Thanks @adamrher. We're in the process of testing and generating new surface datasets again for an upcoming CTSM5.3 tag that will be the final science tag for CESM3 development. Do you need these polar grids for work now, or can we wait to bring them in with the 5.3 series?

Also, I wonder if the tools used to generate this grid should be incorporated into the mksurfdata_esmf workflow?

By the grids document we should support: 1979,1850,2000,1979-2026 surface datasets. But, we should separate the mksurfdata_esmf part of this from this part for CTSM support.

Things to do here:

• Rebase to the cesm3_0_beta04_changes branch and work on this PR after Fix failing ne0CONUS, ne0ARCTIC tests #2805
• Resolve conflicts in namelist_defaults_ctsm
• Remove the clm5_1 physics support
• Remove the clm5_0 physics support
• Add the new grid to the build-namelist tester
• Add some tests for this grid to ctsm_sci
• Move new datasets into CESM inputdata and rimport them
• Make an issue to make these datasets in ctsm5.3.0 -- Make polarcap surface datasets in ctsm5.3.0, remove VR 1850 datasets #2720
• Run aux_clm and ctsm_sci test lists

@adamrher I'm thinking we should just support this for clm6_0 with cam7.0. With support for cam6.0 being next in priority. This would mainly be for IC file support, and users can still use other options, they'll just need to figure out what IC file to use.

By the grids document we should support: 1979,1850,2000,1979-2026 surface datasets.

@ekluzek in hindsight, I don't think we need 1850 support for the VR grids. And I noticed in namelist_defaults_ctsm.xml, that the fsurdat entries for 1850 are set to the 1979 fsurdat file for the ARCTIC,ARCTICGRIS VR grids, e.g.,

<fsurdat hgrid="ne0np4.ARCTIC.ne30x4"     sim_year="1850">
lnd/clm2/surfdata_esmf/ctsm5.2.0/surfdata_ne0np4.ARCTIC.ne30x4_hist_1979_78pfts_c240425.nc</fsurdat>

Which I replicated with the POLARCAP grid in this PR. Should we just modify these entries to sim_year="1979" so they are consistent with the file pointed to?

@adamrher thanks for letting us know about not needing 1850. I removed it for the other VR grids in the grids spreadsheet. #2498 gets into supporting these 1979 datasets in a clearer more sustainable way. So we'll leave it like it is here for now.

@adamrher we have new surface dataset files for this grid as part of ctsm5.3.0 here:

/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/surfdata_esmf/ctsm5.3.0/surfdata_ne0np4.POLARCAP.ne30x4_hist_1979_78pfts_c240908.nc
/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/surfdata_esmf/ctsm5.3.0/landuse.timeseries_ne0np4.POLARCAP.ne30x4_SSP2-4.5_1979-2026_78pfts_c240908.nc

We haven't tested them though, and won't until I get back to this PR after ctsm5.3.0 is done. I have some important cesm3_0_beta03 work that will need to be before this for next week. But, if you'd like to try it yourself before I get to it, we could talk about how you could do that after ctsm5.3.0 is finished. Let me know next week if you'd like to do that...

@adamrher would you be ok giving me permission to push to this PR? You would have to:

• Click Settings
• Click Collaborators
• Click Add people

If not, I would open a new PR to finish this work. Let me know your preference. Thanks, Adam.

@ekluzek are we still removing clm50 support here? I see a comment that I introduced very recently in our testlist:
Clm50 because that is what CAM tests currently; remove when CAM tests for Clm60

Testing on derecho:
PASS ./build-namelist_test
PASS python testing
PEND aux_clm and ctsm_sci after corrections that I'm about to push

derecho testing complete:
OK aux_clm
OK ctsm_sci

izumi:
OK aux_clm: roundoff diffs from the baseline due to this week's node upgrades; also a few tests decided that cesm.exe does not exist though it does; this may or may not have something to do with the upgrade; removing cesm.exe and rebuilding in one of the tests did not help.