Optionally produce coeffs_reconstruct stream for MPAS-O, link coeffs file from database for Omega by cbegeman · Pull Request #448 · E3SM-Project/polaris

cbegeman · 2026-01-12T22:18:06Z

The following changes are needed because only MPAS-O produces the coefficients needed to reconstruct zonal and meridional components from edge fields:

Use config option to determine whether to add coeffs_reconstruct to the stream for forward steps with MPAS-O (ensure this capability can only be turned on with MPAS-O)
Add the netcdf files containing the coeffs_reconstruct stream to the input database (see Optionally produce coeffs_reconstruct stream for MPAS-O, link coeffs file from database for Omega #448 (comment))
Link the netcdf files containing the coeffs_reconstruct stream within the forward step when the model is Omega
Analysis or viz involving *Zonal, *Meridional fields should use the output stream if present or reconstruct the fields using coeffs

Checklist

User's Guide has been updated
Developer's Guide has been updated
API documentation in the Developer's Guide (api.md) has any new or modified class, method and/or functions listed
Documentation has been built locally and changes look as expected
Testing comment in the PR documents testing used to verify the changes
New tests have been added to a test suite

cbegeman · 2026-01-12T22:20:35Z

@xylar Let me know what you think.

Pro of using this approach: no need to run a separate forward step besides those being run. (coeffs_reconstruct doesn't have a time dimension so I think this stream is only written once but it's worth double checking.) I was also thinking we could set write_coeffs_reconstruct = True in polaris/ocean/mpas_ocean.cfg and run whichever cases we want to have cached.

Con of using this approach: if there are multiple forward steps for a single test case (corresponding to the config file for which write_coeffs_reconstruct = True, all will produce the coeffs_reconstruct stream.

cbegeman · 2026-01-13T17:24:03Z

The plan will be to

Cache all mesh steps for test cases that support Omega (https://docs.e3sm.org/polaris/main/developers_guide/api.html#cache)
Use cached mesh steps in Omega PR and nightly suites
Add coeffs_reconstruct.nc files corresponding to each mesh to the input_database with date stamps
For any viz steps that support Omega and use reconstructed velocity fields, load coeffs_reconstruct.nc and use reconstruct_variable https://github.com/MPAS-Dev/MPAS-Tools/blob/297a1b0bf08c0d9e191c0b0e1d35ae315ee49617/conda_package/mpas_tools/vector/reconstruct.py#L2-L18

cbegeman · 2026-01-13T17:24:23Z

This PR addresses #445

…he stream

cbegeman · 2026-04-03T17:49:08Z

Input database additions

When tests have multiple resolutions, the coefficient files are saved as {res}_coeffs.nc otherwise they are named coeffs.nc

Spherical mesh files are stored at {database_root}/mesh/{prefix} where prefix is either qu or icos. A helper function parse_mesh_path has been added to polaris.mesh.base
Planar mesh files are stored at {database_root}/ocean/{task_name}

…sphere_transport viz

cbegeman · 2026-04-04T15:32:37Z

icos/rotation_2d 480km results produced by polaris/mpas_tools with coeffs for Omega:

results produced internally by MPAS-O:

cbegeman · 2026-04-07T14:32:11Z

@xylar Ready for review when you have a chance

xylar · 2026-04-07T14:36:52Z

Great, I'll make this a priority.

xylar

This looks good!

Just two suggestions for keeping the code clean.

Other than that, I think it just needs a little bit of documentation. I'll go ahead and approve on the assumption that goes in.

It doesn't look like more testing is needed on my part but let me know if you'd like some.

xylar · 2026-04-09T06:55:45Z

polaris/ocean/model/ocean_model_step.py

+            self.write_coeffs_reconstruct = False
+        if self.write_coeffs_reconstruct:
+            model = self.config.get('ocean', 'model')
+            if not model == 'mpas-ocean':


For whatever reason, I've been lead to believe this is preferred:

Suggested change

if not model == 'mpas-ocean':

if model != 'mpas-ocean':

xylar · 2026-04-09T06:56:22Z

polaris/tasks/ocean/__init__.py

+        if reconstruct_variables is None:
+            reconstruct_variables = []
+        for variable in reconstruct_variables:
+            if variable in ds.keys():
+                if 'normal' in variable:
+                    out_var_name = variable.replace('normal', '').lower()
+                else:
+                    out_var_name = variable
+                if (
+                    f'{out_var_name}Zonal' in ds.keys()
+                    and f'{out_var_name}Meridional' in ds.keys()
+                ):
+                    return ds
+                if mesh_filename is None:
+                    raise ValueError(
+                        'mesh_filename must be provided to '
+                        f'open_model_dataset for reconstruction of {variable}'
+                    )
+                ds_mesh = xr.open_dataset(mesh_filename)
+                if coeffs_filename is None:
+                    raise ValueError(
+                        'coeffs_filename must be provided to '
+                        f'open_model_dataset for reconstruction of {variable}'
+                    )
+                ds_coeff = xr.open_dataset(coeffs_filename)
+                coeffs_reconstruct = ds_coeff.coeffs_reconstruct
+                reconstruct_variable(
+                    out_var_name,
+                    ds[variable],
+                    ds_mesh,
+                    coeffs_reconstruct,
+                    ds,
+                    quiet=True,
+                )
+            else:
+                raise ValueError(
+                    f'User requested vector reconstruction for {variable} but '
+                    f"it isn't present in the dataset."
+                )


Could we move this to a helper function, just to keep things tidy?

Good idea. I'll make these changes and flesh out the docs. Thanks for reviewing!

cbegeman force-pushed the ocn/add-optional-coeffs-reconstruct-output branch from 3da3ae4 to b548c38 Compare January 12, 2026 22:21

cbegeman force-pushed the ocn/add-optional-coeffs-reconstruct-output branch from b548c38 to 36bad9c Compare January 27, 2026 00:56

cbegeman added 2 commits April 2, 2026 11:20

Use config option to determine whether to add coeffs_reconstruct to t…

8581703

…he stream

Remove later: test write coeffs

771813c

cbegeman self-assigned this Apr 3, 2026

cbegeman added ocean Related to the ocean component mesh Related to the mesh component labels Apr 3, 2026

cbegeman changed the title ~~Use config option to determine whether to add coeffs_reconstruct to the stream~~ Optionally produce coeffs_reconstruct stream for MPAS-O, link coeffs file from database for Omega Apr 3, 2026

cbegeman added 3 commits April 3, 2026 16:43

If model is Omega, link coeffs file from database

f8e5d4a

Clean-up omega_pr suite order

64da2c1

Use mpas_tools reconstruct utility in open_model_dataset and test in …

2e84e77

…sphere_transport viz

cbegeman force-pushed the ocn/add-optional-coeffs-reconstruct-output branch from 36bad9c to 2e84e77 Compare April 3, 2026 21:43

Ensure coeff reconstruction is not used when variables present

9fee43c

cbegeman marked this pull request as ready for review April 4, 2026 15:29

cbegeman requested a review from xylar April 4, 2026 15:29

xylar approved these changes Apr 9, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Optionally produce coeffs_reconstruct stream for MPAS-O, link coeffs file from database for Omega#448

Optionally produce coeffs_reconstruct stream for MPAS-O, link coeffs file from database for Omega#448
cbegeman wants to merge 6 commits intoE3SM-Project:mainfrom
cbegeman:ocn/add-optional-coeffs-reconstruct-output

cbegeman commented Jan 12, 2026 •

edited

Loading

Uh oh!

cbegeman commented Jan 12, 2026 •

edited

Loading

Uh oh!

cbegeman commented Jan 13, 2026

Uh oh!

cbegeman commented Jan 13, 2026

Uh oh!

cbegeman commented Apr 3, 2026

Uh oh!

cbegeman commented Apr 4, 2026

Uh oh!

cbegeman commented Apr 7, 2026

Uh oh!

xylar commented Apr 7, 2026

Uh oh!

xylar left a comment

Uh oh!

xylar Apr 9, 2026

Uh oh!

xylar Apr 9, 2026

Uh oh!

cbegeman Apr 9, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

cbegeman commented Jan 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cbegeman commented Jan 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cbegeman commented Jan 13, 2026

Uh oh!

cbegeman commented Jan 13, 2026

Uh oh!

cbegeman commented Apr 3, 2026

Input database additions

Uh oh!

cbegeman commented Apr 4, 2026

Uh oh!

cbegeman commented Apr 7, 2026

Uh oh!

xylar commented Apr 7, 2026

Uh oh!

xylar left a comment

Choose a reason for hiding this comment

Uh oh!

xylar Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

xylar Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

cbegeman Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

cbegeman commented Jan 12, 2026 •

edited

Loading

cbegeman commented Jan 12, 2026 •

edited

Loading