Implement POLARCAP & NATL grids + L58 inic for all other SE-VR grids

[adamrher]

Fixes #979
Fixes #1330

This PR also adds L58 ncdata files for the three other SE var-res grids (ARCTIC, ARCTICGRIS, CONUS).

This PR is awaiting the resolution of this CTSM issue: ESCOMP/CTSM#3304

[cacraigucar]

I know you haven't asked for review yet, but in case we decide to try to commit this quickly - I looked in here. You probably would have remembered these changes without me saying anything

[cacraigucar]

These files need to be moved into the /glade/p/cesmdata/inputdata section of derecho and make sure they have the "mandatory metadata" https://www2.cesm.ucar.edu/working_groups/Atmosphere/amwg_datasets.html

[adamrher]

@rcjwills please check the inputdata sets I've set up for the NATL grid (click "Files changed" tab). In particular, the ncdata file for L32. Note I generated an L58 ncdata file, interpolated from ERA5.

[rcjwills]

@adamrher With that ncdata file (cami-mam4_0000-01-01_ne0np4.NATL.ne30x8_L32_c201109.nc), an extensive spinup with substepping was needed to get the configuration running stably. The other file in the same directory (NATL_F2000_initial.cam.i.0001-01-06-00000.nc) is from after that spinup, but it has some unusual anomalies in the stratosphere. Would it make sense to output a new cam.i file from later in one of the existing simulations? I'm not sure what's done for the other grids.

[adamrher]

@rcjwills I can generate an L32 file from ERA5 using betacast. The ERA5 top is ~80 km, so it should have reasonable temperatures all the way to the model top (~40 km).

[rcjwills]

@adamrher I think that's a good idea considering that the cami-mam4_0000-01-01_ne0np4.NATL.ne30x8_L32_c201109.nc initial condition took quite some effort to get running.

[adamrher]

@cacraigucar I needed to update the externals ccs_config_cesm1.0.48 and cmeps1.1.2, which has the new NATL grid support. However, these new externals are breaking CLM at the compile stage:

Building lnd with output to /glade/derecho/scratch/aherring/cam-polarcap_FHISTC_LTso_ne0ARCTICne30x4_ne0ARCTICne30x4_mt12_2304pes_250620_test/bld/lnd.bldlog.250620-163528
/glade/campaign/cgd/amp/aherring/src/cam-polarcap/components/clm/src/biogeochem/VOCEmissionMod.F90(613): error #6460: This is not a component name that is defined in the encompassing structure. [COEFF]

Is this something to bring up with Erik at the CSEG meeting, or should we include him in this thread?

[cacraigucar]

@cacraigucar I needed to update the externals ccs_config_cesm1.0.48 and cmeps1.1.2, which has the new NATL grid support. However, these new externals are breaking CLM at the compile stage:

Building lnd with output to /glade/derecho/scratch/aherring/cam-polarcap_FHISTC_LTso_ne0ARCTICne30x4_ne0ARCTICne30x4_mt12_2304pes_250620_test/bld/lnd.bldlog.250620-163528 /glade/campaign/cgd/amp/aherring/src/cam-polarcap/components/clm/src/biogeochem/VOCEmissionMod.F90(613): error #6460: This is not a component name that is defined in the encompassing structure. [COEFF]

Is this something to bring up with Erik at the CSEG meeting, or should we include him in this thread?

@ekluzek - Do you have a suggestion on how to solve the problem that @adamrher has encountered?

[ekluzek]

Oh this is easy. You just need to update to a later CMEPS version.

There is a dependency between CTSM and CMEPS for this. Look at the cesm3_0_alpha07a alpha tag for the dependencies between things.

[adamrher]

I believe I'm running with the latest cmeps tag. Perhaps I need to update my ctsm tag (currently ctsm5.3.029)? I dont't see a cesm3_0_alpha07a tag to sync with. The latest tag I see is beta06.

[ekluzek]

@adamrher yes, to use the latest CMEPS you need at least ctsm5.3.052.

Actually alpha07a was just barely made after you checked. But, more importantly you should be familiar with the testdb to know how to look at planned tags. So go to...

https://cseg.cgd.ucar.edu/testdb/cgi-bin/login.cgi

Then you'll see the planned tags. To see alpha07a, you go to the "Alpha tags" Tab at the top navigation. For this particular alpha tag there were a bunch of submodules dependent on each other. So you might have to coordinate several submodules together.

The planned tags page can be really useful to see this information for CESM tags that are starting to be planned out. And the dependency information can be really important to figure these type of things out.

[adamrher]

Thanks for the add'l info, @ekluzek. I'll make sure to bookmark the testdb for future reference.

[adamrher]

Updating the gitmodule to ctsm5.3.052 seems to have done the trick. Everything is running (I've tested all the new grids and new inic files).

This PR is going to now be in a holding pattern until a new ctsm tag is made with the NATL grid support.