Update coupled tests to use P7 configuration, add standalone P7 test suite

[DeniseWorthen]

PR Checklist

• Ths PR is up-to-date with the top of all sub-component repositories except for those sub-components which are the subject of this PR. Please consult the ufs-weather-model wiki if you are unsure how to do this.

• This PR has been tested using a branch which is up-to-date with the top of all sub-component repositories except for those sub-components which are the subject of this PR

• An Issue describing the work contained in this PR has been created either in the subcomponent(s) or in the ufs-weather-model. The Issue should be created in the repository that is most relevant to the changes in contained in the PR. The Issue and the dependent sub-component PR
  are specified below.

• If new or updated input data is required by this PR, it is clearly stated in the text of the PR.

Instructions: All subsequent sections of text should be filled in as appropriate.

The information provided below allows the code managers to understand the changes relevant to this PR, whether those changes are in the ufs-weather-model repository or in a subcomponent repository. Ufs-weather-model code managers will use the information provided to add any applicable labels, assign reviewers and place it in the Commit Queue. Once the PR is in the Commit Queue, it is the PR owner's responsiblity to keep the PR up-to-date with the develop branch of ufs-weather-model.

Description

• Updates coupled regression tests to use P7 configuration. Retains a single benchmark configuration test cpld_bmark_p7 for C384mx025 + waves using benchmark ICs (2013-04-01). Replaces all other coupled tests with the P7 configuration using 2021-03-22-06 ICs without waves.

• Adds a suite of P7 tests for the standalone app=ATM.

control_p7
control_restart_p7
control_debug_p7
control_decomp_p7
control_2threads_p7

The standalone and coupled P7 tests share the namelist cpld_control.nml.IN.

• Uses a single compile (app=S2SW) for all coupled tests (except debug) and controls non-wave tests using namelist variables.

• Removes ccpp suite files which are now unused (FV3_GFS_2017_coupled, suite_FV3_GFS_2017_couplednsst.xml and suite_FV3_GFS_2017_satmedmf_coupled.xml).

• Uses a new input data directory input-data-20210930. Adds new ICs for MOM6 and CICE6 for 2021-03-22-06. Adds Frac grid consistent ICs for the coupled model from /scratch2/NCEPDEV/ensemble/Bing.Fu/p7rt/merge on 20210825. These are placed into INPUT_L127_mxRES directories within the correct FV3_input_data directory where RES=ocean/ice resolution (400,100,050,025). These were updated with files from /scratch2/NCEPDEV/ensemble/Bing.Fu/p7rt/merge_fix/CRES_mxRES/2021032206/gfs/CRES/INPUT/ with the ceil/floor bug fix on 09/13

• Changes MOM6 setting MOM6_REPRO_LA to MOM6_USE_LI2016. These are equivalent, but the new setting is more descriptive of the setting.

• Adds ice_in setting sec_init to accommodate initial time of 6 hours.

• Sets FNSMCC = global_soilmgldas.statsgo.t1534.3072.1536.grb for all configurations

• Removes NEMSDatm tests; associated settings and files in default_vars and parm will be removed in a follow-up PR.

• Removes standalone feature tests for features which have been incorporated into the P7 configuration (e.g. noahmp, merra2, ugwpv1).

• For pre-commit testing, the new coupled and standalone tests can be tested with rt.p7.conf using rt.sh -l rt.p7.conf .... on platforms which have the inputdata directory available.

 

New input-data-directory (input-data-20210930) transfers:

• hera
• orion
• cheyenne
• gaea
• jet
• wcoss_cray
• wcoss_dell_p3

Issue(s) addressed

• fixes Update coupled tests to use P7 configuration #766
• fixes 1deg CICE fix files are inconsistent with mesh and MOM6 runtime mask #763
• fixes Add control_p7 test #817
• fixes point to icepack consortium master NOAA-EMC/CICE#39

Testing

How were these changes tested? What compilers / HPCs was it tested with? Are the changes covered by regression tests? (If not, why? Do new tests need to be added?) Have regression tests and unit tests (utests) been run? On which platforms and with which compilers? (Note that unit tests can only be run on tier-1 platforms).

• The input.nml created for the cpld_control_c384_p7 test was compared to the input.nml for the existing cpld_bmark_wave_v16_p7b test. The fv_core_nml, gfdl_cloud_microphysics_nml, and cires_ugwp_nml are functionally equivalent (e.g. .T. vs .true.)

• The gfs_nml differences are shown graphically here:

• All the non-global coupled RTs (DATM, HAFS, ATM, ATMW, ATMAERO) were run using the input-data-20210930 at commit 19fa289. All tests pass against the develop-20210915 baseline with the exception of the DATM configurations. These do not pass because of the fix to the 1deg CICE grid file (1deg CICE fix files are inconsistent with mesh and MOM6 runtime mask #763) and updated CICE grid files created for the mx025 resolution. The currently used files were created on Theia on 20180927. The new files have been created using the updated gen_grid routine which will be added to UFS-Utils. The values of ulat,ulon changed because the conversion factor rad2deg now uses a specified constant for pi rather than pi= 4.0*atan(1.0).

• Testing against the develop-20210929 baseline at commit 3d2b0a6 shows the following tests fail

Hera.gnu

Missing baselines:
cpld_control_p7 
cpld_debug_p7
control_p7
control_debug_p7

Hera.intel

all datm_cdeps tests (due to change in CICE grid file and/or fix for land mask at 1deg noted above)
all P7 tests (missing baseline)

 

Baseline Testing

• hera.intel
• hera.gnu
• orion.intel
• cheyenne.intel
• cheyenne.gnu
• gaea.intel
• jet.intel
• wcoss_cray
• wcoss_dell_p3
• CI ; not run, changes in this PR need to be propagated to CI tests

Dependencies

• Requires fixes for these issues:

1. Issue CA fails debug mode with GNU compiler #770
2. Issue control_ca test is not reproducible with 2-threads #743
3. Issue control_ugwpv1 fails when changing the decomp #742
4. Issue ulwrf_ave calculated by ATM is not reproducible on restart #774

• FV3 PR #819
• FV3 PR #397
• FV3 PR #402
• FV3 PR #396
• Cice PR #40
• FV3 PR #404

[climbfuji]

There are too many changes in this PR to go through every line. I don't see anything wrong and will approve once all the regression tests passed and the submodules are merged and updated correctly.

[BrianCurtis-NOAA]

Automated RT Failure Notification
Machine: orion
Compiler: intel
Job: BL
Repo location: /work/noaa/nems/emc.nemspara/autort/pr/718079436/20210930210010/ufs-weather-model
Please manually delete: /work/noaa/stmp/bcurtis/stmp/bcurtis/FV3_RT/rt_305967
Baseline creation and move successful
Repo location: /work/noaa/nems/emc.nemspara/autort/pr/718079436/20211001053330/ufs-weather-model
Please manually delete: /work/noaa/stmp/bcurtis/stmp/bcurtis/FV3_RT/rt_350928
Test cpld_control_c192_p7 006 failed failed
Test cpld_control_c192_p7 006 failed in run_test failed
Please make changes and add the following label back:
orion-intel-BL

[MinsukJi-NOAA]

CI tests break because of the changes made to the regression test-related files in this PR. They will be fixed in the next PR via #829

[climbfuji]

fv3atm was merged, new hash is 799b157