search

NCAR / ccpp-scm

CCPP Single Column Model
Other
13 stars 50 forks source link

UFS/dev PR#34 (RRTMGP refactoring) #367

Closed grantfirl closed 1 year ago

grantfirl commented 1 year ago

update typedefs, SDFs, submodule pointers, and ccpp_prebuild_config.py to work with ccpp-physics ufs-dev/PR#34

grantfirl commented 1 year ago

ccpp_prebuild CI test fails due to known issue (flag_for_vertical_ordering_in_rrtmgp is accidentally used in RRTMGP and is fixed in https://github.com/NCAR/ccpp-physics/pull/1001)

grantfirl commented 1 year ago

@dustinswales It looks like some RTs unexpectedly differ. How are we supposed to update the baselines again? I'm wondering if they just weren't updated from the last merged PR?

dustinswales commented 1 year ago

@dustinswales It looks like some RTs unexpectedly differ. How are we supposed to update the baselines again? I'm wondering if they just weren't updated from the last merged PR?

@grantfirl

ccpp_prebuild CI test fails due to known issue (flag_for_vertical_ordering_in_rrtmgp is accidentally used in RRTMGP and is fixed in NCAR/ccpp-physics#1001)

Correct, I changed flag_for_vertical_orderinginrrtmgp flag_for_vertical_orderinginradiaiton in ufs/dev #38

dustinswales commented 1 year ago

@dustinswales It looks like some RTs unexpectedly differ. How are we supposed to update the baselines again? I'm wondering if they just weren't updated from the last merged PR?

Yeah, it's the same two tests that differ from the last PR. To update the baselines I upload the "github artifacts" to the DTC ftp site when they need updating (This is the only piece not controlled by CI, which I can't figure out how to automate).