Moving FV3 internal state from PT to T

[wmputman]

FVdycoreCubed_GridComp/DynCore_GridCompMod.F90

Line 2208 in c159b3c

call MAPL_AddInternalSpec ( gc, &

When we move to the NH core as default I'd like to make the internal FV3 temperature state Dry Temperature not potential temperature. This will simplify the interface between hydrostatic and non-hydrostatic dynamics. But will of course complicate the compatibly with older restarts. The issue is that NH PKZ and the PKZ used to make scaled potential temperature in the physics differ. And the logic in DynGridComp becomes complicated.

It may be that we just require a regridded FV3 restart for all NH runs. And the regrinding routine fills T not PT.

[bena-nasa]

Pardon my ignorance but is "dry temperature" just the plain air temperature as we understand it in normal life?

[mathomp4]

Ooh boy. This might trigger GEOSgcm 11 if you change the fvcore restart! I suppose we'd need to discuss with @sdrabenh...

[wmputman]

@bena-nasa : yes
@mathomp4 : and the uproar from this would be historic

[wmputman]

The trouble is the NH pkz in FV3 used to make PT, is not the same as the hydrostatic PKZ used throughout the physics. It's best to just interface with the dynamics in dry T, and let the physics/DAS create their own PT base on a hydrostatic PKZ

[wmputman]

Initially I'm just adding T, and making the PT based on hydrostatic PKZ.