Create coarser grids for SE dycore and use for regression testing

cacraigucar commented 1 year ago

Issue Type

Other (please describe below)

Issue Description

Currently we have ne30pg3_ne30pg3_mg17 which is good for science runs, but not coarse enough in both time and output file sizes for routine regression testing.

It has been determined that the coarsest grid that can be made is ne3pg3 and this could be used to perform regression testing, much like the way f10 was used for testing the FV dycore.

@adamrher suggested that we have the following grids: ne3pg3_ne3pg3_mg37 <---we have to get this installed on our end first ne5pg3_ne5pg3_mg37 ne16pg3_ne16pg3_mg17

Other than the ne3pg3 grid, we only need CTSM support (i.e, a CTSM PR), which requires two additional files for each grid. @adamrher will take care of generating all the ne3pg3 files -- this will require a CIME PR (or whatever externals is responsible for ccs_config/) and a CAM PR, in addition to the CTSM PR for land model support.

Refer to the spreadsheet https://docs.google.com/spreadsheets/d/1Osq56e423CF107zhoNQ0VS7-iH_JXLF9AtCvBdXyfJ4/edit#gid=0 for possible grids and CTSM support

Will this change answers?

Yes

Will you be implementing this yourself?

No

adamrher commented 1 year ago

All ne3pg3 files have been created to complete the CAM side of this work and are located here:

/glade/work/aherring/grids/uniform-res/ne3np4.pg3/

PeterHjortLauritzen commented 1 year ago

Here is a new ne3pg3 topo file with the topo bug fix:

/glade/p/cgd/amp/pel/topo/files/ne3pg3_gmted2010_modis_bedmachine_nc0540_Laplace1000_noleak_20230209.nc

adamrher commented 1 year ago

Thanks, peter. I've added that to my ne3np4.pg3 repo in the comment above.

I think we need https://github.com/ESMCI/ccs_config_cesm/issues/76 completed and then update our new externals before we can issue a CAM PR for this new grid.

cacraigucar commented 10 months ago

Coarse grid files were completed some time ago - closing

ESCOMP / CAM