search

NCAR / ccpp-scm

CCPP Single Column Model
Other
13 stars 50 forks source link

UFS-dev PR#64 #381

Closed grantfirl closed 1 year ago

grantfirl commented 1 year ago

Contains changes analogous to https://github.com/NCAR/fv3atm/pull/94 for working with https://github.com/NCAR/ccpp-physics/pull/1015

dustinswales commented 1 year ago

@grantfirl Looks like the RAP twpice test differs, which is expected, but not the other RAP tests?

grantfirl commented 1 year ago

@grantfirl Looks like the RAP twpice test differs, which is expected, but not the other RAP tests?

Ya, it is a bit surprising that the arm_sgp_summer_1997_A case does not also differ with the RAP suite because it is the only other case that features deep convection and would have a chance at triggering the GF code changes. But, I'm assuming that whatever convection occurs in the SGP case is not strong enough to trigger the code changes, perhaps?

dustinswales commented 1 year ago

That's what I suspect. I don't have a sense for the forcing in the all of the test cases or how this change manifests itself within the convection. I know we are only running 0.1x of these cases for our tests, and TWPICE is the longest test, which maybe is why it's triggering the convection?

grantfirl commented 1 year ago

That's what I suspect. I don't have a sense for the forcing in the all of the test cases or how this change manifests itself within the convection. I know we are only running 0.1x of these cases for our tests, and TWPICE is the longest test, which maybe is why it's triggering the convection?

Ya, that's yet another reason for the ARM case to not differ that I always forget (that we're only running 10% of the case model time), since I think the deep convection probably doesn't happen right away.