RRTMGP w/ GFS SDFs + some houskeeping#181
Conversation
| @@ -1850,12 +1850,6 @@ module GFS_typedefs | |||
| real (kind=kind_phys), pointer :: cld_resnow(:,:) => null() !< Cloud snow effective radius | |||
There was a problem hiding this comment.
@dustinswales I don't see these same changes in NCAR/fv3atm#46. Were they previously committed there? We try to keep the two GFS_typedefs as close to each other as possible.
There was a problem hiding this comment.
@grantfirl
You are correct. I did some cleanup in here to the RRTMGP interstitial fields. I will do the same on the fv3atm side.
| @@ -4,5 +4,5 @@ | |||
| branch = dtc/develop | |||
There was a problem hiding this comment.
After testing, we'd like to see this reverted.
|
SCM results, both SDFs, twpice configuration. RRTMGP (blue), RRTMG (red) |
…se_LW_jacobian Conflicts: ccpp/config/ccpp_prebuild_config.py ccpp/physics
…updated correctly.
OK, let me know if you need anything. I'm happy to help, since this is required to have the master SCM code work with ccpp-physics/master. |
Conflicts: scm/src/default_namelists.py
…velop Conflicts: ccpp/config/ccpp_prebuild_config.py scm/src/GFS_typedefs.F90 scm/src/default_namelists.py scm/src/gmtb_scm_input.F90
|
@grantfirl |
OK, I'll pull in your changes and see what I can do. If I find something that I can fix, I'll submit a PR from my fork to yours. I'm also just about ready to merge in #191, so I'll update your branch for that too. It shouldn't affect anything from your PR since it's just adding new cases. |
…GP update from Dustin
…to run successfully
…with ccpp_prebuild
|
@dustinswales A quick update: starting from this branch, I now have a branch that is working OK. I'm building/compiling one last time to double-check, and then I'll submit a PR from my branch to this one for you to review. Where necessary, I took my lead from the associated changes in FV3 (like SDFs and ccpp_prebuild_config.py). You may want to pay particular attention to changes in those to make sure that I updated them as you intended. |
|
If you accept the PR from my fork to yours, I'd like to have you remove the Apollo setup script before I accept/merge. I think that the code managers want to keep only setup scripts for officially-supported machines. |
Grant's changes to PR#181
|
@grantfirl |
grantfirl
left a comment
There was a problem hiding this comment.
Thanks @dustinswales, approved. I think that this is good to merge on my side. I'll ask @climbfuji if he wants to take a look before merging.
climbfuji
left a comment
There was a problem hiding this comment.
Two minor comments, but probably just my lack of understanding.
| ! | ||
| if (Model%do_RRTMGP) then | ||
| if (Model%use_LW_jacobian) then | ||
| !Interstitial%fluxlwUP_jac = clear_val |
There was a problem hiding this comment.
Why commented out? Same for line 6994.
|
I'll go ahead and merge. If @climbfuji would rather me remove the ACM SDF altogether, I can do that in a followup. |
RRTMGP w/ GFS SDFs + some houskeeping
This PR contains the necessary changes to work w/ ccpp-physics PR#446 (NCAR/ccpp-physics#446).
Similar to that PR, there were some changes to the names of the radiation heating rates variables, so there is some cleaning up in GFS_typedefs addressing this.