MOM_interface changes for regional setups by manishvenu · Pull Request #316 · ESCOMP/MOM_interface

manishvenu · 2026-03-25T16:59:12Z

Details:
Changes to the configuration for regional setups. This includes new PE layouts, new compsets, and a new xml variable to turn on regional defaults, which currently only changes MOM_input. The bulk of the new changes are to MOM_input.

Included in this change is the regional testing suite which currently does not pass restarts.

… mom_input changes

alperaltuntas · 2026-04-30T16:36:41Z

    MLE_USE_PBL_MLD:
        description: |
-            "TODO"
+            "TODO


Remove TODO?

Hey Alper, I wasn't sure if this was just for the description, if it is, happy to remove it

yes, i think it's just for the description.

alperaltuntas · 2026-04-30T16:38:04Z

@manishvenu Is this PR ready for testing? Any reason to keep it draft, or is it ready to be reviewed?

mnlevy1981

If the regular expressions in config_pes.xml get to be too much, we can punt on that... but I do think we want to update config_component.xml

mnlevy1981 · 2026-04-30T23:35:15Z

+    <valid_values>REGIONAL,GLOBAL</valid_values>
+    <default_value>GLOBAL</default_value>
+    <values>
+      <value compset="_MOM6%REGIONAL">REGIONAL</value>


If you change this to

-<value compset="_MOM6%REGIONAL">REGIONAL</value> +<value compset="_MOM6%[^_]*REGIONAL">REGIONAL</value>

Then you are not required to put %REGIONAL as the first compset modifier (MOM6%REGIONAL%MARBL-BIO would be equivalent to MOM6%MARBL-BIO%REGIONAL)

mnlevy1981 · 2026-04-30T23:36:46Z

      </pes>
    </mach>
+    <mach name="derecho">
+      <pes pesize="any" compset="_DATM.+_SICE.*_MOM6%REGIONAL(?!.*%MARBL-BIO).+_SWAV">


Same thing, I think you want MOM6%[^_]*REGIONAL here

hmm, you might need additional logic to rule out MOM6%MARBL-BIO%REGIONAL. What a mess

This worked if you added the exclusion before the Regional.

mnlevy1981 · 2026-04-30T23:38:46Z

+          <rootpe_glc>0</rootpe_glc>
+        </rootpe>
+      </pes>
+      <pes pesize="any" compset="_DATM.+_SICE.*_MOM6%REGIONAL[^_]*MARBL-BIO.+_SWAV">


this one is trickier, because you need an or statement to capture when MARBL-BIO precedes REGIONAL, but I would start with MOM6%[^_]*REGIONAL[^_]*MARBL-BIO. The right answer might be MOM6%([^_]*REGIONAL[^_]*MARBL-BIO|[^_]*MARBL-BIO[^_]*REGIONAL) but I'm not super confident in that

Done - the OR worked.

mnlevy1981 · 2026-04-30T23:40:08Z

+          <rootpe_glc>0</rootpe_glc>
+        </rootpe>
+      </pes>
+      <pes pesize="any" compset="_DATM.+_CICE.*_MOM6%REGIONAL(?!.*%MARBL-BIO).+_SWAV">


same as one of the above comments

mnlevy1981 · 2026-04-30T23:40:15Z

+          <rootpe_glc>0</rootpe_glc>
+        </rootpe>
+      </pes>
+      <pes pesize="any" compset="_DATM.+_CICE.*_MOM6%REGIONAL[^_]*MARBL-BIO.+_SWAV">


same as one of the above comments

…r_croc

manishvenu · 2026-05-01T16:38:53Z

Hey @MNLevy,

I updated the regex for the PEs and Claude wrote a quick test script that means it potentially works: /glade/u/home/manishrv/documents/croc/dev/random/test_regex_script.py/glade/u/home/manishrv/documents/croc/dev/random/test_regex_script.py.

It's super confusing! Happy to back out the regex changes (last two commits)

This reverts commit ebc2315.

mnlevy1981 · 2026-06-18T21:18:09Z

+  <compset>
+    <alias>GR_JRA</alias>
+    <lname>1850_DATM%JRA_SLND_CICE_MOM6%REGIONAL_SROF_SGLC_SWAV</lname>
+  </compset>
+
+  <compset>
+    <alias>GR1850MARBL_JRA</alias>
+    <lname>1850_DATM%JRA_SLND_CICE_MOM6%REGIONAL%MARBL-BIO_SROF_SGLC_SWAV</lname>
+  </compset>
+
+  <compset>
+    <alias>GR_JRA_GLOFAS</alias>
+    <lname>1850_DATM%JRA_SLND_CICE_MOM6%REGIONAL_DROF%GLOFAS_SGLC_SWAV</lname>
+  </compset>
+
+  <compset>
+    <alias>GR1850MARBL_JRA_GLOFAS</alias>
+    <lname>1850_DATM%JRA_SLND_CICE_MOM6%REGIONAL%MARBL-BIO_DROF%GLOFAS_SGLC_SWAV</lname>
+  </compset>


@dabail10 and I were talking, and we probably want a CICE modifier as well (e.g. CICE%REGIONAL). Now that open boundary conditions are available in CICE, he's going to test that code out but eventually CICE will want to be able to configure regional domains.

Two notes:

I assume changing CICE to CICE%<something_else> will not change answers if CICE's config_component.xml doesn't know what to do with the new modifier, but it's worth testing that hypothesis

CICE%REGIONAL is my first thought, but @dabail10 may propose a better modifier

Maybe this should be another PR? Or can CICE use the MOM6 regional? Because you can't have one without the other?

manishvenu · 2026-06-24T21:58:00Z

Merged #315 into this for simplicity

update regional compsets to JRA v1.5 and add wave regional compsets

…to JRA-1p5-2023, use DICE%SSMI in all C*R*

manishvenu · 2026-10-01T17:41:53Z

Need to update the input nml number of fields to 400 because 300 is too small for MARBL cases on one processor (it's split per processor but there are 330 fields in regional OBC (4) models now)

cice_ic_panama1_1c3aaa.nc came from the old reference_ice, which wrote a single ice category, so CICE aborted reading it. cice_ic_panama1_6c83a5.nc is regenerated with the fixed reference_ice (CrocoDash ESCOMP#301).

river_nutrients_panama1_1c3aac_nnsm.nc's time axis is 2 days late in FMS (0001-01-01 reference encoded with a Julian-before-1582 calendar), so MARBL runs aborted at their 1900-01-01 start. river_nutrients_panama1_e09757_nnsm.nc is regenerated with the fix (CrocoDash ESCOMP#297); the data is unchanged.

manishvenu added 5 commits March 25, 2026 10:09

New tests

001e6a1

MOM_interface changes for regional, only todo is the mom6_domian tupe…

4ec21f8

… mom_input changes

MOM_input changes

e7d3055

Fix formatting

bd7a2b1

Update MOm6 commit & add the input_nl change back

3a86a50

alperaltuntas reviewed Apr 30, 2026

View reviewed changes

manishvenu marked this pull request as ready for review April 30, 2026 17:28

mnlevy1981 requested changes Apr 30, 2026

View reviewed changes

manishvenu added 5 commits April 30, 2026 18:46

Review COmments + Update MOM6 Commit

e0f871f

Merge remote-tracking branch 'upstream/main' into regional_changes_fo…

c78ed29

…r_croc

Duplicate Param

458432b

PE Layouts

7a5c1a8

Update

28deca3

manishvenu added 14 commits May 5, 2026 09:26

Line Add

2373b84

Reformatted BGC

38b9aff

in terms of din loc root

947569c

Try Adding DIN_LOC_ROOT

a34efb3

Changes

1a29180

Reduced FIle

49462f7

Back:

f74950f

Relative Path

57d19bd

Fix Riv nutri Path

5d51097

Change the max files/fields

5423719

INPUTDIR Changes

53ca846

Bracket

e27a674

Try This

ebc2315

Revert "Try This"

795f8b5

This reverts commit ebc2315.

mnlevy1981 reviewed Jun 18, 2026

View reviewed changes

Merge branch 'ct_from_main' into regional_changes_for_croc

d9e6b95

manishvenu mentioned this pull request Jun 24, 2026

Regional Tests #315

Closed

ashjbarnes mentioned this pull request Aug 3, 2026

Set up M_regional_template branch ACCESS-NRI/access-om3-configs#1493

Merged

6 tasks done

manishvenu and others added 9 commits August 31, 2026 09:57

Add named pesize tiers (XS/S/M/L) for regional MOM6 on Derecho (#10)

1a7369b

Merge branch 'main' into regional_changes_for_croc

661d723

CHange MAXTRUNC to Global Vals

97aceda

Update MOm6 to regional tag

eb92c31

new MOM6 commit (merged regional to main)

5957bf4

New WW3 compsets

9bd5808

update regional compsets to JRA v1.5 and add wave regional compsets

2523e16

Merge pull request #12 from CROCODILE-CESM/new_regional_compsetes

d73fa52

update regional compsets to JRA v1.5 and add wave regional compsets

Regional compsets: drop duplicate CWR_JRA/GWR_JRA, move WW3 variants …

cd732d3

…to JRA-1p5-2023, use DICE%SSMI in all C*R*

manishvenu and others added 3 commits October 6, 2026 10:29

Add regional CICE, WW3 and all-component tests (#13)

5419651

Point regional cice testmods at the ncat=5 Panama CICE initial state

6fb0c16

cice_ic_panama1_1c3aaa.nc came from the old reference_ice, which wrote a single ice category, so CICE aborted reading it. cice_ic_panama1_6c83a5.nc is regenerated with the fixed reference_ice (CrocoDash ESCOMP#301).

Conversation

manishvenu commented Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alperaltuntas commented Apr 30, 2026

Uh oh!

mnlevy1981 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

manishvenu commented May 1, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

manishvenu commented Jun 24, 2026

Uh oh!

manishvenu commented Oct 1, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

manishvenu commented Mar 25, 2026 •

edited

Loading