Add CRUJRA2024 as an option to DATM_MODE

[slevis-lmwg]

Description of changes

Change the default datm input from GSWP3v1 to CRUJRA2024 for Clm6.
Add a CRUJRA2024 compset for clm5. <-- This gets addressed in ctsm exclusively as far as I can tell.

Connected to the changes in ESCOMP/CTSM#2956 and the corresponding issue ESCOMP/CTSM#1895

See definition of DONE in the issue ESCOMP/CTSM#1895

Specific notes

Contributors other than yourself, if any:

CDEPS Issues Fixed (include github issue #):
#323
#324

Are there dependencies on other component PRs (if so list):
ESCOMP/CTSM#2956

Are changes expected to change answers (bfb, different to roundoff, more substantial):
Yes, see note in ESCOMP/CTSM#2956

Any User Interface Changes (namelist or namelist defaults changes):
Yes.

Testing performed (e.g. aux_cdeps, CESM prealpha, etc):
None, yet.

Hashes used for testing:

[slevis-lmwg]

@jedwards4b
I ran ./run_sys_tests -s aux_cdeps -c /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55 --skip-generate
and got one error that seems expected, but I wanted to run it by you:

SMS_D_Ld3.f10_f10_ais8gris4_mg37.DATAMODELTEST.derecho_gnu
ERROR: Invalid compset name, DATAMODELTEST, all stub components generated

Also I didn't see a more recent baseline to compare to in cesm_baselines and did not know if I should have looked elsewhere.

[jedwards4b]

@slevis-lmwg Yes this is an expected failure. I will generate an update to baseline tests this week, we can merge this if you are ready. Thank you for adding tests specific to this new datm mode.

[slevis-lmwg]

Thanks @jedwards4b. I will discuss my corresponding CTSM mods with @ekluzek in the next day or two in order to get his comments. Meanwhile @ekluzek I'm ok with Jim merging this PR but let me know if you disagree.

@jedwards4b a heads up though: My branch here describes as cdeps1.0.61-1-g2385f03 which I think may not include the very last commit to cdeps main. Do you prefer that I handle merging main back to this branch and retesting, or is that unnecessary?

[jedwards4b]

@slevis-lmwg I did the merge here @jedwards4b
Merge branch 'main' into mk_crujra_default
bfb435b. Did you check after pulling that change?

[slevis-lmwg]

Looks good, thank you for pointing that out, I hadn't noticed that you had done that. My mods seem orthogonal to those of the merge. Since you were going to generate a new baseline, I will let you run aux_cdeps this time, and I would expect it to work fine.

[ekluzek]

@slevis-lmwg and I went over this together and we noted the things to do. I'm marking this as "requesting changes", but once the conversations are resolved this will be good to go.

[ekluzek]

@slevis-lmwg you'll need to update you share version in your CTSM clone in order to use the updated form for shr_sys_abort.

[jedwards4b]

@slevis-lmwg you just need to update the cesm_share external to share1.1.8

[slevis-lmwg]

@ekluzek thank you.
I think you're saying that I do not need the "bug fix" that I introduced and, instead, I should update my /share.

If I update /share, will shr_sys_abort still work in the other components that I'm pointing to and in clm?

[jedwards4b]

@slevis-lmwg share1.1.8 and everything should work.

[slevis-lmwg]

@slevis-lmwg Yes this is an expected failure. I will generate an update to baseline tests this week, we can merge this if you are ready. Thank you for adding tests specific to this new datm mode.

Thank you, @jedwards4b
I am now ready

[slevis-lmwg]

@ekluzek CRUJRA is second in line in CTSM's upcoming tags, which requires this cdeps PR merged pending your approval (just a reminder).

[slevis-lmwg]

@jedwards4b should I run aux_cdeps against cdeps1.0.55 again given the latest merge of main?

[slevis-lmwg]

...and should I generate a new baseline in the process?

[jedwards4b]

@slevis-lmwg that would be great - thank you

[slevis-lmwg]

Thanks @jedwards4b.
I accidentally named the new baseline cdeps1.0.67, so rename to cdeps1.0.68 DONE.

[slevis-lmwg]

aux_cdeps results

These are new tests, so this result is expected:

    FAIL SMS_Ld5.f10_f10_mg37.1850_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel BASELINE /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55: ERROR BFAIL baseline directory '/glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55/SMS_Ld5.f10_f10_mg37.1850_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel' does not exist
    FAIL SMS_Ld5.f10_f10_mg37.2010_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel BASELINE /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55: ERROR BFAIL baseline directory '/glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55/SMS_Ld5.f10_f10_mg37.2010_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel' does not exist
    FAIL SMS_Ld5.f10_f10_mg37.HIST_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel BASELINE /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55: ERROR BFAIL baseline directory '/glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55/SMS_Ld5.f10_f10_mg37.HIST_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel' does not exist
    FAIL SMS_Ld5.f10_f10_mg37.SSP585_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel BASELINE /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55: ERROR BFAIL baseline directory '/glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55/SMS_Ld5.f10_f10_mg37.SSP585_DATM%CRUJRA2024_SLND_SICE_SOCN_SROF_SGLC_SWAV_SESP.derecho_intel' does not exist

Otherwise, NLCOMP diffs, which I believe are expected.

[ekluzek]

Awesome to see this coming in. I make a few suggestions, so marked it so. But, they shouldn't be hard to do. But, feel free to ping me to discuss...

[slevis-lmwg]

@ekluzek
I made the recommended changes and reran
./run_sys_tests -s aux_cdeps -c /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.55 -g /glade/campaign/cesm/cesmdata/cseg/cesm_baselines/cdeps1.0.68
successfully, so this PR is ready for your approval.

@jedwards4b I think you would then make the new cdeps1.0.68 tag? Or would Erik or I do it?

[jedwards4b]

I can merge and make the new tag as soon as @ekluzek follows through on approval.

[ekluzek]

I ask for some changes to comments. But, marking as approve, so you don't have to ask me again. Thanks for the work here @slevis-lmwg, I'm really glad to see this come in.

[slevis-lmwg]

@jedwards4b I will rerun aux_cdeps since we changed one of the tests. I will be in touch as soon as it is done.

[ekluzek]

@slevis-lmwg in terms of process, I think you've seen what to do here. But, to make it concrete here are what the steps are for here (as I view it) and other repositories like it (CMEPS, cime, etc..). This is how I view it, and @jedwards4b can confirm or suggest changes to anything below...

• Make an issue in the repository to bring awareness of the desired change
• Make a PR
• Add new tests to the repository testlist if there is new capability added
• Run the baseline for the testlist (this might be OK to shortcut for smaller changes, but should ask repo maintainers)
• Compare the baseline to with the updated PR (make sure you are doing baselines with your branch updated to the latest)
• Have a repository maintainer approve it (Jim for sure in CDEPS, Bill Sacks as a secondary) (look at contributors to help gauge this)
• Also have a CTSM person approve it, if there's a lot of CTSM specific content
• Ask repository maintainer to do the merge
• Repository maintainer pushes the merge button, and a tag is made automatically when that happens

[slevis-lmwg]

@jedwards4b aux_cdeps worked as expected, and I generated a new cdeps1.0.68 baseline. This is ready.

[slevis-lmwg]

@jedwards4b thank you for the merge. A heads up that in reviewing ESCOMP/CTSM#2956, @ekluzek found a few new issues that I will revise in cdeps. This means that I will open a new small cdeps PR, likely this afternoon.