NOAA-EMC / global-workflow

Global Superstructure/Workflow supporting the Global Forecast System (GFS)
https://global-workflow.readthedocs.io/en/latest
GNU Lesser General Public License v3.0
74 stars 164 forks source link

update fix files for CICE and MOM6/post #2483

Closed DeniseWorthen closed 3 months ago

DeniseWorthen commented 4 months ago

Target fix directory

cice, mom6

Type of change

Any workflow changes needed?

No change needed

Related issues

The cice grid files (/glopara/fix/cice) need to be updated after a fix for the incorrect rotation angle for CICE was merged in UFS-UTILS/cpld_gridgen. The 4deg (400) can be removed. This was actually the 5-deg resolution which was initially incorrectly named 400 instead of 500 in UWM.

The mom6 post files also need updating.

As part of https://github.com/ufs-community/UFS_UTILS/issues/929, I also require an two additional files be added:

These can be added to the existing mom6/post directories. Note this directory already contains files used for ice post-processing as well as ocean post.

Associated issues:

https://github.com/ufs-community/UFS_UTILS/issues/921 https://github.com/ufs-community/ufs-weather-model/issues/2228

The PR to update the files for use by UWM is https://github.com/ufs-community/ufs-weather-model/pull/2235

Pickup location

/scratch1/NCEPDEV/nems/role.ufsutils/ufs_utils/reg_tests/cpld_gridgen/baseline_data

Size

~1000 MB for the 025 resolution

KateFriedman-NOAA commented 4 months ago

@DeniseWorthen I have made a copy of the latest 20231219 cice timestamp subfolder to make the updates within. I have copied 20231219 to 20240416 and removed the 400 folder within the new 20240416 set. Since you have the new mom6/post files within your 025/050/100/500 folders, which files do you want me to copy into the new cice/20240416? Similarly, are there files that can/should be removed within the 20240416 copy?

As part of https://github.com/ufs-community/UFS_UTILS/issues/929, I also require an two additional files be added:

I think I'm clear on the mom6/post files to pull into the existing 20231219 set. Will do that now.

DeniseWorthen commented 4 months ago

@KateFriedman-NOAA In the 20240416/cice directories, there should be the three files, each w/ a resolution identifier, where [] = 025,050,100,500

grid_cice_NEMS_mx[].nc
kmtu_cice_NEMS_mx[].nc
mesh.mx[].nc

The cice/100 directory still contain the mom6 fix files by mistake (issue #2480).

In the mom6/post subdirectories I need any file containing "tripole" in the file name.

KateFriedman-NOAA commented 4 months ago

@DeniseWorthen Thanks for the clarification! How does the updated cice/20240416 look?

[role.glopara@hfe08 20240416]$ pwd
/scratch1/NCEPDEV/global/glopara/fix/cice/20240416
[role.glopara@hfe08 20240416]$ ll *
025:
total 157972
-r--r--r-- 1 role.glopara global 68429828 Apr  3 18:09 grid_cice_NEMS_mx025.nc
-r--r--r-- 1 role.glopara global  6221520 Apr  3 18:09 kmtu_cice_NEMS_mx025.nc
-r--r--r-- 1 role.glopara global 87107397 Apr  3 18:09 mesh.mx025.nc

050:
total 42144
-r--r--r-- 1 role.glopara global 18248708 Apr  3 18:10 grid_cice_NEMS_mx050.nc
-r--r--r-- 1 role.glopara global  1659600 Apr  3 18:10 kmtu_cice_NEMS_mx050.nc
-r--r--r-- 1 role.glopara global 23238629 Apr  3 18:10 mesh.mx050.nc

100:
total 5660
-r--r--r-- 1 role.glopara global 5069828 Apr  3 18:10 grid_cice_NEMS_mx100.nc
-r--r--r-- 1 role.glopara global  461520 Apr  3 18:10 kmtu_cice_NEMS_mx100.nc
-r--r--r-- 1 role.glopara global 6463973 Apr  3 18:10 mesh.mx100.nc

500:
total 276
-r--r--r-- 1 role.glopara global 111908 Apr  3 18:10 grid_cice_NEMS_mx500.nc
-r--r--r-- 1 role.glopara global  10800 Apr  3 18:10 kmtu_cice_NEMS_mx500.nc
-rw-r--r-- 1 role.glopara global 154762 Jul 14  2022 mesh.mx500.nc
DeniseWorthen commented 4 months ago

@KateFriedman-NOAA Those are the correct files for the fix/cice subdirectory. Thanks.

KateFriedman-NOAA commented 4 months ago

@DeniseWorthen Thanks for confirming the new cice timestamp copy is good.

Since we're overwriting existing files in the mom6/post folder, I also created a copy of the current 20231219 mom6 and placed the updated files within it. I also removed some symlinks in the new copy that redirected folks to updated naming conventions, since we shouldn't need them in the new copy anymore. The other resolution folders/files at the same level as the post folder are untouched and simply copied from the 20231219 set.

Please see the new mom6/20240416/post copy: /scratch1/NCEPDEV/global/glopara/fix/mom6/20240416/post.

Let me know if the new mom6/20240416/post is as requested. Thanks!

DeniseWorthen commented 4 months ago

@KateFriedman-NOAA This all seems correct. Thanks!

KateFriedman-NOAA commented 4 months ago

@DeniseWorthen Great, thanks for confirming! I will sync these updates to the fix sets on the other platforms. I am doing a similar fix update in issue #2493, so will wait for confirmation from that user before doing the syncs. Will confirm when the syncs are complete.

We will now need to update mom6_ver=20240416 in versions/fix.ver to use the updated files. Do you have plans to do so in an upcoming PR, perhaps with related script updates?

DeniseWorthen commented 4 months ago

The new CICE fix files will be switched over to in an upcoming UFS PR (https://github.com/ufs-community/ufs-weather-model/pull/2235). As for how the post files are used, I don't really know anything about the versioning in G-W or even where I would make that change.

KateFriedman-NOAA commented 4 months ago

Oh right, we'll need to update both cice_ver and mom6_ver in versions/fix.ver. Noted about the UFS PR. We'll need to update to cice_ver=20240416 when we move to the hash associated with that PR.

Whenever users want to start using the updated mom6/20240416/post files the mom6_ver setting in versions/fix.ver will need to be updated in global-workflow.

KateFriedman-NOAA commented 4 months ago

The CICE and MOM6/post fix updates have now been synced to both WCOSS2, Orion, and Jet.

KateFriedman-NOAA commented 4 months ago

The CICE and MOM6/post fix updates are also now synced to our global AWS bucket.

KateFriedman-NOAA commented 4 months ago

Work is done but awaiting updates to relevant fix.ver versions. Will do this and close issue as complete when we update to a UWM hash at or beyond the result of https://github.com/ufs-community/ufs-weather-model/pull/2235.

KateFriedman-NOAA commented 3 months ago

@DeniseWorthen Shall we also update to the newer mom6/20240416 version when we update to the newer cice/20240416? Or do we need to wait for a component update?

DeniseWorthen commented 3 months ago

@KateFriedman-NOAA Yes, the mom6/20240416 should be updated as well because that is where the post weights, which would have been updated, reside.

KateFriedman-NOAA commented 3 months ago

Yes, the mom6/20240416 should be updated as well

@DeniseWorthen got it, I will update that version in the same PR as the cice_ver.