Move Noah MP init to CCPP and add new variables for Noah MP

[climbfuji]

Description

This PR does several things:

1. Move the Noah MP initialization of surface fields from fv3atm (io/FV3GFS_io.F90) to CCPP from @climbfuji.
2. Updates to Noah MP and Noah MP - radiation coupling from @HelinWei-NOAA. The changes in ccpp-physics require adding four additional variables in GFS_typedefs.F90.
3. Revert a change in ccpp/data/GFS_typedefs.F90, made as part of the UGWP update PR (UGWP v0 v1 combined #241), that permanently disabled the old GFS v12-v15 convective gravity wave drag scheme. @SMoorthi-emc @ValeryYudin-NOAA please check, around line 4030 in the updated GFS_typedefs.F90.
4. Cleanup work in GFS_typdefs.F90 for the recently merged RRTMGP updates: Add Interstitial%gas_concentrations%concs(:)%conc to the Interstitial%rad_reset call, remove RRTMGP variables that no longer exist from GFS_typedefs.meta, add missing active flags to RRTMGP variables in GFS_typedefs.meta

Issue(s) addressed

This finally completes step 2 of #214.

Fixes #206.

Testing

For regression testing, see ufs-community/ufs-weather-model#425.

Dependencies

NCAR/ccpp-physics#575
#249
ufs-community/ufs-weather-model#425

[junwang-noaa]

In the compile log file, the ccpp_static_api.F90 with ccpp_physics_init and ccpp_physics_run are build in FV3/ccpp/driver/libccppdriver.a, not in FV3/ccpp/framework/src/libccpp.a. So it seems not from ccpp framework

…

On Wed, Feb 24, 2021 at 9:20 AM Dom Heinzeller ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In CMakeLists.txt <#249 (comment)>: > @@ -109,7 +108,9 @@ target_include_directories(fv3dycore INTERFACE $<BUILD_INTERFACE:${CMAKE_CURRENT $<INSTALL_INTERFACE:mod>) target_link_libraries(fv3dycore PUBLIC fms - gfsphysics The dycore does depend explicitly on ccpp=ccppframework, because it uses call ccpp_physics_init etc. — You are receiving this because your review was requested. Reply to this email directly, view it on GitHub <#249 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TJYA4JPL3KPDXFBO7TTAUDK5ANCNFSM4X67L6LQ> .

[climbfuji]

But they use the ccpp type ccpp_t, which is defined in FV3/ccpp/framework/src/ccpp_types.F90 !

…

On Feb 24, 2021, at 7:34 AM, Jun Wang ***@***.***> wrote: In the compile log file, the ccpp_static_api.F90 with ccpp_physics_init and ccpp_physics_run are build in FV3/ccpp/driver/libccppdriver.a, not in FV3/ccpp/framework/src/libccpp.a. So it seems not from ccpp framework On Wed, Feb 24, 2021 at 9:20 AM Dom Heinzeller ***@***.***> wrote: > ***@***.**** commented on this pull request. > ------------------------------ > > In CMakeLists.txt > <#249 (comment)>: > > > @@ -109,7 +108,9 @@ target_include_directories(fv3dycore INTERFACE $<BUILD_INTERFACE:${CMAKE_CURRENT > $<INSTALL_INTERFACE:mod>) > > target_link_libraries(fv3dycore PUBLIC fms > - gfsphysics > > The dycore does depend explicitly on ccpp=ccppframework, because it uses call > ccpp_physics_init etc. > > — > You are receiving this because your review was requested. > Reply to this email directly, view it on GitHub > <#249 (comment)>, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AI7D6TJYA4JPL3KPDXFBO7TTAUDK5ANCNFSM4X67L6LQ> > . > — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#249 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RJGNGPE6CXGEY5NQQLTAUE7PANCNFSM4X67L6LQ>.

[climbfuji]

use CCPP_data,         only: ccpp_suite,          &
                                cdata => cdata_tile, &
                                CCPP_interstitial

where cdata is from ccpp-framework (ccpp_t). Is this a direct or indirect dependency? Whatever, I'll remove the one line.

[junwang-noaa]

I think that is implicit, fv3 dycore itself only needs ccpp_data and ccpp_static_api

…

On Wed, Feb 24, 2021 at 9:39 AM Dom Heinzeller ***@***.***> wrote: use CCPP_data, only: ccpp_suite, & cdata => cdata_tile, & CCPP_interstitial where cdata is from ccpp-framework (ccpp_t). Is this a direct or indirect dependency? Whatever, I'll remove the one line. — You are receiving this because your review was requested. Reply to this email directly, view it on GitHub <#249 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TNO4KF3UT7IGNHISJDTAUFQ3ANCNFSM4X67L6LQ> .

[climbfuji]

I think that is implicit, fv3 dycore itself only needs ccpp_data and ccpp_static_api
…

Agreed, I'll make this change and address @barlage's comments to the ccpp-physics PR when it these PRs are the next to commit.

[climbfuji]

@dustinswales I added the rad_reset call for the gas concentrations that was missing in the last PR, and fixed a few RRTMGP issues in GFS_typedefs.meta. Please check commit 1d9c848. Thanks!

[SMoorthi-emc]

I assume the changes in radiation_surface are applicable only to NoahMP.

[climbfuji]

I assume the changes in radiation_surface are applicable only to NoahMP.

Yes, they are only used when the ialbflg parameter is set to 2. All regression tests that do not use Noah MP pass against the existing baselines with those changes.