Add new IcosXISC30E3r7 ocean and sea-ice mesh

[xylar]

This mesh is identical to the IcoswISC30E3r5 mesh except that ice-shelf cavities have been culled (eXcluded).

The spun-up initial condition files (mpaso.IcosXISC30E3r7.rstFromPiControlSpinup-chrysalis.20240314.nc and mpassi.IcosXISC30E3r7.rstFromPiControlSpinup-chrysalis.20240314.nc) are culled versions of the restart files from 2001-01-01 of the 20231209.v3.LR.piControl-spinup.chrysalis simulation (4032626741/20231209.v3.LR.piControl-spinup.chrysalis) with the xtime variable removed. These files have been culled with the tools added in https://github.com/MPAS-Dev/MPAS-Tools/pull/557 to MPAS-Tools and in https://github.com/MPAS-Dev/compass/pull/794 to Compass. The intention is to continue from these files so that we can compare with the v3.LR.piControl run.

The remaining support files have been created by running files_for_e3sm from Compass on the culled MPAS-Ocean restart file. Thus, the ocean initial condition in mpaso.IcosXISC30E3r7.20240314.nc should be identical, while the sea-ice initial condition in mpassi.IcosXISC30E3r7.20240314.nc will just be the mesh variables and so on that allow sea ice to start from either a circular disk or no sea ice. We do not intend to make use of these initial conditions, just the "spun-up" versions.

[xylar]

@jonbob, @proteanplanet, @cbegeman and @darincomeau, I have staged these files in /lcrc/group/e3sm/data/inputdata. Please have a look in the ocn/mpas-o/IcosXISC30E3r7 and ice/mpas-seaice/IcosXISC30E3r7 subdirectories for the corresponding files.

@jonbob, unfortunately, this mesh will need new mapping and domain files. This does not need to be a fully supported mesh as far as I understand it. We only need mapping and domain files to be able to run a B-case that is essentially identical to the v3.LR.piControl run for this mesh.

[cbegeman]

@xylar Is velocityTidalRMS_CATS2008.IcosXISC30E3r7.20240314.nc included for completeness? My understanding was that these files are only used for thermodynamically active cavities

[xylar]

@cbegeman, I believe that file is defined both inside and outside of cavities. It is not yet used anywhere but my understanding was that it would be useful to generate that file for all meshes, not just those with cavities, in case we wanted to use the tidal information elsewhere in the future.

And, yes, in this case, I just generated all the Compass files for E3SM support because that was easier than trying to figure out which were needed and which were not.

[cbegeman]

@xylar Sounds good. Just checking

[xylar]

Hmm, I don't see any mention of that discussion in https://github.com/MPAS-Dev/compass/pull/770, so maybe it was a mistake to include it for meshes without cavities. I'll check with @irenavankova to see what she thinks.

[cbegeman]

@xylar I looked through the restart files, specifically the min and max values of each ocean variable in the culled and unculled restarts. All the landIce* variables look good. I noticed high salinity values of 52 PSU in a shallow water column in the tropics in both meshes, so presumably just due to evap.

I'll wait on the rest of my review until after we have at least a short simulation with the new mesh.

[irenavankova]

The tidal file is only used in thermodynamically active cavities, but it is true that if other tidal parameterizations are implemented the same tidal file would likely be used in those cases as well and that would apply outside of cavities. However, the tidal file would have to be modified, as this one only includes Antarctica from the CATS2008 model

[xylar]

@xylar I looked through the restart files, specifically the min and max values of each ocean variable in the culled and unculled restarts. All the landIce* variables look good. I noticed high salinity values of 52 PSU in a shallow water column in the tropics in both meshes, so presumably just due to evap.

@cbegeman, thanks for having a look. That high salinity value sounds odd but outside the scope of this particular project so we probably just want to keep it in the back of our minds as a possible issue to chase in the future.

I'll wait on the rest of my review until after we have at least a short simulation with the new mesh.

Yes, sounds good.

[xylar]

The tidal file is only used in thermodynamically active cavities, but it is true that if other tidal parameterizations are implemented the same tidal file would likely be used in those cases as well and that would apply outside of cavities. However, the tidal file would have to be modified, as this one only includes Antarctica from the CATS2008 model

@irenavankova, thanks, those are great points. I made a Compass PR that I'll merge sortly to only produce these files in the future for meshes with cavities: https://github.com/MPAS-Dev/compass/pull/798

[jonbob]

OK, I added domain and mapping files to support this new mesh in limited configurations:

• TL319_IcosXISC30E3r7
• ne30pg2_r05_IcosXISC30E3r7

I ran a one-month G-case as a test and used the restart files to make alternate IC files for the B-case, if needed. All necessary files are staged in the local lcrc inputdata directory. Please let me know if you need anything else

[xylar]

Wonderful, thank you @jonbob! We'll keep you posted and ask questions as they arise.

[jonbob]

Sorry about the mess with my scripts -- the new mapping and domain files are replaced in this PR and on lcrc

[jonbob]

I ran ncks --chk_map on all the new mapping files and they all look good except

map_TL319_to_IcosXISC30E3r7_trbilin.20240326.nc

I checked some other mapping files for TL319 and they all seem to have a similar issue, so maybe conservation metrics make no sense for this situation

[xylar]

Yes, I think that would be expected for bilinear maps both because they're not conservative by design and because we do need to be renormalizing them.

[darincomeau]

Comparison analysis of 10 year run using the culled ocn/ice restarts from the end of the v3 picontrol-spinup with v3 piControl: MPAS-Analysis: https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.dcomeau/E3SMv3/v3.LR-IcosXISC30E3r7_vs_LR.piControl.chrysalis/yrs1-10/ E3SM-diags: https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.dcomeau/E3SMv3/e3sm_diags_output/v3.LR-IcosXISC30E3r7_vs_LR.piControl.chrysalis.YRS1-10/viewer/

[xylar]

closed in favor of https://github.com/E3SM-Project/E3SM/pull/6312