelist.pl - point to physics specific config_definitions.xml file M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml - replace xml special characters with alternatives so file can be parsed. Add new build-namelist failures. M models/lnd/clm/bld/build-namelist - add logic for all new clm4_5 namelist variables, commandline options, switch defaults to use_N. M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml - add new namelist variables M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 - add CNNDynamicsInit() and logic to set nfix_timeconst from use_nitrif_denitrif namelist instead of CPP. M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 - convert ifdef'd parameters to variables M models/lnd/clm/src/clm4_5/main/controlMod.F90 - final conversion of CPP flags to namelist variables Switch the following files to use new namelist variables for attributes: M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml M models/lnd/clm/bld/namelist_files/use_cases/2000_control.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/2000-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/2000_glacierMEC_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/20thC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850_glacierMEC_control.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp2.6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp6_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp4.5_glacierMEC_transient.xml M models/lnd/clm/bld/namelist_files/use_cases/1850-2100_rcp8.5_glacierMEC_transient.xml CLM testing: build-namelist tests: yellowstone - most tests will fail, tests need to be updated in future tag frankfurt regular tests (aux_clm): - nlcomp tests fail for all clm4_5 because of the new namelist variables. - SSP tests should now be BFAIL, pass on next tag - All hist comp are bit for bit. yellowstone_intel ok yellowstone_pgi ok frankfurt_intel ok frankfurt_pgi ok frankfurt_nag ok tools testing: yellowstone interactive - smiS4 (getregional) also fails in clm4_5_40 frankfurt interactive - N/A CLM tag used for the baseline comparisons: clm4_5_40 Changes answers relative to baseline: None. =============================================================== =============================================================== Tag name: clm4_5_40 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Thu Oct 24 07:54:46 MDT 2013 One-line Summary: fix Bug 1752 - urban conductances depend on weights in an undesirable way Purpose of changes: Three parts to this tag. Description from Oleson: 1) fix Bug 1752 - modified UrbanMod.F90 to calculate 2) add 2 new diagnostic history fields (FIRE_U, FIRE_R) the conductances correctly. I created new variables to more clearly distinguish between scaled and unscaled conductances. 3) fix small bug in which the history field output of some of the anthropogenic heat flux variables are not bfb on restart when finidat is blank. I found this when verifying bfb for the original bug fix. The cause of this is initialization which sets non-urban to special value, instead of zero, which is what is desired. On restart, this initialization is not done and the history file set_nourb=0 has precedent. A related issue is that eflx_building_heat is not zero for pervious/impervious road when nlevurb /= nlevgrnd. Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): 1752 long test with irrigate=.true. (because irrigation likely never took effect in the old test). Requirements for tag: Test level of tag: regular & build_namelist Bugs fixed (include bugzilla ID): - 1827 / 1830: testmods don't work right for multi-instance tests (fix via scripts update) - 1829: PCT_SAND, PCT_CLAY and SOIL_COLOR are incorrect for some grid cells (fix via new surface datasets) - 1831: turning on irrigation leads to death in initialization (fix via a new initial conditions file) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: see changes in boundary datasets List any changes to the defaults for the boundary datasets: - new surface datasets for all resolutions for CLM4.5: same as before except for PCT_SAND, PCT_CLAY and SOIL_COLOR: these are no longer zeroed out under points that are believed to be 100% glacier ("believed to be" because this previously zeroed out some points that ended up having some other special landunits, such as lake) - new initial conditions file for CLM4.5, irrigate=.true., f10 Describe any substantial timing or memory changes: none Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts: scripts4_131001 -> scripts4_131003 List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= new logic for irrigation for finidat M models/lnd/clm/bld/build-namelist ========= new surface datasets; distinguish finidat based on value of 'irriagte' M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml CLM testing: build-namelist tests: caldera: OK The following failures were expected due to new surface datasets: 413/439 < FAIL> 418/439 < FAIL> 423/439 < FAIL> 428/439 < FAIL> 433/439 < FAIL> 438/439 < FAIL> regular tests (aux_clm): edison_intel (aux_clm_ys_intel & aux_clm_ys_pgi lists): OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: edison_intel: aux_clm_ys_intel list: OK edison_intel: aux_clm_ys_pgi list: OK CLM tag used for the baseline comparisons: clm4_5_35 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45 - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Not investigated carefully, but probably larger than roundoff/same climate. Answer changes are due to new surface datasets. This is due to fixing bug 1829 (PCT_SAND, PCT_CLAY and SOIL_COLOR are incorrect for some grid cells); it looks like this just affects answers over a small number of lake points (e.g., 9 lake points in an f19 run). If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_35 Originator(s): sacks (Bill Sacks) Date: Tue Oct 1 09:47:45 PDT 2013 One-line Summary: get CLM running on edison Purpose of changes: Update scripts and Machines externals to get the CLM test suite running on edison; this will be our replacement for yellowstone while yellowstone is down. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: N/A List any svn externals directories updated (csm_share, mct, etc.): -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_130929 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130927 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_131001 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130930b List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Remove PEM test that should pass now; add xFails for edison M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: NOTE: Testing was done on tag port_to_edison_02_clm4_5_34. Since then, scripts has been updated from scripts4_130930a to scripts4_131001. However, the only difference is the removal of some duplicated tests from the test list. regular tests (aux_clm): edison_intel (aux_clm_ys_intel & aux_clm_ys_pgi lists): OK All pass except: See "ERH_D.f19_g16.I1850CLM45CN.yellowstone_pgi.clm-default Can't find user datasets Seg fault while writing h1 file See ERS_Ld211_D_P112x1.f10_f10.ICNCROP in yellowstone intel list #1 and #4 have been failing on yellowstone, #2 and #3 are new failures on edison frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: NOTE: only generate done here, because no baselines existed edison_intel: aux_clm_ys_intel list: OK edison_intel: aux_clm_ys_pgi list: OK CLM tag used for the baseline comparisons: clm4_5_34 NOTE: Baseline comparisons only done for frankfurt tests; no baseline comparisons done with component_gen_comp, because no baselines existed on edison. Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_34 Originator(s): erik (Erik Kluzek) Date: Mon Sep 30 01:30:25 MDT 2013 One-line Summary: Get PTCLM working, fix a few small bugs Purpose of changes: Get PTCLM fully working. Allow PTCLM to work with ALL I compsets including clm4_0 compsets. Add a new option to PTCLM from Keith Oleson --cycle_forcing to set it up to cycle over the forcing. Another option is to build datasets in the "-mydatafiles" directory (by default under PTCLM directory). The datasets now have creation time-stamps in them as well. Rename QIAN_tower_yrs to -use_tower_yrs and remove QIANforcing (now chosen by compset). scripts now has four different I1PT compsets two new ones for CLM40CN and CLM45BGC. datm CLM_USRDAT domain file for CLM1PT forcing points to the ATM_DOMAIN_FILE/PATH. Add some new datasets to the siteDIR from Keith Oleson. Update documenation, remove unused template dir. Add a script to rename creation dates for map files, so you don't have to regenerate them each day. Fix a few small bugs. Allow clm4_5 to have suplnitro and bgc_spinup to only give a warning rather than die. Fix a corrupted rawdata PFT file. Fix mkscripgrid.ncl for regional SCRIP grid creation. Remove some leftover fine-mesh variables that aren't needed anymore. Remove reference to scaled_harvest in CLM build-namelist which was removed a long time ago. Remove a mapping file that didn't seem to be needed for clm4_0 mkmapdata (a default didn't exist for it either). Add -usr_mapdir option to clm4_0 mksurfdata.pl which is needed for PTCLM for clm4_0 compsets. Requirements for tag: Update scripts and get PTCLM working Test level of tag: regular Bugs fixed (include bugzilla ID): 1826 (allow clm4_5 and suplnitro to be ALL for bgc_spinup) 1818 (two new options to PTCLM) 1762 (Fix corrupted rawdata PFT file) 1757 (Bug in mkscripgrid.ncl for regional/global SCRIP grid creation) 1623 (Remove some leftover fine-mesh variables _a arrays) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Allow two options to go without dying List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts, Machines, datm, and csm_share scripts to scripts4_130929 Machines to Machines_130927 datm to datm8_130919 csm_share to share3_130918 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/tools/shared/mkmapdata/mkmapdata.sh ------- Remove 10x10min_IGBPmergeICESatGIS for clm4_0 M models/lnd/clm/tools/shared/mkmapgrids/mkscripgrid.ncl --- Fix bug 1757 for regional grid creation M models/lnd/clm/tools/clm4_0/mksurfdata_map/mksurfdata.pl - Add usr_mapdir option M models/lnd/clm/bld/build-namelist - Allow missing clm_usrdat files to continue, remove scaled_harvest (long gone) allow bgc_spinup and suplnitro to coexist with warning M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0_tools.xml - Fix corrupted 856 raw PFT file M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5_tools.xml - Fix corrupted 856 raw PFT file, remove 1000-1004 testyrs M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 - Remove finemesh _a variables M models/lnd/clm/src/clm4_5/main/clmtype.F90 -------- Remove finemesh _a variables M models/lnd/clm/src/clm4_0/main/clmtypeInitMod.F90 - Remove finemesh _a variables M models/lnd/clm/src/clm4_0/main/clmtype.F90 -------- Remove finemesh _a variables CLM testing: regular, build_namelist, tools build-namelist tests: yellowstone yes frankfurt yes regular tests (aux_clm): yellowstone_intel yes yellowstone_pgi yes frankfurt_intel yes frankfurt_pgi yes frankfurt_nag yes tools testing: yellowstone interactive yes CLM tag used for the baseline comparisons: clm4_5_33 Changes answers relative to baseline: no bit-for-bit =============================================================== =============================================================== Tag name: clm4_5_33 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Thu Sep 26 10:42:56 MDT 2013 One-line Summary: clean up from mistakes in previous tag Purpose of changes: clean up time-stamps and a mistake in clm4_5_32 Requirements for tag: N/A Test level of tag: N/A Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: N/A List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: N/A CLM testing: N/A CLM tag used for the baseline comparisons: N/A Changes answers relative to baseline: N/A =============================================================== =============================================================== Tag name: clm4_5_32 Originator(s): muszala (Stefan Muszala,UCAR/TSS,303-497-1320) Date: Thu Sep 26 10:07:14 MDT 2013 One-line Summary: bug fix tag - 1798, 1810 Purpose of changes: fix bug 1798 and 1810. http://bugs.cgd.ucar.edu/show_bug.cgi?id=1798 http://bugs.cgd.ucar.edu/show_bug.cgi?id=1810 Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID): 1798 and 1810 Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: - for bug 1798 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 - for bug 1810 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_0/main/controlMod.F90 CLM testing: regular tests (aux_clm): yellowstone_intel - OK yellowstone_pgi - OK frankfurt_intel - OK frankfurt_pgi - OK frankfurt_nag - OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK CLM tag used for the baseline comparisons: clm4_5_31 Changes answers relative to baseline: no =============================================================== =============================================================== Tag name: clm4_5_31 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Wed Sep 25 10:12:47 MDT 2013 One-line Summary: fix bug 1820: incomplete conditional in CNSoyfix leads to buggy results and decomposition dependence Purpose of changes: Fix bug 1820: incomplete conditional in CNSoyfix leads to buggy results and decomposition dependence. Fix for this is based on analysis of the original Agro-IBIS code. Requirements for tag: fix bug 1820, the following tests should now pass: PET_P15x2_Lm25.f10_f10.ICLM45BGCDVCROP.yellowstone_intel PET_P15x2_Ly3.f10_f10.ICLM45BGCCROP.yellowstone_intel Test level of tag: regular Bugs fixed (include bugzilla ID): 1820: incomplete conditional in CNSoyfix leads to buggy results and decomposition dependence Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: Beth Drewniak, Sam Levis List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 ========= Remove now-passing PET tests M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK CLM tag used for the baseline comparisons: clm4_5_30 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 with CROP - what platforms/compilers: ALL - nature of change (roundoff; larger than roundoff/same climate; new climate): larger than roundoff, but not investigated in detail Note that no changes were observed in the test suite, but this is due to a limitation of the test suite (there are very few multi-year crop tests; the only global multi-year tests are the newly-passing PET tests, which don't have baselines) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_30 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Tue Sep 24 13:08:01 MDT 2013 One-line Summary: fix performance bug in decomposition initialization Purpose of changes: Fix performance bug in decomposition initialization (bug 1771). Code mods from Tony Craig. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1771 ( Fix for an initialization performance bug) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: Improves timing of initialization for high resolution casse Code reviewed by: tcraig List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/main/decompInitMod.F90 M models/lnd/clm/src/clm4_0/main/decompInitMod.F90 CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK CLM tag used for the baseline comparisons: clm4_5_29 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_29 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Tue Sep 24 10:55:42 MDT 2013 One-line Summary: fix threading in CLM4.5, and other misc fixes Purpose of changes: (Mostly) fix threading in CLM4.5. This consisted of: (1) Rework initGridCellsMod to keep all points in a clump contiguous (2) Add info in bounds derived type (not necessary, but this allows for more error checking and simplifies some code) (3) Fix which bounds are passed to reweightWrapup in initialization (4) Get rid of syntax like foo(:) = 0, instead using explicit bounds (5) Rework bounds declarations for subroutine array arguments, both in caller (explicitly subset argument by bounds) and callee (use assumed-shape array arguments rather than declaring upper bounds), and add assertions on array sizes. See https://wiki.ucar.edu/display/ccsm/Community+Land+Model+Developers+Guide ("Guidelines for passing array arguments to subroutines") for the new conventions that are implemented here. (6) Fix crop threading bug, related to nyrs (bug 1598), both in clm4.5 and clm4.0 However, note that there is still a crop threading bug (bug 1820), which will need to be fixed in a separate tag. Also, some unrelated changes: (1) Fix size of a megan variable, both in clm4.5 and clm4.0. (2) Remove some unused variables from Hydrology2Mod / SoilHydrologyMod (3) Fix some bugs in histFileMod / histFldsMod (4) Reorder a loop in SurfaceAlbedo to get better performance (especially with expanded memory allocation for dynamic landunits, in an upcoming tag) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1598 (crop threading in clm4.0 and clm4.5) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: Added f10 initial file for CLM4.5 BGCCROP, for testing purposes Describe any substantial timing or memory changes: Timing is currently 5-10% worse, due to calls to shr_log_errMsg, within shr_assert calls. This should return to previous timings in non-debug runs once shr_assert calls are ifdef'ed out in non-debug runs (this requires a csm_share update that Sean Santos is working on). Code reviewed by: portions reviewed by erik List any svn externals directories updated (csm_share, mct, etc.): scripts4_130912 -> scripts4_130916 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Significant rework to work with multiple clumps per proc M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 ========= Add info in bounds derived type M models/lnd/clm/src/clm4_5/main/decompMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 - also fix crop threading bug (1598) M models/lnd/clm/src/clm4_5/main/reweightMod.F90 - also use explicit bounds instead of things like foo(:) ========= Remove some unused variables M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 - also other bounds-related changes as above ========= Change 'bounds' to 'bounds_proc', use clump bounds for call to reweightWrapup, get rid of abort if running with openMP M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ========= Get rid of syntax like foo(:), instead using explicit bounds M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 ========= Rework bounds declarations for subroutine array arguments, both in caller (explicitly subset argument by bounds) and callee (use assumed-shape array arguments rather than declaring upper bounds), and add assertions on array sizes M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/main/dynlandMod.F90 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/main/pft2colMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 - also remove some now-unneeded temporary arrays - also use explicit bounds instead of things like foo(:) M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 - also reorder a loop to get better performance M models/lnd/clm/src/clm4_5/biogeophys/BandDiagonalMod.F90 ========= Fix crop threading bug (1598), in both clm4.5 and clm4.0, by reworking where nyrs is updated M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_0/main/clm_driver.F90 ========= Add comments M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 ========= Fix size of a megan variable M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 - also initialize rootr, as it was in clm4_5_20 and prior M models/lnd/clm/src/clm4_0/main/clmtypeInitMod.F90 ========= Fix some hist file bugs: - increase max number of characters allowed for hist field names - when adding a field, make it work to say default='active' -- previously, explicitly setting default='active' did the same thing as setting default='inactive' - change ptr_pft to ptr_col for a few column-level history variables - remove two duplicate hist_addfld calls M models/lnd/clm/src/clm4_5/main/histFileMod.F90 - also add some bounds to array arguments M models/lnd/clm/src/clm4_5/main/histFileMod.F90 ========= Add f10 initial file for CLM45 BGCCROP, for testing M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Remove some PET tests from the xFail list; note that some PET tests still fail due to bug 1820 M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK CLM tag used for the baseline comparisons: clm4_5_28 Changes answers relative to baseline: NO, with the following minor exceptions: In general, no answer changes for non-threaded runs (changes answers for threaded runs due to significant bug fixes!) Changes answers for CLM45 BGC CROP at f10 due to new initial conditions (instead of cold start) =============================================================== =============================================================== Tag name: clm4_5_28 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Sep 20 21:29:39 MDT 2013 One-line Summary: fix FracH2oSfc bug Purpose of changes: Fix bug 1811: FracH2oSfc is called from within a loop over all points. Sean Swenson realized that the offending block of code is no longer needed, so we have removed it. Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1811 (FracH2oSfc is called from within a loop over all points) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: Sean Swenson List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Remove the offending (and no longer needed) block of code M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 ========= Remove a now-unused variable M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 ========= Move xFail test to the right location M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK yellowstone_pgi: OK See notes below on answer changes CLM tag used for the baseline comparisons: clm4_5_27 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45 cases - what platforms/compilers: All - nature of change: larger than roundoff/same climate If bitwise differences were observed, how did you show they were no worse than roundoff? Sean Swenson performed two 20-year runs with and without the fix. His report is: There are some differences, mainly in runoff, and it looks like at the 1% level mostly, with a few scattered points showing up on the significance plots. Other fields like latent heat, soil moisture/temperature, or water table show even less differences. http://www.cgd.ucar.edu/staff/swensosc/public/diagnostics/test_frach2o-test_no_frach2o/setsIndex.html =============================================================== =============================================================== Tag name: clm4_5_27 Originator(s): sacks (Bill Sacks,UCAR/CSEG,303-497-1762) Date: Fri Sep 20 20:43:16 MDT 2013 One-line Summary: fix crop nyrs bug Purpose of changes: Fix bug 1815 (nyrs is incorrect at the start of a crop run, leading to incorrect GDD values for the first 20 years or so of a crop simulation) Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): - 1815 (nyrs is incorrect at the start of a crop run, leading to incorrect GDD values for the first 20 years or so of a crop simulation) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None (but see changes to boundary datasets below) List any changes to the defaults for the boundary datasets: New crop initial conditions for CLM4.5 BGCCROP @ f19 - same as old dataset, but with restyear changed from 1 to 0 Describe any substantial timing or memory changes: None Code reviewed by: slevis List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Don't increment nyrs on the first timestep of a startup run, so that nyrs is correctly 0 rather than 1 for the first year M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNPhenologyMod.F90 ========= New crop initial conditions for CLM4.5 BGCCROP @ f19 - same as old dataset, but with restyear changed from 1 to 0 M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= add test that sometimes runs out of time, move test from yellowstone_intel to yellowstone_pgi M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml CLM testing: build-namelist tests: yellowstone: OK regular tests (aux_clm): yellowstone_intel: OK yellowstone_pgi: OK frankfurt_intel: OK frankfurt_pgi: OK frankfurt_nag: OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: OK - expected diffs in SMS_Ly1.f19_g16.ICLM45BGCCROP.yellowstone_intel - baselines messed up for ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel, so comparison not done for that test yellowstone_pgi: NO BASELINES, SO COMPARISONS NOT RUN CLM tag used for the baseline comparisons: clm4_5_27 Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All crop cases (clm4.0 or clm4.5) that either use arbitrary initial conditions or use the clm4.5 out-of-the-box initial conditions for BGCCROP @ f19 - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) Sam Levis examined a 3-year run with arbitrary initial conditions, and verified that crop LAI is much more realistic in the new simulation (previously, crop LAI was near-zero for the first few years) =============================================================== =============================================================== Tag name: clm4_5_26 Originator(s): muszala (Stefan Muszala) Date: Thu Sep 19 17:07:11 MDT 2013 One-line Summary: water balance and SMS_Ly1.f19_g16.ICLM45BGCCROP fix Purpose of changes: 1) Fix water balance error in f09_g16 I1850CRUCLM45BGC simulation 2) Get all machine/compiler combinations of SMS_Ly1.f19_g16.ICLM45BGCCROP working - this fix required (1) and a fix to fthresh in RtmFloodInit - new RTM tag rtm1_0_32 to go along with this - PGI+frankfurt version of this test only work with 16 MPI processes Requirements for tag: N/A Test level of tag: regular Bugs fixed (include bugzilla ID):1808 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: S. Swenson, D. Lawrence List any svn externals directories updated (csm_share, mct, etc.): rtm1_0_32 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 Index: models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 =================================================================== --- models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 (revision 51190) +++ models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 (working copy) @@ -1110,15 +1110,8 @@ do j = 1, nlevsoi if(h2osoi_liq(c,j)<0._r8)then qflx_deficit(c) = qflx_deficit(c) - h2osoi_liq(c,j) - h2osoi_liq(c,j) = 0._r8 endif enddo - !reduce qcharge if necessary - !ideally, I can set qflx_deficit as a local variable, but it is helpful - !to diagnose the problem associated with the solver for the richards' equation. - if(qflx_deficit(c)>0._r8)then - qcharge(c) = qcharge(c) - qflx_deficit(c)/dtime - endif enddo end associate @@ -1892,9 +1885,12 @@ do fc = 1, num_hydrologyc c = filter_hydrologyc(fc) - xs1(c) = max(max(h2osoi_liq(c,1),0._r8)-max(0._r8,(pondmx+watsat(c,1)*dzmm(c,1)-h2osoi_ice(c,1))),0._r8) - h2osoi_liq(c,1) = min(max(0._r8,pondmx+watsat(c,1)*dzmm(c,1)-h2osoi_ice(c,1)), h2osoi_liq(c,1)) + !scs: watmin addition to fix water balance errors + xs1(c) = max(max(h2osoi_liq(c,1)-watmin,0._r8)- & + max(0._r8,(pondmx+watsat(c,1)*dzmm(c,1)-h2osoi_ice(c,1)-watmin)),0._r8) + h2osoi_liq(c,1) = h2osoi_liq(c,1) - xs1(c) + if (urbpoi(clandunit(c))) then qflx_rsub_sat(c) = xs1(c) / dtime else CLM testing: - general note: for clm45 compsets-both clm and cpl history files change - specfic testing for these bug fixes: 1) Water balance fix-ran a clone of run from Dave Lawrence: -- create_newcase -compset I1850CRUCLM45BGC -res f09_g16 -mach yellowstone -case /glade/u/home/dlawren/expts/clm4.5/clm45bgc_1deg4519_1850spin_bd -- this ran for over 25 years with no water balance errors. 2) SMS_Ly1.f19_g16.ICLM45BGCCROP PASS SMS_Ly1.f19_g16.ICLM45BGCCROP.yellowstone_intel.clm-reduceOutput.115612 PASS SMS_Ly1.f19_g16.ICLM45BGCCROP.yellowstone_pgi.clm-reduceOutput.115522 PASS SMS_Ly1.f19_g16.ICLM45BGCCROP.frankfurt_intel.clm-reduceOutput.115217 PASS SMS_Ly1_Mmpich.f19_g16.ICLM45BGCCROP.frankfurt_nag.clm-reduceOutput.120824 PASS SMS_Ly1.f19_g16.ICLM45BGCCROP.frankfurt_pgi.clm-reduceOutput.016 -- only with 16 MPI tasks 3) To make sure the RTM refactor did not create any BFB changes, I ran SMS_D.f19_g16.ICLM45BGCCROP.yellowstone_intel with rtm1_0_31 and rivrtm/branch_tags/bcf_tags/bcf_02_rtm1_0_31. - With flood_mode='NULL' - Coupler and land history files were BFB. - With flood_mode='ACTIVE' - Coupler, rtm and land history files were BFB. build-namelist tests: N/A regular tests (aux_clm): yellowstone_intel - OK - changes in cpl. hist files expected yellowstone_pgi - OK - changes in cpl. hist files expected frankfurt_intel - OK - changes in cpl. hist files expected frankfurt_pgi - OK - changes in cpl. hist files expected frankfurt_nag - OK - changes in cpl. hist files expected history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel - OK - changes in clm fields expected. CLM tag used for the baseline comparisons: clm4_5_25 Changes answers relative to baseline: Yes. For CLM45 compsets. All changes are from the SoilHydrology mods as the RTM refactor and fthresh fix are BFB. Coupler history fields that change: l2x_Sl_avsdr, l2x_Sl_anidr, l2x_Sl_avsdf, l2x_Sl_anidf, l2x_Sl_tref, l2x_Sl_qref, l2x_Sl_t, l2x_Sl_fv, l2x_Sl_ram1, l2x_Sl_snowh, l2x_Sl_u10, l2x_Fall_swnet, l2x_Fall_taux, l2x_Fall_tauy, l2x_Fall_lat, l2x_Fall_sen, l2x_Fall_lwup, l2x_Fall_evap, l2x_Fall_flxdst1, l2x_Fall_flxdst2, l2x_Fall_flxdst3, l2x_Fall_flxdst4, l2x_Flrl_rofl, l2x_Fall_voc001, l2x_Fall_voc002, l2x_Fall_voc003, l2x_Fall_voc004, l2x_Fall_voc005, l2x_Fall_voc006, l2x_Fall_voc007, l2x_Fall_voc008, x2l_Flrr_volr, r2x_Forr_rofl, r2x_Forr_rofi, r2x_Flrr_volr, x2r_Flrl_rofl If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: CLM45 compsets - what platforms/compilers: All =============================================================== =============================================================== Tag name: clm4_5_25 Originator(s): erik (Erik Kluzek) Date: Fri Sep 13 13:49:45 MDT 2013 One-line Summary: Bring in Tony's changes to kick sno all the way up to the coupler layer, makes all CESM components more similar to each other Purpose of changes: Bring in Tony's cplupa branch (cplupa_n06_clm4_5_24) to trunk. This branch moves sno fields all the way to the top coupler layer rather than being inside of CLM. This makes all CESM components more similar to each other. There was also some small fixes on the side that allow some more tests to work. Test level of tag: regular Bugs fixed (include bugzilla ID): 1789 (fix NoVSNoNI test) 1788 (fix US-UMB test) 1779 (fix RTM multi-instance) 1777 (fix RTM branch cases) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: Add new CPP token to CLM USE_ESMF_METADATA (with no direct machanism to use) We hate CPP tokens (but had to let this go, only needed by ESMF development folks) Describe any changes made to the namelist: drv namelist changes, no changes to CLM namelist List any changes to the defaults for the boundary datasets: CLM_USRDAT fsurdat files different directory for clm4_0 than clm4_5 remove missing ne16np4 fpftdyn file Describe any substantial timing or memory changes: Code reviewed by: self, tcraig List any svn externals directories updated (csm_share, mct, etc.):, scripts, drv, cism, rtm, csm_share, data and stub models scripts to scripts4_130912 drv to drvseq4_3_03 datm to datm8_130424 socn/sice/sglc/swav to stubs1_4_02 rtm to rtm1_0_31 cism to cism1_130905 csm_share to share3_130906 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/build-namelist_test.pl ------------ Remove ne16 20thC test M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml ---- Remove missing ne16 fpftdyn file M models/lnd/clm/bld/namelist_files/namelist_defaults_usr_files.xml - Different directory for clm4_0/clm4_5 surface datasets M models/lnd/clm/src/cpl_share/clm_cpl_indices.F90 - Field names change M models/lnd/clm/src/cpl_mct/lnd_comp_mct.F90 ------ Remove "sno" and "s" fields M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 ---- Remove rofi/rofl M models/lnd/clm/src/cpl_esmf/lnd_comp_esmf.F90 ---- Remove "sno" and "s" fields, add USE_ESMF_METADATA #ifdef M models/lnd/clm/src/cpl_esmf/lnd_comp_mct.F90 ----- Remove "sno" and "s" fields, add USE_ESMF_METADATA #ifdef M models/lnd/clm/src/clm4_0/main/clm_glclnd.F90 ---- Remove rofi/rofl CLM testing: regular build-namelist tests: bluefire yes frankfurt yes regular tests (aux_clm): yellowstone_intel yes yellowstone_pgi yes frankfurt_intel yes frankfurt_pgi yes frankfurt_nag yes history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel partial (not completed before tag made) yellowstone_pgi partial (not completed before tag made) CLM tag used for the baseline comparisons: clm4_5_24 Changes answers relative to baseline: None (bit-for-bit) Although my "I compset" testing showed no changes, fully coupled changes do show differences. Coupler namelists also change. =============================================================== =============================================================== Tag name: clm4_5_24 Originator(s): sacks (sacks) Date: Tue Sep 3 21:36:13 MDT 2013 One-line Summary: update externals to cesm1_3_beta02 or later Purpose of changes: Update externals to cesm1_3_beta02 or later Requirements for tag: Test level of tag: regular Bugs fixed (include bugzilla ID): 1722 (Test failure with VIC and more_vertlayers) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: Not investigated Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): scripts : scripts4_130816 -> scripts4_130830a Machines : Machines_130529 -> Machines_130830 drv : drvseq4_2_33 -> drvseq4_2_35 cism : cism1_130502 -> cism1_130624 csm_share : share3_130528 -> share3_130723 timing : timing_130417 -> timing_130506 mct : compiler_fixes_n01_MCT_2.8.3 -> compiler_fixes_n03_MCT_2.8.3 mapping : mapping_130509 -> mapping_130716 gen_domain : mapping_130509/gen_domain_files -> mapping_130716/gen_domain_files List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Remove trailing whitespace M .ChangeLog_template Index: models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml =================================================================== --- models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml (revision 50759) +++ models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml (working copy) @@ -51,17 +51,14 @@ use before define in history. runs with frankfurt_pgi, yellowstone_intel and yellowstone_pgi - floating point exception. problem with VIC combined with vertical layers once threading fixed, this should pass once threading fixed, this should pass - core dumps in ref1. Problems with vertical layers. Run without clm-vrtlay and clm-default and it runs starting in clm4_5_07--The cpl.hi.nc file is not being copied on a generate like it should. Cannot turn clm4me mode on -- without clm4_5 physics! problem in scripts4_130809b - Problem with scripts and testId string length. This passes with a long testId + Problem with scripts and testId string length. This passes with a long testId Restart not BFB. Runs as ERH_D.f19_g16.I1850CLM45CN.yellowstone_pgi - floating point exception. problem with VIC combined with vertical layers once threading fixed, this should pass once threading fixed, this should pass once threading fixed, this should pass @@ -78,7 +75,7 @@ Water balance errors followed by "negative conc. in ch4tran", then tries "-10^-12 < smin_nh4 < 0. resetting to zero.", then it exits at approximately 9 months. This same test passes with yellowstone_intel and frankfurt_intel - CMake 2.8.6 or higher is required. You are running version 2.6.4 + Problem with cism build Fails after reading clmforc.Li_2012_hdm_0.5x0.5_AVHRR_simyr1850-2010_c130401.nc. This same test passes with yellowstone_intel, yellowstone_pgi, frankfurt_intel and frankfurt_nag Fails after reading clmforc.Li_2012_hdm_0.5x0.5_AVHRR_simyr1850-2010_c130401.nc. This same test passes with yellowstone_intel, yellowstone_pgi, frankfurt_intel and frankfurt_nag Fails after reading clmforc.Li_2012_hdm_0.5x0.5_AVHRR_simyr1850-2010_c130401.nc. This same test passes with yellowstone_intel CLM testing: build-namelist tests: yellowstone: YES All PASS or xFAIL regular tests (aux_clm): yellowstone_intel: YES yellowstone_pgi: YES frankfurt_intel: YES frankfurt_pgi: YES frankfurt_nag: YES All PASS or xFAIL history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel: YES All PASS except for the following newly-passing tests (for which these failures are unsurprising): BFAIL2 ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_23.clm2.h0 (baseline history file does not exist) BFAIL2 ERI_D.ne30_g16.ICLM45BGC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_23.clm2.h1 (baseline history file does not exist) FAIL ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_23.clm2.h0 FAIL ERS_D.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_23.clm2.h1 yellowstone_pgi: YES Some answer changes (presumably due to compiler change) CLM tag used for the baseline comparisons: clm4_5_23 Note: renamed baselines for frankfurt nag because test names have changed Changes answers relative to baseline: YES If a tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: - what platforms/compilers: - nature of change (roundoff; larger than roundoff/same climate; new climate): compsets using cism change answers for g2x gields many tests change answers with pgi, both on yellowstone & frankfurt, presumably due to new compiler nature of change not investigated If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: N/A URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_23 Originator(s): muszala (Stefan Muszala) Date: Thu Aug 22 09:42:43 MDT 2013 One-line Summary: refactor to allow CH4 params. to be read from netcdf file and clean up clm4_5_20 Purpose of changes: The second of two tags that brings in parameters that are read from netcdf file (ch4 parameters). Please see the ChangeLog entry for clm4_5_20. - Bring in ch4 parameters - Combine fconsts file and fpftcon file. New file name is paramfile (clm_params.c130821.nc) - Refactor so that types, subroutine names and type instances have the names params in them (instead of consts) - Remove many ch4 namelist vars. since they are now read from the param file - Add new namelist called use_aereoxid_prog to control old aereoxid namelist A bulk of this work was completed by Rajendra Paudel. Requirements for tag: N/A Test level of tag: regular and build_namelist Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: remove many optional ch4 namelists. add new namelist valled use_aereoxid_prog. List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, dlawren, Rajendra Paudel List any svn externals directories updated (csm_share, mct, etc.): scripts4_130730 -> scripts4_130816 List all files eliminated: - these were renamed models/lnd/clm/src/clm4_5/biogeochem/CNSharedConstsMod.F90 models/lnd/clm/src/clm4_5/main/readConstantsMod.F90 List all files added and what they do: - renamed models/lnd/clm/src/clm4_5/biogeochem/CNSharedParamsMod.F90 models/lnd/clm/src/clm4_5/main/readParamsMod.F90 List all existing files that have been modified, and describe the changes: - refactor to remove old namelist vars. for ch4 and add new functionality M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml M models/lnd/clm/bld/build-namelist M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - put in ch4 parameters and refactor const->params names M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4varcon.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/main/initSoilParVICMod.F90 M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 M models/lnd/clm/src/clm4_5/biogeophys/initSLakeMod.F90 CLM testing: build-namelist tests: yellowstone: OK. Some changes to phys45 and phys45-crop. Should be OK in next tag. regular tests (aux_clm): A few nlcomp differences that will go away in the next tag. Other than a few expected failures due to new scripts entries, all OK. yellowstone_intel OK yellowstone_pgi OK frankfurt_intel OK frankfurt_pgi OK frankfurt_nag OK history file comparison with component_gen_comp and summarize_cprnc_diffs: yellowstone_intel OK yellowstone_pgi OK CLM tag used for the baseline comparisons: clm4_5_22 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_22 Originator(s): muszala (Stefan Muszala) Date: Tue Jul 30 15:22:51 MDT 2013 One-line Summary: aux_clm testlist reorganization Purpose of changes: Reorganize all aux_clm tests and fix new failing tests. As part of this, bring in a change from Maoyi for VIC w/vertical layers. A few bug fixes to get new tests working. Why did we do this? 1) better balance between frankfurt, yellowstone and various compilers 2) faster turn around time for development 3) make sure current science functionality is properly tested 4) removed outdated / irrelevant tests Other points: 1) Introduce regular and short test list. Testing now can consist of: a) regular (must be run before handing off a tag to SEs and must be run before committing a tag) b) build_namelist (if namelists and/or build_system changed)) c) tools (only if tools are modified and no CLM source is modified) d) short (for use during development and in rare cases where only a small change with known behavior is added ... eg. a minor bug fix) e) doc (no source testing required) 2) PET tests will fail until threading is fixed in CLM Requirements for tag: N/A Test level of tag: regular. ran tests with old testlists to double check any new mods. Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: add no-vert:no-nitrif option to configure Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self. Extensive discussion regaring list content with Lawrence, Sacks, Kluzek and Andre. List any svn externals directories updated (csm_share, mct, etc.): new scripts List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: - add no-vert:no-nitrif option M models/lnd/clm/bld/configure M models/lnd/clm/bld/config_files/config_definition.xml - Update expected failures M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml - Maoyi VIC+vertical layers fix M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 - Update change log template M .ChangeLog_template - Change intent out to inout for ciso_flux M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 - Remove crop_prog check to get rid of unassociated pointer with NAG M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 - Fix bounds type error (should be intent=in) M models/lnd/clm/src/clm4_5/main/histFileMod.F90 Machines testing ran on: (Tests in priority order) Note: 1) All PET tests will fail until openMP is fixed. 2) All IG compsets fail with the NAG compiler due to non-compliant f77 code. 3) If one combination fails, it's backed up with a another combination that passes. There are one or two exceptions to this. I) aux_clm tests with old testlists: yellowstone/aux_clm intel - OK yellowstone/aux_clm pgi - OK frankfurt/aux_clm intel - OK frankfurt/aux_clm pgi - OK frankfurt/aux_clm nag - OK CESM history file comparison: yellowstone/aux_clm intel - OK II) aux_clm tests with new testlists: yellowstone/aux_clm intel - OK yellowstone/aux_clm pgi - OK frankfurt/aux_clm intel - OK frankfurt/aux_clm pgi - OK frankfurt/aux_clm nag - OK III) aux_clm_short tests with new testlists: yellowstone/aux_clm intel - OK yellowstone/aux_clm pgi - OK frankfurt/aux_clm intel - OK frankfurt/aux_clm pgi - OK frankfurt/aux_clm nag - OK CESM history file comparison: Not run since no baseline comparisons. CLM tag used for the baseline comparison tests if applicable: CLM4_5_21 with old testlist. Only ran generate with new testlists. Changes answers relative to baseline: Only for VIC with vertical layers. =============================================================== =============================================================== Tag name: clm4_5_21 Originator(s): muszala (Stefan Muszala) Date: Wed Jul 24 14:23:19 MDT 2013 One-line Summary: ifdef and bounds refactor Purpose of changes: - Almost all implementation by Mvertens - Refactor ifdef use so that a majority are now in controlMod.F90. This is the first step to removing them competely. - Introduction of bounds_type and clump_type - Refactor interfaces to support bounds_type and clump_type - Bug fix from Sacks Also: Changed layout of landunit, column and pft-level arrays: Previously, all points for a given grid cell were grouped together. Now, all points for a given landunit type are grouped together. This improves performance of loops over filters, because it leads to more memory locality – this will be especially true when we add more 0-weight points to arrays for the purpose of dynamic landunits. For example, if a processor has 2 grid cells and there are 3 landunit types: Old layout in memory: (G1L1, G1L2, G1L3, G2L1, G2L2, G2L3) New layout in memory: (G1L1, G2L1, G1L2, G2L2, G1L3, G2L3) Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: mvertens, sacks, self List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: D models/lnd/clm/src/util_share/decompInitMod.F90 D models/lnd/clm/src/util_share/ndepStreamMod.F90 D models/lnd/clm/src/util_share/decompMod.F90 D models/lnd/clm/src/clm4_5/main/initParametersMod.F90 List all files added and what they do: A + models/lnd/clm/src/clm4_5/main/decompInitMod.F90 A + models/lnd/clm/src/clm4_5/main/ndepStreamMod.F90 A + models/lnd/clm/src/clm4_5/main/decompMod.F90 A + models/lnd/clm/src/clm4_0/main/decompInitMod.F90 A + models/lnd/clm/src/clm4_0/main/ndepStreamMod.F90 A + models/lnd/clm/src/clm4_0/main/decompMod.F90 List all existing files that have been modified, and describe the changes: M models/lnd/clm/doc/ChangeLog M models/lnd/clm/doc/ChangeSum M models/lnd/clm/src/cpl_share/clm_cpl_indices.F90 M models/lnd/clm/src/cpl_mct/lnd_comp_mct.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/initch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/MEGANFactorsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4RestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNWoodProductsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVLightMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4varcon.F90 M models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEcosystemDynIniMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVegStructUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSharedConstsMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 M models/lnd/clm/src/clm4_5/main/CNiniTimeVar.F90 M models/lnd/clm/src/clm4_5/main/dynlandMod.F90 M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/subgridMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/pft2colMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/clm_varsur.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/initSoilParVICMod.F90 M models/lnd/clm/src/clm4_5/main/CNiniSpecial.F90 M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/main/readConstantsMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 M models/lnd/clm/src/clm4_5/main/mkarbinitMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/QSatMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CLMVICMapMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/initSLakeMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FracWetMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BandDiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/cpl_esmf/lnd_comp_esmf.F90 M models/lnd/clm/src/clm4_0/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_0/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_0/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_0/main/controlMod.F90 M models/lnd/clm/src/clm4_0/main/clm_varctl.F90 M models/lnd/clm/src/clm4_0/main/clm_driver.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: Not run CESM test lists: yellowstone/aux_clm intel BFB yellowstone/aux_clm pgi BFB frankfurt/aux_clm intel BFB frankfurt/aux_clm pgi BFB frankfurt/aux_clm nag BFB CESM history file comparison: yellowstone/aux_clm intel BFB CLM tag used for the baseline comparison tests if applicable: clm4_5_20 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_20 Originator(s): muszala (Stefan Muszala,UCAR/CGD,303-497-1320) Date: Sat Jul 20 10:54:43 MDT 2013 One-line Summary: refactor to allow CN and BGC params. to be read from netcdf file Purpose of changes: The first of two tags that allows all parameters to be read from a netcdf file to provide sensitivity analysis capability, to increase modularity of code and to remove "magic numbers" from code. This tag introduces a new namelist variable "fconsts" which points to a netcdf file of CN and BGC parameters. In a future tag, this netcdf file will be combined with CH4 parameters and PFT parameters. Values are read in readConstantsMod.F90. Each module that requires a parameter provides a read subroutine. That read subroutine is called from readConstantsMod.F90 and places parameters into a private type for that module. For example, CNDecompMod.F90 provides readCNDecompConsts which is called from readConstantsMod and populates the type instance CNConstShareInst. CHConstShareInst is then used in CNDecompMod as: + sminn_to_denit_decomp_cascade_vr(c,j,k) = -CNDecompConstInst%dnp * pmnf_decomp_cascade(c,j,k which replaces: - dnp = 0.01_r8 ... - sminn_to_denit_decomp_cascade_vr(c,j,k) = -dnp * pmnf_decomp_cascade(c,j,k) A bulk of this work was completed by Rajendra Paudel. Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: new namelist variable called fconsts. List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, D. Lawrence, R. Paudel. (for design: discussion w/ mvertens, sacks, kluzek) List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: Renamed D models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_BGC.F90 D models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_CENTURY.F90 List all files added and what they do: Rename of Deleted files A + models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeBGCMod.F90 A + models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeCNMod.F90 Module to read parameters shared by multiple modules A + models/lnd/clm/src/clm4_5/biogeochem/CNSharedConstsMod.F90 Module that reads shared an private parameters A + models/lnd/clm/src/clm4_5/main/readConstantsMod.F90 List all existing files that have been modified, and describe the changes: Add fconsts namelist variable M models/lnd/clm/bld/build-namelist M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml Add in functionality to read parameters off of netcdf file M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: OK. These FAILs should pass next round: 418/444 < FAIL> 423/444 < FAIL> 428/444 < FAIL> 433/444 < FAIL> 438/444 < FAIL> 443/444 < FAIL> CESM test lists: CLM45 compsets have failures for nlcomp due to the introduction of fconsts namelist variable. For example: FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.test20Pgi.nlcomp FAIL ERH_D.f19_g16.I1850CLM45CN.yellowstone_intel.GC.test20Intel.nlcomp yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CLM history file comparison: yellowstone/aux_clm intel OK CLM tag used for the baseline comparison tests if applicable: clm4_5_19 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_19 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Wed Jul 17 14:29:32 MDT 2013 One-line Summary: fix setting of bd in iniTimeConst Purpose of changes: In iniTimeConst, bd (bulk density) was being set incorrectly, so that, for a given processor, the same value was being put in all (c,j) locations. In addition to being incorrect, this meant that results differed depending on processor count. This tag fixes this problem. This only affects CLM4.5 BGC runs, because the bd array is only used in CNNitrifDenitrifMod.F90. (However, as a side note: This array SHOULD be used in DUSTMod and initSLakeMod, which currently recompute bd.) Requirements for tag: Test level of tag: Bugs fixed (include bugzilla ID): 1736 (bd set incorrectly in iniTimeConst, leads to results that depend on processor count) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: mvertens List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: NO CESM test lists: yellowstone/aux_clm intel yes *** All PASS or xFAIL yellowstone/aux_clm pgi yes All PASS frankfurt/aux_clm intel yes ** All PASS frankfurt/aux_clm pgi yes All PASS frankfurt/aux_clm nag yes All PASS CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes ** All PASS or BFAIL1 except CLM45BGC comparisons test_driver.sh tools testing: yellowstone interactive: no frankfurt interactive: no yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_18 Changes answers relative to baseline: YES IF tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: All CLM45BGC - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Larger than roundoff; still to be determined if this gives new climate (Dave Lawrence will run a simulation to determine this) If bitwise differences were observed, how did you show they were no worse than roundoff? N/A If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) - casename: (Not yet done) URL for LMWG diagnostics output used to validate new climate: (Not yet done) =============================================================== =============================================================== Tag name: clm4_5_18 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Tue Jul 9 10:07:08 MDT 2013 One-line Summary: rework urban indexing Purpose of changes: Change urban derived type and local variables in UrbanMod subroutines to go lbl:ubl rather than 1:num_urbanl. There are a few reasons for this: (1) this works better when the urban filter can change (with dynamic landunits), (2) more consistency with the rest of the CLM code, (3) no longer have to remember whether a given variable should be indexed by fl or l. The downside is that it leads to slightly greater memory use. Along with doing this, I also changed a few loops in UrbanMod to be simpler (which is allowed with the above change). Also, no longer run over 0-weight urban columns - we don't have to do this any more now that I have reworked some loops in UrbanMod. Requirements for tag: Test level of tag: standard Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: Increase in memory use by UrbanMod, which should lead to a small overall memory increase - though this doesn't show up in most memcomp tests, showing that the increase is pretty small. Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Major changes to indexing: local variables now dimensioned lbl:ubl rather than 1:num_urbanl. Also, remove canyon_hwr, wtroad_perv, ht_roof and wtlunit_roof from urban_params, because there are duplicate variables in clmtype. M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 ========= No longer make 0-weight urban columns active M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Remove unused wind_hgt_canyon from clmtype M models/lnd/clm/src/clm4_5/main/clmtype.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 ========= Minor changes for new UrbanMod interfaces M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 Machines testing ran on: (Tests in priority order) Standard Tag Pretag *** Standard Tag Posttag ** build-namelist unit tester: NO CESM test lists: yellowstone/aux_clm intel yes *** All PASS or xFAIL yellowstone/aux_clm pgi yes ** All PASS frankfurt/aux_clm intel yes ** All PASS frankfurt/aux_clm pgi yes All PASS, except the following, which appears to be a system problem: FAIL ERI_D.f19_g16.ICLM45.frankfurt_pgi.GC.214513 I will rerun the above test once the system problem is resolved frankfurt/aux_clm nag yes All PASS CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes ** All PASS or BFAIL, except the following expected failures: *** Expected failures because of failures in the base tests FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_17.clm2.h0 FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_Ld211_D_P112x1.f10_f10.ICNCROP.yellowstone_intel.clm-crop.compare_hist.clm4_5_17.clm2.h0 (no history file in test case) *** Expected differences in cols1d_active and pfts1d_active, as well as FILLDIFFs, due to making 0-weight urban columns no longer active FAIL ERI_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-ch4_set2_ciso.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-ch4_set3_pftroot.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_D.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC.compare_hist.clm4_5_17.clm2.h1 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default.compare_hist.clm4_5_17.clm2.h1 FAIL ERS.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.compare_hist.clm4_5_17.clm2.h1 FAIL ERS_Ld3_D_P64x1.ne30_g16.ICLM45CN.yellowstone_intel.clm-default.compare_hist.clm4_5_17.clm2.h1 FAIL SSP.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default.compare_hist.clm4_5_17.clm2.h1 test_driver.sh tools testing: yellowstone interactive: no frankfurt interactive: no yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_17 Changes answers relative to baseline: NO, except for inconsequential changes in 1-d hist files (cols1d_active, pfts1d_active, and some FILLDIFFS, due to making 0-weight urban columns no longer active) =============================================================== =============================================================== Tag name: clm4_5_17 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Wed Jul 3 10:54:03 MDT 2013 One-line Summary: misc cleanup and bug fixes Purpose of changes: Bit-for-bit cleanup following from tag clm4_5_11. The biggest change is the removal of maxpatch, npatch_* and some related variables from clm_varpar (these were maintenance headaches). Requirements for tag: Test level of tag: standard + tools Bugs fixed (include bugzilla ID): 1747 (need 1x1_tropicAtl datasets) 1754 (mksurfdata_map problem making CH4 parameters for ne240 CLM4.5 surface dataset) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: New 1x1_tropicAtl datasets Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: ========= Remove maxpatch, npatch_* and a few related variables that are no longer needed (these variables were a maintenance headache) M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 ========= Rework code to not require the variables that were removed from clm_varpar M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90 ========= Remove unneeded 'use' statements M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 ========= Allow roundoff-level errors (needed to make ne240 dataset) M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkCH4inversionMod.F90 ========= Fix generation of 1x1_tropicAtl datasets M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl ========= New 1x1_tropicAtl datasets M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml Machines testing ran on: (Tests in priority order) Standard Tag Pretag *** Standard Tag Posttag ** build-namelist unit tester: yes All PASS or xFAIL CESM test lists: yellowstone/aux_clm intel yes *** All PASS or xFAIL yellowstone/aux_clm pgi yes ** All PASS frankfurt/aux_clm intel yes ** All PASS frankfurt/aux_clm pgi yes All PASS frankfurt/aux_clm nag yes All PASS Additional tests (with comparison to clm4_5_16, including component_gen_comp; for the FARM test, used a sandbox corresponding to cesm1_3_alpha01a for components other than CLM): ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel.clm-monthly_noinitial SMS.T42_T42.FARM95C4.yellowstone_intel.clm-daily [the clm-daily nl dir just sets hist_nhtfrq = -24] CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes ** All PASS or BFAIL1, except irrelevant failures from this failing test: FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_16.clm2.h0 FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_16.clm2.h1 test_driver.sh tools testing: yellowstone interactive: yes All PASS except expected failures: 027 smiS4 TSMscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............\c rc=6 FAIL 028 bliS4 TBLscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............\c rc=4 FAIL frankfurt interactive: no yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_16 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_16 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Tue Jul 2 09:22:41 MDT 2013 One-line Summary: only run filters over 'active' points Purpose of changes: In preparation for dynamic landunits, we only want to run most filters over 'active' points. This required changing landunit and column-level filters to only run over active points. In addition, I changed the nourbanp filter to only run over active points (in contrast to other pft-level filters, this filter had previously operated over non-active points, too). In addition, this tag includes some related changes, most of which were required to get the code to run correctly in light of the above changes. Some of these changes - in particular, the changes to reweightMod, filterMod, and the use of the new filter_inactive_and_active in some places - effectively undid that general filter change for select landunits (urban) or subroutine calls. Requirements for tag: Test level of tag: standard Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: none Describe any changes made to the namelist: none List any changes to the defaults for the boundary datasets: none Describe any substantial timing or memory changes: none Code reviewed by: changes to filterMod reviewed by erik, mvertens, stefan, dave lawrence & ben andre; other changes only by self List any svn externals directories updated (csm_share, mct, etc.): none List all files eliminated: none List all files added and what they do: none List all existing files that have been modified, and describe the changes: ========= Change main filters to just run over active points; add new filters that include inactive as well as active points; refactor subroutines to avoid code duplication now that we have two groups of filters M models/lnd/clm/src/clm4_5/main/filterMod.F90 ========= Change filter to just run over active points M models/lnd/clm/src/clm4_5/main/pft2colMod.F90 ========= Change urban columns & pfts to be active whenever their landunit is active (to avoid making urban code messier) M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Get rid of wt > 0 checks, which are no longer appropriate in the code (checks of the active flags should be done instead - and these have been folded in to the filters) M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 ========= Make a loop more consistent in its use of a filter; remove undesirable pactive check (because decomp_vertprofiles now sometimes operates on inactive as well as active points) M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 ========= Use filters that include inactive points in call to decomp_vertprofiles (this is needed because of the unusual placement of this routine in the driver sequence) M models/lnd/clm/src/clm4_5/main/clm_driver.F90 ========= Use filters that include inactive points in calls to SurfaceAlbedo and UrbanAlbedo. For SurfaceAlbedo, this is necessary to avoid floating point exceptions in transient cases; for UrbanAlbedo, this probably isn't necessary now, but likely will be needed when we have dynamic landunits, for the same reason that we need it for SurfaceAlbedo. M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 ========= Remove pactive check in setting up vegsol / novegsol filters - as far as I can tell, this check is now unnecessary, and it led to the odd result that novegsol included all inactive points (e.g., even inactive istsoil points). Also add some comments and remove some obsolete comments. M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 ========= Remove pactive checks that are unnecessary now that the nourbanp definition has changed M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 ========= Only check errsoi_col on active columns (to prevent NaN-related problems in crop runs) M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 ========= Add comments M models/lnd/clm/src/clm4_5/main/CNiniSpecial.F90 M models/lnd/clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: yes All PASS or xFAIL CESM test lists: yellowstone/aux_clm intel yes All PASS or xFail Also ran the following, which PASSes (also PASSes cpl & clm hist comparisons, except for expected failures in .h1 file comparisons, as below): ERS_Ly5.f10_f10.I20TRCRUCLM45BGC.yellowstone_intel.clm-monthly_noinitial yellowstone/aux_clm pgi yes All PASS frankfurt/aux_clm intel yes All PASS frankfurt/aux_clm pgi yes All PASS frankfurt/aux_clm nag yes All PASS CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes All PASS, except: *** Irrelevant, because this test fails FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_15.clm2.h0 FAIL ERS.f09_g16.ICLM45VIC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_15.clm2.h1 *** Expected diffs in h1 files: differences in cols1d_active, pfts1d_active, and related FILLDIFFs in a number of variables FAIL ERI_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-vrtlay.compare_hist.clm4_5_15.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit.compare_hist.clm4_5_15.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-ch4_set2_ciso.compare_hist.clm4_5_15.clm2.h1 FAIL ERS_D.f10_f10.ICLM45BGC.yellowstone_intel.clm-ch4_set3_pftroot.compare_hist.clm4_5_15.clm2.h1 FAIL ERS_D.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC.compare_hist.clm4_5_15.clm2.h1 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default.compare_hist.clm4_5_15.clm2.h1 FAIL ERS.f19_g16.ICRUCLM45BGCCROP.yellowstone_intel.clm-default.compare_hist.clm4_5_15.clm2.h1 FAIL ERS_Ld3_D_P64x1.ne30_g16.ICLM45CN.yellowstone_intel.clm-default.compare_hist.clm4_5_15.clm2.h1 FAIL SSP.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default.compare_hist.clm4_5_15.clm2.h1 test_driver.sh tools testing: yellowstone interactive: no frankfurt interactive: no yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_15 Changes answers relative to baseline: NO, except for changes in fill patterns & active flags in 1-d hist files, as noted above =============================================================== =============================================================== Tag name: clm4_5_15 Originator(s): muszala (Stefan Muszala) Date: Mon Jul 1 10:44:05 MDT 2013 One-line Summary: complete associate refactor for pointers in clm4_5 source Purpose of changes: Refactor all clm4_5 source so that pointers assignements are placed in associate blocks at the start of a subroutine. This allows us to get rid of pointer declarations, makes the code easier to modify, makes the code more robust and sets us up for future interface refactorings. The refactor is explained in more detail in models/lnd/clm/tools/clm4_5/refactorTools/README. - real(r8), pointer :: fdry(:) ! fraction of foliage that is green and dry [-] (new) - fdry => pps%fdry + associate(& + fdry => pps%fdry & ! Output: [real(r8) (:)] fraction of foliage that is green and dry [-] (new) + ) ... - end subroutine FracWet + end associate + end subroutine FracWet Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: BGC compsets will see increases in memory (highwater) use. This can be seen in memcomp portions of testing with one specific example. Something to keep track of. FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.clm-default.GC.4515preIntel.memcomp.clm4_5_14 - highwater goes from 166 MB in clm4_5_14 to 236 MB in clm4_5_15 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.rtm-rtmOff.GC.4515preIntel.memcomp.clm4_5_14 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.rtm-rtmOnFloodOnEffvelOff.GC.4515preIntel.memcomp.clm4_5_14 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.rtm-rtmOnFloodOnEffvelOn.GC.4515preIntel.memcomp.clm4_5_14 FAIL ERS.f19_g16.I1850CLM45BGC.yellowstone_intel.rtm-rtmOnIceOn.GC.4515preIntel.memcomp.clm4_5_14 FAIL ERS.f19_g16_r01.I1850CLM45BGC.yellowstone_intel.rtm-rtmOnFloodOnEffvelOff.GC.4515preIntel.memcomp.clm4_5_14 Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: A models/lnd/clm/tools/clm4_5/refactorTools/associate/refactor_new.pl A models/lnd/clm/tools/clm4_5/refactorTools/associate/README List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/initch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_BGC.F90 M models/lnd/clm/src/clm4_5/biogeochem/MEGANFactorsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVLightMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_CENTURY.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVegStructUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/mkarbinitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CLMVICMapMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/initSLakeMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FracWetMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: OK CESM test lists: yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CESM history file comparison: yellowstone/aux_clm OK CLM tag used for the baseline comparison tests if applicable: clm4_5_14 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_14 Originator(s): muszala (Stefan Muszala) Date: Thu Jun 20 07:51:54 MDT 2013 One-line Summary: preparation for associate refactor in clm4_5_15 Purpose of changes: Most work by mvertens. - prep. work for modifying associate - refactor subgridAveMod.F90 to accept upper and lower bounds - remove duplicate pointer uses - remove inicPerpMod.F90 and is_perpetual use Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: D models/lnd/clm/src/clm4_5/main/inicPerpMod.F90 List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/main/pft2colMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: OK. Fixed generate numbering. CESM test lists: yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel OK CLM tag used for the baseline comparison tests if applicable: clm4_5_13 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_13 Originator(s): andre (Benjamin Andre,LAWRENCE BERKELEY NATIONAL LABORATORY,510-486-4617) Date: Fri Jun 14 15:01:33 MDT 2013 One-line Summary: hydrology reordering from Jinyun Tang Purpose of changes: reordering the operations of the hydrology. hydrology with and without drainage Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: Jinyun Tang, Dave Lawrence List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 - splits out leaching M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 - add icefrac and qflx_deficit M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 - update calls to CNEcosystemDyn M models/lnd/clm/src/clm4_5/main/clm_driver.F90 - update calls to CNEcosystemDyn and Hydrology M models/lnd/clm/src/clm4_5/main/clmtype.F90 - add icefrac and qflx_deficit M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 - splits out drainage calculations M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 - calculate water table before subsurface drainage, icefraction, water deficit Machines testing ran on: (Tests in priority order) build-namelist unit tester: yes/no *** CESM test lists: yellowstone/aux_clm intel yes OK yellowstone/aux_clm pgi yes OK frankfurt/aux_clm intel no frankfurt/aux_clm pgi no frankfurt/aux_clm nag no CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) test_driver.sh tools testing: N/A yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_12 Changes answers relative to baseline: yes IF tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: all clm 4.5 with hydrology - what platforms/compilers: all - nature of change (roundoff; larger than roundoff/same climate; new climate): new climate If bitwise differences were observed, how did you show they were no worse than roundoff? If this tag changes climate describe the run(s) done to evaluate the new climate (put details of the simulations in the experiment database) casename: ERS_D.f19_g16.ICLM45 - an 11-year base line simulation was created with the standard trunk version, then a comparison run was created with the version including hydrology re-ordering. The comparisons were evaluated by looking at the relative differences for hydrologic variables as QDRAI, EFLX_LH_TOT, QRUNOFF. Large relative differences were found for these variables in a few grid cells, but their absolute magnitudes in those grid cells were small. Tests were also conducted with VIC hydrology on, the change in results were similar as that when VIC hydrology was off. URL for LMWG diagnostics output used to validate new climate: N/A =============================================================== =============================================================== Tag name: clm4_5_12 Originator(s): muszala (Stefan Muszala) Date: Thu Jun 13 09:41:56 MDT 2013 One-line Summary: NoVS test, NAG mods and remove TWS from restart file Purpose of changes: -Fix (from jedwards) for ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit test Previous tags using ICLM45BGCNoVS are suspect! -Since I tested this with NAG there are also port mods to CLM that I had to put in. -Remove TWS from BiogeophysRestMod.F90 per sacks request. OK'd by swenson. -Update to rtm1_0_29 Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): 1746 Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, NoVs error: Charlie Koven and jedwards List any svn externals directories updated (csm_share, mct, etc.): rtm1_0_28 -> rtm1_0_29 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: -NoVS fix and NAG mods M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 -Nag mods M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4varcon.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 -remove TWS from restart M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 -update to rtm1_0_29 M SVN_EXTERNAL_DIRECTORIES Machines testing ran on: (Tests in priority order) build-namelist unit tester: no CESM test lists: yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel OK. The only differences are in ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit. These are expected due to the fix in ch4Mod.F90. CLM tag used for the baseline comparison tests if applicable: clm4_5_11 Changes answers relative to baseline: Only for ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit. Previous versions of this compset should be considered suspect. Fields that change (119 in total) for this test are: ACTUAL_IMMOB, CH4STRESS_SAT, CH4STRESS_UNSAT, CH4_AERE_DEPTH_SAT, CH4_AERE_DEPTH_UNSAT, CH4_EBUL_DEPTH_SAT, CH4_OXID_DEPTH_SAT, CH4_OXID_DEPTH_UNSAT, CH4_PROD_DEPTH_SAT, CH4_SURF_AERE_SAT, CH4_SURF_AERE_UNSAT, CH4_SURF_DIFF_SAT, CH4_SURF_DIFF_UNSAT, CH4_SURF_EBUL_SAT, CH4_TRAN_DEPTH_SAT, CH4_TRAN_DEPTH_UNSAT, COL_CTRUNC, COL_NTRUNC, CONC_CH4_SAT, CONC_CH4_UNSAT, CONC_O2_SAT, CONC_O2_UNSAT, CWDC, CWDC_LOSS, CWDC_TO_LITR2C, CWDC_TO_LITR3C, CWDN, CWDN_TO_LITR2N, CWDN_TO_LITR3N, DENIT, ER, FCH4, FCH4TOCO2, FUELC, F_DENIT, F_N2O_DENIT, F_N2O_NIT, F_NIT, GROSS_NMIN, HR, LAND_UPTAKE, LITHR, LITR1C, LITR1C_TO_SOIL1C, LITR1N, LITR1N_TO_SOIL1N, LITR1_HR, LITR2C, LITR2C_TO_SOIL1C, LITR2N, LITR2N_TO_SOIL1N, LITR2_HR, LITR3C, LITR3C_TO_SOIL2C, LITR3N, LITR3N_TO_SOIL2N, LITR3_HR, LITTERC, LITTERC_HR, LITTERC_LOSS, NBP, NEE, NEM, NEP, NET_NMIN, O2STRESS_SAT, O2_AERE_DEPTH_SAT, O2_DECOMP_DEPTH_SAT, O2_DECOMP_DEPTH_UNSAT, POTENTIAL_IMMOB, POT_F_DENIT, POT_F_NIT, SMINN_TO_SOIL1N_L1, SMINN_TO_SOIL1N_L2, SMINN_TO_SOIL1N_S2, SMINN_TO_SOIL1N_S3, SMINN_TO_SOIL2N_L3, SMINN_TO_SOIL2N_S1, SMINN_TO_SOIL3N_S1, SMINN_TO_SOIL3N_S2, SMIN_NH4, SMIN_NO3, SMIN_NO3_LEACHED, SOIL1C, SOIL1C_TO_SOIL2C, SOIL1C_TO_SOIL3C, SOIL1N, SOIL1N_TO_SOIL2N, SOIL1N_TO_SOIL3N, SOIL1_HR_S2, SOIL1_HR_S3, SOIL2C, SOIL2C_TO_SOIL1C, SOIL2C_TO_SOIL3C, SOIL2N, SOIL2N_TO_SOIL1N, SOIL2N_TO_SOIL3N, SOIL2_HR_S1, SOIL2_HR_S3, SOIL3C, SOIL3C_TO_SOIL1C, SOIL3N, SOIL3N_TO_SOIL1N, SOIL3_HR, SOILC, SOILC_HR, SOILC_LOSS, SOMHR, SR, TOTCOLC, TOTCOLCH4, TOTCOLN, TOTECOSYSC, TOTECOSYSN, TOTLITC, TOTLITN, TOTSOMC, TOTSOMN =============================================================== =============================================================== Tag name: clm4_5_11 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Tue Jun 11 20:54:11 MDT 2013 One-line Summary: Change pct_pft and related surface dataset variables to be % of landunit Purpose of changes: Main purpose is to change pct_pft and related surface dataset variables to be % of landunit rather than % of grid cell. This is needed to support transient PFTs with dynamic landunits. This required substantial changes in both mksurfdata_map and CLM. This also required generating all new surface datasets. A very related change is the separation of PCT_PFT in the surface dataset into PCT_NAT_PFT and PCT_CFT; in addition to these two variables, there are also new PCT_NATVEG (% of natural veg landunit on the gridcell) and PCT_CROP (% of crop landunit on the gridcell) variables. Note that the separation of PCT_PFT into natural vs crop was only done on the surface dataset -- raw datasets to mksurfdata_map have not been changed, nor have most of the CLM data structures. In addition, this tag includes the following: (1) Renumbered landunits to (a) add separate landunit numbers for each urban landunit, (b) do away with the obsolete shallow lake, and (c) group together similar landunits (2) In any urban landunit, allocate space for ALL urban columns. Previously, there were some urban landunits with only one of the two road types. This change simplifies the code and only adds a relatively small number of columns in memory. (3) Modified interpinic, partly to have compatibility with (1), partly to fix urban bug (allowed by (1)), and partly to fix an unrelated bug (4) All new initial conditions for CLM4.5, to have compatibility with (1) and (2) (5) Check _OPENMP in initialization rather than driver (6) Tighten error check in reweightMod: checkWeights. It seems like this error check can be stricter with the new pct_pft formulation Requirements for tag: Standard test + tools Test level of tag: standard Bugs fixed (include bugzilla ID): 1621 (normalization issue in ne120np4 datasets and in CLM) 1675 (need to relax error tolerance in reweightMod: weightsOkay) -- note that I have actually TIGHTENED the tolerance, but that seems to be okay now 1702- PARTIAL FIX (clm4.5 interpinic doesn't work right for urban) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ 1747 - need 1x1_tropicAtl surface dataset and pftdyn dataset for clm4_5_11 and later Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Changes to fsurdat and finidat for CLM4.5, as described below List any changes to the defaults for the boundary datasets: For CLM4.5, all surface datasets and initial conditions files have been recreated. For surface datasets, changes result in only roundoff-level differences in the pct_* fields. For initial conditions, the new initial conditions are effectively the same as the old, but bugs in interpinic prevent them from being exactly the same. Describe any substantial timing or memory changes: Slight (probably < 1%) increase in memory for all CLM4.5 cases, due to allocation of ALL urban columns wherever there is an urban landunit Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): NONE List all files eliminated: ========= Remove unneeded modules (iulog moved to fileutils.F90) D models/lnd/clm/tools/clm4_5/mksurfdata_map/src/clm_varpar.F90 D models/lnd/clm/tools/clm4_5/mksurfdata_map/src/clm_varctl.F90 ========= Now differs for clm4_0 and clm4_5, so copied to those two places D models/lnd/clm/src/util_share/clm_varsur.F90 ========= Replaced with new file for testing interpinic D models/lnd/clm/tools/clm4_5/interpinic/clmi.BCN.1949-01-01_10x15_USGS_simyr1850_c121113.nc List all files added and what they do: ========= New file for testing interpinic A models/lnd/clm/tools/clm4_5/interpinic/clmi.I2000CLM45BGC.2000-01-01.10x15_simyr2000_c130607.nc ========= Add tests A models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkpftMod.F90 ========= Moved from util_share A models/lnd/clm/src/clm4_0/main/clm_varsur.F90 ========= Moved from util_share, and modified extensively to support new surface dataset format A models/lnd/clm/src/clm4_5/main/clm_varsur.F90 ========= Add module to do some initialization that doesn't fit well elsewhere, and/or can't go elsewhere because of circular dependencies A models/lnd/clm/src/clm4_5/main/initParametersMod.F90 List all existing files that have been modified, and describe the changes: ========= Change pct_pft and related variables on surface dataset to be % of landunit; this requires significant changes for mkpftMod, mkglcmecMod and the error checks / corrections done in mksurfdat.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkutilsMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/fileutils.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkfileMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkglcmecMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkvarpar.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mksurfdat.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/Srcfiles M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/mkpftMod.F90 ========= Add tests M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mkutilsMod.F90 M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/Srcfiles M models/lnd/clm/tools/clm4_5/mksurfdata_map/unit_testers/test_mksurfdata_map.F90 ========= Update crop landunit numbering, fix urban bug for column-level variables, take code out of a conditional to prevent floating point exceptions M models/lnd/clm/tools/clm4_5/interpinic/src/interpinic.F90 ========= New files for testing interpinic M models/lnd/clm/tools/clm4_5/interpinic/interpinic.runoptions ========= Change landunit and column numbering; delete udenstype M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 ========= Add variables for determining number of natural & crop PFTs M models/lnd/clm/src/clm4_5/main/clm_varpar.F90 ========= Major changes to handle pct_pft being specified as % of landunit rather than % of gridcell M models/lnd/clm/src/clm4_5/main/surfrdMod.F90 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 M models/lnd/clm/src/clm4_5/main/subgridMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 ========= Initialize new surface variables, check _OPENMP here instead of driver M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ========= Update comments, remove udenstype M models/lnd/clm/src/clm4_5/main/clmtype.F90 ========= Change 'use' statements, use ltype instead of udenstype, fix initialization for 0-weight columns M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 ========= Add some consistency checks (moved here from clmtypeInitMod), change others M models/lnd/clm/src/clm4_5/main/clm_varctl.F90 ========= Change 'use' statements; use urbpoi rather than isturb; remove references to 'istslak' M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNEcosystemDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInputMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/clm4_5/main/CNiniTimeVar.F90 M models/lnd/clm/src/clm4_5/main/dynlandMod.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 -- also, remove udenstype, and move some consistency checks elsewhere M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/restFileMod.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 -- also, move _OPENMP check to initialization M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/CNiniSpecial.F90 M models/lnd/clm/src/clm4_5/main/inicPerpMod.F90 M models/lnd/clm/src/clm4_5/main/mkarbinitMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 ========= Tighten tolerance for error check M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ========= Use 'crop_prog' rather than the CROP CPP def M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 ========= New surface datasets and initial conditions M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ========= Change environment variable in component_gen_comp command to something universal M .ChangeLog_template ========= Restore a failing test (see bug 1658) M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml Machines testing ran on: (Tests in priority order) build-namelist unit tester: yes All PASS or xFail except for the following expected baseline failures: 418/444 < FAIL> 423/444 < FAIL> 428/444 < FAIL> 433/444 < FAIL> 438/444 < FAIL> 443/444 < FAIL> CESM test lists: yellowstone/aux_clm intel yes Tests themselves: All PASS or xFail, except ERB.ne30_g16.I_1948-2004.yellowstone_intel, which I have re-added to the xFail list (see bugz 1658) Comparisons: Some nlcomp and compare_hist failures, as expected yellowstone/aux_clm pgi yes All PASS or xFail except for some nlcomp & compare_hist failures (expected) frankfurt/aux_clm intel yes All PASS or xFail except for some nlcomp & compare_hist failures (expected) frankfurt/aux_clm pgi NO frankfurt/aux_clm nag yes All PASS or xFail except for nlcomp failures (expected) CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $CESMDATAROOT/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes Some compare_hist failures for CLM4.5 tests, as expected test_driver.sh tools testing: yellowstone interactive: yes All PASS except for expected baseline failures: 012 blg54 TBLtools.sh clm4_5 mksurfdata_map tools__s namelist ...................................rc=7 FAIL 016 blh54 TBLtools.sh clm4_5 interpinic tools__ds runoptions ....................................rc=7 FAIL 020 bli24 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_T31_crpglc_2000^tools__ds rc=7 FAIL 022 bli53 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__o ....rc=7 FAIL 024 bli54 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__ds ...rc=7 FAIL 026 bli57 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_10x15_1850^tools__do ...rc=7 FAIL 032 bliT4 TBLscript_tools.sh clm4_5 mksurfdata_map mksurfdata.pl mksrfdt_1x1_numaIA_mp24_2000^tools_rc=7 FAIL and expected failures: 027 smiS4 TSMscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............rc=6 FAIL 028 bliS4 TBLscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional .............rc=4 FAIL frankfurt interactive: NO yellowstone/PTCLM: NO CLM tag used for the baseline comparison tests if applicable: clm4_5_10 Changes answers relative to baseline: YES IF tag changes answers relative to baseline comparison the following should be filled in (otherwise remove this section): Summarize any changes to answers, i.e., - what code configurations: Changes in all CLM4.5 configurations. See below for details - what platforms/compilers: All - nature of change (roundoff; larger than roundoff/same climate; new climate): Mostly roundoff, but some larger than roundoff -- see below for details. Changes are limited to CLM4.5. Where the below notes refer to "all configurations", this is really limited to CLM4.5 configurations. For all configurations, there are roundoff-level changes due to roundoff-level differences in subgrid weights. These changes can quickly grow to greater than roundoff (which I believe is due to nonlinear feedbacks with snow variables), but as described below, I have verified that the root cause of differences is this roundoff-level change. For cases that use initial conditions, where these initial conditions were previously interpinic'ed, there are greater-than-roundoff level changes due to various bugs and limitations of interpinic (for example, some fields, like tsai, are skipped). I took pains to ensure that, for cases using original (non-interpinic'ed) initial conditions in clm4_5_10 and prior, the new initial conditions are nearly identical to the old (but not entirely identical, due to bug 1702 - see comment 2); this applies to most f09 initial conditions. However, this was not practical for cases that used interpinic'ed files; this applies to f19, ne30 and hcru initial conditions, as well as f09 BGCDV initial conditions. So for this latter set of cases, there can be large differences from clm4_5_10, especially at the start of the simulation. There are also greater than roundoff-level changes for some glc_mec virtual columns, because we now use information on topo_glc_mec whenever we can. There are also greater than roundoff-level changes in subgrid weights in virtual (0-weight) glc_mec and crop landunits, now that we no longer use arbitrary subgrid weights there; I don't think this will affect anything important, though. Some tests that exhibited larger-than-usual changes from baseline, in cpl hist and/or clm hist files, were the following (ignoring changes that can be explained by the above notes): ERS_Lm3.f19_g16.IGRCP60CLM45CN.yellowstone_intel ERS_D.f19_g16.ICLM45GLCMEC.yellowstone_intel.clm-glcMEC ERS_D.f10_f10.ICLM45BGCNoVS.yellowstone_intel.clm-rootlit ERS_Ld3_D_P64x1.ne30_g16.ICLM45CN.yellowstone_intel.clm-default From these results, it seems that large changes may occur more often in glc_mec runs, even above and beyond the virtual column changes that are expected, as noted above. For the four above tests, I verified that differences were attributable to the roundoff-level changes in subgrid weights, using the procedure documented below. If bitwise differences were observed, how did you show they were no worse than roundoff? In order to confirm that answers only differed by roundoff, I ran 7 additional tests (these tests all had nl_dirs, but that was not central to these tests, so I'm not listing them here): SMS_Lm1.f19_g16.I_1850_CLM45_BGC SMS_Ly2.T31_g37.IG1850CLM45CN SMS_Ly5.T31_g37.I20TRCRUCLM45BGC SMS_Ly5.T31_g37.ICLM45BGCDV SMS_Ly5.T31_g37.I_2000_CLM45_BGC SMS_Ly5.f10_f10.ICLM45BGCDVCROP SMS_Ly5.f10_f10.I_2000_CLM45_BGC_CROP Each test was done as follows: (1) Created baselines from clm4_5_04 (2) Ran tests from a branch (allocate_all_urban_cols), where all I changed from the baseline was (a) wherever we have an urban landunit, create ALL urban landunits, and (b) for f19, change finidat files to use a new, interpinic'ed file based on the original. Confirmed that this was bfb with (1) except for (a) 1-d history files (now have extra urban columns), and (b) any CLM4.5 test that uses initial conditions, since interpinic is currently broken for urban. Side-note: I actually confirmed bfb behavior for the full yellowstone aux_clm test suite, in addition to the above 7 tests This extra branch was necessary because I cannot compare 1-d history files directly between my main branch and the trunk, because of the extra urban columns present in the new code. (3) Ran these 7 tests from my main branch, off of clm4_5_04, comparing with (2). For this comparison, I only confirmed that the subgrid weights were the same within roundoff (up to about 1e-12 differences for the transient case; smaller for other cases). Note that greater than roundoff-level changes are seen in many other fields, presumably because small differences in subgrid pft weights can cause differences in how variables are averaged from pft to column. This, in turn, can lead to larger changes due to nonlinearities in the system (e.g., snow). The following steps were taken to confirm that other differences between my branch and the trunk were only due to these small differences in subgrid weights. That is, I am confirming that: (a) the only differences in the branch are subgrid weights (b) these subgrid weights only differ by roundoff (4) Reran (2), but with extra code to write out subgrid weights (including writing these weights at every time step for pftdyn) -- from branch allocate_all_urban_cols_writeWeights (5) Reran (3), but with extra code to read the subgrid weights written in (4). Confirmed that, with this one-off, my branch was bfb with (4). =============================================================== =============================================================== Tag name: clm4_5_10 Originator(s): muszala (Stefan Muszala) Date: Mon Jun 10 13:10:31 MDT 2013 One-line Summary: refactor clmtype Purpose of changes: Refactor clmtype so that there is only one level of indirection. call hist_addfld1d (fname='SNOOCFRCL', units='W/m^2', & avgflag='A', long_name='surface forcing of OC in snow (land) ', & - ptr_pft=clm3%g%l%c%p%pef%sfc_frc_oc, set_urb=spval) + ptr_pft=pef%sfc_frc_oc, set_urb=spval) There is a README (with more detailed information) and a script to help with future merges in: models/lnd/clm/tools/clm4_5/refactorTools/clmType/{README & renameClmType.pl} Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, mvertens List any svn externals directories updated (csm_share, mct, etc.): N/A List all files eliminated: N/A List all files added and what they do: - script and README for refactoring clmType A models/lnd/clm/tools/clm4_5/refactorTools A models/lnd/clm/tools/clm4_5/refactorTools/associate A models/lnd/clm/tools/clm4_5/refactorTools/clmType A models/lnd/clm/tools/clm4_5/refactorTools/clmType/renameClmType.pl A models/lnd/clm/tools/clm4_5/refactorTools/clmType/README List all existing files that have been modified, and describe the changes: - major refactor in these to flatten clmtype M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/clmtype.F90 - change derived type access to match those of clmtype M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGapMortalityMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNitrifDenitrifMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNGRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/initch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNBalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNMRespMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_BGC.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVerticalProfileMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4RestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNWoodProductsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPrecisionControlMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVLightMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate3Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/VOCEmissionMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEcosystemDynIniMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAnnualUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNStateUpdate2Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompCascadeMod_CENTURY.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNNDynamicsMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSetValueMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNVegStructUpdateMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSoilLittVertTranspMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 M models/lnd/clm/src/clm4_5/main/clm_varcon.F90 M models/lnd/clm/src/clm4_5/main/CNiniTimeVar.F90 M models/lnd/clm/src/clm4_5/main/dynlandMod.F90 M models/lnd/clm/src/clm4_5/main/subgridRestMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/pftdynMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/pft2colMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/initSurfAlbMod.F90 M models/lnd/clm/src/clm4_5/main/filterMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 M models/lnd/clm/src/clm4_5/main/initGridCellsMod.F90 M models/lnd/clm/src/clm4_5/main/initSoilParVICMod.F90 M models/lnd/clm/src/clm4_5/main/CNiniSpecial.F90 M models/lnd/clm/src/clm4_5/main/inicPerpMod.F90 M models/lnd/clm/src/clm4_5/main/reweightMod.F90 M models/lnd/clm/src/clm4_5/main/mkarbinitMod.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SnowHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FrictionVelocityMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/TridiagonalMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeHydrologyMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BiogeophysRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/ActiveLayerMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/clm_driverInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CLMVICMapMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BareGroundFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/initSLakeMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Biogeophysics2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/FracWetMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/UrbanInitMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SLakeRestMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology2Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 Machines testing ran on: (Tests in priority order) Standard Tag Pretag *** Standard Tag Posttag ** build-namelist unit tester: no CESM test lists: yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CESM history file comparison: yellowstone/aux_clm intel CLM tag used for the baseline comparison tests if applicable: clm4_5_09 Changes answers relative to baseline: No. Everything in this refactor should be BFB. =============================================================== =============================================================== Tag name: clm4_5_09 Originator(s): muszala (Stefan Muszala,UCAR/CGD,303-497-1320) Date: Tue Jun 4 15:59:07 MDT 2013 One-line Summary: volr and vic fix, update mct and rtm Purpose of changes: add volr area correction, minor vic fix from maoyi, update mct and rtm externals Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, swenson List any svn externals directories updated (csm_share, mct, etc.): -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_27 +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_28 -models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.3 +models/utils/mct https://github.com/quantheory/MCT/tags/compiler_fixes_n01_MCT_2.8.3 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: - fix for VIC hydrology M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 - volr area correction M models/lnd/clm/src/cpl_mct/lnd_comp_mct.F90 M models/lnd/clm/src/cpl_esmf/lnd_comp_esmf.F90 - mct and rtm update M SVN_EXTERNAL_DIRECTORIES - clean up M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml Machines testing ran on: (Tests in priority order) build-namelist unit tester: OK CESM test lists: yellowstone/aux_clm intel OK yellowstone/aux_clm pgi OK frankfurt/aux_clm intel OK frankfurt/aux_clm pgi OK frankfurt/aux_clm nag OK CLM tag used for the baseline comparison tests if applicable: clm4_5_08 Changes answers relative to baseline: only for VIC compsets. VOLR diagnostic changes. =============================================================== =============================================================== Tag name: clm4_5_08 Originator(s): muszala (Stefan Muszala,UCAR/CGD,303-497-1320) Date: Mon Jun 3 13:29:30 MDT 2013 One-line Summary: port for NAG compiler Purpose of changes: Bring in Sean Santos mods, port clm4_5 and test with the NAG compiler on Frankfurt. Requirements for tag: N/A Test level of tag: standard Bugs fixed (include bugzilla ID): 1721 - Jim Edwards fixed problem in PIO Known bugs (include bugzilla ID): 1722 - Error in some VIC tests starting in clm4_5_07 http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, Santos List any svn externals directories updated (csm_share, mct, etc.): PIO - update to pio1_7_2 List all files eliminated: D models/lnd/clm/src/util_share/nanMod.F90 List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/xFail/expectedFail.pm M models/lnd/clm/src/util_share/clm_time_manager.F90 M models/lnd/clm/src/util_share/accumulMod.F90 M models/lnd/clm/src/util_share/ndepStreamMod.F90 M models/lnd/clm/src/util_share/ncdio_pio.F90 M models/lnd/clm/src/util_share/spmdMod.F90 M models/lnd/clm/src/util_share/domainMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCIsoFluxMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNSummaryMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/ch4varcon.F90 M models/lnd/clm/src/clm4_5/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDecompMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNCStateUpdate1Mod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_5/biogeochem/CNC14DecayMod.F90 M models/lnd/clm/src/clm4_5/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_5/main/subgridMod.F90 M models/lnd/clm/src/clm4_5/main/accFldsMod.F90 M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_5/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_5/main/histFileMod.F90 M models/lnd/clm/src/clm4_5/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_5/main/controlMod.F90 M models/lnd/clm/src/clm4_5/main/clm_driver.F90 M models/lnd/clm/src/clm4_5/main/pftvarcon.F90 M models/lnd/clm/src/clm4_5/main/histFldsMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/BalanceCheckMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilTemperatureMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/Hydrology1Mod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SNICARMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/H2OSfcMod.F90 M models/lnd/clm/src/clm4_5/biogeophys/SoilHydrologyMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CropRestMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/DUSTMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNPhenologyMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/STATICEcosysDynMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNAllocationMod.F90 M models/lnd/clm/src/clm4_0/biogeochem/CNDVEstablishmentMod.F90 M models/lnd/clm/src/clm4_0/main/clm_glclnd.F90 M models/lnd/clm/src/clm4_0/main/subgridMod.F90 M models/lnd/clm/src/clm4_0/main/accFldsMod.F90 M models/lnd/clm/src/clm4_0/main/clmtypeInitMod.F90 M models/lnd/clm/src/clm4_0/main/pftvarcon.F90 M models/lnd/clm/src/clm4_0/main/iniTimeConst.F90 M models/lnd/clm/src/clm4_0/main/histFileMod.F90 M models/lnd/clm/src/clm4_0/main/clm_atmlnd.F90 M models/lnd/clm/src/clm4_0/biogeophys/SNICARMod.F90 M SVN_EXTERNAL_DIRECTORIES Machines testing ran on: (Tests in priority order) build-namelist unit tester: - OK. yellowstone/aux_clm intel - OK. yellowstone/aux_clm pgi - OK. One BFAIL for hcru_hcru which should pass next time around. Bug fixed when upgrading to pio1_7_2. BFAIL ERS_D.hcru_hcru.I_2000_CRUFRC_CLM45_CN.yellowstone_pgi.GC.08testPgi.compare_hist.clm4_5_07 frankfurt/aux_clm intel - OK. frankfurt/aux_clm pgi - OK. frankfurt/aux_clm nag - OK. No baselines to compare against. CLM tag used for the baseline comparison tests if applicable: clm4_5_07 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_07 Originator(s): erik (Erik Kluzek) Date: Fri May 31 02:49:45 MDT 2013 One-line Summary: New spinup files for CLM45 AND RTM, work on PTCLM, turn drydep off by default, update externals Purpose of changes: Bring in new spinup finidat files (f09_g16@1850 for SP and BGC). interpinic to 2deg, hcru_hcru and ne30. New spinup finidat files for BGCCROP and BGCDV (f19 and f09 respectively) New spinup finidat files for 2000 (f09_g16 for SP and BGC) Update RTM to bring in finidat_rtm files for either 1850 or 2000. Update scripts, Machines, pio scripts includes update for CLM40CRU hybrid startup Turn drydep namelist off by default Do a lot of work on getting PTCLM working and tools working for single-point. Requirements for tag: New spinup files, fix bugs: 1708, 1700 Test level of tag: standard Bugs fixed (include bugzilla ID): 1715 (rmdups.ncl fails for no-overlap case) 1714 (mkscripgrid.ncl doesn't calculate corners correctly.) 1708 (Need Initial conditions for RTM) 1706 (VIC tests fail) 1700 (Memory leak in MPI layer on yellowstone) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: drydep namelist now OFF by default List any changes to the defaults for the boundary datasets: New initial conditions Native initial conditions for f09 for: I1850CLM45SP, I1850CLM45BGC, ICLM45SP, ICLM45BGC Interpinic for: I1850CLM45 & I1850CLM45BGC: f19, hcru_hcru, ne30 ICLM45BGCCROP @ f19 ICLM45BGCDB @ f09 ALSO NOTE THAT NOW RTM HAS INITIAL CONDITIONS FOR R05 -- SO RIVERFLOW CHANGES FOR BOTH CLM45 AND CLM40 Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): many csm_share, pio, scripts, Machines, RTM, cprnc, mapping scripts to scripts4_130529 (update PTCLM, send simyr to RTM, new IC for CLM40CRUCN) csm_share to share3_130528 rtm to rtm1_0_27 (Set startup initial condition files by -simyr flag) Machines to Machines_130529 (Set hcru_hcru PE-layout, and PE-layout on yellowstone for f09 I cases) pio to pio1_7_1 cprnc to cprnc_130425 mapping to mapping_130509 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/test/tools/TBLCFGtools.sh --- Correctly point to TSMCFGtools rather than TSMtools.sh. M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl - Add -usr_mapdir option M models/lnd/clm/tools/shared/mkmapdata/rmdups.ncl ------ Exit early if n_s==0 M models/lnd/clm/tools/shared/mkmapdata/mkmapdata.sh ---- Skip if file already exists, give directory for rmdups.ncl M models/lnd/clm/tools/shared/mkmapdata/mknoocnmap.pl --- Don't hide NCL output M models/lnd/clm/tools/shared/mkmapgrids/mkscripgrid.ncl Explicitly calculate corners M models/lnd/clm/bld/build-namelist - Set drydep to off by default, check crop setting for finidat files M models/lnd/clm/bld/clm.buildnml.csh - Add back logic in about ignoring IC year or date M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml - New initial conditions for: Native initial conditions for f09 for: I1850CLM45SP, I1850CLM45BGC, ICLM45SP, ICLM45BGC Interpinic for: I1850CLM45SP & I1850CLM45BGC: f19, hcru_hcru, ne30 ICLM45BGCCROP @ f19 ICLM45BGCDB @ f09 M models/lnd/clm/bld/namelist_files/namelist_defaults_usr_files.xml - Use surfdata_map rather than surfdata for CLM_USRDAT_NAME fsurdat files M models/lnd/clm/bld/namelist_files/use_cases/stdurbpt_pd.xml - Remove setting of dtime, adjust hist output More work on readme files... M README M models/lnd/clm/doc/IMPORTANT_NOTES M models/lnd/clm/doc/Quickstart.GUIDE M models/lnd/clm/doc/Quickstart.userdatasets Machines testing ran on: (Tests in priority order) build-namelist unit tester: yes CESM test lists: yellowstone/aux_clm intel yes yellowstone/aux_clm pgi yes frankfurt/aux_clm intel yes frankfurt/aux_clm pgi yes CESM history file comparison: (ccsm_utils/Tools/component_gen_comp -compare $oldtag -generate $newtag -testid $testid -baselineroot $GLDCSEG/ccsm_baselines/ -model clm2) yellowstone/aux_clm intel yes test_driver.sh tools testing: yellowstone interactive: yes frankfurt interactive: yes yellowstone/PTCLM: yes! CLM tag used for the baseline comparison tests if applicable: clm4_5_06 Changes answers relative to baseline: Yes -- due to new initial condition files for I1850CLM45SP and I1850CLM45BGC @ f09, f19, hcru, ne30 ICLM45SP and ICLM45BGC @ f09 ICLM45BGCCROP @ f19 and ICLM45BGCDB @ f09 and ICLM40CRUCN @ f09 AND new initial conditions for RTM for ALL R05 grids And turning drydep namelist off in the driver causes answers to appear to be different when comparing coupler history files. =============================================================== =============================================================== Tag name: clm4_5_06 Originator(s): erik (Erik Kluzek) Date: Wed May 15 13:52:43 MDT 2013 One-line Summary: A few small bug fixes, more updates to README files Purpose of changes: More work on README files and documentation. Fix from Danica/Bill for transient simulations. Fix from Zack for Lake output variables Another multi-instance script fix. Fix tropixAtl pftdyn filename. Remove models/lnd/clm/bld/config_query as doesn't work with new CESM scripts. Requirements for tag: Requirements: Fix bug: 1697, 1691, 1675, fix tropicAtl fpftdyn file, minimal testing on frankfurt Test level of tag: critical Bugs fixed (include bugzilla ID): 1675 (need to relax error tolerance in reweightMod: weightsOkay) 1691 (Scripts issue for multi-instance for CLM/RTM) 1697 (ZLAKE and DZLAKE are NOT set) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: Fix 1x1_tropicAtl fpftdyn file Describe any substantial timing or memory changes: None Code reviewed by: self, sacks (transient fix), dlawren/subin (lake fix), jedwards (multi-instance scripts) List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: testing namelist files for old CLM standalone D models/lnd/clm/bld/config_query --- no longer works with new CESM scripts D models/lnd/clm/test/tools/nl_files/nl_ch4_set2_ciso D models/lnd/clm/test/tools/nl_files/nl_ch4_set3_pftroot D models/lnd/clm/test/tools/nl_files/nl_rootlit D models/lnd/clm/test/tools/nl_files/nl_ciso D models/lnd/clm/test/tools/nl_files/nl_anoxia_wtsat D models/lnd/clm/test/tools/nl_files/nl_vrtlay D models/lnd/clm/test/tools/nl_files/nl_oldhyd List all files added and what they do: A models/lnd/clm/tools/clm4_5/interpinic/addmetadata --- Add script to add important meta-data to finidat files. List all existing files that have been modified, and describe the changes: ---------------- Work on README files documentation M models/lnd/clm/test/tools/config_files/README M models/lnd/clm/test/tools/README M models/lnd/clm/test/tools/README.testnames M models/lnd/clm/tools/README M models/lnd/clm/doc/IMPORTANT_NOTES M models/lnd/clm/doc/Quickstart.GUIDE M models/lnd/clm/doc/README ---------------- M models/lnd/clm/bld/config_files/config_definition.xml ------------ Document experimental settings / fix syntax error M models/lnd/clm/bld/clm.buildnml.csh ------------------------------ Multi-instance fix M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml --- Fix 1x1_tropicAtl fpftdyn filename M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml - Document experimental settings ---------------- M models/lnd/clm/src/clm4_5/main/histFileMod.F90 ---- ZLAKE/DZLAKE fix M models/lnd/clm/src/clm4_5/main/subgridAveMod.F90 -- ZLAKE/DZLAKE fix M models/lnd/clm/src/clm4_5/main/reweightMod.F90 ---- Increase tolerance to 1.e-7 so transient simulations can run their full course. Machines testing ran on: (Tests in priority order) build-namelist unit tester: yes CESM test lists: (limited testing on yellowstone/aux_clm/intel) frankfurt/aux_clm pgi yes frankfurt/aux_clm intel yes CLM tag used for the baseline comparison tests if applicable: clm4_5_04 Changes answers relative to baseline: no (bit-for-bit) =============================================================== =============================================================== Tag name: clm4_5_05 Originator(s): muszala (Stefan Muszala,UCAR/CGD,303-497-1320) Date: Tue May 14 13:15:12 MDT 2013 One-line Summary: hcru bug fixes Purpose of changes: update pio tag and nfire init. mod Requirements for tag: N/A Test level of tag: Only run hcru_hcru tests Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self, Fang and Erik for nfire problem List any svn externals directories updated (csm_share, mct, etc.): update pio to 1_7_0 List all files eliminated: N/A List all files added and what they do: N/A List all existing files that have been modified, and describe the changes: -- nfire init. changed from nan to spval to fix problem with hcru_hcru debug + intel runs M models/lnd/clm/src/clm4_5/main/clmtypeInitMod.F90 Machines testing ran on: Only testing hcru_hcru resolutions Yellowstone Tests: The following were run with DEBUG=TRUE and for 1 day initial + 1 day restart hcru_hcru_I_2000_CRUFRC_CLM45_CN_yellowstone_gnu_pioFixed/ PASS hcru_hcru_I_2000_CRUFRC_CLM45_CN_yellowstone_intel_pioFixed/ PASS hcru_hcru_I_2000_CRUFRC_CLM45_CN_yellowstone_pgi_pioFixed/ PASS ERS_D.hcru_hcru.I_2000_CRUFRC_CLM45_CN.yellowstone_intel.125102 PASS ERS_D.hcru_hcru.I_2000_CRUFRC_CLM45_CN.yellowstone_pgi.125128 PASS Frankfurt Tests: The following were run with DEBUG=TRUE and for 1 day initial + 1 day restart hcru_hcru_I_2000_CRUFRC_CLM45_CN_frankfurt_pgi_pioFixed/ PASS hcru_hcru_I_2000_CRUFRC_CLM45_CN_frankfurt_intel_pioFixed/ FAIL initial run (this is likely related to other existing MPI problems on Frankfurt). CLM tag used for the baseline comparison tests if applicable: N/A Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_04 Originator(s): erik (Erik Kluzek) Date: Mon May 13 12:25:14 MDT 2013 One-line Summary: Fix the previous broken tag Purpose of changes: Fix the problems in the clm4_5_03 untested tag. Requirements for tag: Fix bug 1692, 1693 Test level of tag: standard Bugs fixed (include bugzilla ID): 1693 (Misc. issues with clm4_5_03) 1692 (externals screwed up in clm4_5_03) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): externals updated to those in SVN_EXTERNAL_DIRECTORIES List all files eliminated: Remove test/system as replaced by CESM testing models/lnd/clm/test/system -- Delete the whole directory tree List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/src/clm4_0/main/clm_initializeMod.F90 -- fixed screwed up code Machines testing ran on: (Tests in priority order) Standard Tag Pretag *** Standard Tag Posttag ** build-namelist unit tester: yes CESM test lists: yellowstone/aux_clm intel yes yellowstone/aux_clm pgi yes frankfurt/aux_clm intel yes frankfurt/aux_clm pgi yes CLM tag used for the baseline comparison tests if applicable: clm4_5_01 Changes answers relative to baseline: No =============================================================== =============================================================== Tag name: clm4_5_03 Originator(s): erik (Erik Kluzek) Date: Fri May 10 17:29:56 MDT 2013 One-line Summary: Several bug fixes for release, urban and test single point surface datasets Purpose of changes: Some work on IMPORTANT_NOTES file. Fix PTS_MODE restarts from John Truesdale. (implimented, but there are still issues) Fix history change number of tapes on startup issue. Bring in urban single pt surface datasets and single pt test: mexicocityMEX, vancouverCAN , urbanc_alpha, 1x1_tropicAtl, 1x1_smallvilleIA Drydep use before defined problem. Always bypass first two time-steps for CN/BGC. Fix gregorian calendar on history files. Remove two fields on clm45 fpftdata file as per Gordon Bonan. ncd_pio fix from Jim Edwards/Mariana V. set nsegspc=20 for HOMME and high resolution grids. Change documentation on CLM build-namelist -drydep, but keep it default on (will change to off in next tag) Remove a bunch of datm/drv fields in namelist_definition. Fix some issues with Crop and DV that Sam found. Fix a scripts issue with multi-instance. Update RTM (multi-instance fix, allow null grid). Update test list so that CLM45/DV/CROP are exercised. Update scripts/machines tag because of multiple problems. Requirements for tag: fix bug 1488, 1673, 1677, 1682, 1653, 1689, 1690, 1687, 1688, 1685, 1691 Test level of tag: limited! Bugs fixed (include bugzilla ID): 1025 (partial -- implement changes from John Truesdale so SCAM can read global IC files) 1488 (HOMME grids can not use nsegspc=20) 1653 (Calls to PIO are not properly done) 1673 (B compset gregorian calendar not reflected in CLM history) 1677 (Remove bypass_CN_balance_check_on_restart in CLM45) 1682 (Problem starting up CLM with no history files) 1685 (use before define issue in DryDeposition) 1687 (SBN scripts bug) 1688 (misc. issues with new create_test) 1689 (CLM45 dgvm does not build) 1690 (CLM45 CNDV lightning namelist is missing) 1691 (Scripts issue for multi-instance) Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None Describe any changes made to the namelist: Set nsegspc=20 for HOMME and f05/f02 resolutions Set stream_fldfilename_lightng for CLM45/CNDV List any changes to the defaults for the boundary datasets: New single-point test and urban datasets New surface datsets for: mexicocityMEX, vancouverCAN, urbanc_alpha, 1x1_tropicAtl, 1x1_smallvilleIA New fpftdyn for: 1x1_tropicAtl 1850-2005 New pft-physiology file for CLM45 with three fields removed that were NOT being read in (qe25, mp, and resist) Describe any substantial timing or memory changes: None Code reviewed by: self, jedwards/mvertens (fix for bug 1653), jet (fix for bug 1025), slevis (fixes for DV) List any svn externals directories updated (csm_share, mct, etc.): Machines, scripts, rtm Machines to Machines_130509 scripts to scripts4_130510 rtm to rtm1_0_25 List all files eliminated: None List all files added and what they do: None List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/build-namelist ---- Document drydep as if it's off (will actually become off in next tag) M models/lnd/clm/bld/clm.buildnml.csh -- Multi-instance bug fix. M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_0.xml ----- nsegspc for ALL grids is 20 M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_0.xml --- Remove datm/drv namelist crap M models/lnd/clm/bld/namelist_files/namelist_defaults_clm4_5.xml ----- nsegspc for ALL grids is 20 New pft-physiology file, new surface/fpftdyn datasets for single point test and urban Set stream_fldfilename_lightng for CNDV. M models/lnd/clm/bld/namelist_files/namelist_definition_clm4_5.xml --- Remove datm/drv namelist crap M models/lnd/clm/doc/IMPORTANT_NOTES -- updates M models/lnd/clm/src/util_share/clm_time_manager.F90 - Set parameters for calendar type. M models/lnd/clm/src/util_share/ncdio_pio.F90 -------- Fix so that type of data output on read is based on the variable type of the data rather than the type of data on the input file. (from mvertens/jedwards) M models/lnd/clm/src/clm4_5/biogeochem/CNFireMod.F90 ------ Fix so CNDV can build. M models/lnd/clm/src/clm4_5/biogeochem/CNrestMod.F90 ------ Remove bypass_CN_balance_check_on_restart M models/lnd/clm/src/clm4_5/biogeochem/DryDepVelocity.F90 - Fix use before define error. M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ---- On restart set fieldlist and later compare to make sure not screwed up. M models/lnd/clm/src/clm4_5/main/histFileMod.F90 ---------- Make htapes_fieldlist public, check calendar for output files, check that namelist didn't change number of tapes or fields on restart M models/lnd/clm/src/clm4_5/main/clm_driver.F90 ----------- Remove bypass_CN_balance_check_on_restart, NEVER do balance check on first time-step M models/lnd/clm/src/clm4_0/biogeochem/DryDepVelocity.F90 - Fix use before define error. M models/lnd/clm/src/clm4_0/main/histFileMod.F90 ---------- Make htapes_fieldlist public, check calendar for output files, check that namelist didn't change number of tapes or fields on restart M models/lnd/clm/src/clm4_5/main/clm_initializeMod.F90 ---- On restart set fieldlist and later compare to make sure not screwed up. Machines testing ran on: Limited! (watch out for this tag!) I ran preliminary testing, with versions on the cbugfixclm450 ranch. We will fix other issues with the entire package as we find them. CLM tag used for the baseline comparison tests if applicable: clm4_5_02 Changes answers relative to baseline: No bit-for-bit =============================================================== =============================================================== Tag name: clm4_5_02 Originator(s): sacks (Bill Sacks,UCAR/CGD,303-497-1762) Date: Tue May 7 21:04:35 MDT 2013 One-line Summary: make 'shared' tools directory, and other minor tools fixes Purpose of changes: - Make separate 'shared' tools directory, move some tools from the clm4_5 directory into there. - Change interpinic so that htop and hbot are skipped - Change Makefile.common files in tools to use ifort by default on yellowstone, so users can just type 'gmake' without needing to do 'gmake USER_FC=ifort'. For simplicity, this has been implemented by defaulting to ifort for ALL Linux machines. - Fix minor mksurfdata.pl bugs (1669, 1681). Requirements for tag: - fix bug 1669, 1681* - only tools testing needed Test level of tag: tools only Bugs fixed (include bugzilla ID): - 1669: change needed for mksurfdata.pl for smallville (or crop PFT override anyway) - Changes to get mksurfdata.pl working with urban single point datasets Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: None for CLM; tools builds changed to use ifort by default on Linux machines Describe any changes made to the namelist: None List any changes to the defaults for the boundary datasets: None Describe any substantial timing or memory changes: None Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): None List all files eliminated: ========= Moved to 'shared' directory D models/lnd/clm/tools/clm4_5/mkmapdata D models/lnd/clm/tools/clm4_5/mkprocdata_map D models/lnd/clm/tools/clm4_5/ncl_scripts D models/lnd/clm/tools/clm4_5/mkmapgrids List all files added and what they do: ========= Tools moved from clm4_5 directory to shared directory A models/lnd/clm/tools/shared A models/lnd/clm/tools/shared/mkmapdata/mvNimport.sh A models/lnd/clm/tools/shared/mkmapdata/rmdups.ncl A models/lnd/clm/tools/shared/mkmapdata/regridbatch.sh A models/lnd/clm/tools/shared/mkmapdata/createXMLEntries.pl A models/lnd/clm/tools/shared/mkmapdata/mkmapdata.sh A models/lnd/clm/tools/shared/mkmapdata/mkunitymap.ncl A models/lnd/clm/tools/shared/mkmapdata/mknoocnmap.pl A models/lnd/clm/tools/shared/mkmapdata/README A models/lnd/clm/tools/shared/mkmapdata A models/lnd/clm/tools/shared/mkprocdata_map/clm4054_ne30g16_I2000.clm2.h0.2000-01_c121107.nc A models/lnd/clm/tools/shared/mkprocdata_map/map_ne30np4_nomask_to_fv1.9x2.5_nomask_aave_da_c121107.nc A models/lnd/clm/tools/shared/mkprocdata_map/mkprocdata_map_functions.bash A models/lnd/clm/tools/shared/mkprocdata_map/src/mkprocdata_map.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/gridmapMod.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/constMod.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/Makefile.common A models/lnd/clm/tools/shared/mkprocdata_map/src/fmain.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/shr_file_mod.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/nanMod.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/Mkdepends A models/lnd/clm/tools/shared/mkprocdata_map/src/Srcfiles A models/lnd/clm/tools/shared/mkprocdata_map/src/Filepath A models/lnd/clm/tools/shared/mkprocdata_map/src/Makefile A models/lnd/clm/tools/shared/mkprocdata_map/src/fileutils.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src/shr_kind_mod.F90 A models/lnd/clm/tools/shared/mkprocdata_map/src A models/lnd/clm/tools/shared/mkprocdata_map/mkprocdata_map_in A models/lnd/clm/tools/shared/mkprocdata_map/mkprocdata_map_all A models/lnd/clm/tools/shared/mkprocdata_map/clm A models/lnd/clm/tools/shared/mkprocdata_map/mkprocdata_map_wrap A models/lnd/clm/tools/shared/mkprocdata_map/clm4054_f19g16_I2000.clm2.h0.2000-01_c121107.nc A models/lnd/clm/tools/shared/mkprocdata_map/README A models/lnd/clm/tools/shared/mkprocdata_map A models/lnd/clm/tools/shared/ncl_scripts/cprnc.pl A models/lnd/clm/tools/shared/ncl_scripts/getco2_historical.ncl A models/lnd/clm/tools/shared/ncl_scripts/cprnc.ncl A models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.pl A models/lnd/clm/tools/shared/ncl_scripts/getregional_datasets.ncl A models/lnd/clm/tools/shared/ncl_scripts/README A models/lnd/clm/tools/shared/ncl_scripts A models/lnd/clm/tools/shared/mkmapgrids/src/Makefile.common A models/lnd/clm/tools/shared/mkmapgrids/src/domainMod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/shr_sys_mod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/shr_file_mod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/nanMod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/shr_log_mod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/Mkdepends A models/lnd/clm/tools/shared/mkmapgrids/src/Srcfiles A models/lnd/clm/tools/shared/mkmapgrids/src/mkmapgrids.F90 A models/lnd/clm/tools/shared/mkmapgrids/src/Filepath A models/lnd/clm/tools/shared/mkmapgrids/src/Makefile A models/lnd/clm/tools/shared/mkmapgrids/src/shr_kind_mod.F90 A models/lnd/clm/tools/shared/mkmapgrids/src A models/lnd/clm/tools/shared/mkmapgrids/mkmapgrids.namelist A models/lnd/clm/tools/shared/mkmapgrids/mkscripgrid.ncl A models/lnd/clm/tools/shared/mkmapgrids/mkmapgrids.csh A models/lnd/clm/tools/shared/mkmapgrids/README A models/lnd/clm/tools/shared/mkmapgrids ========= Add test for mkmapdata using '-p clm4_0' A models/lnd/clm/test/tools/nl_files/mkmapdata_ne30np4_clm4_0 List all existing files that have been modified, and describe the changes: ========= Point to new 'shared' tools directory where appropriate M models/lnd/clm/tools/clm4_5/mksurfdata_map/mksurfdata.pl - also fix mksurfdata.pl for crop PFT override (bug 1669) - also changes to get mksurfdata.pl working with urban single point datasets (bug 1681) M models/lnd/clm/tools/clm4_0/mksurfdata_map/mksurfdata.pl - also fix mksurfdata.pl for crop PFT override (bug 1669) M models/lnd/clm/test/tools/TBLCFGtools.sh M models/lnd/clm/test/tools/TOPtools.sh M models/lnd/clm/test/tools/TBLscript_tools.sh M models/lnd/clm/test/tools/TBLtools.sh M models/lnd/clm/test/tools/input_tests_master - also add test for mkmapdata using '-p clm4_0' M models/lnd/clm/tools/README M models/lnd/clm/tools/clm4_5/mksurfdata_map/README.developers ========= Put gen_domain in 'shared' tools directory M SVN_EXTERNAL_DIRECTORIES ========= Use ifort by default on yellowstone (and other Linux machines) M models/lnd/clm/tools/clm4_5/interpinic/src/Makefile.common M models/lnd/clm/tools/clm4_5/mksurfdata_map/src/Makefile.common M models/lnd/clm/tools/clm4_0/interpinic/src/Makefile.common M models/lnd/clm/tools/clm4_0/mksurfdata_map/src/Makefile.common ========= Change interpinic so that htop and hbot are skipped M models/lnd/clm/tools/clm4_5/interpinic/src/interpinic.F90 Machines testing ran on: (Tests in priority order) build-namelist unit tester: no CESM test lists: yellowstone/aux_clm intel no frankfurt/aux_clm_int intel no yellowstone/aux_clm pgi no frankfurt/aux_clm intel no test_driver.sh tools testing: lynx interactive: no yellowstone interactive: yes All PASS except for the following expected failures (note that 006 & 008 baselines are expected to always fail): 006 ble14 TBLCFGtools.sh shared gen_domain CFGtools__ds T31.runoptions .......................... rc=4 FAIL 008 ble@4 TBLCFGtools.sh shared gen_domain CFGtools__ds ne30.runoptions ......................... rc=4 FAIL 027 smiS4 TSMscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional ............. rc=6 FAIL 028 bliS4 TBLscript_tools.sh shared ncl_scripts getregional_datasets.pl getregional ............. rc=4 FAIL yellowstone/PTCLM: no CLM tag used for the baseline comparison tests if applicable: clm4_5_01 Changes answers relative to baseline: NO =============================================================== =============================================================== Tag name: clm4_5_01 Originator(s): muszala (Stefan Muszala,UCAR/CGD,303-497-1320) Date: Mon May 6 16:52:27 MDT 2013 One-line Summary: update externals Purpose of changes: update externals to alpha08b Requirements for tag: N/A Test level of tag: std-test Bugs fixed (include bugzilla ID): N/A Known bugs (include bugzilla ID): N/A http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: N/A Describe any changes made to the namelist: N/A List any changes to the defaults for the boundary datasets: N/A Describe any substantial timing or memory changes: N/A Code reviewed by: self List any svn externals directories updated (csm_share, mct, etc.): < scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_130502 < scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130502 < models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_33 --- > scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_130422 > scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130412 > models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_29 13c13 < models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_130502 --- > models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_130405 16c16 < models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_130423 --- > models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_130226 18c18 < models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_130417 --- > models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_130214 20c20 < models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_6_9/pio --- > models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_6_7/pio 23,25c23,25 < tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_130425 < tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130426a < models/lnd/clm/tools/clm4_5/gen_domain https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130426a/gen_domain_files --- > tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_130411 > tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130403 > models/lnd/clm/tools/clm4_5/gen_domain https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130403/gen_domain_files List all files eliminated:N/A List all files added and what they do:N/A List all existing files that have been modified, and describe the changes: M models/lnd/clm/bld/unit_testers/xFail/expectedClmTestFails.xml - clean up test list Machines testing ran on: (Tests in priority order) build-namelist unit tester: OK CESM test lists: yellowstone/aux_clm intel OK (detail of fails that should pass next time) BFAIL ERI_D.f10_f10.I20TRCN.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round FAIL ERS.f19_g16_r01.I1850CLM45CN4Me.nldir_rtmOnFloodOnEffvelOff.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- No matching time found in cprnc? should pass next time. FAIL ERS_D.f19_g16.ICLM45GLCMEC.nldir_glcMEC.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERS_D.f19_g16.ICLM45GLCMEC.nldir_glcMEC.yellowstone_intel.GC.170226.nlcomp -- changes in cism namelist and cism config FAIL ERS_D.f19_g16.IGRCP26CLM45CN.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERS_D.f19_g16.IGRCP26CLM45CN.yellowstone_intel.GC.170226.nlcomp -- changes in cism namelist and cism config FAIL ERS_D.f19_g16.IGRCP26CN.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERS_D.f19_g16.IGRCP26CN.yellowstone_intel.GC.170226.nlcomp -- changes in cism namelist and cism config BFAIL ERS_Ld3_D_P64x1.ne30_g16.ICLM45CN.nldir_default.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL ERS_Ld3_D_P64x16.ne30_g16.ICN.nldir_default.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round FAIL ERS_Lm3.f19_g16.IGRCP60CN.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERS_Lm3.f19_g16.IGRCP60CN.yellowstone_intel.GC.170226.nlcomp -- changes in cism namelist and cism config BFAIL PET_D_P1x30.ne30_g16.ICN.nldir_default.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL SMS_RLA.f45_f45.I.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL SMS_RLA.f45_f45.ICLM45.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL SMS_ROA.f45_f45.I.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL SMS_ROA.f45_f45.ICLM45.yellowstone_intel.GC.170226.compare_hist.clm4_0_81 -- no baseline, should pass next round yellowstone/aux_clm pgi OK (detail of fails that should pass next time) FAIL ERI.f19_g16.IG1850.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERI.f19_g16.IG1850.yellowstone_pgi.GC.170137.nlcomp -- changes in cism namelist and cism_config FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 FAIL ERI.f19_g16.IG1850CLM45.yellowstone_pgi.GC.170137.nlcomp -- changes in cism namelist and cism config BFAIL ERI_D.f10_f10.I20TRCN.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 -- no baseline, should pass next round FAIL SMS.T31_g37.IG4804.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 FAIL SMS.T31_g37.IG4804.yellowstone_pgi.GC.170137.nlcomp -- changes in cism namelist and cism config -- changes in drv_in (ocn_ntreades=2) FAIL SMS.T31_g37.IG4804CLM45.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 -- changes in g2x_Sg_frac01 and g2x_Sg_topo01 BFAIL SMS_RLB.f45_f45.I.yellowstone_pgi.GC.170137.compare_hist.clm4_0_81 -- no baseline, should pass next round BFAIL SMS_RLB.f45_f45.ICLM45.yellowstone_pgi.GC.170137.compare_hist.clm4_0_8 -- no baseline, should pass next round frankfurt/aux_clm intel OK CLM tag used for the baseline comparison tests if applicable: clm4_0_81 Changes answers relative to baseline: yes IF tag changes answers relative to baseline comparison the following should be filled in: Summarize any changes to answers, i.e., - what code configurations: only changes in g2x_Sg_frac01 and g2x_Sg_topo01 - what platforms/compilers: all =============================================================== =============================================================== Tag name: clm4_5_00 Originator(s): erik (Erik Kluzek) Date: Thu May 2 00:20:17 MDT 2013 One-line Summary: Official end to CLM4.5 development for CLM offline Purpose of changes: Changes from clm4_0_54 to now... Compsets and Scripts Changes: Remove ability to set compset file on command line, and use a new expanded compset file definition that allows user to create many compsets on the fly by defining a long name with the "-user_compset" option to "create_newcase". "-user_compset" is in the form of... TIME_DATM[%phys]_CLM[40|45][%phys]_SICE_SOCN_RTM[%phys]_GLC[%phys]_SWAV[_BGC%phys] Where TIME = Time period (e.g. 2000, 20TR, RCP8...) GLC = [CISM1, SGLC] BGC = optional BGC scenario The OPTIONAL %phys attributes specify submodes of the given system So for example ./create_newcase -user_compset 1850_DATM%CRU_CLM45%BGC_SICE_SOCN_RTM_SGLC_SWAV -case cru1850 -res f19_g16 -mach yellowstone -compiler intel will setup a 1850 case at f19 resolution with CRUNCEP forcing with CLM4.5-BGC. Changes for both clm4.0 and clm4.5: * Bug fixes in MEGAN VOC emission fluxes and dry deposition velocities * CRUNCEP is now an option for atmospheric forcing * Change from Sam Levis for CROP to pft-physiology file so that CROP parameter is in Kelvin rather than Celsius. * Don't re-weight pftdyn if weights are essentially identical. CLM4.5 includes the following: * Bring in flood capability to RTM. * Bring LBNL-merge branch on with: vertical soil, Methane, CENTURY, split nitrification, new-lake model. * Modifications to GPP, on gppdev branch, multilayer canopy and then single-layer version that reproduces it. * Crop model updates. Irrigation included with crop model as an option. Fix CNDV-CROP. * Urban model updates, multi-density, urban depth seperate from soil depth, wasteheat to zero. * Bring in permafrostsims09 branch with Sean Swensons's flooding changes. * Update pft-physiology file, change some CN defaults, change min flow slightly in RTM. * Set ponding to zero, acclimation mods from Keith Oleson, a hydrology change from Sean Swenson. * Add active flags, change subgrid weighting convention. * Turn off subgrid topography snow parameterization for glc_mec landunits. * Jinyun photosynthesis change impacting arid regions. * Keith Oleson's photosynthesis change, changes canopy top: triose phosphate util. rate to be dependent on vcmax. * VIC hydrology is an option. * Update mksurfdata_map for CLM4.5 (also add support for glc_nec=36 although we have no datasets for this). * Snow depth averaged over grid-cell (SNOWDP) on history file changed in favor of SNOW_DEPTH (averaged only over snow covered area). * Spinup changes from Charlie Koven from build-time to run-time (spinup now option added to CLM_BLDNML_OPTS as "-spinup on|off"). * Bring the F. Li and S. Levis Fire model for CLMCN and CLMBGC based on Li et al. (2012a,b; 2013). * BSW calculation changed affecting drought phenology and frozen temperature sensitivity (SP, CN, and BGC as well as DV) Test level of tag: doc Bugs fixed (include bugzilla ID): Known bugs (include bugzilla ID): http://bugs.cgd.ucar.edu/ Known Limitations: 1147 (mkgriddata can't straddle over Greenwich) 1025 (SCM mode can NOT use a global finidat file) 1017 (SCM mode can NOT restart) 896 (T62 mode does not work) 701 (svn keyword) 452 (Problem with support of soil-colors != 8 or 20) Describe any changes made to build system: Changes to CLM configure: -phys option to specify clm4_0 or clm4_5 -pergro and -c13 option removed -spinup option removed for CLM4_0 New options for clm4_5: -clm4me, -vichydro, -exlaklayers, -vsoilc_centbgc Describe any changes made to the namelist: For CLM4.0: WRF and 360x720cru resolutions added For CLM4.5: new namelists: popd_streams light_streams clm_hydrology1_inparm clm_soilhydrology_inparm irrigate is a namelist option rather than using different surface datasets New namelist items for clm_inparm: anoxia no_frozen_nitrif_denitrif atm_c14_filename override_bgc_restart_mismatch_dump cryoturb_diffusion_k perchroot decomp_depth_efolding perchroot_altk deepmixing_depthcrit pftspecific_rootingprofile deepmixing_mixfact rootprof_exp exponential_rooting_profile rootprof_exp froz_q10 som_adv_flux hist_wrtch4diag som_diffus lake_melt_icealb spinup_state max_altdepth_cryoturbation surfprof_exp max_depth_cryoturb use_c13 more_vertlayers use_c14 nfix_timeconst use_c14_bombspike List any changes to the defaults for the boundary datasets: All CLM4.5 datasets are new. For CLM4.0, new ne120, ne240, and 360x720cru surface datasets (ne120 ne120 finidat files) new pft-physiology file New history fields: Dozens of new fields for clm4_5. Three new fields for clm4_0: +