(answer changing for DGLC compsets) Update the model_datafiles_list file used for DGLC Greenland grids to remove the ice shelves

[ekluzek]

@gunterl will update the model_datafiles_list file used for Greenland from

$DIN_LOC_ROOT/glc/cism/Greenland/[greenland_4km_epsg3413_c251001.nc

to a file that removes the floating ice shelves. Once, that's done @Katetc can make a PR here pointing to that file.

[mvertens]

@ekluzek - can we please try to make this backwards compatible with the current values - which are being used by NorESM? @billsacks and I are trying to keep the two repositories close together so that merging can happen more easily.

[ekluzek]

@ekluzek - can we please try to make this backwards compatible with the current values - which are being used by NorESM? @billsacks and I are trying to keep the two repositories close together so that merging can happen more easily.

I talked to @Katetc about this last week and she'll bring it up with you the next time they talk with the NorESM group. So probably in August when Norwegians are back from July holiday.

I think what could be done is that cime_model could be added as a attribute to check for in the name list definition and a different filename used for CESM vs. NorESM.

I can see how to do that, and it's not too hard but it would be adding something new in. I can explain it to @Katetc when the time comes.

[billsacks]

I think what could be done is that cime_model could be added as a attribute to check for in the name list definition and a different filename used for CESM vs. NorESM.

While I think that's the right long-term solution, my understanding is that NorESM is still using cime_model=cesm. @mvertens is that correct? I'm struggling myself to see how to keep this backwards compatible before that change is made to have NorESM use cime_model=noresm; @mvertens do you have any suggestions? Or this may need to be another place where we have a divergence in the xml between NorESM and CESM until that flag is changed.

[mvertens]

@billsacks @ekluzek - I have just introduced a PR to ESCOMP/CMEPS - ESCOMP/CMEPS#677 - that should permit distinguishing between noresm and cesm. I am happy to carry out testing on derecho if you just tell me what tests need to be run. This is a change that was already introduced in noresm - but I did not want to bring in all of the other changes from noresm at this point.

This adds a noresm valid value to the CMEPS COUPLING_MODE xml variable and can be accessed by any component. This is already in place in noresm CMEPS.

[ekluzek]

@billsacks @ekluzek - I have just introduced a PR to ESCOMP/CMEPS - ESCOMP/CMEPS#677 - that should permit distinguishing between noresm and cesm. I am happy to carry out testing on derecho if you just tell me what tests need to be run. This is a change that was already introduced in noresm - but I did not want to bring in all of the other changes from noresm at this point.

This adds a noresm valid value to the CMEPS COUPLING_MODE xml variable and can be accessed by any component. This is already in place in noresm CMEPS.

This looks great @mvertens. Then @Katetc would just mimic the changes you made in the CMEPS PR above to CDEPS/dglc for buildnml, and namelist_definition for this issue.