Updates for component land model #682
Conversation
|
Testing now. |
|
@haiqinli If you don't mind could you also test in your side. I have just merged your ccpp/physics and FV3 forks to this branch. |
|
Sure, you are welcome! - Haiqin
…On Wed, Jan 24, 2024 at 1:51 PM Ufuk Turunçoğlu ***@***.***> wrote:
@haiqinli <https://github.com/haiqinli> If you don't mind could you also
test in your side. I have just merged your ccpp/physics and FV3 forks to
this branch.
—
Reply to this email directly, view it on GitHub
<#682 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AJHAAOFPKWFKMF4TT4K2ZULYQFX43AVCNFSM6AAAAAA23ATSFWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMBYHA4TINZTG4>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
ufs-community/ccpp-physics#94 is merged. This is ready to update the ccpp/physics submodule pointer and revert .gitmodules. |
|
@grantfirl okay. done it. |
|
|
@zach1221 let me check |
|
@zach1221 sorry, I think it is done. |
|
@jkbk2004 this fv3atm sub-pr should be good to go. Can you please merge for us? |
|
@jkbk2004 We need to ask ufuk to first merge the top of FV3, which contains documentation updates. |
@uturuncoglu there was a recent commit at fv3 develop. Can you sync up one more time? |
|
@jkbk2004 Okay, i synced. BTW, I still see changes in ccpp/physics.If you don't mind could you double check that all are fine. |
I think changes just show diff against fv3/dev. Looks ok. I will let CI done and merge. |
|
@jkbk2004 Okay. That is great. BTW, thanks for all your help. |
|
@uturuncoglu Sorry to resurrect this discussion from the dead, but I had a question about your updates to the CCPP suite suite_FV3_GFS_v17_p8.xml: should the |
|
@mkavulich I am not sure which suite file is used by GFS v17 but currently we have only in . That is right. You could add it to any suite file and it will update the land fluxes onlycpllnd2atm=.true. case.
|
Description
This PR aims to add support for fully coupled atmosphere-land component configuration.
cpllnd2atmto keep supporting side-by-side configuration (both component NoahMP and CCPP/NoahMP runs and component NoahMP does not send information to FV3, it is important particularly to compare the results coming from component land with the one that runs under CCPP/Physics).Issue(s) addressed
N/A
Testing
How were these changes tested?
The full regression test suite is run on Orion and all tests passed except land component related ones since new land component rely on newer version of ESMF (8.5.0) to read/write multi-tile netcdf files.
What compilers / HPCs was it tested with? Intel / MSU's Orion
Are the changes covered by regression tests? (If not, why? Do new tests need to be added?)
Following new RTs are added,
control_p8_atmlnd: fully coupled ATM-LND configurationcontrol_restart_p8_atmlnd: tests restart capability of fully coupled ATM-LND configurationdatm_cdeps_lnd_era5: component land model forced with ERA5datm_cdeps_lnd_era5_rst: tests restart capability of component land model forced withHave the ufs-weather-model regression test been run? On what platform?
Once UFS Weather model switches to new version of ESMF (8.5.0) there might have some answer changes due to some changes in ESMF interpolation routines. The full release note can be seen in the following link: https://github.com/esmf-org/esmf/releases/tag/v8.5.0
Dependencies
If testing this branch requires non-default branches in other repositories, list them.
Those branches should have matching names (ideally) N/A
Do PRs in upstream repositories need to be merged first?
If so add the "waiting for other repos" label and list the upstream PRs