Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes #525

ytakano3 · 2023-02-08T06:56:41Z

This merge introduces 2 meshes used in Yohei's research on the Kuroshio and the northeast Pacific.

It also includes some reorganization of how config options are handled in the global_ocean test group to handle a set of shared config options for the 2 meshes as well as a small number of unique config options for each one. Rather than having a configure() function that was called in various configure() methods, we now call the mesh's configure() method from the init's configure() method to get additional config options (e.g. metadata) related to the mesh, and we call init's configure() from forward test cases (performance, dynamic adjustment, restart test, etc.) in order to get metadata about both the mesh and initial condition. This turns out to be a more elegant and flexible approach where the info supplied is in a more intuitive location (mesh info in the Mesh test case; init info in the Init test case).

Checklist

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

xylar · 2023-02-08T23:27:06Z

@ytakano3, I have completed all the test cases for the 12to60 mesh here:

/lcrc/group/e3sm/ac.xylar/compass_1.2/chrysalis/test_20230207/kuroshio12/ocean/global_ocean/Kuroshio12to60

Could you take a look at the results and make sure they look quite similar to what you have seen previously? Ideally, we would make sure the relevant config and namelist options are the same as what you ran with before but that may be more work than either of us has time for. So I'm hoping a look at the mesh and the output after dynamic adjustment will be sufficient.

I am still working on the 8to60 dynamic adjustment, which takes quite a bit longer.

xylar · 2023-02-08T23:28:51Z

docs/developers_guide/ocean/test_groups/global_ocean.rst

+The ``Kuroshio8to60`` and ``Kuroshio12to60`` mehses are designed to explore
+dynamics of the Kuroshio Current.
+
+The class
+:py:class:`compass.ocean.tests.global_ocean.mesh.kuroshio.KuroshioBaseMesh`
+defines the resolution for the meshes, where the finest resolution comes from
+the ``min_res`` config option in the ``[global_ocean]`` section of the config
+file.
+
+The ``compass.ocean.tests.global_ocean.mesh.kuroshio8to60`` and
+``compass.ocean.tests.global_ocean.mesh.kuroshio12to60`` modules include
+namelist options appropriate for forward simulations with split-explicit (but
+not RK4) time integration on this mesh.  These set the time step and default
+run duration for short runs with this mesh.
+
+Except for ``min_res``, default config options for these meshes come from a
+shared config file in the ``compass.ocean.tests.global_ocean.mesh.kuroshio``
+module:


@ytakano3, please check this documentation and suggest modifications here if you would like (using the "Add suggestions" button with the +/-).

@xylar Minor comment (change) added.

xylar · 2023-02-08T23:29:34Z

docs/users_guide/ocean/test_groups/global_ocean.rst

+The Kuroshio 8- and 12- to 60-km meshes (Kuroshio8to60 and Kuroshio12to60) are
+designed to explore dynamics of the Kuroshio Current.
+
+The meshes have 8 and 12 km resolution, respectively, in the eastern
+North Pacific, tapering to 60 km at mid latitudes, 30 km at the equator, and
+35 km in polar regions (the same as the EC30to60).


@ytakano3, same here. Please make any suggested changes here with the "Add a suggestion" button.

@xylar Minor comment (change) added.

docs/users_guide/ocean/test_groups/global_ocean.rst

xylar · 2023-02-09T16:54:51Z

Testing

I ran all test cases for both meshes on Chrysalis with Intel and OpenMPI. Results are in these directories:

/lcrc/group/e3sm/ac.xylar/compass_1.2/chrysalis/test_20230207/kuroshio12/ocean/global_ocean/Kuroshio12to60
/lcrc/group/e3sm/ac.xylar/compass_1.2/chrysalis/test_20230207/kuroshio8/ocean/global_ocean/Kuroshio8to60

I also checked the pr test suite using main as a baseline (and using the same build of MPAS-Ocean from the E3SM-Project submodule from main). Everything (importantly the global_ocean tests) passed and were bit-for-bit.

xylar · 2023-02-09T16:56:35Z

@ytakano3, can you take a look at the output in the directories above and make sure it looks similar to (hopefully even identical to) what you have already been using as an initial condition? That's the main thing I need from you as a review. Can you do that in the next week or so?

I would like to get this in.

docs/developers_guide/ocean/test_groups/global_ocean.rst

docs/users_guide/ocean/test_groups/global_ocean.rst

xylar · 2023-02-09T23:25:55Z

@ytakano3, thank you for your comments. I believe I have incorporated all of them. I will rerun the full set of tests now and hopefully they will be there for you to look at tomorrow:

/lcrc/group/e3sm/ac.xylar/compass_1.2/chrysalis/test_20230209/kuroshio12/ocean/global_ocean/Kuroshio12to60
/lcrc/group/e3sm/ac.xylar/compass_1.2/chrysalis/test_20230209/kuroshio8/ocean/global_ocean/Kuroshio8to60

xylar · 2023-02-17T20:52:42Z

@ytakano3, any headway on looking at this output?

ytakano3 · 2023-02-17T21:54:24Z

@xylar, thank you for updating. I am looking at the restart.nc from compass run now for both 12km and 8km meshes. Will update on this soon.

xylar · 2023-02-21T21:53:22Z

@ytakano3, just a warning that I'm going to do one more rebase on this branch to take care of some conflicts with #474

Only the mesh generation, need to work on dynamical adjustment and namelist.

Instead of having a shared function, have forward test cases call `init.configure` and have `init` call `mesh.configure`. Each can pass an config object to modify.

ytakano3 · 2023-02-21T22:29:39Z

@xylar Thank you for the rebase warning. I have checked the sea surface temperature and salinity (sst and sss from compass (your version and my version). The global cell numbers differ slightly, here is the summary (cell numbers no bathymetry)

           Kuroshio 12km     Kuroshio 8km
 Xylar       322685             457263
 Yohei       322637             457732

and this is the comparison of sst and sss.

Here is a closer look in Kuroshio regions, regional cell numbers and figures

           Kuroshio 12km     Kuroshio 8km
 Xylar       114747             229907
 Yohei       114791             230231

Despite slight differences in cell numbers, I think two fields looks very close but what do you think?

xylar · 2023-02-21T22:37:00Z

@ytakano3, that looks very good to me. I'm going to run one more test following the rebase to make sure I didn't break anything. Then, I'm going to approve and merge.

xylar

I did one last test with this branch and the pr suite is bit-for-bit with main so I think the rebase went fine.

xylar · 2023-02-21T23:58:31Z

Thanks for working on this @ytakano3!

xylar changed the title ~~Add kuroshio 8 and 12 meshes~~ Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes Feb 8, 2023

xylar self-assigned this Feb 8, 2023

xylar self-requested a review February 8, 2023 15:22

xylar added new mesh An E3SM mesh for special review ocean labels Feb 8, 2023

xylar mentioned this pull request Feb 8, 2023

Fix nonhydro documentation #526

Merged

5 tasks

xylar force-pushed the add-kuroshio-8-and-12-meshes branch 3 times, most recently from e657a92 to 974884d Compare February 8, 2023 23:06

xylar reviewed Feb 8, 2023

View reviewed changes

docs/users_guide/ocean/test_groups/global_ocean.rst Outdated Show resolved Hide resolved

ytakano3 commented Feb 9, 2023

View reviewed changes

docs/developers_guide/ocean/test_groups/global_ocean.rst Outdated Show resolved Hide resolved

ytakano3 commented Feb 9, 2023

View reviewed changes

docs/users_guide/ocean/test_groups/global_ocean.rst Outdated Show resolved Hide resolved

xylar force-pushed the add-kuroshio-8-and-12-meshes branch from 4de483a to e3f1347 Compare February 9, 2023 23:41

xylar force-pushed the add-kuroshio-8-and-12-meshes branch 2 times, most recently from 66b06fb to dfe7d95 Compare February 17, 2023 23:32

ytakano3 added 4 commits February 21, 2023 14:57

First test example for Kuroshio mesh.

acf04ee

Only the mesh generation, need to work on dynamical adjustment and namelist.

Add Kuroshio mesh to global ocean and mesh test case

86b49b9

Update on mesh parameters

938c68d

Add remaining steps for Kuroshio mesh

cc9b450

xylar force-pushed the add-kuroshio-8-and-12-meshes branch from dfe7d95 to c54c947 Compare February 21, 2023 22:09

ytakano3 and others added 15 commits February 21, 2023 15:22

Generalize Kuroshio for 12km and 8km meshes

bf084ff

Changing timestep and del2 and del4.

a45533a

Updated new Kuroshio mesh for 12km and 8km

41f4e51

Add new geojson files for Kuroshio mesh

579f378

Change how global ocean test cases are configured

64d1415

Instead of having a shared function, have forward test cases call `init.configure` and have `init` call `mesh.configure`. Each can pass an config object to modify.

Fix names of init and forward config options

e4c840a

Fix API docs following reorg of global ocean config

1b57c76

Fix several class names in global ocean docs

8d870db

Put in pull request number for inclusion in metadata

7712cc6

Add Kuroshio meshes to user's and developer's guides

1595155

Add test suites for Kuroshio meshes

ce515f6

Fix some issues with the docs

f6f041b

Change Kuroshio image

695da11

Update Kuroshio namelist options

8eace9d

Fix formatting with pre-commit

07af372

xylar force-pushed the add-kuroshio-8-and-12-meshes branch from c54c947 to 07af372 Compare February 21, 2023 22:35

xylar approved these changes Feb 21, 2023

View reviewed changes

xylar merged commit 01a1a41 into MPAS-Dev:main Feb 21, 2023

xylar deleted the add-kuroshio-8-and-12-meshes branch February 21, 2023 23:58

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes #525

Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes #525

ytakano3 commented Feb 8, 2023 •

edited by xylar

Loading

xylar commented Feb 8, 2023

xylar Feb 8, 2023

ytakano3 Feb 9, 2023

xylar Feb 8, 2023

ytakano3 Feb 9, 2023

xylar commented Feb 9, 2023

xylar commented Feb 9, 2023

xylar commented Feb 9, 2023

xylar commented Feb 17, 2023

ytakano3 commented Feb 17, 2023

xylar commented Feb 21, 2023

ytakano3 commented Feb 21, 2023 •

edited

Loading

xylar commented Feb 21, 2023

xylar left a comment

xylar commented Feb 21, 2023

Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes #525

Add Kuroshio8to60E2r4 and Kuroshio12to60E2r4 meshes #525

Conversation

ytakano3 commented Feb 8, 2023 • edited by xylar Loading

xylar commented Feb 8, 2023

xylar Feb 8, 2023

Choose a reason for hiding this comment

ytakano3 Feb 9, 2023

Choose a reason for hiding this comment

xylar Feb 8, 2023

Choose a reason for hiding this comment

ytakano3 Feb 9, 2023

Choose a reason for hiding this comment

xylar commented Feb 9, 2023

Testing

xylar commented Feb 9, 2023

xylar commented Feb 9, 2023

xylar commented Feb 17, 2023

ytakano3 commented Feb 17, 2023

xylar commented Feb 21, 2023

ytakano3 commented Feb 21, 2023 • edited Loading

xylar commented Feb 21, 2023

xylar left a comment

Choose a reason for hiding this comment

xylar commented Feb 21, 2023

ytakano3 commented Feb 8, 2023 •

edited by xylar

Loading

ytakano3 commented Feb 21, 2023 •

edited

Loading