Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks #569

gdicker1 · 2025-06-12T22:37:20Z

Description of changes

This PR edits config_*.xml files and fields exchanged to enable MPAS Ocean and MPAS SeaIce to be used as components in CESM. These components have been forked from E3SM and developed in the EarthWorks project (EarthWorksOrg/EarthWorks on GitHub).

Specific notes

Contributors other than yourself, if any: @dazlich (Don Dazlich)

Are changes expected to change answers? bfb

Any User Interface Changes (namelist or namelist defaults changes)? N/A

Testing performed

I tested a CESM sandbox with this and other relevant changes compared against the baseline version, cesm3_0_beta06 and the externals cloned there. I mainly ran the aux_cmeps via create_test on Derecho. I did not notice any new fails, just many NLFAILS reported since model_version was compared between the namelists.

Baseline: ESCOMP/CESM:cesm3_0_beta06 tag (637b59332d7ff3d339906116942a334ed5537d18)
Test: gdicker1/CESM:eworg_mpaso_mpassi branch (e6698083a8f7199dedc1cc50204dd4b324808d45)

Baseline aux_cmeps FAILS on Derecho, 5 of 24 tests:

  ERP_Ln9.f09_f09_mg17.F2000climo.derecho_nvhpc.cam-outfrq9s (Overall: FAIL) details:
    FAIL ERP_Ln9.f09_f09_mg17.F2000climo.derecho_nvhpc.cam-outfrq9s MODEL_BUILD time=393
  ERS_Ld5.T62_t232.G_JRA.derecho_nvhpc (Overall: FAIL) details:
    FAIL ERS_Ld5.T62_t232.G_JRA.derecho_nvhpc MODEL_BUILD time=175
  ERS_Lm13_P32.f10_f10_mg37.I1850Clm50SpG.derecho_nvhpc (Overall: FAIL) details:
    FAIL ERS_Lm13_P32.f10_f10_mg37.I1850Clm50SpG.derecho_nvhpc RUN time=0
  ERS_Ln9_C3.f19_g17_rx1.A.derecho_intel (Overall: FAIL) details:
    FAIL ERS_Ln9_C3.f19_g17_rx1.A.derecho_intel CREATE_NEWCASE
  SMS_Ln11_D.f19_g17_rx1.A.derecho_gnu (Overall: FAIL) details:
    FAIL SMS_Ln11_D.f19_g17_rx1.A.derecho_gnu CREATE_NEWCASE

Test (this PR) aux_cmeps FAILS on Derecho, 5 of 24 tests (plus 19 NLFAILS):

  ERP_Ln9.f09_f09_mg17.F2000climo.derecho_nvhpc.cam-outfrq9s (Overall: FAIL) details:
    FAIL ERP_Ln9.f09_f09_mg17.F2000climo.derecho_nvhpc.cam-outfrq9s MODEL_BUILD time=349
    FAIL ERR_Ld5.ne30pg3_t232.BMT1850.derecho_intel.allactive-defaultio TPUTCOMP Could not get default throughput
  ERS_Ld5.T62_t232.G_JRA.derecho_nvhpc (Overall: FAIL) details:
    FAIL ERS_Ld5.T62_t232.G_JRA.derecho_nvhpc MODEL_BUILD time=165
  ERS_Lm13_P32.f10_f10_mg37.I1850Clm50SpG.derecho_nvhpc (Overall: FAIL) details:
    FAIL ERS_Lm13_P32.f10_f10_mg37.I1850Clm50SpG.derecho_nvhpc RUN time=0
  ERS_Ln9_C3.f19_g17_rx1.A.derecho_intel (Overall: FAIL) details:
    FAIL ERS_Ln9_C3.f19_g17_rx1.A.derecho_intel CREATE_NEWCASE
  SMS_Ln11_D.f19_g17_rx1.A.derecho_gnu (Overall: FAIL) details:
    FAIL SMS_Ln11_D.f19_g17_rx1.A.derecho_gnu CREATE_NEWCASE

…cesm_mod.F90 Since EarthWorksOrg/cmeps has been based on 'cmeps0.13.68' the routine signatures are different to those used in 'cmeps0.14.2'. So, logic added for mpas-ocean and mpas-seaice fields did not automatically update to the new implementations. This commit manually does that update.

As the resolution of simulation increases, it makes sense to also increase the coupling intervals to maintain stability and realism.

If the compset name contains "_MPAS" then NCPL_BASE_PERIOD changes to hour instead of day. The new values are equivalent to the day-based values. Also add an entry for ROF, so that it is set to ATM_NCPL if the compset includes "_MPAS" in the name.

Necessary coupling fields associated with the new version of MPAS-Ocean have been added. These fields are associated with Stokes drift and may not actually be used.

Catch up to development, especially to cesm3_0_beta06 version.

gdicker1 · 2025-06-12T22:39:31Z

Some help reviewing from @jedwards4b and optionally @fischer-ncar would be appreciated! (Can't assign reviewers.)

gdicker1 · 2025-06-12T22:43:33Z

This is part of a collection of PRs to enable MPAS-O and MPAS-SI in CESM:

jedwards4b

I am concerned that these changes are being applied to all cases but may only be relevant for earthworks. Are you certain that they have no adverse effects on other compsets?

gdicker1 · 2025-06-16T18:35:48Z

I am concerned that these changes are being applied to all cases but may only be relevant for earthworks. Are you certain that they have no adverse effects on other compsets?

I'm not exhaustively certain. I have not noticed any answer differences from cesm3_0_beta06 after adding these changes for what I've tested (especially aux_cmeps runs for this PR). @dazlich any thoughts?

@jedwards4b, would it help your concern to add some logic so this new code is only executed if MPAS-O is the ocean model (`ocn_name == 'mpaso')?

jedwards4b · 2025-06-16T18:50:30Z

Yes, I think that would address my concern. Thanks

dazlich · 2025-06-16T18:55:39Z

@gdicker1 , @jedwards4b I believe the concern might be that mpasa framework might get confused with the framework used by mpaso and mpassi. We have modified the mpaso/si framework so that all modules carry a different name from those in mpasa.
One place that might need addressing is the mpas-seaice column physics - they are closely related to the CICE column physics, but if only one seaice model is built that shouldn't be an issue.

dazlich · 2025-06-16T19:06:05Z

@gdicker1 , @jedwards4b I should add that means maintenance of two copies of mpas-framework code, not a good practice. The effort to merge them is likely non-trivial, but likely not too onerous.

gdicker1 · 2025-06-16T19:25:19Z

@dazlich, noted on the mpas-framework. I likely won't be able to address it before these PRs, but that would be good to address soon-ish (i.e. before August).

dazlich · 2025-06-16T19:27:26Z

@gdicker1 - agreed. And I can make time available to work on this.

gdicker1 · 2025-06-16T19:30:18Z

@gdicker1 , @jedwards4b I believe the concern might be that mpasa framework might get confused with the framework used by mpaso and mpassi. We have modified the mpaso/si framework so that all modules carry a different name from those in mpasa. One place that might need addressing is the mpas-seaice column physics - they are closely related to the CICE column physics, but if only one seaice model is built that shouldn't be an issue.

@dazlich, could you help my understanding a bit? Is this comment a concern overall with adding MPAS-O and MPAS-SI to CESM? Or am I not following how this comment relates to the changes in esmFldsExchange_cesm_mod.F90?

dazlich · 2025-06-16T19:39:57Z

@gdicker1 - that’s an overall comment on porting MPAS-O and MPAS-SI into CESM. I neglected to note this was in the CMEPS discussion when my name popped up. Is there another place to put my framework comments? As to esmFldsExchange_cesm_mod.F90 - I recall adding mpaso/si variables as necessary. I don’t have a good idea what, if any, conflicts they might introduce into cesm, so the mpaso specific logic comment is a good idea.

gdicker1 · 2025-06-16T19:59:27Z

@gdicker1 - that’s an overall comment on porting MPAS-O and MPAS-SI into CESM. I neglected to note this was in the CMEPS discussion when my name popped up. Is there another place to put my framework comments?

@dazlich, I think this works for now since I don't have an ESCOMP/CESM PR. I'll get some of these points copied into EarthWorksOrg/EarthWorks Issue#40.

As to esmFldsExchange_cesm_mod.F90 - I recall adding mpaso/si variables as necessary. I don’t have a good idea what, if any, conflicts they might introduce into cesm, so the mpaso specific logic comment is a good idea.

Thanks!

This change tries to ensure that adding MPAS-O and MPAS-SI won't affect CESM results.

gdicker1 · 2025-06-18T17:56:20Z

@jedwards4b, could you make sure I've satisfied requested edits?

And @dazlich, your opinion would be valued too.

jedwards4b · 2025-06-18T18:00:03Z

mediator/esmFldsExchange_cesm_mod.F90

+    !-----------------------------
+    ! to ocn:
+    !-----------------------------
+    if (ocn_name == 'mpaso') then


I think that you can bracket all of this code with one if instead of having one for each variable.

Understood, will do.

My recent commit just did the big 'mpaso' section. Let me know if I need to consolidate more.

dazlich · 2025-06-18T18:41:22Z

I agree with Jim, otherwise looks good

jedwards4b

thanks

Don Dazlich and others added 11 commits August 27, 2022 17:03

Add mpas-ocean and seaice to the components

e187b54

Merge tag ESCOMP/CMEPS 'cmeps0.13.68' into 'eworg_mpaso_mpassi'

add92e9

Merge tag ESCOMP/CMEPS 'cmeps0.14.45' into 'eworg_mpaso_mpassi'

3f20fc7

Configure ATM_NCPL based on mpasa grids used

c057001

As the resolution of simulation increases, it makes sense to also increase the coupling intervals to maintain stability and realism.

Merge tag ESCOMP/CMEPS 'cmeps1.0.16' into 'eworg_mpaso_mpassi'

1d7b135

MPAS-Ocean update

be66c78

Necessary coupling fields associated with the new version of MPAS-Ocean have been added. These fields are associated with Stokes drift and may not actually be used.

Typo fix

decbf4e

Merge tag ESCOMP/CMEPS 'cmeps1.0.33' into 'eworg_mpaso_mpassi'

ce2f732

Merge tag 'cmeps1.0.47' into branch 'eworg_mpasso_mpassi'

e44ab51

Catch up to development, especially to cesm3_0_beta06 version.

This was referenced Jun 12, 2025

Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks ESMCI/ccs_config_cesm#241

Merged

Feature request: Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks ESMCI/cime#4806

Closed

Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks ESMCI/cime#4807

Merged

gdicker1 mentioned this pull request Jun 12, 2025

Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks ESCOMP/CAM#1326

Open

fischer-ncar requested review from jedwards4b and fischer-ncar and removed request for fischer-ncar June 13, 2025 17:21

jedwards4b reviewed Jun 13, 2025

View reviewed changes

gdicker1 mentioned this pull request Jun 16, 2025

Duplicate compilation of shared MPAS infrastructure EarthWorksOrg/EarthWorks#40

Open

Review request, guard field exchanges added by MPAS-O and MPAS-SI

5209116

This change tries to ensure that adding MPAS-O and MPAS-SI won't affect CESM results.

jedwards4b requested changes Jun 18, 2025

View reviewed changes

Reveiw req., consolidate exchange MPAS guards

832a085

jedwards4b approved these changes Jun 18, 2025

View reviewed changes

Merge branch 'main' into eworg_mpaso_mpassi

dcd1c7c

Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks #569

Are you sure you want to change the base?

Add MPAS-O and MPAS-SI from EarthWorksOrg/EarthWorks #569

Uh oh!

Conversation

gdicker1 commented Jun 12, 2025

Description of changes

Specific notes

Testing performed

Uh oh!

gdicker1 commented Jun 12, 2025

Uh oh!

gdicker1 commented Jun 12, 2025

Uh oh!

jedwards4b left a comment

Choose a reason for hiding this comment

Uh oh!

gdicker1 commented Jun 16, 2025

Uh oh!

jedwards4b commented Jun 16, 2025

Uh oh!

dazlich commented Jun 16, 2025

Uh oh!

dazlich commented Jun 16, 2025

Uh oh!

gdicker1 commented Jun 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dazlich commented Jun 16, 2025 via email

Uh oh!

gdicker1 commented Jun 16, 2025

Uh oh!

dazlich commented Jun 16, 2025 via email

Uh oh!

gdicker1 commented Jun 16, 2025

Uh oh!

gdicker1 commented Jun 18, 2025

Uh oh!

jedwards4b Jun 18, 2025

Choose a reason for hiding this comment

Uh oh!

gdicker1 Jun 18, 2025

Choose a reason for hiding this comment

Uh oh!

gdicker1 Jun 18, 2025

Choose a reason for hiding this comment

Uh oh!

dazlich commented Jun 18, 2025

Uh oh!

jedwards4b left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

gdicker1 commented Jun 16, 2025 •

edited

Loading