new global_nest_v1 suite and #1965

[SamuelTrahanNOAA]

PR Author Checklist:

• I have linked PR's from all sub-components involved in section below.
• I am confirming reviews are completed in ALL sub-component PR's.
• I have run the full RT suite on either Hera/Cheyenne AND have attached the log to this PR below this line:
  • LOG: Committed to branch. Tested on hera and jet.
• I have added the list of all failed regression tests to "Anticipated changes" section.
• I have filled out all sections of the template.

Description

This is a suite of physics schemes nominally planned for the RRFS v2 upgrade. It's being used for the Water in the West project to simulate atmospheric rivers using a global model with a static nest. Hence "global_nest_v1."

Includes #1965 which fixes issue #1964

Linked Issues and Pull Requests

Associated UFSWM Issue to close

• fixes Updates to C3 and SAS cumulus convection schemes #1964
• closes Updates to C3 and SAS convection schemes #1965

Subcomponent Pull Requests

• FV3: new global_nest_v1 suite and #715 NOAA-EMC/fv3atm#709
• ccpp-physics: Updates to C3 and SAS convection schemes ccpp-physics#122
• Lisa ccpp-physics: Merge PR #121 and resolve conflicts lisa-bengtsson/ccpp-physics#3

Blocking Dependencies

None.

Subcomponents involved:

• AQM
• CDEPS
• CICE
• CMEPS
• CMakeModules
• FV3
• GOCART
• HYCOM
• MOM6
• NOAHMP
• WW3
• stochastic_physics
• none

Anticipated Changes

Input data

• No changes are expected to input data.

Regression Tests:

• No changes are expected to any regression test.
• Changes are expected to the following tests:

There are two new tests:

1. gnv1_c96_no_nest_debug - The standard C96 initial conditions used for most physics suites, configured to run global_nest_v1. This is run in debug mode for both compilers
2. gnv1_nested - Run off of the HAFS global static nest test initial conditions, this tests the suite in a static nest.

Libraries

• Not Needed

Code Managers Log

• This 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.
• Move new/updated input data on RDHPCS Hera and propagate input data changes to all supported systems.
  • N/A

Testing Log:

• RDHPCS
  • Hera
  • Orion
  • Hercules
  • Jet
  • Gaea
  • Cheyenne
• WCOSS2
  • Dogwood/Cactus
  • Acorn
• CI
  • Completed
• opnReqTest
  • N/A
  • Log attached to comment

[SamuelTrahanNOAA]

@SamuelTrahanNOAA thanks! @BrianCurtis-NOAA @zach1221 @FernandoAndrade-NOAA I think we can create and add new baseline cases gnv1_c96_no_nest_debug and gnv1_nested to current baseline BL_DATE=20231108.

I already updated my BL_DATE to 20231109 so it would not clash with the other baseline during my own testing. If you want the old date, you'll need to revert my bl_date.conf.

Also: I generated the baselines on Hera and Jet already. You should consider using those, to save time.

[jkbk2004]

Member

That's fine. We can copy over to new one.

[SamuelTrahanNOAA]

The OTR and CI tests have said they're "Pending" for a while. Is something wrong?

[zach1221]

The OTR and CI tests have said they're "Pending" for a while. Is something wrong?

No, it's running currently. No issues yet.

[jkbk2004]

@SamuelTrahanNOAA I see time out issue on /scratch1/NCEPDEV/stmp2/Jong.Kim/FV3_RT/rt_72722/gnv1_nested_intel
@zach1221 can you pick up from there?

[zach1221]

/scratch1/NCEPDEV/stmp2/Jong.Kim/FV3_RT/rt_72722/gnv1_nested_intel

I think Jong tried increasing the wall clock. I'll try to get it to run on my end.

[epic-cicd-jenkins]

Jenkins-ci ORTs passed

[zach1221]

Baseline creation on Hera for the new tests went fine, without timeouts. Running the matching now.

[SamuelTrahanNOAA]

@SamuelTrahanNOAA I see time out issue on /scratch1/NCEPDEV/stmp2/Jong.Kim/FV3_RT/rt_72722/gnv1_nested_intel @zach1221 can you pick up from there?

It's running, but it's not finishing in time.

234: THVSK_lnd(I) 181 3 360.094395904550 234: 351.458657668197 305.341506797005 305.341506797005 234: 4.043075031881894E-002 0: PASS: fcstRUN phase 1, n_atmsteps = 180 time is 8.939060

From my regression test logs, it finished in four minutes:

0: The total amount of wall time = 242.467459

Yours is running about ten times slower than mine which is an implausible difference. I'll rerun now to see if I can reproduce the problem. Most likely, you got a bad node and rerunning will fix it.

[SamuelTrahanNOAA]

I've run the gnv1_nested_intel test under my account five times, and all of them finished in under five minutes.

Edit: Those tests were on Hera, the machine where another user's gnv1_nested_intel hit its wallclock limit.

[FernandoAndrade-NOAA]

Ok, testing is complete. This should be ready to continue the merge process, I'll leave a note in the sub PRs.

[SamuelTrahanNOAA]

The FV3 submodule points to the head of develop and I've reverted .gitmodules. This PR is ready for final review and merge.