================================================================================ This file describes what main-trunk tags were created and why ================================================================================ ================================================================================ Originator: sacks Date: May 03, 2017 Model: cism Version: cism2_1_33 One-line summary: Point to new 4km input file Purpose of changes: Joe Kennedy made a new version of the 4km input file that has a slightly extended grid. The point of this is to have a number of grid cells that can be evenly divided by 2 a number of times, to facilitate changing resolution. This tag points to the new file. There are also some miscellaneous other changes in this tag: (1) Remove support for gland10 and gland5 - gland5UM should be used in place of gland5 - gland10 was just being used for testing; I am now using gland20 for those tests - Point is: These grids weren't being used, and were just a maintenance burden, particularly when we need to make new mapping files. (2) Remove unused GLC_SMB variable (used to be used in CLM, no longer is) (3) Move prealpha & prebeta tests from yellowstone to cheyenne - For the trilinos test: I have replaced it with a non-trilinos test, since trilinos isn't installed on cheyenne, and we don't feel a need to support trilinos for cesm2. - For the pgi test: I have replaced it with an intel test since pgi isn't supported on cheyenne. (4) When using test_coupling, tell CLM that it should allow non-annual dynbal adjustments. This is needed when running CLM in CN/BGC mode. Changes answers relative to previous tag: YES The new file is effectively identical to the old one in the grid cells that overlap. One potential difference is in exactly where topg is submerged to -200m, since we now use a slightly different mask for that submerging; however, even in that case, the new version appears to be effectively the same as the old. Answers change for the following reasons: (1) Variables on the CISM grid cannot be compared for runs at 4km (2) The new 4km mapping files have roundoff-level differences from the old ones, resulting in changes in CLM (which are presumably initially roundoff-level, but then grow; I think the differences arise from small differences in the topographic heights of CLM's columns, leading to differences in downscaling of atmospheric fields). These differences will affect even cases with CISM evolution turned off. (3) There are 3 CLM cells with greater-than-roundoff-level differences in icemask. These CLM cells lay partly outside the old CISM grid, though lie entirely within the new grid. Because icemask is currently mapped with normalization, this led to differences in the remapped icemask field. Currently, I'm pretty sure this doesn't matter, because icemask in CLM is only checked for whether it's > 0. (4) With 2-way coupling, I expect answer changes in the runoff fields after remapping to the ocean, because we're using new mapping files for ALL GRIDS in this respect, to incorporate a fix to the runoff mapping tool. (This change comes from the new version of cime rather than this version of cism per se.) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): - https://github.com/ESMCI/cime/issues/1262 - Remake all glc to ocean mapping files List all modified files, and describe the changes: A cime_config/testdefs/testmods_dirs/cism/test_coupling/user_nl_clm M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M cime_config/testdefs/testlist_cism.xml M cime_config/config_component.xml Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed except for expected baseline comparison failures for 4-km tests, and expected NLCOMP failures. Externals used for testing: cesm2_0_alpha06g, but cime at 4e2cd84 (git describe: billsacks/glc_tavg_v2_n03-2-g4e2cd84) (cism_new_mapping_files branch) cism tag used for baseline comparisons: cism2_1_32 Any other externals that differed in baseline: Baseline had cime at billsacks/glc_tavg_v2_n03 ================================================================================ Originator: sacks Date: May 01, 2017 Model: cism Version: cism2_1_32 One-line summary: Set forcing_start_time correctly when test_coupling is true Purpose of changes: ERI tests that stopped and started mid-year and used test_coupling (e.g., via the allactive-defaultio testmod) were dying because forcing_start_time was being set incorrectly in this case. The changes here fix that problem. In addition to some code changes, this also required moving the test_coupling namelist option into the glc layer (rather than being defined in glad): This is now needed before the glad configuration information is read. I changed glint similarly, even though we don't use glint any more. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: *** Main changes as described above M SVN_EXTERNAL_DIRECTORIES M source_glc/glc_InitMod.F90 M source_glc/glc_constants.F90 M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml *** Add an ERI test that uses the test_coupling testmod: Changes often break the test_coupling testmod, so include an ERI test with this testmod to help catch these problems M cime_config/testdefs/testlist_cism.xml Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed except for expected NLCOMP failures. Externals used for testing: cesm2_0_alpha06g, but cime at billsacks/glc_tavg_v2_n03 (8ce40ab7) cism tag used for baseline comparisons: cism2_1_31 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Apr 29, 2017 Model: cism Version: cism2_1_31 One-line summary: Use correct 'time' on history and restart files Purpose of changes: The time value was wrong in some cases, such as when the REFDATE/STARTDATE was not year 1. These changes should fix that problem. This also changes the baseline date for the 'time' variable to be 0000-01-01, rather than 1-1-1, at Gary Strand's suggestion. There are now two time variables on output files: - 'internal_time' stores CISM's internal time variable, relative to the starting point of this simulation plus the amount of time that elapsed leading up to the restart file (if any) - 'time' stores a more externally-useful version of time; for climate model runs, this should agree with the climate model's time The reason for having these two different variables is: There can be roundoff-level differences upon restart if the internal time variable is modified upon restart. So it is important that this variable is maintained exactly as is when reading a restart file. However, maintaining this value as is isn't always the right thing to do in terms of the climate model's time, so we sometimes need a separate variable to represent that time. It's possible that, with some more thought, these two variables could be combined into one. But having two variables seemed like the easiest and least error-prone solution for now. Changes answers relative to previous tag: YES - just changes answers for the 'time' variable Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES M source_glc/glc_io.F90 Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed except for BASELINE failures in 'time'. Externals used for testing: cesm2_0_alpha06g, but cime at billsacks/glc_tavg_v2_n03 (8ce40ab7) cism tag used for baseline comparisons: cism2_1_30 Any other externals that differed in baseline: none ================================================================================ Originator: mvertens, sacks Date: Apr 24, 2017 Model: cism Version: cism2_1_30 One-line summary: Allow coupler to pass annual-average fields Purpose of changes: Until now, the CESM coupler has been sending CISM daily-average fields, and then CISM has been averaging these fields to annual-averages in order to do its mass balance time step using annual-average SMB (and temperature). This has a number of problems: (1) We could not restart mid-year, because CISM doesn't write the necessary partial accumulation information to its restart files. (2) With Bill Lipscomb's new remapping (bilinear + conservation correction), there were too-large corrections needed when doing the remapping from CLM to CISM on daily fields. (3) This led to a discrepancy between TG and BG/IG compsets: In BG/IG, CISM received the annual average of remapped daily fields; in TG, CISM received the remapped annual average fields. The remapping is not a linear operation, so there would typically be differences in these two runs. This tag, together with some CIME changes, allows the coupler to make annual averages before remapping fields to CISM. This was a bit tricky: CISM still needs to be called once per day in order to support mid-year restarts (CESM can only restart at a frequency when all components are called). So now the GLC calling frequency has been separated from the GLC averaging frequency. Then GLC/CISM is passed a flag saying whether the inputs are currently valid; CISM only accumulates averages of the inputs when they are valid. This tag also includes some other changes that were needed to support mid-year restarts, as well as a check to confirm that you're not trying to write a restart file in a time/configuration when it would lead to non-bfb restarts. Changes answers relative to previous tag: YES Changes answers for non-TG runs Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): - 1611: CISM restarts can only occur on year boundaries This is not completely fixed, but is fixed well enough for standard use cases: if we use the default glc averaging period of yearly, then we can restart mid-year. Eventually, we may want more general support for mid-year restarts even if using a different glc averaging period in the coupler. List all modified files, and describe the changes: *** Main changes, as described above M SVN_EXTERNAL_DIRECTORIES M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_io.F90 M drivers/cpl/glc_comp_mct.F90 *** Add some mid-year restart tests, and change IG test to f10 resolution for efficiency M cime_config/testdefs/testlist_cism.xml *** For test_coupling testmod, need to set GLC_AVG_PERIOD=glc_coupling_period A cime_config/testdefs/testmods_dirs/cism/test_coupling/shell_commands Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed except for expected NLCOMP failures and expected BASELINE failure in IG without test_coupling. (Actually, the only baseline failure is ERS_D_Ld9.f19_g16_gl4.IGCLM45.yellowstone_pgi.cism-override_glc_frac, which just changes answers in x2g_Sl_tsrf and x2g_Flgl_qice, as expected. There would be answer changes in other IG tests, but the one other relevant test [the 5-year ERS test] is changed in this tag, so it shows up as a BFAIL.) Also ran these three tests: ERS_Lm24.T31_g37.B1850Gcheap.yellowstone_intel ERS_Lm30.T31_g37.B1850Gcheap.yellowstone_intel ERS_Lm36.T31_g37.B1850Gcheap.yellowstone_intel They passed except for differences in time and time_bound in the POP history file, which presumably is unrelated to these changes. Also ran aux_glc test suite on yellowstone with GLC_AVG_PERIOD=glc_coupling_period, plus this test which I have removed from the test list: ERS_Ly5.T31_g37_gl4.IGCLM45.yellowstone_intel.clm-glcMEC_long. Tests pass and are bfb with trunk, as expected, except for expected failures in these three tests due to trying to write a mid-year restart file: ERS_D_Ld9.f19_g16_gl4.IGCLM45.yellowstone_pgi.cism-override_glc_frac ERI_Lm24.f10_f10.IGCLM45.yellowstone_intel.clm-glcMEC_long ERS_Lm24.f10_f10.IGCLM45.yellowstone_intel.clm-glcMEC_long Externals used for testing: cesm2_0_alpha06g, but cime at billsacks/glc_tavg_v2_n03 (8ce40ab7). cism tag used for baseline comparisons: cism2_1_29 Any other externals that differed in baseline: Baseline used cesm2_0_alpha06g ================================================================================ Originator: sacks Date: Dec 30, 2016 Model: cism Version: cism2_1_29 One-line summary: Add two tests to the aux_glc test suite Purpose of changes: Add two tests with ice evolution turned off, since we weren't testing this configuration, which is the most common configuration in coupled CESM runs. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: A cime_config/testdefs/testmods_dirs/cism/noevolve/shell_commands A cime_config/testdefs/testmods_dirs/cism/noevolve/README A cime_config/testdefs/testmods_dirs/cism/noevolve M cime_config/testdefs/testlist_cism.xml Summary of testing: Just ran the two new tests: ERS_D_Ly3.f09_g16_gl5.TG1.yellowstone_gnu.cism-noevolve ERS_D_Ly3.f09_g16_gl4.TG.yellowstone_pgi.cism-noevolve Externals used for testing: cesm2_0_alpha05b with cime at cism4km_new_grid cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Dec 29, 2016 Model: cism Version: cism2_1_28 One-line summary: Just update ChangeLog entry for cism2_1_27 Purpose of changes: I forgot to mention a few things in the ChangeLog entry for cism2_1_27 Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M ChangeLog Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Dec 29, 2016 Model: cism Version: cism2_1_27 One-line summary: Use new 4-km input file, with epsg3413 projection Purpose of changes: Joe Kennedy has created a new 4-km input file, which uses the epsg3413 projection. This projection is more commonly used, so facilitates comparisons with observations, etc. This tag switches to this new input file. In addition to changing the projection, this input file also updates the glacier / ice sheet cover - particularly, including peripheral glaciers. In addition, this new file has: (1) spatially-varying bheatflx; (2) no tempstag field, so it is not spun-up in this respect; (3) no beta field. Some related changes in this tag are: (1) Change default which_ho_babc to 4 for 4km runs, because the new input file does not have a beta field. (2) No longer set projection section in cism.config file. This seems to no longer be needed now that glint is gone, and the default values are wrong for this new 4km file. And some unrelated changes in this tag are: (1) Fix from Jim Edwards to allow build-namelist to work with latest version of cime, where NTASKS and NINST have been reworked (2) Rename xmlchange_cmnds to shell_commands in a testmod: xmlchange_cmnds has been deprecated This tag depends on a new version of cime (currently on the cism4km_new_grid branch, which will soon have a PR to master). This cime version has new mapping files to go with the new 4km grid. Changes answers relative to previous tag: YES Changes answers for all cases that use the 4km grid Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M bld/namelist_files/namelist_definition_cism.xml M bld/namelist_files/namelist_defaults_cism.xml M bld/build-namelist D cime_config/testdefs/testmods_dirs/cism/oneway/xmlchange_cmnds A cime_config/testdefs/testmods_dirs/cism/oneway/shell_commands Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed except for expected NLCOMP failures, and expected BASELINE failures in 4km tests: FAIL ERS_Ly3.f09_g16_gl4.TG.yellowstone_intel BASELINE FAIL ERS_Ly5.T31_g37_gl4.IGCLM45.yellowstone_intel.clm-glcMEC_long BASELINE FAIL ERS_D_Ld9.f19_g16_gl4.IGCLM45.yellowstone_pgi.cism-override_glc_frac BASELINE In addition, ran B compset tests: (1) SMS_D_Ld5.T31_g37_gl4.B1850C5L45BGCRG.yellowstone_intel.allactive-defaultio (2) ./create_newcase -case test_b_mapping_1228a -compset B1850 -res f09_g16_gl4 with CISM_EVOLVE=TRUE, and test_coupling = .true. Externals used for testing: cesm2_0_alpha05b with cime at cism4km_new_grid cism tag used for baseline comparisons: cism2_1_25 Any other externals that differed in baseline: Baseline used out-of-the-box cesm2_0_alpha05b (so an earlier version of cime) ================================================================================ Originator: sacks Date: Dec 6, 2016 Model: cism Version: cism2_1_26 One-line summary: Change time units to common_year Purpose of changes: CF/Udunits defines common_year to be 365 days, which is what we use. ('year' means exactly 365.242198781 days.) This change is needed for the post-processing tools. Also change time_vars.def for this and the previous change, although it doesn't look like that file is used anywhere. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: *** remove_zero_gcm_fluxes_n03 -> remove_zero_gcm_fluxes_n04 M SVN_EXTERNAL_DIRECTORIES Summary of testing: aux_glc tests just on hobart-nag Externals used for testing: clm4_5_12_r203 cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Nov 15, 2016 Model: cism Version: cism2_1_25 One-line summary: Fix calendar attribute on hist files; remove debug print statements Purpose of changes: (1) Fix calendar attribute on hist files; this is needed by post-processing tools. For now, assume a noleap calendar, since the use of a 365-day year is hard-coded throughout CISM. (This change came in via an update in the glimmer-cism external.) (2) Remove some write statements that were being done by all tasks (both in the glimmer-cism external and in the glc code), in order to reduce output to cesm.log. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: *** remove_zero_gcm_fluxes_n02 -> remove_zero_gcm_fluxes_n03 M SVN_EXTERNAL_DIRECTORIES *** This module was not being used consistently - e.g., the error flag was never being set. It was just leading to unnecessary prints in finalization from all tasks. D source_glc/glc_ErrorMod.F90 *** Remove uses of glc_ErrorMod and a few other write statements that were called from all tasks M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_FinalMod.F90 Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed. Externals used for testing: clm4_5_12_r203 cism tag used for baseline comparisons: cism2_1_24 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Oct 29, 2016 Model: cism Version: cism2_1_24 One-line summary: Remove problematic diagnostic prints Purpose of changes: Updates the glimmer-cism external to remove some diagnostic prints that could lead to array out-of-bounds errors when restarting in debug mode. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Full aux_glc test suite on yellowstone and hobart. All tests passed. Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cism2_default_tags/cism2_default_n02_clm4_5_12_r198 cism tag used for baseline comparisons: cism2_1_23 Any other externals that differed in baseline: essentially the same: https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cism2_default_tags/cism2_default_n02_clm4_5_12_r198 ================================================================================ Originator: sacks Date: Oct 27, 2016 Model: cism Version: cism2_1_23 One-line summary: Change TG compset aliases, make CISM2 the default Purpose of changes: (1) Change TG compset aliases to follow the new CESM convention, putting 'G' at the end - so, e.g., rather than TG1850 we now have T1850G (2) Make CISM2 the default in compset aliases - so, e.g., TG means CISM2, whereas TG1 means CISM1 Changes answers relative to previous tag: NO - but caveat: due to compset renaming, comparisons against baselines either BFAIL or FAIL, unless you do the appropriate sym linking. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M cime_config/config_component.xml M cime_config/config_compsets.xml M cime_config/testdefs/testlist_cism.xml Summary of testing: Ran aux_glc test suite, just on yellowstone All tests passed except for expected MEMLEAKs and BFAILs Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cism2_default_tags/cism2_default_n01_clm4_5_12_r198 Cism tag used for baseline comparisons: cism2_1_22 Any other externals that differed in baseline: Used clm4_5_12_r198 Made sym links in baseline directory, applying the compset renames. However, IG tests that used to use CISM1 still BFAILed (as expected), because the new CLM branch has been changed to use CISM2-4km rather than CISM1-5km in its out-of-the-box compsets. ================================================================================ Originator: sacks Date: Oct 27, 2016 Model: cism Version: cism2_1_22 One-line summary: Use new 4km input file, tweak test list Purpose of changes: (1) Point to a new input file for cism2 4km runs. This new file is based on the file we are currently using for the branch that will become cism2.1. In particular, it uses thk and topg from that new file, and the set of fields present on the file match that new file pretty closely. However, some aspects of the old file are maintained for now, so that the only differences in an out-of-the-box 4km run are due to the differences in thk and topg. This file was created like this: ncks -x -v bheatflx greenland_4km_2015_06_03.mcb_c160418.nc greenland_4km_2015_06_03.mcb_trunk_c161025.nc ncks -A -v bheatflx,beta,tempstag greenland_4km.glissade.10kyr.beta6.SSacab_c150415a.nc greenland_4km_2015_06_03.mcb_trunk_c161025.nc (2) Tweak the xml test list (a) Update to v2 testlist format (b) Add wallclock times to tests (c) Add memleak_tolerance for a couple of tests (d) A few small tweaks to the aux_glc and other test lists: remove some unneeded tests, remove titan, change hopper to bluewaters Changes answers relative to previous tag: YES Changes answers for cism2 at 4km; other configurations are bit-for-bit Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Point to new 4km input file M bld/namelist_files/namelist_defaults_cism.xml *** Tweak test list, as noted above M cime_config/testdefs/testlist_cism.xml *** Update ChangeLog for last tag with some additional test results M ChangeLog Summary of testing: Ran full aux_glc test suite: yellowstone and hobart. All tests passed except for expected MEMLEAKs (now added to the testlist as new memleak_tolerance values) and expected answer change for CISM2 at 4km. Also ran SMS_Ly3_P24x1.f09_g16_gl4.TGIS2.hobart_intel with comparison to baseline for a tweaked input file, which was similar to the new one, but with thk and topg copied from the old file: ncks -x -v bheatflx,thk,topg greenland_4km_2015_06_03.mcb_c160418.nc greenland_4km_2015_06_03.mcb_trunk_temporary_c161024a.nc ncks -A -v bheatflx,thk,topg,beta,tempstag greenland_4km.glissade.10kyr.beta6.SSacab_c150415a.nc greenland_4km_2015_06_03.mcb_trunk_temporary_c161024a.nc This was bit-for-bit with the trunk, confirming that the differences in this new tag arise only from differences in thk and topg. Externals used for testing: cism2_1_21 cism tag used for baseline comparisons: cesm2_0_beta02 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Sept 27, 2016 Model: cism Version: cism2_1_21 One-line summary: Change extension on initial history file Purpose of changes: The glc code is set up to output a history file in initialization, giving the results of the initial, diagnostic solve. With cime5, this causes failures to be reported for some tests (particularly ERI tests, and maybe some others), where there is an initial history file for one run that has no counterpart in the other run of a multi-run test. This tag implements a work-around for this problem by changing the extension of the initial history file to .initial_hist rather than .h: The test system only picks up .h files in doing its comparisons. Changes answers relative to previous tag: NO However, this test is reported as a failure: FAIL ERS_D_Ld9.f19_g16_gl5.IGCLM45.yellowstone_pgi.cism-override_glc_frac COMPARE_baseline simply because there is no longer a cism.h file in the new code. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 2355 - need to change extension on initial history file so it's not picked up in CESM testing List all modified files, and describe the changes: M source_glc/history_tape_base.F90 M source_glc/glc_history.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_io.F90 Summary of testing: Ran full aux_glc test suite: yellowstone and hobart. All tests passed, except for these expected failures: FAIL ERS_Ly5.T31_g37_gl5.IGCLM45.yellowstone_intel.clm-glcMEC_long MEMLEAK memleak detected, memory went from 504.810000 to 590.900000 in 11129 days - seems to be a false positive (i.e., memory was expected to increase like this) FAIL SMS_D_Ld5_P24x1.T31_g37_gl20.IGCLM45IS2.hobart_nag.cism-test_coupling MEMLEAK memleak detected, memory went from 195.220000 to 219.040000 in 3 days - I'm not concerned about this memleak failure for a test_coupling test (a testing-only configuration) on nag (a testing-only compiler) and this MEMCOMP failure, which I'm ignoring since it's the only one: FAIL ERS_Ly11.f09_g16_gl20.TG1850IS2.yellowstone_gnu MEMCOMP Error: Memory usage increase > 10% from baseline Externals used for testing: cesm2_0_beta01, with cime at db03e895f4c54428ff207dc1b85a57e2ecb73be9 cism tag used for baseline comparisons: cism2_1_20 Any other externals that differed in baseline: same, but cime at f55da30e17efbbae8cb1c026fc8f32e406ee9934 ================================================================================ Originator: sacks Date: Sept 13, 2016 Model: cism Version: cism2_1_20 One-line summary: Exclude some history fields when ice evolution is turned off Purpose of changes: A few diagnostic history fields do not restart properly when ice evolution is turned off: uflx and vflx for cism1, and beta_internal for cism2 (see also https://github.com/ACME-Climate/cism-piscees/issues/59 for the beta_internal issue). Since these diagnostic fields are not important with ice evolution turned off, this tag implements a workaround where we simply do not add these fields to the default list of history fields if running with the CISM_EVOLVE xml variable set to FALSE. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml Summary of testing: Ran aux_glc test suite just on yellowstone. All tests passed, except for this BFAIL: BFAIL (baseline directory /glade/scratch/sacks/cesm_baselines/cesm1_5_beta07/SMS_Ly5_N2_P2_D.f09_g16_gl5.TG1850.yellowstone_intel does not exist) Externals used for testing: cesm1_5_beta07 cism tag used for baseline comparisons: cism2_1_17 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Sept 12, 2016 Model: cism Version: cism2_1_19 One-line summary: Remove apply_to_multiinstance testmods Purpose of changes: The apply_to_multiinstance testmods are no longer needed with cime5 Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: D cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance D cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance/README D cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance/shell_commands D cime_config/testdefs/testmods_dirs/cism/override_glc_frac/include_user_mods D cime_config/testdefs/testmods_dirs/cism/test_coupling/include_user_mods D cime_config/testdefs/testmods_dirs/cism/trilinos/include_user_mods Summary of testing: None! Externals used for testing: n/a cism tag used for baseline comparisons: n/a Any other externals that differed in baseline: n/a ================================================================================ Originator: sacks Date: June 30, 2016 Model: cism Version: cism2_1_18 One-line summary: Remove some tests Purpose of changes: Remove ESMF and PEA tests. We are no longer supporting ESMF, so I am removing CME and _E tests. In most cases, the fundamental configuration covered by these tests was already covered by one or more other tests. In one case, though, I changed a CME test to an SMS test: CME_Ly5_N2_P2_D.f09_g16_gl5.TG1850.yellowstone_intel -> SMS_Ly5_N2_P2_D.f09_g16_gl5.TG1850.yellowstone_intel # include a multi-instance debug test Also, temporarily removing PEA tests because they are not supported by ESMCI's CIME. I will add them back in later: PEA_P1.f09_g16_gl5.TGHIST.yellowstone_pgi PEA_P1_Ly2.f09_g16_gl20.TGIS2.yellowstone_intel # needs to be at least 2 years for there to be enough cpl fields for the cpl log comparison to work Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M cime_config/testdefs/testlist_cism.xml Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: May 19, 2016 Model: cism Version: cism2_1_17 One-line summary: Split override_delay into 3 separate options Purpose of changes: Split override_delay into 3 separate options: increase, decrease and rearrange override delays. This allows us to do tests like: increase glacier area to 100% over 5 days, then wait 5 days, then decrease glacier area to 0% Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): none List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M cime_config/testdefs/testmods_dirs/cism/override_glc_frac/user_nl_cism M source_glc/glc_override_frac.F90 Summary of testing: Ran full aux_glc test suite: yellowstone & hobart. All tests passed. Externals used for testing: cesm1_5_beta05 cism tag used for baseline comparisons: cism2_1_16 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: April 12, 2016 Model: cism Version: cism2_1_16 One-line summary: Make allow_leapyear depend on CALENDAR Purpose of changes: Logic for allow_leapyear setting in cism_in was wrong. This tag fixes it. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 1562: should allow_leapyear default be based on CALENDAR? List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Ran aux_glc test suite on yellowstone - NOT on hobart. All tests passed. Externals used for testing: cesm1_5_beta05 cism tag used for baseline comparisons: cism2_1_15 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: March 5, 2016 Model: cism Version: cism2_1_15 One-line summary: Set glcocn_present and glcice_present to false for one-way-coupling Purpose of changes: For one-way coupling (with zero_gcm_fluxes = .true.), fluxes sent to ocn and ice are all zero. Thus, there is no reason for the coupler to do glc -> ocn or glc -> ice mapping. By setting glcocn_present and glcice_present to .false. in this case, we make it easier for users to run a one-way-coupled case with some new ocean grid, for which they do not have glc -> ocn mapping files. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M drivers/cpl/glc_coupling_flags.F90 Summary of testing: Ran full aux_glc test suite: yellowstone & hobart. All tests passed. Also ran the following tests, with comparison against cesm1_5_beta05 (using cism2_1_13 but with glimmer-cism at the head of piscees/develop): SMS_Ld5.T31_g37_gl20.B1850C5L45BGCRG2.yellowstone_intel.allactive-cism-test_coupling - passes and bit-for-bit SMS_D_Ld5.T31_g37_gl20.B1850C5L45BGCRG2.yellowstone_intel.allactive-cism-test_coupling with zero_gcm_fluxes - passes and bit-for-bit SMS_D_Ld5.T31_g37_gl20.B1850C5L45BGCRG2.yellowstone_intel.allactive-cism-test_coupling with zero_gcm_fluxes, and in my new code, removed all instances of GLC2OCN_RMAPNAME from config_grids.xml (resulting in glc2ocn_rmapname = idmap) - passes and bit-for-bit Externals used for testing: cesm1_5_beta05 cism tag used for baseline comparisons: cism2_1_14 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: March 4, 2016 Model: cism Version: cism2_1_14 One-line summary: Move zero_gcm_fluxes handling to GLC Purpose of changes: Previously, CISM (particularly glad) was responsible for handling zero_gcm_fluxes: zeroing out flux variables and setting icemask_coupled_fluxes. I have now moved that handling into the GLC layer. The main motivation for doing this is that, in the next tag, I plan to add more logic based on zero_gcm_fluxes: has_ocn_coupling and has_ice_coupling will be false if zero_gcm_fluxes is true. It will be easier to make sure everything is done consistently if all of the zero_gcm_fluxes handling is moved into GLC. Specifically, it seems this will make it less likely that we would set has_ocn_coupling and has_ice_coupling to false when they should really be true for some reason. Possible downside: If we have multiple ice instances (e.g., Greenland + Antarctica): Previously, I think I had been imagining that we would have one cism.config for each of them, but potentially only a single cism_in file. But if we want zero_gcm_fluxes to be able to differ for each instance, then we would need to have separate cism_in files for each instance (or have an array of zero_gcm_fluxes values in the single cism_in file). I'll need to come back to this. Also refactored the setting of indexing-related stuff: Moved behavior into glc_indexing.F90 Also updated CISM external to match latest version of piscees/develop. Changes answers relative to previous tag: YES Roundoff-level changes just in the acab field in the cism history file. I have confirmed that these changes are due to previous changes in piscees/develop, not due to the changes in this refactoring. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Renamed, and moved behavior into here D source_glc/glc_indexing_info.F90 A source_glc/glc_indexing.F90 M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 M drivers/cpl/glc_import_export.F90 M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M bld/build-namelist M source_glc/glc_route_ice_runoff.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_fields.F90 M source_glc/glc_constants.F90 *** Updates needed for moving zero_gcm_fluxes into GLC. While I was at it, updated to latest version of piscees/develop. M SVN_EXTERNAL_DIRECTORIES *** Added a one-way coupling test with CISM2, to make sure logic works in parallel cases M cime_config/testdefs/testlist_cism.xml Summary of testing: Ran full aux_glc test suite: yellowstone & hobart. All tests passed. CME_Ly3.f09_g16_gl20.TGRCP85IS2.yellowstone_intel CFAILed as part of a test suite, but passed when run alone (with build on login node) - chalking it up to a cime issue Externals used for testing: cesm1_5_beta05 cism tag used for baseline comparisons: cism2_1_13 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Jan 13, 2016 Model: cism Version: cism2_1_13 One-line summary: Clean up ChangeLog entry for last tag Purpose of changes: Just minor ChangeLog cleanup Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M ChangeLog Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Jan 12, 2016 Model: cism Version: cism2_1_12 One-line summary: Port CISM to NAG Purpose of changes: Port CISM to NAG Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Point to new external, which is ported to NAG M SVN_EXTERNAL_DIRECTORIES *** Introduce hobart_nag tests, and fix test that has been CFAILing for a while M cime_config/testdefs/testlist_cism.xml Summary of testing: Ran full aux_glc test suite: yellowstone & hobart. All tests passed. Externals used for testing: cesm1_5_beta04, with cime at branch cism_nag_port (git describe: cime4.3.10-3-g3de3528) cism tag used for baseline comparisons: cism2_1_11 Any other externals that differed in baseline: cime may have been the version out-of-the-box in cesm1_5_beta04 ================================================================================ Originator: sacks Date: Jan 7, 2016 Model: cism Version: cism2_1_11 One-line summary: Update external, tweak some namelist items Purpose of changes: (1) Update to latest version of piscees/develop (2) For cism2, exclude uflx and vflx history fields (these aren't set) (3) For cism2, include beta_internal history field (4) Add ho_beta_const to namelist_definition, so it can be changed via user_nl_cism Changes answers relative to previous tag: NO however, there are namelist changes Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M SVN_EXTERNAL_DIRECTORIES Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed, EXCEPT this expected failure: CFAIL ERS_Ly20_E.f09_g16_gl5.TGHIST.yellowstone_intel Externals used for testing: cesm1_5_beta02, but with cime at 1e5b15f424d9d7bc57ef5158c3da7bcc0ef02fc3 (return_reverted_drv branch) cism tag used for baseline comparisons: cism2_1_10 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Nov 13, 2015 Model: cism Version: cism2_1_10 One-line summary: Update external, change some higher-order config settings Purpose of changes: Update to a version of CISM and config setting changes which, together, allow for long runs of CISM2. Changes answers relative to previous tag: YES Changes answers for all CISM2 runs (but not CISM1) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: ========= Point to Bill Lipscomb's branch (not yet merged to 'develop') M SVN_EXTERNAL_DIRECTORIES ========= New config settings, and changed defaults for others M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml ========= In aux_glc test list, change 1-year SMS test of 4-km CISM2 to 3-year ERS test: performance of CISM2 has improved so that this is feasible (just about 2-min run time). Also add 4-km tests back to prealpha / prebeta test suites. M cime_config/testdefs/testlist_cism.xml Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed, EXCEPT this expected failure: CFAIL ERS_Ly20_E.f09_g16_gl5.TGHIST.yellowstone_intel.GC.1113-0634.intel Externals used for testing: cesm1_5_beta02, but with cime at 1e5b15f424d9d7bc57ef5158c3da7bcc0ef02fc3 (return_reverted_drv branch) cism tag used for baseline comparisons: cism2_1_08 Any other externals that differed in baseline: cime at slightly different version ================================================================================ Originator: sacks Date: Nov 8, 2015 Model: cism Version: cism2_1_09 One-line summary: Add a single-processor TG test to the prealpha test suite Purpose of changes: I realized we don't have any single-processor TG tests in the prealpha or prebeta test suites, and recent CIME changes have broken these. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M cime_config/testdefs/testlist_cism.xml Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Oct 15, 2015 Model: cism Version: cism2_1_08 One-line summary: Add a compset option for turning off ice evolution Purpose of changes: Allow turning off ice evolution (setting evolve_ice to 0 and zero_gcm_fluxes to 1) via a compset option (CISM1%NOEVOLVE or CISM2%NOEVOLVE in the compset long name). The purpose of this is to provide out-of-the-box settings for the CESM2-dev compset, without needing to use a user_mods_dir. Longer-term, I have been planning to rework the possible out-of-the-box options, as part of turning CISM on by default. So, longer-term, I should either: (1) Back this out, replacing it with whatever my new scheme is OR (2) Keep something essentially like this long-term, as the mechanism to turn ice evolution on/off. In this case, I should probably make the default to NOT evolve, and then allow CISM%EVOLVE in the name. (or maybe require explicitly setting CISM%EVOLVE or NOEVOLVE in the compset long name???) I think my earlier thinking was to not make the setting of ice evolution be a compset-level option (because this is simple enough to change after setting up a compset), but now that I have this compset-level thing in place, I *may* want to revisit that thinking. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: ========= Main change described here M bld/namelist_files/namelist_defaults_cism.xml M bld/build-namelist M cime_config/config_component.xml ========= Unrelated change: fix for cime3 M cime_config/buildnml Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed, EXCEPT these expected failures: CFAIL ERS_Ly20_E.f09_g16_gl5.TGHIST.yellowstone_intel - See note under cism2_1_04 RUN SMS_Ly1.f09_g16_gl4.TGIS2.yellowstone_intel - 4km Greenland runs are now dying due to a CFL violation RUN ERS_Ly11.f09_g16_gl20.TG1850IS2.yellowstone_gnu.GC.1015-1613.gnu - Dies with MCT attribute array length mismatch error, which has been fixed in a more recent cime tag than the one used here Externals used for testing: cime https://github.com/CESM-Development/cime/tags/cime3.0.5 components/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/cime_refactor_tags/cime_refactor_n03_cam5_4_11/components/cam components/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cime_refactor_tags/cime_refactor_n02_clm4_5_1_r119/components/clm components/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice5_20150921 components/pop https://svn-ccsm-models.cgd.ucar.edu/pop2/branch_tags/cime_refactor_tags/cime_refactor_n02_cesm_pop_2_1_20150810 components/aquap https://svn-ccsm-models.cgd.ucar.edu/aquap/branch_tags/cime_refactor_tags/cime_refactor_n01_aquap_150301 components/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/branch_tags/cime_refactor_tags/cime_refactor_n01_rtm1_0_52 components/ww3 https://svn-ccsm-models.cgd.ucar.edu/ww3/branch_tags/cime_refactor_tags/cime_refactor_n01_ww3_150311 cism tag used for baseline comparisons: cism2_1_06 Any other externals that differed in baseline: (same) ================================================================================ Originator: sacks Date: Sept 25, 2015 Model: cism Version: cism2_1_07 One-line summary: Remove failing tests from prealpha / prebeta test lists Purpose of changes: 4km CISM2 tests are currently failing. Thus, temporarily replace them with 20-km tests. I am using _D tests for the replacement tests so that the replacement tests match tests that I have in the aux_glc test suite. Changes answers relative to previous tag: Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M cime_config/testdefs/testlist_cism.xml prealpha: 6c6 < SMS_Ly1.f09_g16_gl4.TGIS2.yellowstone_intel # include one short test of the typical production resolution for CISM2 --- > SMS_D_Ly1.f09_g16_gl20.TGIS2.yellowstone_intel prebeta: 8c8 < SMS_Ly1.f09_g16_gl4.TGIS2.hopper_pgi # included simply to have a test on hopper --- > SMS_D_Ly1.f09_g16_gl20.TGIS2.hopper_pgi # included simply to have a test on hopper Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Sept 25, 2015 Model: cism Version: cism2_1_06 One-line summary: Update external, change some higher-order config settings Purpose of changes: This brings in a version of the CISM code with extensive changes, largely from Bill Lipscomb to fix problems with glissade and make it more robust. The version we had previously been pointing to was about 10 months out-of-date with respect to the PISCEES repository. The new version is up-to-date with 'develop' in the PISCEES repository. In addition, this changes some higher-order config settings - both adding some new options and changing default values for some options, to values suggested by Steve Price, Lauren Vargo and Jeremy Fyke. Finally, this fixes a bug in the cismIO generation script. Changes answers relative to previous tag: YES Changes answers for CISM1 as well as CISM2 compsets Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M cime_config/cism.template M SVN_EXTERNAL_DIRECTORIES move_glint_to_cpl_n02 -> cism2_150925 Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed, EXCEPT: CFAIL ERS_Ly20_E.f09_g16_gl5.TGHIST.yellowstone_intel - See note under cism2_1_04 RUN SMS_Ly1.f09_g16_gl4.TGIS2.yellowstone_intel - 4km Greenland runs are now dying due to a CFL violation Externals used for testing: cime https://github.com/CESM-Development/cime/tags/cime3.0.5 components/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/cime_refactor_tags/cime_refactor_n03_cam5_4_11/components/cam components/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cime_refactor_tags/cime_refactor_n02_clm4_5_1_r119/components/clm components/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice5_20150921 components/pop https://svn-ccsm-models.cgd.ucar.edu/pop2/branch_tags/cime_refactor_tags/cime_refactor_n02_cesm_pop_2_1_20150810 components/aquap https://svn-ccsm-models.cgd.ucar.edu/aquap/branch_tags/cime_refactor_tags/cime_refactor_n01_aquap_150301 components/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/branch_tags/cime_refactor_tags/cime_refactor_n01_rtm1_0_52 components/ww3 https://svn-ccsm-models.cgd.ucar.edu/ww3/branch_tags/cime_refactor_tags/cime_refactor_n01_ww3_150311 cism tag used for baseline comparisons: cism2_1_04 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Sept 23, 2015 Model: cism Version: cism2_1_05 One-line summary: Amend ChangeLog for cism2_1_04 Purpose of changes: Amend ChangeLog for cism2_1_04 Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: ========= Note that a test failed M ChangeLog Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: mvertens (brought to trunk by sacks) Date: Sept 23, 2015 Model: cism Version: cism2_1_04 One-line summary: Updates for cime3 Purpose of changes: Rework of directory structure and build scripts to work with cime3 (de-CESMization of CIME). Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: D cimetest/testmods_dirs/cism/oneway/README D cimetest/testmods_dirs/cism/oneway/xmlchange_cmnds D cimetest/testmods_dirs/cism/oneway D cimetest/testmods_dirs/cism/trilinos/README D cimetest/testmods_dirs/cism/trilinos/user_nl_cism D cimetest/testmods_dirs/cism/trilinos/include_user_mods D cimetest/testmods_dirs/cism/trilinos/shell_commands D cimetest/testmods_dirs/cism/trilinos D cimetest/testmods_dirs/cism/apply_to_multiinstance/README D cimetest/testmods_dirs/cism/apply_to_multiinstance/shell_commands D cimetest/testmods_dirs/cism/apply_to_multiinstance D cimetest/testmods_dirs/cism/override_glc_frac/include_user_mods D cimetest/testmods_dirs/cism/override_glc_frac/user_nl_cism D cimetest/testmods_dirs/cism/override_glc_frac D cimetest/testmods_dirs/cism/test_coupling/user_nl_cism D cimetest/testmods_dirs/cism/test_coupling/include_user_mods D cimetest/testmods_dirs/cism/test_coupling D cimetest/testmods_dirs/cism D cimetest/testmods_dirs D cimetest/testlist_cism.xml D cimetest D bld/cism.buildlib D bld/cism.buildnml D bld/user_nl_cism D bld/cism.template M bld/README.build-namelist M bld/build-namelist A cime_config/cism.template A cime_config/buildlib A cime_config/buildnml A cime_config/config_component.xml A cime_config/config_pes.xml A cime_config/config_compsets.xml A cime_config/testdefs/testmods_dirs/cism/oneway/xmlchange_cmnds A cime_config/testdefs/testmods_dirs/cism/oneway/README A cime_config/testdefs/testmods_dirs/cism/oneway A cime_config/testdefs/testmods_dirs/cism/trilinos/user_nl_cism A cime_config/testdefs/testmods_dirs/cism/trilinos/include_user_mods A cime_config/testdefs/testmods_dirs/cism/trilinos/shell_commands A cime_config/testdefs/testmods_dirs/cism/trilinos/README A cime_config/testdefs/testmods_dirs/cism/trilinos A cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance/README A cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance/shell_commands A cime_config/testdefs/testmods_dirs/cism/apply_to_multiinstance A cime_config/testdefs/testmods_dirs/cism/override_glc_frac/user_nl_cism A cime_config/testdefs/testmods_dirs/cism/override_glc_frac/include_user_mods A cime_config/testdefs/testmods_dirs/cism/override_glc_frac A cime_config/testdefs/testmods_dirs/cism/test_coupling/user_nl_cism A cime_config/testdefs/testmods_dirs/cism/test_coupling/include_user_mods A cime_config/testdefs/testmods_dirs/cism/test_coupling A cime_config/testdefs/testmods_dirs/cism A cime_config/testdefs/testmods_dirs A cime_config/testdefs/testlist_cism.xml A cime_config/testdefs A cime_config/user_nl_cism A cime_config Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed, EXCEPT: CFAIL ERS_Ly20_E.f09_g16_gl5.TGHIST.yellowstone_intel - Failed because single-processor tests are now using mpi-serial, which is incompatible with ESMF. Waiting on a resolution of this (either go back to using a real mpi library for single-processor tests, or else change this test to explicitly specify a MPI library). Externals used for testing: cime https://github.com/CESM-Development/cime/tags/cime3.0.5 components/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/cime_refactor_tags/cime_refactor_n03_cam5_4_11/components/cam components/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cime_refactor_tags/cime_refactor_n02_clm4_5_1_r119/components/clm components/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice5_20150921 components/pop https://svn-ccsm-models.cgd.ucar.edu/pop2/branch_tags/cime_refactor_tags/cime_refactor_n02_cesm_pop_2_1_20150810 components/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branch_tags/cime_refactor_tags/cime_refactor_n02_cism2_1_03 components/aquap https://svn-ccsm-models.cgd.ucar.edu/aquap/branch_tags/cime_refactor_tags/cime_refactor_n01_aquap_150301 components/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/branch_tags/cime_refactor_tags/cime_refactor_n01_rtm1_0_52 components/ww3 https://svn-ccsm-models.cgd.ucar.edu/ww3/branch_tags/cime_refactor_tags/cime_refactor_n01_ww3_150311 cism tag used for baseline comparisons: cism2_1_03 Any other externals that differed in baseline: baseline used cesm1_4_beta07 ================================================================================ Originator: sacks Date: Sept 13, 2015 Model: cism Version: cism2_1_03 One-line summary: Write history file at initialization Purpose of changes: Write a history file giving the initial state, in any run that is NOT a restart run. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 M source_glc/history_tape_base.F90 M source_glc/glc_history.F90 M source_glc/glc_InitMod.F90 Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed. Externals used for testing: cesm1_4_beta07 cism tag used for baseline comparisons: cism2_1_02 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Aug 4, 2015 Model: cism Version: cism2_1_02 One-line summary: Move prebeta goldbach tests to hobart Purpose of changes: With hobart replacing goldbach, all components are moving prealpha & prebeta tests from golbach to hobart. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M cimetest/testlist_cism.xml Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: June 3, 2015 Model: cism Version: cism2_1_01 One-line summary: Update glimmer-cism external to add a 'use' statement Purpose of changes: Two needed variables were missing from a 'use' statement. For some reason, this did not cause problems on the previous yellowstone testing, but showed up as a problem in standalone CISM testing. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES -glimmer-cism https://github.com/CESM-Development/cism/tags/move_glint_to_cpl_n01 +glimmer-cism https://github.com/CESM-Development/cism/tags/move_glint_to_cpl_n02 Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed. Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp/branch_tags/move_glint_into_cpl2_tags/move_glint_into_cpl2_n04_cesm1_4_beta04 (I *think*), with mpi-serial patch as noted below cism tag used for baseline comparisons: cism2_1_00 Any other externals that differed in baseline: see below ================================================================================ Originator: sacks Date: May 19, 2015 Model: cism Version: cism2_1_00 One-line summary: Move GLC <-> LND coupling to CESM coupler Purpose of changes: Up until now, the GLC <-> LND coupling has been done by CISM's glint library. This is problematic for a few reasons: - glint can only handle regular land grids - not (for example) the CAM-SE grid - glint can only perform bilinear interpolation, not area-conservative remapping, for mapping from LND -> GLC - glint’s remapping is done in serial, on the master processor - any alternative ice sheet model that wants to couple to CESM (e.g., MPAS-Land Ice) needs to reimplement glint - coupling between the ocean and GLC would need to go through the LND grid, leading to loss of accuracy (especially for high-resolution ocean grids) To address these limitations, we have moved the GLC <-> LND coupling out of CISM, into the CESM coupler. This required significant changes to the coupler and scripts. In addition, for GLC/CISM, it required: - Writing a replacement for glint in CISM (called glad, for "glimmer already downscaled"), which works with inputs and outputs on the CISM grid rather than on the LND grid. - Changing GLC so that its grid is now the CISM grid, not the LND grid. Along with this, removed the CISM_GRID xml variable (now, just use the GLC_GRID variable, which has the value that CISM_GRID used to have). - Changing various namelist options to accommodate these changes I have also enabled the glc -> ocn coupling. Previously, this was disabled because we didn't have the necessary mapping files. Now I have generated the necessary mapping files, and they are available out-of-the-box in the upcoming CIME tag. The new code currently does not have any PDD support. I have pulled out some never-quite-working code related to the PDD option, because now the PDD option is definitively NOT supported. Changes answers relative to previous tag: YES Since the coupling method has changed - including using area-conservative remapping rather than bilinear remapping for the downscaling - answers change significantly. However, I have done many eyeball comparisons to verify that the new code is operating roughly the same as before. Dependencies: Depends on an upcoming CIME tag (cime2_0_00?) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Point to glimmer-cism branch M SVN_EXTERNAL_DIRECTORIES *** glc_global_fields renamed to glc_fields for accuracy; this is significantly different now D source_glc/glc_global_fields.F90 A source_glc/glc_fields.F90 *** Operation of glc overrides fundamentally reworked: It no longer works to scale fractions. Instead, the fractions specify topographic height cutoffs below / above which ice_covered is set to 0 / 1. M source_glc/glc_override_frac.F90 *** Point to new input files, which contain lat/lon values that agree with the values on Jeremy's new SCRIP grid files. These SCRIP grid files were needed for creating mapping files for the coupler. Lat/lon values are now also needed on the input files in order for CISM/GLC to tell the coupler about its domain information. It seemed good (although maybe not completely necessary) for these two sets of files to agree in terms of lat/lon values. Also, removed some no-longer-needed namelist items. M bld/namelist_files/namelist_defaults_cism.xml *** Fix some documentation M tools/glc2scripConvert.ncl M tools/README.glc_overlap_tools *** Other changes; see "Purpose of changes", above, for details. D source_glc/glc_global_grid.F90 A source_glc/glc_indexing_info.F90 M source_glc/glc_io.F90 M source_glc/glc_constants.F90 M source_glc/history_tape_base.F90 M source_glc/glc_FinalMod.F90 M source_glc/glc_history.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M cimetest/testmods_dirs/cism/override_glc_frac/user_nl_cism M drivers/cpl/glc_comp_esmf.F90 M drivers/cpl/glc_import_export.F90 M drivers/cpl/glc_coupling_flags.F90 M drivers/cpl/glc_cpl_indices.F90 M drivers/cpl/glc_comp_mct.F90 M bld/cismIO/cism.buildIO.template.csh M bld/README.build-namelist M bld/cism.template M bld/build-namelist M bld/namelist_files/namelist_definition_cism.xml M bld/trilinosOptions/README M bld/cism.buildnml Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed. Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp/branch_tags/move_glint_into_cpl2_tags/move_glint_into_cpl2_n03_cesm1_4_beta04 with mpi-serial patch as noted below cism tag used for baseline comparisons: cism2_0_09 Any other externals that differed in baseline: see externals notes under cism2_0_09 ================================================================================ Originator: mvertens (brought to trunk by sacks) Date: Apr 29, 2015 Model: cism Version: cism2_0_09 One-line summary: Updates for latest version of cime Purpose of changes: Updates to make (a) location of test directory and (b) variable names consistent with the latest version of CIME. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Change variable names to be consistent with latest version of CIME M bld/README.build-namelist M bld/build-namelist M bld/cism.buildlib M bld/cism.buildnml M bld/cism.template *** Rename cesmtest to cimetest D cesmtest D cesmtest/testlist_cism.xml D cesmtest/testmods_dirs D cesmtest/testmods_dirs/cism D cesmtest/testmods_dirs/cism/apply_to_multiinstance D cesmtest/testmods_dirs/cism/apply_to_multiinstance/README D cesmtest/testmods_dirs/cism/apply_to_multiinstance/shell_commands D cesmtest/testmods_dirs/cism/oneway D cesmtest/testmods_dirs/cism/oneway/README D cesmtest/testmods_dirs/cism/oneway/xmlchange_cmnds D cesmtest/testmods_dirs/cism/override_glc_frac D cesmtest/testmods_dirs/cism/override_glc_frac/include_user_mods D cesmtest/testmods_dirs/cism/override_glc_frac/user_nl_cism D cesmtest/testmods_dirs/cism/test_coupling D cesmtest/testmods_dirs/cism/test_coupling/include_user_mods D cesmtest/testmods_dirs/cism/test_coupling/user_nl_cism D cesmtest/testmods_dirs/cism/trilinos D cesmtest/testmods_dirs/cism/trilinos/README D cesmtest/testmods_dirs/cism/trilinos/include_user_mods D cesmtest/testmods_dirs/cism/trilinos/shell_commands D cesmtest/testmods_dirs/cism/trilinos/user_nl_cism A + cimetest Summary of testing: Ran full yellowstone aux_glc test suite. All tests passed. Note, however, that two tests needed to be rerun to pass: PEA_P1_M_Ly2.f09_g16_gl20.TGIS2.yellowstone_intel - seemed to have a system problem the first time SMS_D_Ly1.f09_g16_gl20.TGHISTIS2.yellowstone_gnu - died with: MCT::m_AttrVect::lsize_: attribute array length mismatch error, stat =892988089 - passed when I reran it - likely a compiler bug Externals used for testing: cesm1_4_beta02, with mpi-serial patch as noted below cism tag used for baseline comparisons: cism2_0_08 Any other externals that differed in baseline: see externals notes under cism2_0_08 ================================================================================ Originator: mvertens, sacks Date: Feb 20, 2015 Model: cism Version: cism2_0_08 One-line summary: Updates for Mariana's new testing infrastructure Purpose of changes: This tag updates CISM to fit in with Mariana's major overhaul of the test system and build scripts. Specifically: (1) Tests are now distributed with their components, rather than being centralized in 'scripts' (2) Build scripts have been converted to perl, and implicit dependencies on the environment have been removed. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Build scripts converted to perl; implicit dependencies on environment removed D bld/cism.buildexe.csh D bld/cism.cpl7.template D bld/cism.buildnml.csh M bld/trilinosOptions/README A bld/cism.buildlib A bld/cism.buildnml M bld/README.build-namelist A bld/cism.template M bld/build-namelist M bld/README *** Tests now distributed into components, rather than being centralized in scripts A cesmtest/testmods_dirs/cism/override_glc_frac/user_nl_cism A cesmtest/testmods_dirs/cism/override_glc_frac/include_user_mods A cesmtest/testmods_dirs/cism/override_glc_frac A cesmtest/testmods_dirs/cism/test_coupling/user_nl_cism A cesmtest/testmods_dirs/cism/test_coupling/include_user_mods A cesmtest/testmods_dirs/cism/test_coupling A cesmtest/testmods_dirs/cism/oneway/xmlchange_cmnds A cesmtest/testmods_dirs/cism/oneway/README A cesmtest/testmods_dirs/cism/oneway A cesmtest/testmods_dirs/cism/trilinos/user_nl_cism A cesmtest/testmods_dirs/cism/trilinos/include_user_mods A cesmtest/testmods_dirs/cism/trilinos/shell_commands A cesmtest/testmods_dirs/cism/trilinos/README A cesmtest/testmods_dirs/cism/trilinos A cesmtest/testmods_dirs/cism/apply_to_multiinstance/shell_commands A cesmtest/testmods_dirs/cism/apply_to_multiinstance/README A cesmtest/testmods_dirs/cism/apply_to_multiinstance A cesmtest/testmods_dirs/cism A cesmtest/testmods_dirs A cesmtest/testlist_cism.xml A cesmtest Summary of testing: Ran full yellowstone test suite from Mariana's experimental tag (https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/newtesting_cesm1_3_beta18c). All tests PASSed. Since the CME _D tests failed in the baseline, I ran SMS versions of these, as noted below (cism2_0_07). These baseline comparisons PASSed. (The CME tests themselves now pass, too.) Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/newtesting_cesm1_3_beta18c - but since this points to branches rather than tags, it is hard to say exactly what was used. (I think it was at r68306 for most components.) Also, mpi-serial changes, as noted below. cism tag used for baseline comparisons: cism2_0_07 Any other externals that differed in baseline: See below for notes on externals used for cism2_0_07. ================================================================================ Originator: sacks Date: Feb 20, 2015 Model: cism Version: cism2_0_07 One-line summary: Separate cism history frequency from coupler history frequency Purpose of changes: With Mariana's new testing framework, some exact restart tests were failing because CISM was writing history files at the end of a few-day run, and these didn't restart properly (specifically, IG/FG/BG tests without test_coupling). This provided motivation for something I have been wanting to do for a long time: separating the cism history frequency from the coupler history frequency. With this tag, the default cism history frequency is annual. You can still tie cism history to cpl history if you want, by setting hist_option = 'coupler'. However, this is not recommended for production runs, because it does not write frequency metadata to the history file. Along with this change, I have added a global metadata field to the history files saying the frequency of history writes. This is in accordance with a new requirement from the CESM post-processing/workflow group. As noted above, this does NOT work for hist_option = 'coupler'. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml A source_glc/glc_history.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 A source_glc/history_tape_standard.F90 M source_glc/glc_io.F90 A source_glc/history_tape_coupler.F90 A source_glc/history_tape_base.F90 M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 Summary of testing: Ran full yellowstone test suite, using the new aux_glc test suite that I have developed... this test suite is not on the scripts trunk (it is on Mariana's testing branch), so I'm listing it below: CME_D.T31_g37.IGCLM45.yellowstone_intel.cism-test_coupling # Note that this is more about testing CLM (specifically the esmf code in CLM) than it is about testing CISM ERS_Ly5.T31_g37.IGCLM45.yellowstone_intel.clm-glcMEC_long # Need IG ERS test to catch problems with fields sent before the end of the first year after restart. Also use glcMEC_long testmods to get shorter snow_persistence_max - the main motivation here is to have a long ERS test that checks restart of the snow_persistence stuff (which more belongs in the aux_clm test list, but was added here to avoid needing to add a long ERS test to that test list)... this is mainly needed because we do not have an IG test that uses spun-up initial conditions, so we currently need a long test to completely test the restartability of the snow_persistence stuff. Also note that this is the only multi-year non-TG test in the test list, so this is the one test that a production-like configuration can run for a few years. Consider moving this to the aux_clm45 test list once this test can be made shorter, either through use of initial conditions and/or moving to f10 resolution. ERS_D_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac # Make sure glc override options work correctly, and restart properly SMS_D.T31_g37_gl20.IGCLM45IS2.yellowstone_gnu.cism-test_coupling SMS_D.f09_g16.TG.yellowstone_intel ERS_Ly20_N2_P2.f09_g16_gl10.TG.yellowstone_pgi SMS_D.f09_g16_gl10.TG.yellowstone_gnu SMS_D.f09_g16_gl10.TG.yellowstone_pgi CME_Ly5_N2_P2_D.f09_g16.TG1850.yellowstone_intel ERS_Ly20.f09_g16.TG1850.yellowstone_gnu.cism-oneway ERS_Ly11.f09_g16_gl20.TG1850IS2.yellowstone_gnu SMS_D_Ly1.f09_g16_gl20.TG1850IS2.yellowstone_pgi ERS_Ly20_E.f09_g16.TGHIST.yellowstone_intel PEA_P1_M.f09_g16.TGHIST.yellowstone_pgi ERI_Ly15.f09_g16_gl20.TGHISTIS2.yellowstone_pgi SMS_D_Ly1.f09_g16_gl20.TGHISTIS2.yellowstone_gnu PEA_P1_M_Ly2.f09_g16_gl20.TGIS2.yellowstone_intel # needs to be at least 2 years for there to be enough cpl fields for the cpl log comparison to work SMS_D_Ly1.f09_g16_gl20.TGIS2.yellowstone_intel SMS_D_Ly1.f09_g16_gl20.TGIS2.yellowstone_intel.cism-trilinos SMS_Ly1.f09_g16_gl4.TGIS2.yellowstone_intel # include one short test of the typical production resolution for CISM2 ERI_Ly44.f09_g16.TGRCP85.yellowstone_intel CME_Ly3.f09_g16_gl20.TGRCP85IS2.yellowstone_intel NCK_Ly3.f09_g16_gl20.TGRCP85IS2.yellowstone_pgi These all passed, except for these, which are currently expected to fail for issues unrelated to CISM: PEND CME_D.T31_g37.IGCLM45.yellowstone_intel.cism-test_coupling.GC.150219-095007 PEND CME_Ly5_N2_P2_D.f09_g16.TG1850.yellowstone_intel.GC.150219-095007 Also, this comparison failed due to having no history file in the test case. This is expected and okay - note that this test is about testing the fields sent to the coupler, so it's fine for this to not have a CISM history file (and, as a side-note, this was the main test that provided motivation for this tag, because in Mariana's branch, the ERS comparison of the CISM history file fails for this test) FAIL ERS_D_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.cism.h.compare_hist.cism2_0_05_alpha17d Since the CME_D tests currently fail (because of issues unrelated to CISM), I ran SMS versions of these tests: SMS_D.T31_g37.IGCLM45.yellowstone_intel.cism-test_coupling SMS_Ly5_N2_P2_D.f09_g16.TG1850.yellowstone_intel Externals used for testing: cesm1_3_alpha17d, with these mods: Index: mpi-serial/mpi.h =================================================================== --- mpi-serial/mpi.h (revision 1224) +++ mpi-serial/mpi.h (working copy) @@ -152,6 +152,8 @@ } MPI_Status; +#define MPI_STATUS_IGNORE (MPI_Status *)1 +#define MPI_STATUSES_IGNORE (MPI_Status *)1 /* * Collective operations Index: mpi-serial/mpif.master.h =================================================================== --- mpi-serial/mpif.master.h (revision 1224) +++ mpi-serial/mpif.master.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: mpi-serial/mpif.real4double8.h =================================================================== --- mpi-serial/mpif.real4double8.h (revision 1224) +++ mpi-serial/mpif.real4double8.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: mpi-serial/mpif.real8double16.h =================================================================== --- mpi-serial/mpif.real8double16.h (revision 1224) +++ mpi-serial/mpif.real8double16.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: mpi-serial/mpif.real8double8.h =================================================================== --- mpi-serial/mpif.real8double8.h (revision 1224) +++ mpi-serial/mpif.real8double8.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: ccsm_utils/Testlistxml/testmods_dirs/cism/test_coupling/user_nl_cism =================================================================== --- ccsm_utils/Testlistxml/testmods_dirs/cism/test_coupling/user_nl_cism (revision 68086) +++ ccsm_utils/Testlistxml/testmods_dirs/cism/test_coupling/user_nl_cism (working copy) @@ -1,3 +1,6 @@ ! This option changes the ice sheet dynamics time step to 1 day rather than 1 year ! Thus, the ice sheet dynamics can be exercised in a few-day run test_coupling = .true. + +! This option gives us a history file at the end of the run, even for a few-day run +history_option = 'coupler' cism tag used for baseline comparisons: cism2_0_05 Any other externals that differed in baseline: same, but without the mod in test_coupling/user_nl_cism ================================================================================ Originator: sacks Date: Feb 13, 2015 Model: cism Version: cism2_0_06 One-line summary: Update glimmer-cism external to fix cray compilation problem Purpose of changes: Fix compilation problem with a string continuation line (bug 2145) Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 2145 List all modified files, and describe the changes: *** cism2_141202 -> cism2_150213 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Just tested the build for these three tests: SMS_D.f09_g16_gl4.TGIS2.yellowstone_intel SMS_D.f09_g16_gl4.TGIS2.yellowstone_pgi SMS_D.f09_g16_gl4.TGIS2.yellowstone_gnu Externals used for testing: cesm1_3_alpha17d cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Dec 02, 2014 Model: cism Version: cism2_0_05 One-line summary: update glimmer-cism external Purpose of changes: Update to latest development version. Among other things, this includes a bug fix for basal melt for the higher-order code. Changes answers relative to previous tag: YES Changes answers for CISM2 compsets Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update from cism2_141119a -> cism2_141202 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Ran full aux_glc test suite. All tests PASSed. Externals used for testing: cesm1_3_beta12 with these diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 65603) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/rollback_timing_updates_tags/rollback_timing_updates_n08_scripts4_141023 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/piscees_testing_tags/piscees_testing_n02_Machines_140923 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,14 +10,14 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc -models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 +models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140918b models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r084//models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd cism tag used for baseline comparisons: cism2_0_04 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Nov 19, 2014 Model: cism Version: cism2_0_04 One-line summary: update glimmer-cism external Purpose of changes: Point to new external location, in the CESM-Develop github area. Update that repo to pull in some minor bug fixes from the piscees repo. Also add some error-checking to ensure that beta is present on the input file if using which_ho_babc = 5. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES -glimmer-cism https://github.com/CISM/cism/tags/v2.0 +glimmer-cism https://github.com/CESM-Development/cism/tags/cism2_141119a Summary of testing: Ran full aux_glc test suite (which now just runs on yellowstone). All tests PASSed. Externals used for testing: cesm1_3_beta12 with these diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 65521) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/rollback_timing_updates_tags/rollback_timing_updates_n08_scripts4_141023 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/piscees_testing_tags/piscees_testing_n02_Machines_140923 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,14 +10,14 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc -models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 +models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140918b models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r084//models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd Also, diffs for mpi-serial, as noted below cism tag used for baseline comparisons: cism2_0_03 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Nov 18, 2014 Model: cism Version: cism2_0_03 One-line summary: tweak namelist options for cism2 Purpose of changes: - remove 10km support for CISM2: we'll just do testing using the 20km grid - add which_ho_approx namelist option - for gland20 with cism2, change default dt to 0.5, and default which_ho_babc to 4 - decrease dt for gland4 to achieve stability in some of my tests (note that we need to use a non-round value because dt needs to translate into an integer number of hours) Changes answers relative to previous tag: YES Answers change for cism2 runs due to changes in default namelist options Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Ran two versions of the aux_glc yellowstone tests. (Note that I am removing the aux_glc titan tests). The final set used externals as listed below. The earlier set used rollback_timing_updates_n05_scripts4_141023 (up-to-date with rework_glc_compsets_n04_scripts4_141112a). All tests passed in the final set of tests. In the earlier set, ERS_Ly3.f09_g16_gl4.TGIS2.yellowstone_intel failed simply because there were no lines compared in the cpl log file (cpl hist files were bfb). There were lots of BFAILs, because I have totally reworked the test list. However, there were still quite a few compare_hist PASSes. These failed compare_hist, due to changes in default length of TG tests: SMS_D.f09_g16_gl10.TG.yellowstone_pgi SMS_D.f09_g16.TG.yellowstone_intel Externals used for testing: cesm1_3_beta12 with these diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 65521) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/rollback_timing_updates_tags/rollback_timing_updates_n08_scripts4_141023 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/piscees_testing_tags/piscees_testing_n02_Machines_140923 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,14 +10,14 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc -models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 +models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140918b models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r084//models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd Also, diffs for mpi-serial, as noted below cism tag used for baseline comparisons: cism2_0_00 Any other externals that differed in baseline: see below for externals used for cism2_0_00 ================================================================================ Originator: sacks Date: Nov 5, 2014 Model: cism Version: cism2_0_02 One-line summary: modify tools for creating glcmask files Purpose of changes: The toolchain for creating glcmask files is out-dated and didn't work exactly how I wanted it to. I have made the following modifications: (1) Create mapping file using the shared CESM tools, rather than the outdated tool that used to be contained here. (2) It was getting awkward to support all combinations of CLM grid x CISM grid. Thus, I have rewritten the documentation to just do this glcmask generation for a single CISM grid. From a quick look at the different existing glcmask files, as well as the different CISM grids that we currently support, I don't think the choice of CISM grid should make a significant difference in this glcmask file. I have (somewhat arbitrarily) chosen the (old) 5km CISM grid for this purpose. (3) Remove dependence on CLM's landfrac. This dependence seems unnecessary, and fragile in the face of potentially changing ocean grids. With this change, the overlap file contains a 1 value in any CLM grid cell that overlaps a CISM grid cell, regardless of whether this grid cell has any land according to landfrac. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: D tools/scrip_make_wgts_CCSM_to_GLC_bilin.csh M tools/README.glc_tools M tools/scrip2CLMoverlap.ncl M tools/glc2scripConvert.ncl M tools/README.glc_overlap_tools Summary of testing: Did NOT run any aux_glc tests, because I have only changed these offline tools. I have tested the new tool-chain by creating glcmask files for the 3 supported CLM resolutions. These new glcmask files have been committed to the inputdata repository. Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Oct 31, 2014 Model: cism Version: cism2_0_01 One-line summary: Update ChangeLog entry for last tag Purpose of changes: Add stuff to ChangeLog that was accidentally excluded from the last tag Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M ChangeLog Summary of testing: none Externals used for testing: n/a cism tag used for baseline comparisons: n/a Any other externals that differed in baseline: n/a ================================================================================ Originator: sacks Date: Oct 31, 2014 Model: cism Version: cism2_0_00 One-line summary: Point to cism2 release code base Purpose of changes: Update glimmer-cism external to point to recently-released cism2 code base. This pulls in developments that have been made over the last year in the private PISCEES repo. In addition, a number of misc. changes to namelist defaults, build options, etc. to work smoothly with new code base. Changes answers relative to previous tag: No answer changes for SIA. Higher-order code (using glissade dycore) finally works! Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Point to cism2 release code base M SVN_EXTERNAL_DIRECTORIES *** Change default config options for cism2 runs, including using glissade dycore and allowing runs at 4km resolution (trilinosOptions for 4km simply copies the 5km file). Also add some error checking for the match between cism_phys (cism1/cism2) and resolution. M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M bld/build-namelist A bld/trilinosOptions/trilinosOptions_gland4.xml *** Some changes to accommodate changes made in the glimmer-cism code or build system. Also, switch to always using parallel_mpi (rather than parallel_slap), because that's easier than trying to come up with a rule for when we need one or the other. M bld/cism.buildexe.csh M bld/cism.cpl7.template M source_glc/glc_io.F90 Summary of testing: Ran full aux_glc test suite, on yellowstone and titan. All tests PASS! In addition, ran these tests from the prebeta test list: PASS SMS.T31_g37_gl10.BGCNIS2.yellowstone_intel PASS SMS.T31_g37_gl10.BGCNIS2.yellowstone_intel.memleak PASS SMS_Ly2.f09_g16_gl10.TGIS2.yellowstone_intel PASS SMS_Ly2.f09_g16_gl10.TGIS2.yellowstone_intel.memleak Externals used for testing: cesm1_3_beta12 with the following diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 64812) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/rollback_timing_updates_tags/rollback_timing_updates_n02_scripts4_141023 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/piscees_testing_tags/piscees_testing_n01_Machines_140923 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,14 +10,14 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc -models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 +models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140918b models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r084//models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd In addition, the following diffs in my yellowstone sandbox, which are needed for the PEA tests: Index: models/utils/mct/mpi-serial/mpi.h =================================================================== --- models/utils/mct/mpi-serial/mpi.h (revision 1122) +++ models/utils/mct/mpi-serial/mpi.h (working copy) @@ -152,6 +152,8 @@ } MPI_Status; +#define MPI_STATUS_IGNORE (MPI_Status *)1 +#define MPI_STATUSES_IGNORE (MPI_Status *)1 /* * Collective operations Index: models/utils/mct/mpi-serial/mpif.master.h =================================================================== --- models/utils/mct/mpi-serial/mpif.master.h (revision 1122) +++ models/utils/mct/mpi-serial/mpif.master.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: models/utils/mct/mpi-serial/mpif.real4double8.h =================================================================== --- models/utils/mct/mpi-serial/mpif.real4double8.h (revision 1122) +++ models/utils/mct/mpi-serial/mpif.real4double8.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: models/utils/mct/mpi-serial/mpif.real8double16.h =================================================================== --- models/utils/mct/mpi-serial/mpif.real8double16.h (revision 1122) +++ models/utils/mct/mpi-serial/mpif.real8double16.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) Index: models/utils/mct/mpi-serial/mpif.real8double8.h =================================================================== --- models/utils/mct/mpi-serial/mpif.real8double8.h (revision 1122) +++ models/utils/mct/mpi-serial/mpif.real8double8.h (working copy) @@ -132,6 +132,8 @@ INTEGER MPI_SOURCE, MPI_TAG, MPI_ERROR PARAMETER(MPI_SOURCE=1, MPI_TAG=2, MPI_ERROR=3) + INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE) + INTEGER MPI_STATUSES_IGNORE(MPI_STATUS_SIZE,1) cism tag used for baseline comparisons: cism1_141006 and piscees_n13_cism1_141006 yellowstone testing: - CISM1 tests bfb with cism1_141006, according to both cism & cpl hist files; CISM2 tests BFAIL with this comparison because the tests were failing on the trunk until now - All tests (CISM1 & CISM2) bfb with piscees_n13_cism1_141006, according to both cism & cpl hist files titan testing: - just compared with piscees_n13_cism1_141006; all tests bfb Any other externals that differed in baseline: externals were similar, but possibly not identical in the baseline ================================================================================ Originator: sacks Date: Oct 6, 2014 Model: cism Version: cism1_141006 One-line summary: update glimmer-cism to relax tolerances in glimmer_sparse.F90 Purpose of changes: Merge r1571 from piscees repo for libglimmer-solve/glimmer_sparse.F90. This relaxes the tolerances from 1e-11 to 1e-8. My purpose in merging this change into the cesm repo is so that SIA answers can remain bfb between the cesm repo and the piscees repo. Changes answers relative to previous tag: YES Changes answers for all runs that are long enough to trigger the main CISM code Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update to r64084 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: All IS2 tests fail (as usual) Baseline comparisons fail for many tests, as expected Externals used for testing: cesm1_3_beta12, with these diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 64083) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140916c scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_17 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,7 +10,7 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 @@ -17,7 +17,7 @@ models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r084//models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd cism tag used for baseline comparisons: cism1_140914 Any other externals that differed in baseline: See ChangeLog entry for cism1_140914 - in particular, externals that differed from cesm1_3_beta12 used branch versions for cism1_140914 ================================================================================ Originator: sacks Date: Sept 16, 2014 Model: cism Version: cism1_140916 One-line summary: Fix ChangeLog entry for cism1_140914 Purpose of changes: I realized that I incorrectly described the reason for answer changes in the IG CLM40 test. I'm fixing that here. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M ChangeLog Summary of testing: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Sept 14, 2014 Model: cism Version: cism1_140914 One-line summary: Add zero_gcm_fluxes option; send icemask_coupled_fluxes field to coupler Purpose of changes: In talking with Dave Lawrence, Bill Lipscomb and Jeremy Fyke, I realized that we need to introduce more logic to distinguish between active icesheets and diagnostic-only icesheets. One aspect of this is the ability to zero out the fluxes CISM sends to the coupler (currently runoff fluxes; later also the heat flux). That way, by zeroing out these fluxes and turning off the connection to CLM, we can continue to run with an evolving ice sheet in diagnostic-only mode (i.e., 1-way coupling). This is achieved via the new zero_gcm_fluxes option. The default for zero_gcm_fluxes is based on the GLC_TWO_WAY_COUPLING xml variable. The current logic in scripts sets GLC_TWO_WAY_COUPLING to true for CLM45 (or later) compsets with CISM, and for TG compsets; otherwise (for CLM40, or for compsets without CISM) this is set to false. In addition, I realized the need to distinguish between areas of active icesheets and areas of either no icesheet or diagnostic-only icesheets in applying CLM's glc_dyn_runoff_routing switch. In particular, we should only use the new runoff routing scheme over areas of icesheet that are generating calving fluxes. From the perspective of CISM, this is achieved by computing and sending a new icemask_coupled_fluxes field to the coupler (which is then passed to CLM). icemask_coupled_fluxes is similar to the existing icemask field, but is zero over icesheet instances for which zero_gcm_fluxes is true - which will be the case for icesheets that are not generating calving fluxes (e.g., diagnostic-only icesheets). Changes answers relative to previous tag: YES Changes icemask at the roundoff-level, for some resolutions and compilers Also, in combination with the new scripts, zeroes out runoff fluxes sent to coupler (g2x_Fogg_rofl and g2x_Fogg_rofi). Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update to r63495: glint changes needed to support the above changes M SVN_EXTERNAL_DIRECTORIES *** add zero_gcm_fluxes namelist option M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml *** send icemask_coupled_fluxes field to the coupler M drivers/cpl/glc_cpl_indices.F90 M drivers/cpl/glc_import_export.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_fields.F90 Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: All IS2 tests fail (as usual) compare_hist failures: FAIL CME_Ly5.T31_g37.IG.yellowstone_intel.clm-reduceOutput.compare_hist.cism1_140602_beta12 FAIL ERS_Ly5.T31_g37.IGCLM45.yellowstone_intel.clm-glcMEC_long.compare_hist.cism1_140602_beta12 FAIL SMS_D.T31_g37.BG1850CN.yellowstone_intel.compare_hist.cism1_140602_beta12 These show roundoff-level changes in icemask. In addition, the CME test has large diffs in g2x_Fogg_rofl and g2x_Fogg_rofi: These fields are identically 0 in the new run. This makes sense, since zero_gcm_fluxes is set to true for this CLM40 run (which arises in combination with the new scripts version, which sets GLC_TWO_WAY_COUPLING to FALSE for this run). Externals used for testing: cesm1_3_beta12, with these diffs: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 63201) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,8 +1,8 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140813 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/glc_runoff_routing_tags/glc_runoff_routing_n01_scripts4_140814a scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/bluewatersfix_Machines_140811_tags/bluewatersfix02_Machines_140811 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_140625 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_140702b -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_15 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/branch_tags/glc_runoff_routing_tags/glc_runoff_routing_n01_drvseq5_0_15 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_140723 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20140708/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_45/models/atm/cam/ @@ -10,7 +10,7 @@ models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_140602 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/glc_runoff_routing models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xglc models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20140711 @@ -17,7 +17,7 @@ models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_1_r079//models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/glc_runoff_routing_tags/glc_runoff_routing_n02_clm4_5_1_r081 models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_8_00/xlnd However, for the PEA test, I merged the changes from scripts4_140903b into the sandbox. cism tag used for baseline comparisons: cism1_140602 Any other externals that differed in baseline: Baseline used cesm1_3_beta12 with updated scripts (scripts4_140905???) ================================================================================ Originator: sacks Date: June 2, 2014 Model: cism Version: cism1_140602 One-line summary: Update glimmer-cism external with some fixes for CISM2 Purpose of changes: Update glimmer-cism external with the following fixes: (1) update CMakeLists.txt to fix the yellowstone build when building with trilinos (2) fix downscaling of fields in elevation class 0 when running with multiple processors (3) set halo values in glint downscaling using parallel_halo calls Changes answers relative to previous tag: NO (at least not for CISM1 [SIA] runs) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update to r60744 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.140530-154133 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.140530-154116 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.140530-154116 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.140530-154116 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.140530-154116 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.140530-154116 Note: The IS2 runs now get further than before, because of problems that have been fixed in this tag. e.g., SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel gets up to a conservation error before stopping (in contrast to dying due to a floating point problem or some other problem previously). However, the IGIS2 test fails with: 0:NOXSolve called 0: 0:************************************************************************ 0:-- Nonlinear Solver Step 0 -- 0:||F|| = inf step = 0.000e+00 dx = 0.000e+00 0:************************************************************************ 0: 35: 35:p=35: *** Caught standard std::exception of type 'Belos::StatusTestError' : 35: 35: /glade/u/home/rory/trilinos-11.0.3-Source/packages/belos/src/BelosStatusTestGenResNorm.hpp:574: 35: 35: Throw number = 1 35: 35: Throw test that evaluated to true: true 35: 35: StatusTestGenResNorm::checkStatus(): NaN has been detected. 35:INFO: 0031-306 pm_atexit: pm_exit_value is 1. Externals used for testing: cesm1_3_beta07, with many diffs - as in cism1_140501 cism tag used for baseline comparisons: cism1_140501 Any other externals that differed in baseline: none ================================================================================ Originator: jfyke, sacks Date: May 1, 2014 Model: cism Version: cism1_140501 One-line summary: Add an elevation class 0, for bare land SMB Purpose of changes: Changes from Jeremy Fyke to add an elevation class 0, for bare land SMB. Also adds ice_sheet_grid_mask. Also some minor, unrelated changes as noted below. This involves extensive changes to glint, as well as some changes to the glc code. In addition, this tag pulls in an unrelated change to the glimmer-cism external: a fix from Stephen Cornford for glint_mbal_io. Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_override_frac.F90 M source_glc/glc_global_fields.F90 M drivers/cpl/glc_cpl_indices.F90 *** Also, clean up interfaces for glc_import & glc_export ('use' things directly rather than having them passed through the interface) M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 M drivers/cpl/glc_import_export.F90 *** Use some constants from shr_const_mod M source_glc/glc_constants.F90 *** Add an option to decrease the mass balance time step, for testing purposes (currently not exercised in any tests) M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures (note: the GEN cases probably would have given CFAIL if I had been patient): CFAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.140430-125442 CFAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.140430-125442 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.140430-125442 GEN SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.140430-125442 GEN SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.140430-125442 CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.140430-125445 Note: I am now getting a CFAIL instead of a FAIL for the IS2 cases... seems to be something having to do with the yellowstone upgrade... I'm guessing it's a problem with the C++/Fortran link. I get this message: cannot find -lmkl_rt Externals used for testing: cesm1_3_beta07, but with quite a few changes: Index: SVN_EXTERNAL_DIRECTORIES =================================================================== --- SVN_EXTERNAL_DIRECTORIES (revision 59711) +++ SVN_EXTERNAL_DIRECTORIES (working copy) @@ -1,23 +1,23 @@ -scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_140113 -scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_140124 +scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/add_dynlu_tests_tags/add_dynlu_tests_n02_scripts4_140305 +scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/arfs_tags/arfs_01_mach140218 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_131224 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_131217a -models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq5_0_08 -models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_131231 +models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/branch_tags/bare_land_smb_tags/bare_land_smb_n02_drvseq5_0_08 +models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/branch_tags/shr_assert_macro_tags/shr_assert_macro_n04_share3_140115 scripts/validation_testing https://svn-ccsm-models.cgd.ucar.edu/validation_testing/trunk_tags/validation_20131108/run_CESM/ models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_3_23/models/atm/cam -models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_131201 +models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_140114 models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/satm models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xatm models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/dead_share -models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_131213 +models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branch_tags/bare_land_smb_tags/bare_land_smb_n06_cism1_140416 models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xglc models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20131002 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_131201 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xice -models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_5_57/models/lnd/clm +models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/bare_land_smb_tags/bare_land_smb_n14_clm4_5_70/models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_131201 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xlnd @@ -27,16 +27,16 @@ models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/socn models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xocn models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_131201 -models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_36 +models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_37 models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/srof models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xrof models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_130213 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_7/pio +models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_8_11/pio models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_131108 models/wav/swav https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_04/swav models/wav/ww3 https://svn-ccsm-models.cgd.ucar.edu/ww3/trunk_tags/ww3_130905 models/wav/xwav https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_09/xwav -scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_140116 +scripts/ccsm_utils/CMake https://github.com/quantheory/CMake_Fortran_utils/tags/CMake_Fortran_utils_140403 scripts/doc https://svn-ccsm-models.cgd.ucar.edu/doc/trunk_tags/doc_131021 tools/unit_testing https://svn-ccsm-models.cgd.ucar.edu/unit_testing/trunk_tags/unit_testing_0_04 cism tag used for baseline comparisons: cism1_140303 Any other externals that differed in baseline: See notes under cism1_140303 below ================================================================================ Originator: muszala Date: April 16, 2014 Model: cism Version: cism1_140416 One-line summary: bring in SHR_ASSERT macros Purpose of changes: Removed 'use shr_assert_mod' statements and replace 'call shr_assert' with SHR_ASSERT or SHR_ASSERT_ALL where appropriate. Add #include "shr_assert.h" to modules where macro is used. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): N/A List all modified files, and describe the changes: M source_glc/glc_override_frac.F90 M drivers/cpl/glc_comp_mct.F90 M drivers/cpl/glc_comp_esmf.F90 M drivers/cpl/glc_import_export.F90 Summary of testing: Tested along with clm4_5_70. This includes yellowstone+{intel,pgi} and goldbach+{nag,pgi,intel} and component comp. gen with ys+intel. Externals used for testing: alpha09b Cism tag used for baseline comparisons: cism1_140303 in clm4_5_69 Any other externals that differed in baseline: Many, including: Pio, Mct, timing, scripts, machines. Please compare clm4_5_69 and clm4_5_70 for all details. ================================================================================ Originator: sacks Date: April 3, 2014 Model: cism Version: cism1_140403 One-line summary: fix allocate statement so it works with gfortran Purpose of changes: I accidentally used a F2008 feature - trying to pick up array dimensions with allocate(..., source=...); I have fixed this. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M drivers/cpl/glc_import_export.F90 Summary of testing: Just ran a subset of tests: PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534.memleak PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534.generate.cism1_140403 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534.compare_hist.cism1_140305 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534.memcomp.cism1_140305 FAIL SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140403-132534.tputcomp.cism1_140305 COMMENT tput_decr = 8.0199999 tput_decr = 3.7669999 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608.memleak PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608.generate.cism1_140403 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608.compare_hist.cism1_140305 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608.memcomp.cism1_140305 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140403-132608.tputcomp.cism1_140305 COMMENT tput_decr = 2.2609999 PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743 PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743.memleak PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743.generate.cism1_140403 PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743.compare_hist.cism1_140305 PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743.memcomp.cism1_140305 PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140403-140743.tputcomp.cism1_140305 And from component_gen_comp: PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.compare_hist.cism1_140305.cism.h PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.generate.cism.h PASS SMS_D.f09_g16.TG.yellowstone_intel.compare_hist.cism1_140305.cism.h PASS SMS_D.f09_g16.TG.yellowstone_intel.generate.cism.h PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.compare_hist.cism1_140305.cism.h PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.generate.cism.h Externals used for testing: cesm1_3_beta07, but scripts4_140305 and Machines_140227 cism tag used for baseline comparisons: cism1_140305 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: March 5, 2014 Model: cism Version: cism1_140305 One-line summary: fix capping of increase_frac override Purpose of changes: The capping of inrease_frac was accidentally capping each elevation class at 100%, when it should be capping the sum of elevation classes at 100%. That is now fixed. Changes answers relative to previous tag: NO (only changes test cases that use increase_frac) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M source_glc/glc_override_frac.F90 Summary of testing: Just ran a subset of tests: PASS ERS_Ld9.f19_g16.IGCLM45.yellowstone_pgi.cism-override_glc_frac.GC.140305-092736 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.memleak PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.generate.cism1_140305 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.compare_hist.cism1_140303 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.memcomp.cism1_140303 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.tputcomp.cism1_140303 PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi.GC.140305-093532.nlcomp PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.memleak PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.generate.cism1_140305 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.compare_hist.cism1_140303 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.memcomp.cism1_140303 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.tputcomp.cism1_140303 PASS SMS_D.f09_g16.TG.yellowstone_intel.GC.140305-093417.nlcomp Externals used for testing: cesm1_3_beta07, but scripts4_140305 and Machines_140227 cism tag used for baseline comparisons: cism1_140303 Any other externals that differed in baseline: baseline used scripts4_140221a ================================================================================ Originator: sacks Date: March 3, 2014 Model: cism Version: cism1_140303 One-line summary: add options to override glc frac sent to cpl, for testing Purpose of changes: Put in place a number of namelist settings that can be used to modify the glc frac sent from GLC -> CPL. The point of this is to push CLM's dynamic landunits in various ways (e.g., going from 0% glacier to non-zero glacier, from 100% glacier down to 0% glacier, etc.). Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml M source_glc/glc_InitMod.F90 A source_glc/glc_override_frac.F90 M source_glc/glc_global_fields.F90 M drivers/cpl/glc_import_export.F90 *** Unrelated change: fix documentation M bld/README.build-namelist Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.140228-162551 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.140228-162551 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.140228-162551 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.140228-162551 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.140228-162551 CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.140228-162555 Externals used for testing: cesm1_3_beta07, but scripts4_140221a and Machines_140227 cism tag used for baseline comparisons: cism1_140203 Any other externals that differed in baseline: baseline used cesm1_3_beta07, unmodified ================================================================================ Originator: sacks Date: Feb 3, 2014 Model: cism Version: cism1_140203 One-line summary: fix exact restart problem with cism -> cpl fields Purpose of changes: Update glimmer-cism external to fix an exact restart problem that appears with a few of the cism -> cpl fields - specifically, the rofi, rofl and hflx fields. This problem showed up with runs with daily (rather than annual) coupling - i.e., all compsets other than TG. For example, this test was failing: ERS_Ly5.T31_g37.IGCLM45.yellowstone_intel Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update from r55984 -> r56989 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.140129-134856 BFAIL ERS_Ly5.T31_g37.IGCLM45.yellowstone_intel.GC.140129-134851.compare_hist.cism1_131213_beta07 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.140129-134851 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.140129-134851 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.140129-134851 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.140129-134851 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.140129-134851 Externals used for testing: cesm1_3_beta07 cism tag used for baseline comparisons: cism1_131213 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Dec 13, 2013 Model: cism Version: cism1_131213 One-line summary: update glimmer-cism external Purpose of changes: Update glimmer-cism external to r55984. The main change here is to remove prevtemp from the computation of the diagnostic quantity dTtop; this changes answers for hflx, but nothing else. Changes answers relative to previous tag: YES Just changes answers for the hflx term sent to the coupler. Diffs are generally about 1e-6 (relative error), though some RMS errors are as large as 1e-4 (relative error). Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.143014 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.143009 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.143009 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.143009 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.143009 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.143009 Externals used for testing: cesm1_3_beta05 cism tag used for baseline comparisons: cism1_131212a Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Dec 12, 2013 Model: cism Version: cism1_131212a One-line summary: update glimmer-cism external Purpose of changes: Update glimmer-cism external from r55935 to r55978. This changes the number of seconds in a year to exactly 365 days, in agreement with the rest of CESM. Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.142829 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.142824 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.142824 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.142824 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.142824 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.142824 Externals used for testing: cesm1_3_beta05 cism tag used for baseline comparisons: cism1_131212 Any other externals that differed in baseline: none ================================================================================ Originator: sacks Date: Dec 12, 2013 Model: cism Version: cism1_131212 One-line summary: update glimmer-cism external, fixing fields sent from glc to cpl Purpose of changes: Update glimmer-cism external from r54059 to r55935. The main changes here involve fixing the fields sent from glc to the CESM coupler - specifically, rofi, rofl and hflx. Also, a few other minor changes. Changes answers relative to previous tag: YES Changes answers for all fields sent from GLC -> CPL: SIGNIFICANT differences for hflx, rofi and rofl; and roundoff-level changes for frac and topo. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.133413 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.133404 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.133404 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.133404 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.133404 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.133404 Externals used for testing: cesm1_3_beta05 cism tag used for baseline comparisons: cism1_131203 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: Dec 3, 2013 Model: cism Version: cism1_131203 One-line summary: call glc_export in mct initialization Purpose of changes: With the changes in cism1_131008, the esmf side had been changed to call glc_export in initialization, but the mct side had not been changed similarly. This tag makes a simple change to the mct side so that CME tests now pass. (This problem was missed in the testing for cism1_131008 because it only shows up in IG/FG/BG compsets, which were excluded from the testing of that tag.) Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M drivers/cpl/glc_comp_mct.F90 Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.103449 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.103445 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.clm-reduceOutput.GC.103445 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.103445 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.103445 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.clm-reduceOutput.GC.103445 Note that this test failed in the baseline: CME_Ly5.T31_g37.IG.yellowstone_intel.clm-reduceOutput However, I manually compared the cpl hist files, and they are identical in the new tag compared with the baseline. Externals used for testing: cesm1_3_beta05 cism tag used for baseline comparisons: cism1_131008 Any other externals that differed in baseline: None (baselines also generated from cesm1_3_beta05) ================================================================================ Originator: sacks, mvertens Date: Oct 8, 2013 Model: cism Version: cism1_131008 One-line summary: Mariana's changes to cpl infrastructure, update external Purpose of changes: (1) Bring in Mariana's refactoring of the cpl infrastructure (2) Point glimmer-cism external to new repository location (3) Small update to glimmer-cism external: Bring in Matt Hoffman's change to the logic for allocating isostasy variables Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** (1) above M bld/cism.buildexe.csh D drivers/cpl_share D drivers/cpl_mct D drivers/cpl_esmf A drivers/cpl/glc_comp_mct.F90 A drivers/cpl/glc_comp_esmf.F90 A drivers/cpl/glc_import_export.F90 A drivers/cpl/glc_coupling_flags.F90 A drivers/cpl/glc_cpl_indices.F90 A drivers/cpl *** (2) and (3) above M SVN_EXTERNAL_DIRECTORIES Summary of testing: Ran just a subset of the yellowstone intel & pgi test lists: Just the TG CISM1 tests. It was not ideal to exclude the IG / FG / BG tests, but I made this compromise because yellowstone is down. All PASS except the following expected failure: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.215132 Update 12-2-13: Ran IG / FG / BG CISM1 tests; all pass EXCEPT: *** Known failure; working on resolving this FAIL CME_Ly5.T31_g37.IG.yellowstone_intel.GC.100040 *** Diffs only in x2l_Flrr_volr FAIL ERI.f19_g16.IGRCP85CN.yellowstone_pgi.GC.100111.compare_hist.cism1_130924 (component_gen_comp: all PASS, including the two above test failures) Externals used for testing: cesm1_3_alpha04c cism tag used for baseline comparisons: cism1_130924 Any other externals that differed in baseline: Baseline used externals at cesm1_3_beta03 ================================================================================ Originator: sacks Date: Sept 24, 2013 Model: cism Version: cism1_130924 One-line summary: Update glimmer-cism external to r1966 Purpose of changes: Bring in latest glimmer-cism external - update from r1950 to r1966 Changes answers relative to previous tag: No diffs according to test suite, although log messages suggest there may have been some small answer changes. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.162146 FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.162141 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.162141 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.162141 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.162141 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.162141 Externals used for testing: cesm1_3_beta03 cism tag used for baseline comparisons: cism1_130905 Any other externals that differed in baseline: None ================================================================================ Originator: sacks, tcraig Date: Sept 5, 2013 Model: cism Version: cism1_130905 One-line summary: rework coupling fields, update glimmer-cism external Purpose of changes: (1) Rework the runoff coupling fields (rofi & rofl) to fit in with Tony Craig's coupler changes. These fields are now routed to the ocean and/or sea ice, instead of to the land. Also, these are no longer separated by elevation class. (2) Update glimmer-cism external from r1936 -> r1950. This brings in a number of changes. The most important are some changes from Bill Lipscomb that fix some of the g2x coupling fields, and remove the elevation class dimension from the g2x rofi & rofl fields. Changes answers relative to previous tag: YES Changes all of the g2x coupling fields. However, CISM history files are bfb Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Changes to coupling fields as described above M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_fields.F90 M drivers/cpl_share/glc_cpl_indices.F90 M drivers/cpl_mct/glc_comp_mct.F90 M drivers/cpl_esmf/glc_comp_esmf.F90 *** Routines to determine whether to send calving flux to ocean or sea ice A source_glc/glc_route_ice_runoff.F90 A drivers/cpl_share/glc_coupling_flags.F90 *** Add a namelist option to choose where calving flux should go M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml *** Update from r1936 -> r1950 M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. Testing was done in the context of Tony Craig's changes, which involved the coupler and many other components. The testing documented here was done on cplupa_n03_cism1_130624. Baseline comparisons fail for g2x fields and the derived x2l fields. CISM history files are bfb according to component_gen_comp. Test failures are: *** Expected failures CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel *** Failures in IG/FG tests that I think are expected (note: these all failed in the baseline, too; also note that I have TG tests that cover the same kind of functionality - e.g., a CME TG test - so this likely points to problems outside of GLC; also, Tony says that the IG failures are expected) FAIL ERI.f19_g16.IGRCP85CN.yellowstone_pgi FAIL SMS_D.f09_g16.IG20TR.yellowstone_pgi FAIL CME_Ly5.T31_g37.IG.yellowstone_intel FAIL SMS.f19_f19.FG20TRCN.yellowstone_intel Externals used for testing: models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/cesmcosp_cam5_3_xx_tags/cesmcosp_n01_cam5_3_04/models/atm/cam models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_130424 models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/satm models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm130325 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_06/xatm models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/branch_tags/cplupa_tags/cplupa_n01_share3_130528 models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/dead_share models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/branch_tags/cplupa_tags/cplupa_n02_drvseq4_2_33 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/cplupa models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/sglc models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xglc models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/branch_tags/cplupa_tags/cplupa_n01_cice4_0_20130524 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/branch_tags/cplupa_tags/cplupa_n01_dice8_130313 models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/sice models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xice models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/cplupa_tags/cplupa_n02_clm4_5_11/models/lnd/clm models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/branch_tags/cplupa_tags/cplupa_n01_dlnd8_130213 models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/slnd models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xlnd models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_130313 models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20130517 models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/socn models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xocn models/ocn/aquap https://svn-ccsm-models.cgd.ucar.edu/aquap/trunk_tags/aquap_130503 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_130503 models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/branch_tags/cplupa_tags/cplupa_n01_rtm1_0_28 models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/srof models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xrof models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_130213 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/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_130506 models/wav/swav https://svn-ccsm-models.cgd.ucar.edu/stubs/branch_tags/cplupa_tags/cplupa_n02_stubs1_4_02/swav models/wav/ww3 https://svn-ccsm-models.cgd.ucar.edu/ww3/trunk_tags/ww3_130314 models/wav/xwav https://svn-ccsm-models.cgd.ucar.edu/dead7/branch_tags/cplupa_tags/cplupa_n02_dead7_7_06/xwav scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/cplupa_tags/cplupa_n02_scripts4_130627 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_130625 tools/cprnc https://svn-ccsm-models.cgd.ucar.edu/tools/cprnc/trunk_tags/cprnc_130529 tools/mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_130509 cism tag used for baseline comparisons: cplupa_n02_cism1_130624 Note: I removed the rofi & rofl coupling stuff from cism so that it would work in the context of the above externals Any other externals that differed in baseline: Same as above ================================================================================ Originator: sacks Date: June 24, 2013 Model: cism Version: cism1_130624 One-line summary: update glimmer-cism external Purpose of changes: Update glimmer-cism external, from r1876 to r1936. Also update documentation of config options. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Yellowstone intel & pgi test lists. All PASS except the following expected failures: *** No ESMF for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.163141 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.163141 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.163141 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.163141 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.163141 Note that nlcomp failed for 10km & 20km, as expected (with comparison to cism1_130514) Externals used for testing: cesm1_2_beta08, with scripts @ scripts4_130513b cism tag used for baseline comparisons: cism1_130514 Any other externals that differed in baseline: None. ================================================================================ Originator: sacks Date: May 24, 2013 Model: cism Version: cism1_130524 One-line summary: change idiag & jdiag for 10km & 20km; allow other sigma values; fix some namelist documentation Purpose of changes: (see summary) Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Ran the following with comparison to baseline (but NOT component_gen_comp): SMS_D.f09_g16_gl10.TG.yellowstone_pgi ERS_Ly20.f09_g16.TG1850.yellowstone_intel SMS_D.f09_g16.TG.yellowstone_intel Ran the following with NO comparison to baseline: SMS_D.f09_g16.IG20TR.yellowstone_intel Also, ran the following with CISM_GRID=gland20, also with comparison to baseline (including component_gen_comp): SMS.f09_g16.TG.yellowstone_pgi Note that nlcomp failed for 10km & 20km, as expected Externals used for testing: cesm1_2_beta08, with scripts @ scripts4_130513b cism tag used for baseline comparisons: cism1_130514 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: May 14, 2013 Model: cism Version: cism1_130514 One-line summary: Update glimmer-cism: use double-precision for all variables Purpose of changes: Bring in changes from Bill L to convert all variables to double precision Changes answers relative to previous tag: YES -- answers change by single-precision roundoff (as expected), along with some slightly unexpected, large changes around the continental margin (at least in acab & temp) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update from r1863 -> r1876 M SVN_EXTERNAL_DIRECTORIES *** Change some variables to double precision M source_glc/glc_InitMod.F90 Summary of testing: Yellowstone intel & pgi test lists. For this testing, I have switched over to using the xml-based test lists, since they have been fixed to include all desired tests in the latest scripts tag. Baseline comparisons fail, as expected. Other than that, failures are: *** No ESMF for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.094304 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.094301 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.094301 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.094301 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.094301 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.094301 Externals used for testing: cesm1_2_beta08, with scripts @ scripts4_130513b cism tag used for baseline comparisons: cism1_130502 Any other externals that differed in baseline: Unsure... I may have used cesm1_2_beta06 for the baseline by accident, but it doesn't really matter because there are known baseline differences in this new tag. ================================================================================ Originator: sacks Date: May 2, 2013 Model: cism Version: cism1_130502 One-line summary: Update glimmer-cism: fix restart bugs, clean-up from Bill L Purpose of changes: Update glimmer-cism from r1849 -> 1863, mainly for two purposes: - Fixes for restart bugs in non-default configurations (from Matt Hoffman) - Clean-up (mostly of comments) from Bill Lipscomb Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. I used the plain text test lists as they existed before the conversion of test lists to xml, because there were some missing tests in the xml All PASS except: *** No ESMF for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.150138 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.150142 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.150142 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.150142 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.150142 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.150142 Externals used for testing: cesm1_2_beta06, but PIO @ pio1_6_8 cism tag used for baseline comparisons: cism1_130430a Any other externals that differed in baseline: No differences ================================================================================ Originator: sacks Date: April 30, 2013 Model: cism Version: cism1_130430a One-line summary: Update glimmer-cism external, including getting restarts to work for evolution=2; also add some config options Purpose of changes: - Update external (1846 -> 1849): - some cleanup from Bill Lipscomb - get restarts to work with evolution=2 - Add some config options to namelist_definition file Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_definition_cism.xml M SVN_EXTERNAL_DIRECTORIES Summary of testing: Yellowstone intel & pgi test lists. I used the plain text test lists as they existed before the conversion of test lists to xml, because there were some missing tests in the xml All PASS except: *** No ESMF for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.084959 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.085004 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.085004 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.085004 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.085004 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.085004 Also ran all CISM1 tests from a sandbox that set evolution=2, dt=0.0083333333333333333333333333333333333 by default (for all resolutions). These all pass except: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.102727 Note, in particular, that all restart tests are now passing for evolution=2. However, note that a small time step is needed for these evolution=2 cases to pass (even using dt=0.025 led to failure of some cases) Externals used for testing: cesm1_2_beta06, but PIO @ pio1_6_8 cism tag used for baseline comparisons: cism1_130429 Any other externals that differed in baseline: No differences ================================================================================ Originator: sacks Date: April 30, 2013 Model: cism Version: cism1_130430 One-line summary: Use new glint interfaces Purpose of changes: Use new glint interfaces from Bill L. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update glimmer-cism external from r1837 to r1846, to use new glint interfaces (this also includes some other changes that should have no effect on the code run within CESM) M SVN_EXTERNAL_DIRECTORIES *** Use new glint interfaces M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 *** Comment out use of deleted itest, jtest (now these are iglint_global and jglint_global, but those are hard-coded values that don't make sense for most runs, so for now it's best not to use them) M source_glc/glc_global_grid.F90 *** Unrelated clean-up of documentation M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Yellowstone intel & pgi test lists. I used the plain text test lists as they existed before the conversion of test lists to xml, because there were some missing tests in the xml All PASS except: *** No ESMF for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.202525 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.202556 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.202556 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.202556 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.202556 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.202556 Externals used for testing: cesm1_2_beta06, but PIO @ pio1_6_8 cism tag used for baseline comparisons: cism1_130429 Any other externals that differed in baseline: No differences ================================================================================ Originator: sacks Date: April 29, 2013 Model: cism Version: cism1_130429 One-line summary: Change some parameter values Purpose of changes: The previous tag kept parameter values at their cism1 defaults, for testing purposes. In this tag, we change parameter values to those that we want moving forward. In particular, this changes: - temp_init: 2 instead of 1 - basal_mass_balance: 1 instead of 0 - sigma: 0 instead of 2 for cism_phys=cism1 Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml *** fix some comments M bld/build-namelist Summary of testing: Yellowstone intel & pgi test lists. I used the plain text test lists as they existed before the conversion of test lists to xml, because there were some missing tests in the xml All PASS except: *** Missing ESMF library CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.G.061715 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.G.061723 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.G.061723 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.G.061723 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.G.061723 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.G.061723 Externals used for testing: cesm1_2_beta06, but PIO @ pio1_6_8 cism tag used for baseline comparisons: N/A -- no baseline comparisons performed Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: April 28, 2013 Model: cism Version: cism1_130428 One-line summary: Cleanup of glimmer-cism external code, with focus on cleaning up config options Purpose of changes: - Bring in latest glimmer-cism external, which cleans up config options, as well as a number of other changes from Bill L - Get cismIO generation working - Run some tests to make sure that new code gives the same answers as the cesm1.1.1 release Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M SVN_EXTERNAL_DIRECTORIES *** Get cismIO generation working. This includes getting files from glimmer-cism external rather than maintaining copies of them D bld/cismIO/glide_lithot_vars.def D bld/cismIO/glint_mbal_vars.def D bld/cismIO/glint_vars.def D bld/cismIO/ncdf_template.F90.in D bld/cismIO/glide_vars.def D bld/cismIO/generate_ncvars.py M bld/cismIO/cism.buildIO.template.csh M bld/cism.cpl7.template *** Handle Bill L's changes to config options M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml *** Add blank line M bld/user_nl_cism Summary of testing: (1) Yellowstone intel & pgi test lists. I used the plain text test lists as they existed before the conversion of test lists to xml, because there were some missing tests in the xml All PASS except: *** Missing ESMF library CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.194914 *** CISM2 failures FAIL ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.GC.194923 FAIL NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel.GC.194923 FAIL PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.GC.194923 FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel.GC.194923 FAIL SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel.GC.194923 (2) Comparisons of latest code base with cesm1.1.1 release code. Expect CISM1 compsets to be the same within roundoff, once some code mods were made to the new and old code to correct for known differences (put in place by Bill L). For each of these, I ran one-year TG compsets from the new code base and from cesm1.1.1; I compared log files (first few time steps, and last time step of the year-long run), looked at cprnc results, and did a visual comparison of history files for any fields with large differences according to cprnc. (a) baseline, out-of-the-box settings (done by Bill L, reproduced by me) (b) basal sliding on (done by Bill L) (c) 10 km (d) 20 km (e) evolution=2 (f) basal_water=1 (g) flow_law: 0 in new, 2 in old (h) sigma=0 (in old code, implemented this by commenting out the call to glide_read_sigma) (i) use Jeremy Fyke's initial conditions, which have been copied to bg40.1850.track1.1deg.006b.cism.r.0863-01-01-00000.nc; along with hotstart=1 (j) slip_coeff: 5 in old, 3 in new; also, use initial conditions as in (i) All of these appeared to differ only by single-precision roundoff, EXCEPT (i) was questionable and (j) had big differences (in cesm1.1.1, (j) was unstable in the old code). The differences in (i) suggest that old cism1 restart files MAY not be compatible with the new code base. The differences in (j) might be connected to the use of initial conditions in this run, rather than a problem with slip_coeff per se. For example, it seems that the old code, with Bill L's mods to make the old & new more similar, didn't read bheatflx from the restart file. Externals used for testing: cesm1_2_beta06, but PIO @ pio1_6_8 cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: April 25, 2013 Model: cism Version: cism1_130425 One-line summary: Allow longer case names, and error-check too-long case names Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M source_glc/glc_io.F90 M source_glc/glc_time_management.F90 Summary of testing: Ran standard cism1 tests on yellowstone, from glc aux test lists (for intel & pgi); took these test lists from an older scripts tag from before the conversion to xml, since the xml list has some problems. All tests PASSed, except the following expected failures: *** No esmf library for pgi CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.GC.145755 *** PIO broken for mpi-serial RUN PEA_P1_M.f09_g16.TG20TR.yellowstone_intel.GC.145802 All baseline comparisons, including component_gen_comp, PASSed or BFAILed in an expected way. Also ran an ERI_Ly44 test with a long testid that resulted in a > 160 character case name (old limit was 80, new is 256 characters). Externals used for testing: cesm1_2_beta06 cism tag used for baseline comparisons: cism1_130405 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: April 5, 2013 Model: cism Version: cism1_130405 One-line summary: Change namelist defaults for cism1 Purpose of changes: I was getting instabilities in cism1. Resolving these required either decreasing the time step or changing evolution back to 0 (rather than the new default of 2). Bill L recommended changing evolution back to 0. So this tag changes these defaults for cism1: - evolution=0 rather than 2 - dt=0.1 rather than 0.025 for gl10 These changes bring the cism1 namelist defaults back to what they were in cism1_121114 Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml Summary of testing: NOTE: The testing was accidentally run on a version of code that did NOT include the changes from cism1_130403. JUST RAN THE CISM1 TESTS FROM THESE LISTS (NO BASELINE COMPARISONS): - yellowstone_intel.glc.auxtest - yellowstone_pgi.glc.auxtest All PASS except the following expected failures: CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.G.205611 FAIL CME_Ly5.T31_g37.IG.yellowstone_intel.G.205554 NOTE THAT THE ERS AND ERI TESTS ARE NOW PASSING! CONFIRMING THAT RESTART TESTS ARE NOW WORKING: Ran the following additional tests: PASS ERS_Ly100.f09_g16.TG.yellowstone_intel.062613 PASS ERS_Ly100.f09_g16.TG.yellowstone_intel.062613.memleak PASS ERI_Ly220.f09_g16.TG20TR.yellowstone_intel.062613 PASS ERI_Ly220.f09_g16.TG20TR.yellowstone_intel.062613.memleak PASS ERS_Ly200.f09_g16_gl10.TG1850.yellowstone_intel.062613 PASS ERS_Ly200.f09_g16_gl10.TG1850.yellowstone_intel.062613.memleak PASS ERI_Ly440.f09_g16_gl10.TGRCP85.yellowstone_intel.062613 PASS ERI_Ly440.f09_g16_gl10.TGRCP85.yellowstone_intel.062613.memleak PASS ERS_Ly20.f09_g16_gl10.TG1850.yellowstone_intel.062613 PASS ERS_Ly20.f09_g16_gl10.TG1850.yellowstone_intel.062613.memleak PASS ERI_Ly44.f09_g16_gl10.TGRCP85.yellowstone_intel.062613 PASS ERI_Ly44.f09_g16_gl10.TGRCP85.yellowstone_intel.062613.memleak SO INDEED, IT SEEMS THAT RESTARTS ARE EXACT NOW, WITH EVOLUTION=0 Externals used for testing: cesm1_2_beta05, except scripts4_130405 and CLM at branches/newcompsets/models/lnd/clm, r45570 cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: April 3, 2013 Model: cism Version: cism1_130403 One-line summary: Add CISM_OBSERVED_IC option for hybrid runs Purpose of changes: Allow cism to ignore the refcase's restart file in a hybrid run by setting the CISM_OBSERVED_IC env_run.xml variable to TRUE. As part of this change, the default value of hotstart is now set in build-namelist rather than being set via namelist_defaults_cism.xml. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: - Test runs where I leave CISM_OBSERVED_IC at its default value (FALSE); compare namelist/config file and cism history file with baseline runs - startup - namelist/config files - hist file - hybrid - namelist/config files - hist file - branch - namelist/config files - hist file - Test runs where I set CISM_OBSERVED_IC to TRUE; compare namelist/config file and cism history file with baseline runs - startup - namelist/config files NOT DOING - hist file - hybrid: should differ from baseline hybrid, be identical to baseline startup (in particular: hostart, cisminputfile) - namelist/config files - hist file: should be identical to baseline startup Note: time coordinate variable differs from baseline hybrid - restart file: should be identical to baseline startup Note: time coordinate variable differs from baseline hybrid - compare netcdf headers for hist & restart files - branch: BUILD-NAMELIST SHOULD DIE Externals used for testing: cesm1_2_beta05, with updated scripts cism tag used for baseline comparisons: cism1_130401 Any other externals that differed in baseline: old scripts used for baseline runs ================================================================================ Originator: sacks Date: April 1, 2013 Model: cism Version: cism1_130401 One-line summary: Update external, including significant glide updates Purpose of changes: Update glimmer-cism external. Among other things, this includes a large set of glide updates from Bill Lipscomb. Changes answers relative to previous tag: YES Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update glimmer-cism from r1773 to r1785 M SVN_EXTERNAL_DIRECTORIES *** Delete bluefire test stubs M ChangeLog_template Summary of testing: yellowstone_intel.glc.auxtest & yellowstone_pgi.glc.auxtest (skipped hopper glc auxtests because for now I just care about testing CISM1 compsets) All PASS except: ***** NEW FAILURE: Problems building with mpi-serial with cmake RUN PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel.G.162507 ***** NEW FAILURE: Problem in CLM or RTM ESMF (should be resolved as of clm4_0_69) FAIL CME_Ly5.T31_g37.IG.yellowstone_intel.G.162507 ***** Expected failures based on failures in cism1_130207 FAIL ERS_Ly20.f09_g16.TG1850.yellowstone_intel.G.162507 FAIL ERI_Ly44.f09_g16.TGRCP85.yellowstone_intel.G.162507 RUN ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel.G.162507 FAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_intel.G.162507 FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.yellowstone_pgi.G.162510 CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi.G.162510 Externals used for testing: cesm1_2_beta05 cism tag used for baseline comparisons: N/A: No baseline comparisons done Any other externals that differed in baseline: N/A ================================================================================ Originator: tcraig Date: March 25, 2013 Model: cism Version: cism1_130325 One-line summary: updates for new CISM_GRID env variable the scripts env variable GLC_GRID was renamed to CISM_GRID and a few changes were needed in cism scripts. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M models/glc/cism/bld/trilinosOptions/README M models/glc/cism/bld/README.build-namelist M models/glc/cism/bld/build-namelist M models/glc/cism/bld/cism.buildnml.csh Summary of testing: Ran a handful of yellowstone prealpha/beta tests Externals used for testing: cesm1_2_alpha04a+ cism tag used for baseline comparisons: cism1_130315 Any other externals that differed in baseline: Several including a dependency on scripts changes and everything else in alpha05a. This tag does depends on the scripts tag in cesm1_2_alpha05a and that scripts tag depends on everything else in alpha05a (add wave model). ================================================================================ Originator: sacks Date: March 15, 2013 Model: cism Version: cism1_130315 One-line summary: update glimmer-cism, including adding support for 36 columns Purpose of changes: Add support for 36 columns. While I'm at it, update glimmer-cism external to latest revision. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Update glimmer-cism from r1746 to r1773 M . M SVN_EXTERNAL_DIRECTORIES Summary of testing: Ran the following test list, including component_gen_comp (note: this is the shortlist.glc.auxtest list, plus a long IG test): SMS_D.f09_g16.TG SMS_D_Ly1.f09_g16_gl10.TGIS2 ERS_Ly20.f09_g16.TG1850 SMS_D.f09_g16.IG20TR SMS_Ly5.T31_g37.IG All pass, except the following, which also failed in the baseline: FAIL ERS_Ly20.f09_g16.TG1850.yellowstone_intel.GC.141629 Externals used for testing: cesm1_2_alpha03d cism tag used for baseline comparisons: cism1_130307 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: March 7, 2013 Model: cism Version: cism1_130307 One-line summary: fix dependency appends for new mkDepends... for real, now Purpose of changes: Yesterday's fix didn't really do the right thing. Now I have removed the awkward awk script, and instead use new functionality in mkDepends to do this appending the right way. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/cism.buildexe.csh Summary of testing: Summary of testing: MINIMAL TESTING DONE: JUST TESTED BUILD FOR A CASE WITH: -mach yellowstone -compset IG -res f09_g16 Externals used for testing: cesm1_2_alpha03b, with Machines at Machines_130307 cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: March 6, 2013 Model: cism Version: cism1_130306 One-line summary: fix dependency appends for new mkDepends Purpose of changes: The awk script that appended libglimmercismfortran.a to the list of dependencies accidentally assumed a space at the end of the lines created by mkDepends. The new version of mkDepends doesn't include a space at the end of each line, so the dependencies weren't created correctly. Now the necessary space is added in the awk script. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Just add a space M bld/cism.buildexe.csh Summary of testing: MINIMAL TESTING DONE: JUST TESTED BUILD FOR A CASE WITH: -mach yellowstone -compset IG -res f09_g16 Externals used for testing: cesm1_2_alpha03b, with Machines at Machines_130304b cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Feb 7, 2013 Model: cism Version: cism1_130207 One-line summary: bring cism2 to trunk Purpose of changes: Bring cism2 to trunk (merge changes from cism2_cesm_bld_wjs branch). This inovles: - bringing in the new glimmer-cism code base as an svn external, rather than using copies of the source files - updating the build system to handle building glimmer-cism as a standalone library, using its native cmake-based build system - modifying some source code so GLC works properly in a parallel environment (see README.parallelization for details) - adding namelist options to support cism2 Changes answers relative to previous tag: YES. There are currently big answer changes (orders of magnitude differences), even for cism1 compsets that are supposed to give roughly the same answer. Bill Lipscomb is looking into this, and a new tag will be made once these changes are resolved. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): None List all modified files, and describe the changes: *** Bring in new glimmer-cism code base as an svn external, delete old copies of *** source files A SVN_EXTERNAL_DIRECTORIES D source_glimmer-cism D source_glimmer-cism/glide.F90 D source_glimmer-cism/glimmer_scales.F90 D source_glimmer-cism/glint_timestep.F90 D source_glimmer-cism/smb_dummy.F90 D source_glimmer-cism/glide_thck.F90 D source_glimmer-cism/isostasy_setup.F90 D source_glimmer-cism/isostasy_el.F90 D source_glimmer-cism/glide_nc_custom.F90 D source_glimmer-cism/glide_lithot_io.F90 D source_glimmer-cism/glint_mbal_io.F90 D source_glimmer-cism/glimmer_ts.F90 D source_glimmer-cism/isostasy_types.F90 D source_glimmer-cism/glimmer_searchcircle.F90 D source_glimmer-cism/glimmer_routing.F90 D source_glimmer-cism/glimmer_daily_pdd.F90 D source_glimmer-cism/xls.F90 D source_glimmer-cism/glint_mbal.F90 D source_glimmer-cism/glide_temp.F90 D source_glimmer-cism/glint_climate.F90 D source_glimmer-cism/glimmer_utils.F90 D source_glimmer-cism/glimmer_map_init.F90 D source_glimmer-cism/glimmer_anomcouple.F90 D source_glimmer-cism/glimmer_filenames.F90 D source_glimmer-cism/glimmer_ncparams.F90 D source_glimmer-cism/glimmer_writestats.F90 D source_glimmer-cism/glimmer_restart_gcm.F90 D source_glimmer-cism/glint_initialise.F90 D source_glimmer-cism/glimmer_paramets.F90 D source_glimmer-cism/glimmer_vers.F90 D source_glimmer-cism/glide_lithot1d.F90 D source_glimmer-cism/glimmer_map_CFproj.F90 D source_glimmer-cism/glide_mask.F90 D source_glimmer-cism/glint_io.F90 D source_glimmer-cism/glide_setup.F90 D source_glimmer-cism/glide_profile.F90 D source_glimmer-cism/glimmer_map_proj4.F90 D source_glimmer-cism/glide_types.F90 D source_glimmer-cism/glide_velo.F90 D source_glimmer-cism/glimmer_global.F90 D source_glimmer-cism/glimmer_map_types.F90 D source_glimmer-cism/glimmer_deriv.F90 D source_glimmer-cism/glimmer_coordinates.F90 D source_glimmer-cism/glimmer_ncdf.F90 D source_glimmer-cism/kelvin.F90 D source_glimmer-cism/glide_stop.F90 D source_glimmer-cism/ncdf_utils.F90 D source_glimmer-cism/glint_example_clim.F90 D source_glimmer-cism/glimmer_log.F90 D source_glimmer-cism/glimmer_integrate.F90 D source_glimmer-cism/glint_precip_param.F90 D source_glimmer-cism/glint_global_grid.F90 D source_glimmer-cism/glint_constants.F90 D source_glimmer-cism/glimmer_sparse.F90 D source_glimmer-cism/glide_diagnostics.F90 D source_glimmer-cism/isostasy.F90 D source_glimmer-cism/glint_global_interp.F90 D source_glimmer-cism/glint_main.F90 D source_glimmer-cism/glimmer_config.F90 D source_glimmer-cism/glimmer_pdd.F90 D source_glimmer-cism/glint_mpinterp.F90 D source_glimmer-cism/glint_interp.F90 D source_glimmer-cism/glide_lithot3d.F90 D source_glimmer-cism/glimmer_physcon.F90 D source_glimmer-cism/glimmer_map_trans.F90 D source_glimmer-cism/profile.F90 D source_glimmer-cism/glimmer_ncio.F90 D source_glimmer-cism/glide_lithot.F90 D source_glimmer-cism/glint_mbal_coupling.F90 D source_glimmer-cism/glide_io.F90 D source_glimmer-cism/glint_type.F90 D source_slap D source_slap/dmset.F D source_slap/dlaputil.F D source_slap/xersla.F D source_slap/dcgs.F D source_slap/blas.F D source_slap/dmvops.F D source_slap/mach.F *** Add xml files giving trilinos options; the correct one (based on resolution) *** is copied to the run directory A bld/trilinosOptions A bld/trilinosOptions/trilinosOptions_gland10.xml A bld/trilinosOptions/trilinosOptions_gland20.xml A bld/trilinosOptions/trilinosOptions_gland5.xml A bld/trilinosOptions/trilinosOptions_gland5UM.xml A bld/trilinosOptions/README M bld/cism.buildnml.csh M bld/README *** New build system for cism2, which includes building a separate glimmer-cism *** library using a cmake-based build system M bld/cism.buildexe.csh *** Support new config options for cism2, including new xml variables *** "CISM_PHYS" and "CISM_USE_TRILINOS". Also change a few defaults for cism1: *** evolution=2 rather than 0, and dt=0.025 rather than 0.1 for 10km (to achieve *** stability when using evolution=2) M bld/build-namelist M bld/README.build-namelist M bld/namelist_files/namelist_definition_cism.xml M bld/namelist_files/namelist_defaults_cism.xml *** Use new sourceMods directory for glimmer-cism-specific code *** (src.cism/glimmer-cism) M bld/cism.cpl7.template *** Support operation in a parallel environment M mpi/glc_communicate.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_io.F90 --------- also, minor changes for new glimmer-cism i/o interface M drivers/cpl_mct/glc_comp_mct.F90 M drivers/cpl_esmf/glc_comp_mct.F90 M drivers/cpl_esmf/glc_comp_esmf.F90 A README.parallelization Summary of testing: ------------------------------------------------------------------------ Core testing ------------------------------------------------------------------------ Ran the following from https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/cism2_bld_tags/cism2_bld_21_cesm1_2_beta01 For each machine, I first give failures, then passes. All of these failures should be resolved at some point (except the titan failures, which basically duplicate the hopper failures: no need to rerun titan tests in the future). Yellowstone: yellowstone_intel.glc.auxtest & yellowstone_pgi.glc.auxtest ***** Restarts currently are not bfb FAIL ERS_Ly20.f09_g16.TG1850.yellowstone_intel FAIL ERI_Ly44.f09_g16.TGRCP85.yellowstone_intel FAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_intel FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.yellowstone_pgi ***** Hangs in initialization of the last run (the continue run) ***** (similar to the ERS IS2 run that fails on hopper) RUN ERI_Ly15.f09_g16_gl10.TG20TRIS2.yellowstone_intel ***** No ESMF library for PGI CFAIL ERS_Ly20_E.f09_g16.TG20TR.yellowstone_pgi PASS SMS_D.f09_g16.TG.yellowstone_intel PASS SMS_D_Ly1.f09_g16_gl10.TGIS2.yellowstone_intel PASS PEA_P1_M.f09_g16.TG20TR.yellowstone_intel PASS PEA_P1_M_Ly2.f09_g16_gl10.TGIS2S.yellowstone_intel PASS NCK.f09_g16.TG.yellowstone_intel PASS NCK_Ly3.f09_g16_gl10.TGRCP85IS2.yellowstone_intel PASS CME_Ly10_N2_P2_D.f09_g16.TG1850.yellowstone_intel PASS CME_Ly5.T31_g37.IG.yellowstone_intel PASS SMS.f19_f19.FG20TRCN.yellowstone_intel PASS SMS_D.T31_g37.BG1850CN.yellowstone_intel PASS SMS_Ly3.T31_g37_gl10.IGIS2.yellowstone_intel PASS SMS_D.f09_g16_gl10.TG.yellowstone_pgi PASS SMS_D.f09_g16.IG20TR.yellowstone_pgi PASS ERI.f19_g16.IGRCP85CN.yellowstone_pgi Hopper: hopper_gnu.glc.auxtest & hopper_pgi.glc.auxtest ***** Hangs in initialization of GLC for the restart run RUN ERS_Ly11.f09_g16_gl10.TG1850IS2.hopper_gnu ***** Dies while writing history file: subscript out of range (Dies ***** while trying to write uflx: The problem is that it's trying to ***** write 1:local_ewn, but for uflx, values only go 1:18, since it's ***** on the flux grid. I think this is due to glint using some glide ***** init routines rather than glissade, so nhalo has the wrong ***** value) FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.hopper_pgi ***** Dies as follows: ***** (shr_mct_sMatReaddnc) reading mapping matrix data decomposed... ***** (shr_mct_sMatReaddnc) * file name : idmap ***** (shr_sys_abort) ERROR: (shr_mct_sMatReaddnc)No such file or directory ***** I believe I have fixed this with a drv change, and have tested ***** SMS.T31_g37_gl10.BGCNIS2.yellowstone_intel, but haven't rerun ***** this exact test FAIL SMS_D.T31_g37_gl10.BGCNIS2.hopper_pgi PASS SMS_D_Ly1.f09_g16_gl10.TGIS2.hopper_gnu PASS SMS.f19_f19_gl10.FGCNIS2.hopper_gnu PASS CME_Ly3.f09_g16_gl10.TGRCP85IS2.hopper_pgi Titan: titan.glc.auxtest: ***** See notes for the similar hopper_pgi test FAIL SMS_D_Ly1.f09_g16_gl10.TGIS2.titan_pgi ***** See notes for the similar hopper_gnu test RUN ERS_Ly11.f09_g16_gl10.TG1850IS2.titan_pgi ***** See notes for the similar hopper_pgi test FAIL SMS_D.T31_g37_gl10.BGCNIS2.titan_pgi PASS CME_Ly3.f09_g16_gl10.TGRCP85IS2.titan_pgi PASS SMS.f19_f19_gl10.FGCNIS2.titan_pgi ------------------------------------------------------------------------ Additional CISM1 testing on multiple machines/compilers ------------------------------------------------------------------------ Ran the following test list: [SMS_Ly5 test is just in case the CME test fails because of lack of an esmf library] [B is to test a NON-GLC compset, to make sure I haven't broken that] SMS_D_Ly2.f09_g16.TGIS1 SMS_D.f09_g16.IG20TRIS1 SMS_Ly5.f09_g16_gl10.TG1850IS1 CME_Ly5.f09_g16_gl10.TG1850IS1 SMS.T31_g37.B on the following machines & compilers: - hopper - pgi - frankfurt - intel - pgi - yellowstone - pgi - intel - gnu - titan - pgi - intrepid - ibm (I did NOT run on Janus, assuming it's close enough to yellowstone. However, I DID test a TG build on Janus) In general, used: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/cism2_bld_tags/cism2_bld_17_cesm1_2_beta01 (on yellowstone, used the following for the B compset test; this updates POP: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/cism2_bld_tags/cism2_bld_18_cesm1_2_beta01) (on yellowstone, used the following for the gnu tests: https://svn-ccsm-models.cgd.ucar.edu/cesm1/exp_tags/cism2_bld_tags/cism2_bld_19_cesm1_2_beta01) IMPORTANT TEST FAILURES (I want to resolve these at some point): FAIL SMS_D_Ly2.f09_g16.TGIS1.frankfurt_intel FAIL SMS_Ly5.f09_g16_gl10.TG1850IS1.frankfurt_intel UNIMPORTANT TEST FAILURES: ***** I think these are just because there are no ESMF libraries for these machines/compilers: CFAIL CME_Ly5.f09_g16_gl10.TG1850IS1.frankfurt_intel CFAIL CME_Ly5.f09_g16_gl10.TG1850IS1.frankfurt_pgi CFAIL CME_Ly5.f09_g16_gl10.TG1850IS1.yellowstone_pgi CFAIL CME_Ly5.f09_g16_gl10.TG1850IS1.yellowstone_gnu CFAIL CME_Ly5.f09_g16_gl10.TG1850IS1.intrepid_ibm ***** Dies in lnd init. Sheri says no I compsets are being tested on ***** intrepid, so I'm not going to worry about this. The following ***** two additional tests passed on intrepid: ***** ERS.f09_f09.FG1850CN.intrepid_ibm and ERS_Ld7.f09_g16.BGRCP26CN FAIL SMS_D.f09_g16.IG20TRIS1.intrepid_ibm ***** Also, Sheri said that she could only get glc to build if she set ***** NTHRDS_GLC=1 (weird!) -- otherwise xlf hung while compiling ***** glide_thck.F90. For now I have changed the PE layouts to set ***** NTHRDS_GLC=1 by default on intrepid, as a workaround. ------------------------------------------------------------------------ Externals used for testing: See above cism tag used for baseline comparisons: N/A -- no baseline comparisons done, because big answer changes were expected (at some point, once some bugs are fixed, we should probably do baseline comparisons against cism1_121114 to confirm that cism1 compsets give [nearly] identical answers) Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Nov 14, 2012 Model: cism Version: cism1_121114 One-line summary: remove docs folder Purpose of changes: Remove duplicated docs folder. Instead, see https://svn-cism-model.cgd.ucar.edu/docs Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: D doc D doc/CESM_ice_sheets_documentation.pdf D doc/glimmer_doc.pdf D doc/README Summary of testing: NONE - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: N/A Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: jwolfe Date: Nov 13, 2012 Model: cism Version: cism1_121113 One-line summary: Add parallel support in source_glc code Purpose of changes: In preparation for bringing cism2 in, some modifications were needed in source_glc so that this interface layer could operate in a parallel environment. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: A mpi/glc_broadcast.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_io.F90 M source_glc/glc_time_management.F90 Summary of testing: (Testing done by Bill Sacks) - bluefire.glc.auxtest: main test results: All PASS Note: I made a mistake in running the baseline comparisons, so ran them after the fact with component_gen_comp -model cpl - bluefire.glc.auxtest: results of component_gen_comp -model cism: All PASS - Other tests: Jon Wolfe ran some tests specific to the new functionality Externals used for testing: cesm1_1_beta18 cism tag used for baseline comparisons: cism1_121012 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: Nov 6, 2012 Model: cism Version: cism1_121106a One-line summary: Change name of documentation file Purpose of changes: Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: A + doc/CESM_ice_sheets_documentation.pdf D doc/cism_doc.pdf M doc/README Summary of testing: NONE - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Nov 6, 2012 Model: cism Version: cism1_121106 One-line summary: Update documentation Purpose of changes: Update documentation for CESM1.1 release Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Add the CISM documentation to the repository A doc/cism_doc.pdf *** Update to version 1.5.1 (rather than 1.0.0) M doc/glimmer_doc.pdf *** Update notes about what's included in the doc directory M doc/README *** Add notes about how to add a new optional section in cism.config M bld/README.build-namelist Summary of testing: NONE (JUST CHANGED DOCUMENTATION) - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: NONE Externals used for testing: N/A cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: sacks Date: Oct 12, 2012 Model: cism Version: cism1_121012 One-line summary: add namelist documentation, validation of dt, add/remove namelist items Purpose of changes: Main purpose is adding namelist documentation. Also added validation of dt in build-namelist (can't add this through simple valid_values). Also added / removed a few namelist parameters: Add scale_factor in cism.config: projection section. Removed namelist parameters which seem not to be used anywhere: isostasy, sliding_law, stress_calc, isos_time Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Add lots of documentation & valid_values; *** add scale_factor namelist parameter; *** remove isostasy, sliding_law, stress_calc, isos_time; *** change standard_parallel to length-2 array *** change categories for html documentation M bld/namelist_files/namelist_definition_cism.xml *** Add validation of dt; remove isostasy, sliding_law, stress_calc, isos_time M bld/build-namelist *** remove isostasy, sliding_law, stress_calc, isos_time M bld/namelist_files/namelist_defaults_cism.xml *** Mostly unrelated: Add documentation M bld/README.build-namelist *** Update ChangeLog entry for cism1_121009: document some more careful *** checks of the reasons for differences from baseline M ChangeLog Summary of testing: - bluefire.glc.auxtest: main test results: All PASS except: (Ignoring failures in NLComp & tputcomp) ***** Test not run in baseline case (this test was deleted in a recent scripts tag) BFAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.C.161152.compare_hist.cesm1_1_alpha18h_cism_smallerDT ***** Diffs just in x2a dust fluxes, due to externals change compared to baseline FAIL SMS.f19_f19.FG20TRCN.bluefire_ibm.GC.063931.compare_hist.cesm1_1_alpha18f_scripts4_121008_clm_newGlacierData2_dlnd_121001_cism_changeTimeToDouble FAIL SMS_D.T31_g37.BG1850CN.bluefire_ibm.GC.063931.compare_hist.cesm1_1_alpha18f_scripts4_121008_clm_newGlacierData2_dlnd_121001_cism_changeTimeToDouble So none of these failures are worrisome - bluefire.glc.auxtest: results of component_gen_comp -model cism: All PASS except: ***** Test not run in baseline case (this test was deleted in a recent scripts tag) BFAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.compare_hist.cesm1_1_alpha18h_cism_smallerDT.cism (baseline directory does not exist) - Other tests: - out-of-the-box TG compset cism_in & cism.config identical to old case, except for the 4 removed parameters (and runid in cism_in) Externals used for testing: https://svn-ccsm-models.cgd.ucar.edu/cesm1/tags/cesm1_1_alpha18g except: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_121008 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_0_54/models/lnd/clm cism tag used for baseline comparisons: cism1_121009 for most tests; cism1_121010 for _gl10 tests Any other externals that differed in baseline: MANY! (used baselines generated from cism1_121009 and cism1_121010, using externals as listed below) ================================================================================ Originator: sacks Date: Oct 10, 2012 Model: cism Version: cism1_121010 One-line summary: Decrease dt for resolutions other than gland5UM Purpose of changes: I found that temperature was blowing up for out-of-the-box gland10 cases, ever since the cism1_121001 tag. Jeremy Fyke & Bill Lipscomb suggested fixing this by decreasing dt to 0.1. I am decreasing dt to 0.1 for gland10 & gland20, and to 0.05 for gland5 (same as what is already the case for gland5UM). Changes answers relative to previous tag: YES, for gland10, gland20 and gland5 (but not gland5UM) Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Change defaults for dt & ndiag M bld/namelist_files/namelist_defaults_cism.xml *** Update ChangeLog entry for cism1_121009: document additional testing M ChangeLog Summary of testing: - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: Ran the single test from the bluefire.glc.auxtest list that tests a resolution other than gland5UM: PASS ERS_Ly20_N2_P2.f09_g16_gl10.TG.bluefire_ibm.G.132320 Also, ran 300-year tests with TG compset, GLC_GRID=gland10 or gland20. Jeremy Fyke confirmed that temperature and other values look reasonable. Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_121009b scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_121009 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120929 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_07 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_121003 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_1_44/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_0_54/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_08 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120927 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120927 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_121008 models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_121001 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120924 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120925 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120921 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.2 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_4/pio cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: lipscomb / sacks Date: Oct 9, 2012 Model: cism Version: cism1_121009 One-line summary: Change time to double precision to fix restart problem Purpose of changes: Starting with cism1_121001, some ERS tests were failing. Bill Lipscomb traced this to the mixture of single and double precision variables (and in one case an integer!) that are used to track time: a double precision time variable was being converted to single precision before being written to the restart file. We have changed these to be double precision universally. Note that this just refers to the time variables that are specified as fractional years: time variables that give number of hours are still integers. Also: Fix GetValueDouble to read in value as double precision Note that Bill L has also implemented these fixes in the seacism branch of cism2. Changes answers relative to previous tag: YES I expected answer changes on the order of single precision roundoff. However, the actual answer changes are larger than that for select grid cells (cprnc's worst decimal digits between 1 and 2 for some variables). I suspect this is just due to non-linear feedbacks in the system, but haven't confirmed that; I am doing more investigation of this, but needed to make this tag before that investigation is complete. - update 10-12-12: diffs in first year are less, then diffs increase. From looking quickly at maps of diffs in first year, I'm not too concerned about this - I can believe that these diffs are really due to single vs double precision, with initial small differences expanding over time due to feedbacks in the system. Also, there are differences in a gland10 run, too; I did not expect these differences, since dt=1 there. This is also undergoing additional investigation. - update 10-11-12: note that geothermal heat flux is now read as double rather than single precision. That could be the explanation. Another explanation could be that time is used in calculations that are now done in double rather than single precision. I'm not too concerned because diffs in first year are very small. Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Main changes: change time variables to double precision M source_glc/glc_io.F90 M source_glimmer-cism/glide_thck.F90 M source_glimmer-cism/glide_lithot_io.F90 M source_glimmer-cism/glint_mbal_io.F90 M source_glimmer-cism/glimmer_ncparams.F90 M source_glimmer-cism/glint_io.F90 M source_glimmer-cism/glide_diagnostics.F90 M source_glimmer-cism/glimmer_ncio.F90 M source_glimmer-cism/glide_io.F90 M source_glimmer-cism/glint_type.F90 M source_glimmer-cism/glimmer_ncdf.F90 - includes change of an integer to double M source_glimmer-cism/glide.F90 - also change unit 6 to stdout, changed some diagnostic output M source_glimmer-cism/glide_types.F90 - also change defaults for tend, tinc - also fix typo in comment M source_glimmer-cism/glint_main.F90 - also add some diagnostic output - also change diagnostic output to write model%numerics%time rather than the local 'time' variable -- latter contains the time at the start of the glint timestep, which can be up to a year behind M source_glimmer-cism/glint_timestep.F90 - also add diagnostic output *** Fix GetValueDouble to read in value as double precision M source_glimmer-cism/glimmer_config.F90 *** Unrelated change: add a comment M bld/namelist_files/namelist_defaults_cism.xml Summary of testing: - bluefire.glc.auxtest: main test results: All PASS except: (Not listing failures in NLComp [which I don't think is working right] or tputcomp) FAIL SMS_D.f09_g16.TG.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL PEA_P1.f09_g16.TG20TR.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 BFAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL ERI_Ly44.f09_g16.TGRCP85.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL NCK_P2.f09_g16.TG.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL CME_Ly10_N2_P2_D.f09_g16.TG1850.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 BFAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 FAIL CME_Ly5_PL.T31_g37.IG.bluefire_ibm.GC.123515.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 These are all failures in history comparisons, which are expected. Note that the previously-failing ERS tests are now passing. - bluefire.glc.auxtest: results of component_gen_comp -model cism: All PASS except: FAIL CME_Ly10_N2_P2_D.f09_g16.TG1850.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL CME_Ly5_PL.T31_g37.IG.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL ERI_Ly44.f09_g16.TGRCP85.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL NCK_P2.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL PEA_P1.f09_g16.TG20TR.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism FAIL SMS_D.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002.cism These are all expected failures, except I did not expect failures in the two gl10 comparisons. I am investigating this further. - Other tests: Note: these two tests below used a user_nl_dir, which set cism's dt=0.05: ERS_D_Ly20.f09_g16_gl10.TG1850.bluefire_ibm ERI_D_Ly44.f09_g16_gl10.TGRCP85.bluefire_ibm PASS ERS_D_Ly200.f09_g16_gl10.TG1850.bluefire_ibm.134929 PASS ERS_Ly100.f09_g16.TG.bluefire_ibm.204752 PASS ERS_D_Ly20.f09_g16_gl10.TG1850.bluefire_ibm.204752 PASS ERI_D_Ly44.f09_g16_gl10.TGRCP85.bluefire_ibm.204752 PASS ERI_D_Ly220.f09_g16_gl10.TGRCP85.bluefire_ibm.204752 Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_121008 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120921 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_02 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120918 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_1_42/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/new_glacier_data2_tags/new_glacier_data2_06_clm4_0_52/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_08 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120927 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/change_time_to_double models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120927 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120925b models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_121001 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120924 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120925 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120921 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.2 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_4/pio cism tag used for baseline comparisons: cism1_121002 Any other externals that differed in baseline: - scripts4_120930 - dlnd8_120918 ================================================================================ Originator: sacks Date: Oct 8, 2012 Model: cism Version: cism1_121008 One-line summary: add some namelist items, remove others Purpose of changes: Add some namelist items to namelist_definition_cism that users may want to change. Remove some items that should not be changed. Requires some new functionality in build-namelist to only create a section in cism.config under certain conditions. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml Summary of testing: - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - shortlist.glc.auxtest on bluefire: main test results: All PASS except: FAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.C.165451 BFAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.C.165451.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 These are expected failures based on cism1_121002 test results - shortlist.glc.auxtest: results of component_gen_comp -model cism: All PASS - Other tests: - Confirmed that build-namelist gives error when trying to define variables in gthf section, without setting do_gthf to .true.; same for isostasy section; same for elastic lithosphere section, but this keys off of lithosphere==1 - Confirmed that GTHF section isn't output when do_gthf is not set to .true.; same for isostasy section; same for elastic lithosphere section, but this keys off of lithosphere==1 - With empty user_nl_cism, gives identical cism.config to that in cism1_121002. - Added all new parameters to user_nl_cism, with values different from their defaults (along with do_gthf=.true., do_isostasy=.true.) - Checked cism.config: made sure each appears in the right section - Ran model, checked output to make sure they are all read in correctly Note: this job died with the error: * FATAL ERROR : (/glade/proj3/cseg/people/sacks/cesm_code/cesm1_1_alpha18f_newCISM/models/glc/cism/source_glimmer-cism/glide_nc_custom.F90:1 29) NetCDF: Start+count exceeds dimension bound I have filed a bug report about this (1560); I am not going to worry about this for now. Externals used for testing: NOTE: perl5lib updated to include my new get_defined_vars_in_group function -- similar, but not identical, to the perl5lib tag I made 10-8-2012 scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_120930 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120921 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_02 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120918 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_1_42/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/new_glacier_data2_tags/new_glacier_data2_06_clm4_0_52/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_08 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120927 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120927 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120925b models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_120918 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120924 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120925 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120921 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.2 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_4/pio cism tag used for baseline comparisons: cism1_121002 Any other externals that differed in baseline: None (except unchanged perl5lib) ================================================================================ Originator: sacks Date: Oct 2, 2012 Model: cism Version: cism1_121002 One-line summary: add some namelist variables, fix others Purpose of changes: Change idiag and jdiag into config parameters rather than hard-coded, following what is done in the seacism branch. Add some optional config options that were missing from namelist_definition_cism.xml (found by searching for GetValue calls in cism). Remove niso (unused). Make ntem & nvel integers Make ndiag 20 for gland5UM Change basal_tract from real(10) to real(5) Fix categories in namelist_definition_cism, for documentation purposes. Changes answers relative to previous tag: NO Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Remove hard-coded idiag, jdiag; instead use values set in config file M source_glimmer-cism/glimmer_paramets.F90 M source_glimmer-cism/glide_setup.F90 M source_glimmer-cism/glide_types.F90 M source_glimmer-cism/glint_main.F90 *** Changes to namelist variables as documented above M bld/build-namelist M bld/namelist_files/namelist_defaults_cism.xml M bld/namelist_files/namelist_definition_cism.xml *** Fix documentation of categories M bld/README.build-namelist Summary of testing: - bluefire.glc.auxtest: main test results: All PASS except: These were all expected failures based on previous tag: FAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.GC.233417 BFAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.GC.233417.generate.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 BFAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.GC.233417.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 FAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.GC.233417 BFAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.GC.233417.generate.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_121002 BFAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.GC.233417.compare_hist.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 tputcomp failures: FAIL SMS_D.f09_g16.TG.bluefire_ibm.GC.233417.tputcomp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 COMMENT tput_decr = 71.718 tput_percent_decr = 21.9 FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.bluefire_ibm.GC.233417.tputcomp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 COMMENT tput_decr = 7477.653 tput_percent_decr = 55.8 FAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.GC.233417.tputcomp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 COMMENT tput_decr = 983.18499 tput_percent_decr = 13.1 NLComp failures; as far as I can tell, this is an error in the test script: FAIL SMS_D.f09_g16.TG.bluefire_ibm.GC.233417.NLComp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 FAIL ERS_Ly20_N2_P2.f09_g16_gl10.TG.bluefire_ibm.GC.233417.NLComp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 FAIL ERS_Ly20_E.f09_g16_gl10.TG20TR.bluefire_ibm.GC.233417.NLComp.cesm1_1_alpha18f_scripts4_120930_clm_newGlacierData2_cism1_120930 - bluefire.glc.auxtest: results of component_gen_comp -model cism: All PASS - Other tests: - added newly-added config options to user_nl_cism, confirmed that they appeared in correct place in cism.config, and that at least some of them were read properly by cism - confirmed that diagnostic output is done at the desired coordinates, based on new idiag & jdiag Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_120930 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120921 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_02 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120918 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_1_42/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/new_glacier_data2_tags/new_glacier_data2_06_clm4_0_52/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_08 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120927 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120927 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120925b models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_120918 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120924 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120925 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120921 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.2 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_4/pio cism tag used for baseline comparisons: cism1_121001 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: Oct 1, 2012 Model: cism Version: cism1_121001 One-line summary: change parameter defaults Purpose of changes: Change some parameters to give better results, based on suggestions from Jeremy Fyke and Bill Lipscomb: - sigma_levels: focus resolution at the bottom - flow_law = 0 - ntem=nvel=niso=1, because we realized these are multipliers, and values < 1 don't make sense Changes answers relative to previous tag: YES! Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Change parameters M bld/namelist_files/namelist_defaults_cism.xml *** Document that ntemp, nvel & niso are multipliers M bld/namelist_files/namelist_definition_cism.xml Summary of testing: Note: baseline comparisons not done: I know this changes answers substantially - bluefire.glc.auxtest: main test results: All pass except the following: FAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.G.140709 BFAIL ERS_Ly20.f09_g16.TG1850.bluefire_ibm.G.140709.generate.cesm1_1_alpha18f_scripts4_120930_cism1_120930 FAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.G.140709 BFAIL ERS_Ly20_E.f09_g16.TG20TR.bluefire_ibm.G.140709.generate.cesm1_1_alpha18f_scripts4_120930_cism1_120930 These failures reveal an old problem which has just come to the surface: under certain circumstatnces, cism isn't restarting exactly. The problem is in the temp field, and this only affects the cpl fields once we set flow_law=0. With flow_law=1 (the old default), there is still a problem, but it is only apparent in the cism history files, so doesn't trigger a failure in the ERS test. I confirmed that the same problem existed in cism1_110418, if you set parameters appropriately (in particular: gland5UM with evolution=0, dt=0.05). - bluefire.glc.auxtest: results of component_gen_comp -model cism: (Just did -generate, not -compare; all PASS) - Other tests: Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_120930 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120921 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_2_02 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120918 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/trunk_tags/cam5_1_42/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/new_glacier_data2_tags/new_glacier_data2_06_clm4_0_52/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_08 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120927 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/trunk models/ocn/pop2 https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120927 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120925b models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_120918 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120924 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120925 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120921 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.2 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_4/pio cism tag used for baseline comparisons: N/A Any other externals that differed in baseline: N/A ================================================================================ Originator: mvertens, tcraig Date: Sept 21, 2012 Model: cism Version: cism1_120921 One-line summary: add esmf interface Purpose of changes: Add esmf interface Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/cism.buildexe.csh A drivers/cpl_esmf/glc_comp_mct.F90 A drivers/cpl_esmf/glc_comp_esmf.F90 Summary of testing: (testing performed by Bill Sacks) *** Full test suite NOT run: Since the changes were isolated to the *** esmf interface, the full test suite seemed unnecessary - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: PASS CME10y.f09_g16.TG1850.bluefire_ibm PASS CME10y_N2_P2_D.f09_g16.TG1850.bluefire_ibm PASS CME10y_N2_P2.f09_g16.TG1850.bluefire_ibm PASS ERS20y_E.f09_g16.TG20TR.bluefire_ibm PASS ERS20y_E.f09_g16.TG20TR.bluefire_ibm.memleak PASS CME5y_PL.T31_g37.IG.bluefire_ibm PASS SMS_D.f09_g16.BGRCP45CN.bluefire_ibm PASS SMS_D.f09_g16.BGRCP45CN.bluefire_ibm.memleak PASS SMS_DE.f09_g16.BGRCP45CN.bluefire_ibm PASS SMS_DE.f09_g16.BGRCP45CN.bluefire_ibm.memleak PASS NCK_E.T31_g37.BG1850CN.bluefire_ibm Ran shortlist.glc.auxtest, with comparison to cesm1_1_alpha18a: PASS SMS_D.f09_g16.TG.bluefire_ibm PASS SMS_D.f09_g16.TG.bluefire_ibm.memleak PASS SMS_D.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha18a PASS SMS_D.f09_g16.TG.bluefire_ibm.memcomp.cesm1_1_alpha18a FAIL SMS_D.f09_g16.TG.bluefire_ibm.tputcomp.cesm1_1_alpha18a COMMENT tput_decr = 79.75 tput_percent_decr = 13.8 PASS ERS20y.f09_g16.TG1850.bluefire_ibm PASS ERS20y.f09_g16.TG1850.bluefire_ibm.memleak PASS ERS20y.f09_g16.TG1850.bluefire_ibm.compare_hist.cesm1_1_alpha18a PASS ERS20y.f09_g16.TG1850.bluefire_ibm.memcomp.cesm1_1_alpha18a PASS ERS20y.f09_g16.TG1850.bluefire_ibm.tputcomp.cesm1_1_alpha18a PASS SMS_D.T31_g37.IG.bluefire_ibm PASS SMS_D.T31_g37.IG.bluefire_ibm.memleak FAIL SMS_D.T31_g37.IG.bluefire_ibm.compare_hist.cesm1_1_alpha18a PASS SMS_D.T31_g37.IG.bluefire_ibm.memcomp.cesm1_1_alpha18a FAIL SMS_D.T31_g37.IG.bluefire_ibm.tputcomp.cesm1_1_alpha18a COMMENT tput_decr = 2.949 tput_percent_decr = 12.1 The failure in compare_hist with the IG case seems just due to runoff changes: [be1105en:/ptmp/sacks/tests/SMS_D.T31_g37.IG.bluefire_ibm.C.111836]$ grep RMS cprnc.out RMS r2x_Forr 1.3464E-06 [be1105en:/ptmp/sacks/tests/SMS_D.T31_g37.IG.bluefire_ibm.C.111836]$ grep FILLDIFF cprnc.out FILLDIFF domr_lat FILLDIFF domr_lon FILLDIFF domr_are FILLDIFF domr_are FILLDIFF domr_mas FILLDIFF domr_fra FILLDIFF r2x_Forr FILLDIFF r2x_Forr So I'm not worried about this Related component_gen_comp: PASS ERS20y.f09_g16.TG1850.bluefire_ibm.compare_hist.cesm1_1_alpha18a.cism PASS SMS_D.T31_g37.IG.bluefire_ibm.compare_hist.cesm1_1_alpha18a.cism PASS SMS_D.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha18a.cism Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/branch_tags/rtmcomp_tags/rtmcomp07_scripts4_120828 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/branch_tags/rtmcomp_tags/rtmcomp03_Machines_120829 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/branch_tags/rtmcomp_tags/rtmcomp15_drvseq4_1_26 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/branch_tags/rtmcomp_tags/rtmcomp02_share3_120803 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/exe_cam5_1_33_tags/exe01_cam5_1_33/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/rtmcomp_tags/rtmcomp14_clm4_0_46/models/lnd/clm models/rof/rtm https://svn-ccsm-models.cgd.ucar.edu/rivrtm/trunk_tags/rtm1_0_04 models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120825 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/rtmcomp models/ocn/pop2/ https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120824 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120626 models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/branch_tags/rtmcomp_tags/rtmcomp02_dlnd8_120626 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120626 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120626 models/rof/drof https://svn-ccsm-models.cgd.ucar.edu/drof/trunk_tags/drof8_120827 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xlnd models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xglc models/rof/xrof https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/xrof models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_7_04/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/slnd models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/sglc models/rof/srof https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_4_01/srof models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.0 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_3/pio cism tag used for baseline comparisons: cism1_120905 Any other externals that differed in baseline: Used cesm1_1_alpha18a for baselines; this involves many differences in externals ================================================================================ Originator: sacks Date: Sept 5, 2012 Model: cism Version: cism1_120905 One-line summary: change basal_tract to real rather than integer Purpose of changes: basal_tract was mistakenly listed as an integer; I am fixing this. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_definition_cism.xml Summary of testing: No system tests; just tested namelist generation - bluefire.glc.auxtest: main test results: NOT DONE - bluefire.glc.auxtest: results of component_gen_comp -model cism: NOT DONE - Other tests: (1) Ran 'create_newcase -compset TG -res f09_g16 -mach bluefire', then 'setup'. Compared default cism.config with baseline version: identical (2) In that case, did the following: echo 'basal_tract = 2.5 10 10 0 1' >> user_nl_cism preview_namelists Confirmed that it worked to use a real value. Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_120828 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120829 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_1_26 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120803 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/exe_cam5_1_33_tags/exe01_cam5_1_33/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_0_46/models/lnd/clm models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120825 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/ensemble models/ocn/pop2/ https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120824 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120626 models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_120626 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120626 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120626 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xlnd models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/slnd models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.0 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_3/pio cism tag used for baseline comparisons: cism1_120829 Any other externals that differed in baseline: None ================================================================================ Originator: sacks Date: Aug 29, 2012 Model: cism Version: cism1_120829 One-line summary: enable ensemble capabilities Purpose of changes: Enable ensemble capabilities for cism. These were already implemented in drv & scripts, but required some changes in the cism code. Also, change flow_factor to real rather than integer. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** New modules to handle variables that differ between ensemble members A source_glc/glc_ensemble.F90 A source_glc/glc_files.F90 A test/unit/unit_test_replacements/glc_files.F90 - simplified version needed because the time management unit test doesn't know anything about ensemble number *** Add ensemble capability M drivers/cpl_mct/glc_comp_mct.F90 M source_glc/glc_io.F90 - Also deleted unused glc_io_create_suffix_cesm, because I think it would have been broken with ensemble capabilities M source_glc/glc_constants.F90 - Move file names that are no longer constant *** Change use statements to reflect movement of some file name variables M source_glc/glc_InitMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_time_management.F90 *** Add use of glc_files.F90 M test/unit/time_management/Srcfiles *** Change flow_factor to real rather than integer M bld/namelist_files/namelist_definition_cism.xml *** Updated changeLog template to give more testing details M ChangeLog_template Summary of testing: - bluefire.glc.auxtest: main test results: All PASS except the following (other than throughput tests, these are all expected failures): FAIL ERS20y.f09_g16.TG1850.bluefire_ibm.tputcomp.cesm1_1_alpha17c COMMENT tput_decr = 1694.229 tput_percent_decr = 44.5 BFAIL ERS20y_N2_P2.f09_g16.TGG10.bluefire_ibm.compare_hist.cesm1_1_alpha17c BFAIL NCK_P2.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha17c FAIL SMS_PL.T31_g37.IGLONG.bluefire_ibm.tputcomp.cesm1_1_alpha17c COMMENT tput_decr = 3.6440000 tput_percent_decr = 2.08 FAIL SMS.f19_f19.FG20TRCN.bluefire_ibm.tputcomp.cesm1_1_alpha17c COMMENT tput_decr = 1.627 tput_percent_decr = 13.5 (Note: test list was that in scripts4_120828, plus ERS20y.f09_g16.TGG10.bluefire_ibm [which was included in the baseline test suite, but has now been replaced by a _N2 version of this test]) - bluefire.glc.auxtest: results of component_gen_comp -model cism: All PASS except the following (these are expected failures): BFAIL ERS20y_N2_P2.f09_g16.TGG10.bluefire_ibm.compare_hist.cesm1_1_alpha17c.cism (baseline directory does not exist) BFAIL NCK_P2.f09_g16.TG.bluefire_ibm.compare_hist.cesm1_1_alpha17c.cism (baseline directory does not exist) - Other tests: - 4-instance TG run with differences in flow_factor: startup with one continue run; compared instances #1 and #3 with analogous single-instance runs (comparisons included cism hist, cism rest & cpl hist files) - as above, but 16 PE run with glc on PEs 4-7: compared with above run to make sure results don't depend on PE location - 4-instance IGLONG test with differences in the albice CLM parameter; compared instance #3 with analogous single-instance run (comparisons included clm hist, cism hist & cpl hist files) Externals used for testing: scripts https://svn-ccsm-models.cgd.ucar.edu/scripts/trunk_tags/scripts4_120828 scripts/ccsm_utils/Machines https://svn-ccsm-models.cgd.ucar.edu/Machines/trunk_tags/Machines_120826 mapping https://svn-ccsm-models.cgd.ucar.edu/tools/mapping/trunk_tags/mapping_120816 models/drv https://svn-ccsm-models.cgd.ucar.edu/drv/seq_mct/trunk_tags/drvseq4_1_26 models/csm_share https://svn-ccsm-models.cgd.ucar.edu/csm_share/trunk_tags/share3_120803 models/atm/cam https://svn-ccsm-models.cgd.ucar.edu/cam1/branch_tags/exe_cam5_1_33_tags/exe01_cam5_1_33/models/atm/cam models/atm/wrf https://svn-ccsm-models.cgd.ucar.edu/wrf/trunk_tags/wrf32_ccsm120726 models/lnd/clm https://svn-ccsm-models.cgd.ucar.edu/clm2/trunk_tags/clm4_0_46/models/lnd/clm models/ice/cice https://svn-ccsm-models.cgd.ucar.edu/cice/trunk_tags/cice4_0_20120825 models/glc/cism https://svn-ccsm-models.cgd.ucar.edu/glc/branches/ensemble models/ocn/pop2/ https://svn-ccsm-models.cgd.ucar.edu/pop2/trunk_tags/cesm_pop_2_1_20120824 models/atm/datm https://svn-ccsm-models.cgd.ucar.edu/datm7/trunk_tags/datm8_120626 models/lnd/dlnd https://svn-ccsm-models.cgd.ucar.edu/dlnd7/trunk_tags/dlnd8_120626 models/ice/dice https://svn-ccsm-models.cgd.ucar.edu/dice7/trunk_tags/dice8_120626 models/ocn/docn https://svn-ccsm-models.cgd.ucar.edu/docn7/trunk_tags/docn8_120626 models/atm/xatm https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xatm models/lnd/xlnd https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xlnd models/ice/xice https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xice models/ocn/xocn https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xocn models/glc/xglc https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/xglc models/dead_share https://svn-ccsm-models.cgd.ucar.edu/dead7/trunk_tags/dead7_6_05/dead_share models/atm/satm https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/satm models/lnd/slnd https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/slnd models/ice/sice https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/sice models/ocn/socn https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/socn models/glc/sglc https://svn-ccsm-models.cgd.ucar.edu/stubs/trunk_tags/stubs1_3_05/sglc models/utils/timing https://svn-ccsm-models.cgd.ucar.edu/timing/trunk_tags/timing_120731 models/utils/esmf_wrf_timemgr https://svn-ccsm-models.cgd.ucar.edu/esmf_wrf_timemgr/trunk_tags/esmf_wrf_timemgr_120427 models/utils/mct https://github.com/MCSclimate/MCT/tags/MCT_2.8.0 models/utils/pio http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_3/pio cism tag used for baseline comparisons: cism1_120823 Any other externals that differed in baseline: - scripts4_120727d: only diff is in bluefire.glc.auxtest list ================================================================================ Originator: sacks Date: Aug 23, 2012 Model: cism Version: cism1_120823 One-line summary: Turn evolution on for gland5UM Purpose of changes: Turn evolution on for gland5UM (evolution=0 rather than -1). Also shorten time steps so ice sheet evolution is stable for gland5UM: dt=0.05, ntem=nvel=niso=0.1. These parameters came from Jeremy Fyke's most recent runs. The values of these (and other) parameters may still change, but I wanted to use something for now that gives ice sheet evolution with gland5UM to facilitate testing. Changes answers relative to previous tag: Yes, for gland5UM Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/namelist_files/namelist_defaults_cism.xml Summary of testing: No system tests! Just created three cases and compared cism.config with previously-generated cases: - gland5UM: diffs as expected - gland5: no diffs - gland10: no diffs ================================================================================ Originator: mvertens Date: June 29, 2012 Model: cism Version: cism1_120629 One-line summary: Change DIN_LOC_ROOT_CSMDATA to DIN_LOC_ROOT Purpose of changes: Change DIN_LOC_ROOT_CSMDATA to DIN_LOC_ROOT Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** simple, one-line change M bld/build-namelist Summary of testing: NONE! ================================================================================ Originator: mvertens Date: May 29, 2012 Model: cism Version: cism1_120529 One-line summary: Modifications to build to fit with latest scripts Purpose of changes: Minor changes to fit in with other system changes Mariana is bringing in Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/cism.cpl7.template Summary of testing: NONE! ================================================================================ Originator: mvertens Date: May 21, 2012 Model: cism Version: cism1_120521 One-line summary: Support removal of env_conf.xml Purpose of changes: Changes to build-namelist to support removal of env_conf.xml, and related cleanup Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: M bld/build-namelist Summary of testing: (All tests except the ERI IGCN test were conducted by Bill S) Main testing: Unless otherwise noted, the following tests were done with a 10-year TG compset, f09_g16, GLC_GRID=gland20. I compared (1) cism_in input file, (2) cism.config input file, (3) year 10 cism history output (using cprnc). The comparison cases were generated from cesm1_1_beta14. - startup run - startup run at f19_g16 - startup run with gland5UM - branch run; 1 year (continuing the above startup run) Other tests: - ERI.T31_g37.IGCN.bluefire_ibm - comparison with cesm1_1_alpha14a - Checked dt_option & dt_count in cism in for the following: - NCPL_BASE_PERIOD=decade, GLC_NCPL=5 (dt_option=steps_per_year, dt_count=0.5) ================================================================================ Originator: mvertens, sacks Date: May 15, 2012 Model: cism Version: cism1_120512 One-line summary: New build-namelist functionality Purpose of changes: - New build-namelist functionality to make cism consistent with other CESM components. Makes cism.buildnml.csh just a thin wrapper to build-namelist. Namelist defaults now set in an xml file. - Set dt_option and dt_count automatically to match the coupling interval. - Multi-instance capability in the build scripts (create multiple cism_in & cism.config files). Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Major rework of build D bld/configure D bld/config_definition.xsl D bld/namelist_defaults_overall.xml D bld/namelist_definition_overall.xml D bld/namelist_defaults.xsl D bld/namelist_definition.xsl D bld/config_definition.xml D bld/namelist_defaults_cism.xml D bld/namelist_definition_cism.xml A bld/user_nl_cism A bld/cismIO/cism.buildIO.template.csh A bld/README.build-namelist A bld/cism.buildexe.csh M bld/build-namelist M bld/cism.cpl7.template A bld/cism.buildnml.csh M bld/README A bld/namelist_files/namelist_defaults_cism.xml A bld/namelist_files/namelist_definition_cism.xml A bld/namelist_files *** Remove now-unnecessary input templates (replaced by *** namelist_files/namelist_defaults_cism.xml) D input_templates *** Add unit test for 2-year time step M test/unit/time_management/restart_testlist A test/unit/time_management/inputs/twoyear/time_management_test_in A test/unit/time_management/inputs/twoyear/cism_in A test/unit/time_management/inputs/twoyear M test/unit/time_management/inputs/README A test/unit/time_management/inputs/twoyear.restart1/time_management_test_in A test/unit/time_management/inputs/twoyear.restart1/cism_in A test/unit/time_management/inputs/twoyear.restart1 A test/unit/time_management/inputs/twoyear.restart2/time_management_test_in A test/unit/time_management/inputs/twoyear.restart2/cism_in A test/unit/time_management/inputs/twoyear.restart2 M test/unit/time_management/testlist *** Fix path M test/unit/unit_test_shr/Makefile.common *** Remove redundant README file D README *** Remove "Known bugs" section M ChangeLog_template Summary of testing: CESM Tests (implemented change off of cesm1_1_beta13, compared with baseline cesm1_1_beta13): - bluefire.glc.auxtest: All tests passed except: CFAIL ERS.T31_g37.IGRCP85CN.bluefire_ibm.C.215553 FAIL SMS.T31_T31.FG20TRCN.bluefire_ibm BFAIL SMS.T31_T31.FG20TRCN.bluefire_ibm.compare_hist.cesm1_1_beta13 FAIL SMS.T31_g37.TG.bluefire_ibm BFAIL SMS.T31_g37.TG.bluefire_ibm.compare_hist.cesm1_1_beta13 These were all expected failures (also failed in cesm1_1_beta13) - ERS.f19_g16.IGCN.bluefire_ibm: PASS - ran this because of failure in ERS.T31_g37.IGRCP85CN.bluefire_ibm Main testing of new build: Unless otherwise noted, the following tests were done with a 10-year TG compset, f09_g16, GLC_GRID=gland20. I compared (1) cism_in input file, (2) cism.config input file, (3) year 10 cism history output (using cprnc). The comparison cases were generated from cesm1_1_beta13. - startup run - Also made sure that list of files in run directory is the same as before - startup run at f19_g16 - startup run with gland5UM - startup run with gland5 - just checked cism_in and cism.config -- didn't actually do the run - startup run with gland10 - just checked cism_in and cism.config -- didn't actually do the run - hybrid run; 1 year (continuing the above startup run) - branch run; 1 year (continuing the above startup run) - continue run; 1 year -- continuing the above startup run - continue run; 1 year -- continuing the above branch run - IG: 1-year test at f19_g16, continuing an older run that had run for 10 years (hybrid) - main point of this was to make sure 1-day coupling still works right Testing automatic setting of dt_option and dt_count, and making sure that it works to use values more general than what we have been using before: - New time manager unit tests, to make sure it works to have fractional dt_count with steps_per_year (as we could get with NCPL_BASE_PERIOD='decade'): - twoyear - looked through output manually (briefly) - compared with output from using dt_option='seconds', dt_count=63072000 - restart_test twoyear - Checked dt_option & dt_count in cism_in for the following: - out-of-the-box TG - out-of-the-box IG - various values of NCPL_BASE_PERIOD & GLC_NCPL - values overridden by user_nl_cism - TG run with NCPL_BASE_PERIOD='year', GLC_NCPL=2 (i.e., two time steps per year): confirmed that there is one cism time step per coupling interval - TG run with NCPL_BASE_PERIOD='decade', GLC_NCPL=5 (i.e., 2-year time step), with mass balance time step = 2 years (requires source mod in glint_mbal.F90), ice dynamics time step = 2 years (dt=2.): confirmed that there is one cism time step per coupling interval - Compared log files in two 1-year TG runs, which have the same time step but specified in different ways: one with NCPL_BASE_PERIOD=day, GLC_NCPL=48 (results in dt_option=steps_per_day, dt_count=48); the other with NCPL_BASE_PERIOD=hour, GLC_NCPL=2 (results in dt_option=seconds, dt_count=1800): glc log files essentially identical (with verbose=.true.) - Note that these runs both died just before the end, with "Unexpected calling of GLINT at time 8759", but I don't think that's a problem that I caused (but it meant I could not compare hist files) Other tests: - modifications via user_nl_cism: just checked the generated cism_in and cism.config files to make sure the modifications are picked up - changed a cism_in variable: cism_debug - changed a cism.config variable: basal_tract - multi-instance test: 4 instances, with no modifications in user_nl_cism_000*: just checked the generated cism_in and cism.config files: paramfile set correctly in cism_in_000*; everything else matches a single-instance case - multi-instance test: 4 instances, with modifications to ihour0 and evolution in user_nl_cism_0001 & 0003 (but no changes in 0002 or 0004): just checked the generated cism_in and cism.config files - cism.buildIO.csh: - confirmed that the generated script is essentially the same as before - added a variable in glide_vars.def, confirmed that all F90 files created by cism.buildIO.csh are identical to those from an old case with the same change in glide_vars.def ================================================================================ Originator: sacks Date: March 22, 2012 Model: cism Version: cism1_120322 One-line summary: fixed memory leak Purpose of changes: Fixed memory leak in glint_downscaling_gcm (~ 40 MB/year for some test runs) Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 1433 (time manager needs additional changes to support longer timesteps) 1441 (time manager doesn't restart properly when last run stopped in the middle of a day) List all modified files, and describe the changes: *** Fix memory leak M https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_120123/source_glimmer-cism/glint_climate.F90 *** Fix comment (unrelated change) M https://svn-ccsm-models.cgd.ucar.edu/glc/trunk_tags/cism1_120123/test/unit/unit_test_shr/Makefile.common Summary of testing: - CESM tests: bluefire.glc.auxtest, with comparison to previous version (implemented change in cesm1_1_beta11, compared with baseline cesm1_1_beta11). All tests passed except: RUN ERS.T31_g37.IGRCP85CN.bluefire_ibm.C.095358 FAIL SMS.T31_T31.FG20TRCN.bluefire_ibm BFAIL SMS.T31_T31.FG20TRCN.bluefire_ibm.compare_hist.cesm1_1_beta11 FAIL SMS.T31_g37.TG.bluefire_ibm BFAIL SMS.T31_g37.TG.bluefire_ibm.compare_hist.cesm1_1_beta11 All of these tests also failed in cesm1_1_beta11 - 10-year TG compset using Bill L's recent instructions: f09_g16, gland5UM, annual coupling interval using Gail's annual-average cpl forcings; evolution=0, dt=ntem=nvel=niso=0.1. Last year of cism hist is identical to code without this fix. NOTE: This change was done off of an earlier version of cism (cism1_111214), but the change is so straightforward that I'm pretty confident that the results should translate to the latest version. ================================================================================ Originator: tcraig Date: Jan 23, 2012 Model: cism Version: cism1_120123 One-line summary: update esmf to 520r M source_glc/glc_io.F90 M drivers/cpl_mct/glc_comp_mct.F90 ================================================================================ Originator: sacks Date: Dec 21, 2011 Model: cism Version: cism1_111221 One-line summary: Restore "correct" glc_time_management, with fix for compiler bug Purpose of changes: Return glc_time_management to its "correct" state - i.e., before the changes made in cism1_111220a; introduce unused module variable in glc_time_management to get around an internal compiler error in xlf. Also, move incorrectly-placed 'implicit none' in glc_time_management_test.F90. Changes answers relative to previous tag: No, but replaces functionality removed in cism1_111220a Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 1433 (time manager needs additional changes to support longer timesteps) 1441 (time manager doesn't restart properly when last run stopped in the middle of a day) List all modified files, and describe the changes: *** Restore to version prior to cism1_111220a, introduce unused module variable to get around internal compiler error M source_glc/glc_time_management.F90 *** Move incorrectly-placed 'implicit none' M test/unit/time_management/glc_time_management_test.F90 Summary of testing: - Tested build for an IGCN case - Full cesm system tests pending, including bluefire.glc.auxtest with comparison to beta07 (needed to create this tag before testing was complete, but I will create a new tag if any fail) ================================================================================ Originator: sacks Date: Dec 20, 2011 Model: cism Version: cism1_111220a One-line summary: Rolled back glc_time_management Purpose of changes: The cesm build was failing with the new version of glc_time_management. I have temporarily rolled this file back to the version in 31393. This should be fixed ASAP! Changes answers relative to previous tag: No, but removes functionality Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): List all modified files, and describe the changes: *** Rolled back M source_glc/glc_time_management.F90 Summary of testing: Just tested build ================================================================================ Originator: sacks Date: Dec 20, 2011 Model: cism Version: cism1_111220 One-line summary: Updated ChangeLog entry for cism1_111214 Purpose of changes: I realized that I hadn't documented the testing done for cism1_111214 entirely correctly; I have now fixed this. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 1433 (time manager needs additional changes to support longer timesteps) 1441 (time manager doesn't restart properly when last run stopped in the middle of a day) List all modified files, and describe the changes: *** Update ChangeLog entry for cism1_111214 M ChangeLog Summary of testing: None ================================================================================ Originator: mvertens Date: Dec 18, 2011 Model: cism Version: cism1_111218 One-line summary: Simplified way that cism field names generated for driver Purpose of changes: Implemented new scheme of determine cism field names that is more flexible and robust Changes answers relative to previous tag: No Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): None List all modified files, and describe the changes: M source_glc/glc_constants.F90 - no longer set glc_nec to 10 - but determine it from fields set in coupler - consistency check is also made in the land - more robust implemention M drivers/cpl_share/glc_cpl_indices.F90 - new index array for multiple elevation classes now is used instead of hard wired index names for each class - much more extensible scheme M drivers/cpl_mct/glc_comp_mct.F90 - using changes in glc_cpl_indices - glc_export_mct and glc_import_mct are greatly simplified (no more hard-wiring) Summary of testing: Ran IGCN and BGCN system tests and verified that results were bit-for-bit ================================================================================ Originator: sacks Date: Dec 14, 2011 Model: cism Version: cism1_111214 One-line summary: Add test driver for glc_time_management Purpose of changes: Add a test driver for glc_time_management, along with a list of standard tests to run. Remove some variables from glc_time_management that aren't being used and weren't properly preserved in restarts. Add code & makefile that can be used for future unit testers. Add ChangeLog_template. Changes answers relative to previous tag: No Bugs fixed (include bugzilla ID) (http://bugs.cgd.ucar.edu/): Known bugs (include bugzilla ID) (http://bugs.cgd.ucar.edu/): 1433 (time manager needs additional changes to support longer timesteps) 1441 (time manager doesn't restart properly when last run stopped in the middle of a day) List all modified files, and describe the changes: *** Remove unused variables that aren't preserved in restarts M source_glc/glc_time_management.F90 *** Add code & makefile that can be used for any cism unit tester A test A test/unit A test/unit/unit_test_replacements A test/unit/unit_test_replacements/glc_communicate.F90 A test/unit/unit_test_replacements/shr_sys_mod.F90 A test/unit/unit_test_replacements/README A test/unit/unit_test_shr/writevar_mod.F90 A test/unit/unit_test_shr/Makefile.common A test/unit/unit_test_shr/README A test/unit/unit_test_shr *** Add test driver for the glc_time_management module A test/unit/time_management A test/unit/time_management/README A test/unit/time_management/Filepath A test/unit/time_management/Srcfiles A test/unit/time_management/Makefile A test/unit/time_management/glc_time_management_test_mod.F90 A test/unit/time_management/glc_time_management_test.F90 *** Script to run exact restart test using the time management test driver A test/unit/time_management/restart_test *** Add list of tests to be run with the time management test driver A test/unit/time_management/restart_testlist A test/unit/time_management/testlist *** Namelists for each of the standard tests to run using the time management test driver and the restart_test script (note: I am not listing all of the individual files added in this directory) A test/unit/time_management/inputs *** Add ChangeLog_template to be used for new ChangeLog entries A ChangeLog_template Summary of testing: No system tests, but many unit tests - test/unit/time_management (tests in testlist & restart_testlist; for testlist, performed regression tests against code similar to cism1_111026 where possible, manually checked output where regression tests were impossible - e.g., where the old code was buggy): Note that there is a known bug in newhour (see bug 1433). I am considering tests to have passed even if they have the wrong newhour value. Also, since I didn't do a careful look at every variable output by the test driver, it's possible that there are errors in other variables that I didn't catch. Here, I am just listing tests that have failures in variables that I checked, other than newhour (variables I checked more carefully were: iyear, imonth, iday, seconds_this_day, eoy, eom, eod, adjust_nyears, elapsed_days, elapsed_months, elapsed_years). All pass except: oneyear_leap: model doesn't run to completion (see bug 1433) threeyear_leap: also has problems related to bug 1433, though here the model runs to completion 219day_leap: also has problems related to bug 1433; the first error I notice is that year 5, day_of_year 73 is translated to 0005-mar-13 rather than 0005-mar-14, but there are likely other errors, too, as for oneyear_leap 511day_leap: also has problems relatd to bug 1433; e.g., both the leap and the non-leap tests have a time step with time stamp 0005-mar-15; the leap should differ ================================================================================ Originator: sacks Date: Nov 22, 2011 Model: cism Version: cism1_111122 One-line summary: Modified time manager to support longer time steps Previously, the time manager did not correctly handle time steps longer than one month. With the changes here, time steps as long as desired should be possible. However, as noted below, I have only tested this for daily and annual time steps so far. M source_glc/glc_time_management.F90 pretag testing: so far I have tested this for daily and annual time steps; more testing is needed of other length time steps - 10-year TG compset with daily time step (i.e., the default), using gland10 and 10 years of forcing from a af19_g16 IG compset: last year of cism history output identical to a similar run set up using old version; cism log files also essentially identical - Similar to the above, but with annual time step & annual coupling interval. This was set up by (1) averaging the 10 years of cpl history files into annual-average files; (2) explicitly setting atm_cpl_dt to 31536000, and the same for other components, in cpl.buildnml.csh; (3) changing dt_option from steps_per_day to steps_per_year in cism.buildnml.csh. Last year of cism history output identical to the daily run; cism log files also show correct time stepping - exact restart, daily time step: similar to daily time step run above, but two 5-year runs; checked last year of history file against the above 10-year run - exact restart, annual time step: similar to annual time step run above, but two 5-year runs; checked last year of history file against the above 10-year run ================================================================================ Originator: sacks Date: Oct 26, 2011 Model: cism Version: cism1_111026 One-line summary: change ntem default to 1000000. for gland5UM This turns off ice temperature evolution (I think) -- at least, as long as you are running for less than 1000000 years. This change should have been made at the same time as the creation of ice.config.gland5UM, when evolution was (correctly) set to -1. M input_templates/ice.config.gland5UM pretag testing: NONE ================================================================================ Originator: jwolfe Date: Model: cism Version: cism1_111007 One-line summary: add support for new GLC_GRID gland5UM M bld/cism.cpl7.template A input_templates/ice.config.gland5UM ================================================================================ Originator: sacks, lipscomb Date: Oct 4, 2011 Model: cism Version: cism1_111004 One-line summary: allow evolution=-1; reformat glimmer_ncio.F90 to satisfy some compilers * Allow evolution=-1 -- i.e., no thickness evolution M source_glimmer-cism/glide.F90 M source_glimmer-cism/glide_setup.F90 * Reformat to satisfy some compilers M source_glimmer-cism/glimmer_ncio.F90 pretag testing: - 10-year TG compsets before and after these modifications (evolution=0 in both cases), using a single year of forcing from a f19_g16 IG compset: last year of cism history output identical - 10-year TG compset with evolution=-1, compared with the above evolution=0 case: last year of cism history output differs, as expected (but didn't evaluate whether the differences make sense) ================================================================================ Originator: jwolfe, lipscomb Date: Mon Apr 18, 2011 Model: cism Version: cism1_110418 One-line summary: changes to make restarts completely functional for branch and hybrid runs, using the glimmer-cism native hotstart capability M bld/cism.cpl7.template M input_templates/ice.config.gland10 M input_templates/ice.config.gland20 M input_templates/ice.config.gland5 ================================================================================ Originator: jwolfe, lipscomb Date: Mon Mar 07, 2011 Model: cism Version: cism1_110307 One-line summary: modify input templates to remove setting native glimmer-cism history output files, which interferes with creating CESM-style history output M input_templates/ice.config.gland10 M input_templates/ice.config.gland20 M input_templates/ice.config.gland5 pretag testing: Results are bfb relative to cesm1_0_beta15 for ERI.f19_f19.FG.bluefire ================================================================================ Originator: jwolfe, lipscomb Date: Sun Feb 20, 2011 Model: cism Version: cism1_110220 One-line summary: reworked the way CISM handles history files, to more closely follow CESM naming conventions and alarms M bld/cism.cpl7.template M source_glc/glc_InitMod.F90 M source_glc/glc_io.F90 M drivers/cpl_mct/glc_comp_mct.F90 pretag testing: Results are bfb relative to cesm1_0_beta15 for ERI.f19_f19.FG.bluefire ================================================================================ Originator: fischer Date: Tue Jan 25, 2011 Model: cism Version: cism1_110125 One-line summary: renamed glint_smb.F90 to smb_dummy.F90 A source_glimmer-cism/smb_dummy.F90 D source_glimmer-cism/glint_smb.F90 ================================================================================ Originator: mvertens Date: Mon Jan 24, 2011 Model: cism Version: cism1_110124 One-line summary: removed use seq_flds_indices, now using local glc_cpl_indices_set A drivers/cpl_share A drivers/cpl_share/glc_cpl_indices.F90 M drivers/cpl_mct/glc_comp_mct.F90 pretag testing: Results are bfb relative to cesm1_0_beta14 for ERI.f19_f19.FG.bluefire ================================================================================ Originator: jwolfe, lipscomb Date: Sep 13, 2010 Model: cism Version: cism1_100913 One-line summary: add more capabilities, clean-up * pull vars.def files together along with template and python script, to allow users to modify the IO list on a per case basis A bld/cismIO A bld/cismIO/glide_lithot_vars.def A bld/cismIO/README.cismIO A bld/cismIO/glint_mbal_vars.def A bld/cismIO/glint_vars.def A bld/cismIO/ncdf_template.F90.in A bld/cismIO/glide_vars.def A bld/cismIO/generate_ncvars.py M bld/cism.cpl7.template * make GLC_DEBUG a variable instead of an IFDEF M source_glimmer-cism/glide.F90 M source_glimmer-cism/glint_timestep.F90 M source_glimmer-cism/glint_climate.F90 M source_glimmer-cism/glimmer_paramets.F90 M source_glimmer-cism/glide_diagnostics.F90 M source_glimmer-cism/glint_main.F90 M source_glimmer-cism/glint_interp.F90 M source_glimmer-cism/glint_type.F90 * reworked restart to keep glc calls out of glimmer-cism M source_glc/glc_InitMod.F90 M source_glc/glc_io.F90 A source_glimmer-cism/glimmer_restart_gcm.F90 M source_glimmer-cism/glint_initialise.F90 M source_glimmer-cism/glimmer_config.F90 * fixed configuration files to make other resolutions work correctly M input_templates/ice.config.gland10 M input_templates/ice.config.gland20 M input_templates/ice.config.gland5 ================================================================================ Originator: jwolfe, lipscomb Date: Jun 17, 2010 Model: cism Version: cism1_100617a One-line summary: keep use statement for shr_file_mod (temporarily) M source_glimmer-cism/glimmer_config.F90 ================================================================================ Originator: jwolfe, lipscomb Date: Jun 17, 2010 Model: cism Version: cism1_100617 One-line summary: pick up mods to glc_comp from Bill L's sandbox M drivers/cpl_mct/glc_comp_mct.F90 ================================================================================ Originator: jwolfe, lipscomb Date: Jun 16, 2010 Model: cism Version: cism1_100616a One-line summary: change Filepath to refer to new drivers directory M bld/cism.cpl7.template ================================================================================ Originator: jwolfe, lipscomb Date: Jun 16, 2010 Model: cism Version: cism1_100616 One-line summary: rearrange driver code for future esmf support, clean-up * rearrange driver code for future esmf support D source_glc/glc_comp_mct.F90 A drivers A drivers/cpl_mct A drivers/cpl_mct/glc_comp_mct.F90 A drivers/cpl_esmf * remove unused code D source_glc/glc_domain_size.F90 D source_glimmer-cism/glimmer_restart_statarr.F90 D source_glimmer-cism/glimmer_restart_pointarr.F90 D source_glimmer-cism/glimmer_restart_common.F90 D source_glimmer-cism/glimmer_restart.F90 D source_glimmer-cism/glimmer_restart_statscal.F90 * fixed input template files for different resolutions M input_templates/ice.config.gland10 M input_templates/ice.config.gland20 M input_templates/ice.config.gland5 * enhanced I/O to fix restart issue M source_glimmer-cism/glint_io.F90 M source_glimmer-cism/glide_lithot_io.F90 M source_glimmer-cism/glint_mbal_io.F90 M source_glimmer-cism/glide_io.F90 * clean-up and fix documentation M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_io.F90 M source_glc/glc_constants.F90 M source_glc/glc_global_fields.F90 M source_glc/glc_FinalMod.F90 M source_glc/glc_time_management.F90 M source_glimmer-cism/glide.F90 M source_glimmer-cism/glimmer_scales.F90 M source_glimmer-cism/glint_timestep.F90 M source_glimmer-cism/glide_thck.F90 M source_glimmer-cism/isostasy_setup.F90 M source_glimmer-cism/isostasy_el.F90 M source_glimmer-cism/glide_nc_custom.F90 M source_glimmer-cism/isostasy_types.F90 M source_glimmer-cism/glimmer_ts.F90 M source_glimmer-cism/glimmer_routing.F90 M source_glimmer-cism/glimmer_searchcircle.F90 M source_glimmer-cism/glimmer_daily_pdd.F90 M source_glimmer-cism/xls.F90 M source_glimmer-cism/glint_mbal.F90 M source_glimmer-cism/glide_temp.F90 M source_glimmer-cism/glint_climate.F90 M source_glimmer-cism/glimmer_utils.F90 M source_glimmer-cism/glimmer_map_init.F90 M source_glimmer-cism/glimmer_anomcouple.F90 M source_glimmer-cism/glimmer_filenames.F90 M source_glimmer-cism/glimmer_ncparams.F90 M source_glimmer-cism/glimmer_writestats.F90 M source_glimmer-cism/glint_initialise.F90 M source_glimmer-cism/glimmer_paramets.F90 M source_glimmer-cism/glimmer_vers.F90 M source_glimmer-cism/glide_lithot1d.F90 M source_glimmer-cism/glimmer_map_CFproj.F90 M source_glimmer-cism/glide_mask.F90 M source_glimmer-cism/glide_profile.F90 M source_glimmer-cism/glide_setup.F90 M source_glimmer-cism/glimmer_map_proj4.F90 M source_glimmer-cism/glide_types.F90 M source_glimmer-cism/glide_velo.F90 M source_glimmer-cism/glimmer_global.F90 M source_glimmer-cism/glimmer_map_types.F90 M source_glimmer-cism/glimmer_deriv.F90 M source_glimmer-cism/glimmer_ncdf.F90 M source_glimmer-cism/glimmer_coordinates.F90 M source_glimmer-cism/kelvin.F90 M source_glimmer-cism/glide_stop.F90 M source_glimmer-cism/ncdf_utils.F90 M source_glimmer-cism/glint_example_clim.F90 M source_glimmer-cism/glimmer_log.F90 M source_glimmer-cism/glimmer_integrate.F90 M source_glimmer-cism/glint_precip_param.F90 M source_glimmer-cism/glint_global_grid.F90 M source_glimmer-cism/glimmer_sparse.F90 M source_glimmer-cism/glint_constants.F90 M source_glimmer-cism/glide_diagnostics.F90 M source_glimmer-cism/isostasy.F90 M source_glimmer-cism/glint_global_interp.F90 M source_glimmer-cism/glimmer_config.F90 M source_glimmer-cism/glint_main.F90 M source_glimmer-cism/glimmer_pdd.F90 M source_glimmer-cism/glint_mpinterp.F90 M source_glimmer-cism/glint_interp.F90 M source_glimmer-cism/glide_lithot3d.F90 M source_glimmer-cism/glimmer_physcon.F90 M source_glimmer-cism/glimmer_map_trans.F90 M source_glimmer-cism/profile.F90 M source_glimmer-cism/glimmer_ncio.F90 M source_glimmer-cism/glide_lithot.F90 M source_glimmer-cism/glint_mbal_coupling.F90 M source_glimmer-cism/glint_smb.F90 M source_glimmer-cism/glint_type.F90 ================================================================================ Originator: jwolfe Date: Jun 08, 2010 Model: cism Version: cism1_100608 One-line summary: Additional resolution support, fix calendar bug M bld/cism.cpl7.template M source_glc/glc_io.F90 M source_glc/glc_time_management.F90 ================================================================================ Originator: jwolfe Date: Jun 03, 2010 Model: cism Version: cism1_100603 One-line summary: Update cpl7 template to work correctly with hybrid and branch runs M bld/cism.cpl7.template ================================================================================ Originator: jwolfe Date: May 25, 2010 Model: cism Version: cism1_100525a One-line summary: patch to new tag, deleting one file that no longer belongs D source_glc/glc_glint_interp.F90 ================================================================================ Originator: jwolfe Date: May 25, 2010 Model: cism Version: cism1_100525 One-line summary: initial commit of cism1, which is more-or-less the same as $SVN/glc/branch_tags/glimmer-cism-gcm_tags/glimmer-cism-gcm03_glc4_100301 except the var.def files have been removed from the bld as well as the scripts to handle them. CSEG decided not to require python, which one of the scripts uses, so the resulting _io.F90 files are included in source_glimmer-cism instead. A source_slap A source_slap/dmset.F A source_slap/dlaputil.F A source_slap/xersla.F A source_slap/dcgs.F A source_slap/blas.F A source_slap/dmvops.F A source_slap/mach.F D bld/namelist_definition_gglc.xml M bld/configure M bld/config_definition.xsl D bld/gglc.cpl7.template M bld/README A bld/namelist_definition_cism.xml D bld/namelist_defaults_gglc.xml A bld/cism.cpl7.template M bld/build-namelist M bld/config_definition.xml A bld/namelist_defaults_cism.xml M source_glc/glc_glint_interp.F90 M source_glc/glc_comp_mct.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_io.F90 M source_glc/glc_constants.F90 M source_glc/glc_global_fields.F90 M source_glc/glc_time_management.F90 A source_glimmer-cism A source_glimmer-cism/glide.F90 A source_glimmer-cism/glimmer_scales.F90 A source_glimmer-cism/glint_timestep.F90 A source_glimmer-cism/glide_thck.F90 A source_glimmer-cism/glide_lithot_io.F90 A source_glimmer-cism/isostasy_setup.F90 A source_glimmer-cism/isostasy_el.F90 A source_glimmer-cism/glide_nc_custom.F90 A source_glimmer-cism/glint_mbal_io.F90 A source_glimmer-cism/glimmer_restart_statarr.F90 A source_glimmer-cism/isostasy_types.F90 A source_glimmer-cism/glimmer_ts.F90 A source_glimmer-cism/glimmer_routing.F90 A source_glimmer-cism/glimmer_searchcircle.F90 A source_glimmer-cism/glimmer_daily_pdd.F90 A source_glimmer-cism/xls.F90 A source_glimmer-cism/glint_mbal.F90 A source_glimmer-cism/glide_temp.F90 A source_glimmer-cism/glint_climate.F90 A source_glimmer-cism/glimmer_utils.F90 A source_glimmer-cism/glimmer_map_init.F90 A source_glimmer-cism/glimmer_anomcouple.F90 A source_glimmer-cism/glimmer_restart_pointarr.F90 A source_glimmer-cism/glimmer_filenames.F90 A source_glimmer-cism/glimmer_ncparams.F90 A source_glimmer-cism/glimmer_writestats.F90 A source_glimmer-cism/glint_initialise.F90 A source_glimmer-cism/glimmer_paramets.F90 A source_glimmer-cism/glimmer_vers.F90 A source_glimmer-cism/glide_lithot1d.F90 A source_glimmer-cism/glimmer_map_CFproj.F90 A source_glimmer-cism/glide_mask.F90 A source_glimmer-cism/glint_io.F90 A source_glimmer-cism/glide_profile.F90 A source_glimmer-cism/glide_setup.F90 A source_glimmer-cism/glimmer_map_proj4.F90 A source_glimmer-cism/glimmer_restart_common.F90 A source_glimmer-cism/glide_types.F90 A source_glimmer-cism/glide_velo.F90 A source_glimmer-cism/glimmer_global.F90 A source_glimmer-cism/glimmer_map_types.F90 A source_glimmer-cism/glimmer_deriv.F90 A source_glimmer-cism/glimmer_coordinates.F90 A source_glimmer-cism/glimmer_ncdf.F90 A source_glimmer-cism/kelvin.F90 A source_glimmer-cism/glide_stop.F90 A source_glimmer-cism/ncdf_utils.F90 A source_glimmer-cism/glint_example_clim.F90 A source_glimmer-cism/glimmer_log.F90 A source_glimmer-cism/glimmer_integrate.F90 A source_glimmer-cism/glint_precip_param.F90 A source_glimmer-cism/glimmer_restart.F90 A source_glimmer-cism/glint_global_grid.F90 A source_glimmer-cism/glimmer_sparse.F90 A source_glimmer-cism/glint_constants.F90 A source_glimmer-cism/glide_diagnostics.F90 A source_glimmer-cism/isostasy.F90 A source_glimmer-cism/glint_global_interp.F90 A source_glimmer-cism/glimmer_config.F90 A source_glimmer-cism/glint_main.F90 A source_glimmer-cism/glimmer_pdd.F90 A source_glimmer-cism/glint_mpinterp.F90 A source_glimmer-cism/glint_interp.F90 A source_glimmer-cism/glide_lithot3d.F90 A source_glimmer-cism/glimmer_restart_statscal.F90 A source_glimmer-cism/glimmer_physcon.F90 A source_glimmer-cism/profile.F90 A source_glimmer-cism/glimmer_map_trans.F90 A source_glimmer-cism/glimmer_ncio.F90 A source_glimmer-cism/glide_lithot.F90 A source_glimmer-cism/glint_mbal_coupling.F90 A source_glimmer-cism/glide_io.F90 A source_glimmer-cism/glint_type.F90 A source_glimmer-cism/glint_smb.F90 M ChangeLog D source_glimmer A input_templates/ice.config.gland10 M input_templates/ice.config.gland20 A input_templates/ice.config.gland5 M README ================================================================================ Originator: jwolfe Date: May 07, 2010 Model: glc Version: glc4_100507 One-line summary: bug fixes and clean up * replace hard-wired unit numbers with calls to shr_file_getunit M source_glimmer/glimmer_ts.F90 M source_glimmer/glimmer_config.F90 * clean up unused files D input_templates/glc_in.jw D input_templates/glc_in * rename "ice.config" file to "cism.config" M bld/gglc.cpl7.template M bld/namelist_defaults_gglc.xml ================================================================================ Originator: jwolfe Date: Mar 30, 2010 Model: glc Version: glc4_100330 One-line summary: CCSMize output filenames, add tools for creating overlap files, bug fix * CCSMize output filenames A source_glc/glc_io.F90 M bld/gglc.cpl7.template M input_templates/ice.config.gland20 M source_glimmer/glint_initialise.F90 * Add tools for creating overlap files A tools A tools/README.glc_tools A tools/README.glc_overlap_tools A tools/glc2scripConvert.ncl A tools/scrip_make_wgts_CCSM_to_GLC_bilin.csh A tools/scrip2CLMoverlap.ncl * Fix bug that closed log file before the end of output M source_glimmer/glint_main.F90 ================================================================================ Originator: jwolfe Date: Mar 16, 2010 Model: glc Version: glc4_100316 One-line summary: Support for fracdata files for downscaling * Changed output for gglc.buildnml.csh to point at fracdata file M bld/gglc.cpl7.template * The following files were changed Bill Lipscomb to support the use of fracdata files for downscaling: M source_glc/glc_InitMod.F90 M source_glc/glc_comp_mct.F90 M source_glc/glc_constants.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_glint_interp.F90 M source_glimmer/glint_global_grid.F90 M source_glimmer/glint_interp.F90 M source_glimmer/glint_main.F90 M source_glimmer/glint_mbal_coupling.F90 M source_glimmer/glint_timestep.F90 ================================================================================ Originator: jwolfe Date: Mar 01, 2010 Model: glc Version: glc4_100301 One-line summary: Set test cell information in code based on global grid size, instead of set as parameters in glc_constants (caused a problem with T31 grid) M source_glc/glc_global_grid.F90 ----- Added test cell settings based on global grid size M source_glc/glc_constants.F90 ------- Removed parameter attribute from test cell constants ================================================================================ Originator: erik Date: Nov 03, 2009 Model: glc Version: glc4_091103 One-line summary: Point to checked in version of topo files, add documentation, add config files to buildnml script so can change, start adding xml support. M bld/gglc.cpl7.template -------------- Point to checked in version of topo files Add documentation and put config file in the buildnml script so can be changed by the user. Start adding flexibility for support for different GLC grids. A bld/README -------------------------- Document files in the build directory A bld/namelist_defaults_gglc.xml ------ GGLC specific namelist defaults A bld/namelist_definition_gglc.xml ---- GGLC specific namelist defintion A bld/namelist_defaults_overall.xml --- Overall namelist defaults A bld/namelist_definition_overall.xml - Overall namelist defintion A bld/namelist_defaults.xsl ----------- Style sheet to display defaults as html A bld/namelist_definition.xsl --------- Style sheet to display definition as html Some scripts starting to add that do NOT work yet. A bld/build-namelist ----------------- Start adding a build-namelist (not working yet) A bld/configure ---------------------- Start adding a configure (not working yet) A bld/configure_definition.xml ------- Definition of configurations A bld/configure_definition.xsl ------- Style sheet to display definition as html ================================================================================ Originator: lipscomb Date: Oct 27, 2009 Model: glc Version: glc4_091027 One-line summary: Add glimmer log file and add greenland 20km config file M source_glc/glc_InitMod.F90 ---------- Change glimmer log file M source_glimmer/glimmer_log.F90 ------ Add subroutine to change glimmer log M input_templates/ice.config.gland20 -- Add greenland 20km config file ================================================================================ Originator: erik Date: Aug 14, 2009 Model: glc Version: glc4_090814 One-line summary: Template changes to allow multiple grid files Set up grid files for 48x96, 1.9x.25, and 0.9x1.25 resolutions. Abort if it isn't one of those. Also remove the coupled_nml namelist as not needed, and set the stop_option to never, since, the coupler will tell it to stop. M bld/gglc.cpl7.template ================================================================================ Originator: lipscomb Date: Aug 13, 2009 Model: glc Version: glc4_090813 One-line summary: Latest changes from Bill Lipscomb Mostly changes to get time-management working correctly. Also make sure proper mask of r8 for some variables. Update some of the documentation and formatting. M README M source_glc/glc_glint.F90 M source_glc/glc_comp_mct.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_grid.F90 M source_glc/glc_FinalMod.F90 M source_glc/glc_time_management.F90 M source_glimmer/glide.F90 M source_glimmer/glint_initialise.F90 M source_glimmer/glimmer_ncio.F90 ================================================================================ Originator: tcraig Date: Fri Jan 16, 2009 Model: glc Version: glc4_090116 One-line summary: Upgrade from ccsm3 to ccsm4 coupling - Try not to touch any glimmer code - Delete unusued code and remove use of unused code from used code as needed - Add bld and gglc.cpl7.template - Delete glc.F90 and add glc_comp_mct.F90 for migration from ccsm3 to ccsm4 - Update the stdout, stderr, and nml unit numbers to be set from shr_file_getunit, eliminate use of local get_unit in favor of shr_file_getunit in glc - Migrate all use of unit 6 to stdout in all source_glc code. includes write, print, and flush statements. this was not done to glimmer. - Add a shr_sys_abort to source_glimmer/glimmer_log.F90 before the stop to prevent ccsm4 hang from glimmer abort. (there are probably more calls to stop that need to be fixed in the code). - Get rid of use and reference to ccsm3 coupling code - Update initialization of glc communicator from ccsm4 driver Other potential issues still not addressed - other unit numbers hardwired in glc/glimmer - use of stop, proper aborts - reference to gland20.input.nc file from ice.config input file - hardwired glc_nec = 10 - hardwired T31 resolution - glimmer mostly writes to unit 6 - restart flag captured in coupling interface but not used A bld A bld/gglc.cpl7.template D mpi/glc_gather_scatter.F90 D mpi/glc_global_reductions.F90 D mpi/glc_broadcast.F90 D mpi/glc_boundary.F90 D mpi/glc_timers.F90 M mpi/glc_communicate.F90 A source_glc/glc_comp_mct.F90 D source_glc/glc_fileunits.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_global_grid.F90 D source_glc/glc_domain.F90 M source_glc/glc_exit_mod.F90 M source_glc/glc_constants.F90 M source_glc/glc_glint.F90 D source_glc/glc_blocks.F90 D source_glc/glc_distribution.F90 D source_glc/glc.F90 M source_glc/glc_FinalMod.F90 M source_glc/glc_ErrorMod.F90 M source_glc/glc_time_management.F90 D source_glc/glc_coupled.F90 D source_glc/glc_io_types.F90 M source_glimmer/glimmer_log.F90 ================================================================================ Originator: tcraig Date: Oct 17, 2009 Model: glc Version: glc_081017 One-line summary: Port to bluefire, update usability M input_templates/glc_in update input file paths to bluefire inputdata area ================================================================================ Originator: Date: Model: glc Version: glc_080817 One-line summary: Changes made: ================================================================================ Originator: jwolfe Date: Mon Nov 5 13:39:33 MST 2007 Model: glc Version: glc_071105 One-line summary: update to include all of Bll Lipscomb's new changes Changes made: ? source_glc/glc_glint.F90 ? source_glc/glc_global_fields.F90 ? source_glc/glc_coupled.F90 ? source_glc/glc_global_grid.F90 ? source_glc/POP_files/grid.F90 M source_glc/glc_domain_size.F90 M source_glc/glc_InitMod.F90 M source_glc/glc_RunMod.F90 M source_glc/glc_domain.F90 M source_glc/glc_io.F90 M source_glc/glc_kinds_mod.F90 M source_glc/glc_constants.F90 M source_glc/glc.F90 M source_glc/glc_FinalMod.F90 M source_glc/glc_ErrorMod.F90 M source_glc/glc_IOUnitsMod.F90 M source_glc/glc_time_management.F90 M source_glimmer/glint_timestep.F90 M source_glimmer/glint_initialise.F90 M source_glimmer/glint_example_clim.F90 M source_glimmer/glint_main.F90 M source_glimmer/glint_interp.F90 M input_templates/glc_in ================================================================================