incorrect topoedits files in /scratch1/NCEPDEV/global/glopara/fix/mom6/20240416/100 (and 20231219/100/)

[DeniseWorthen]

What is wrong?

In the MOM6/100 fix directories, there should be two distinct "topoedits" files: topo_edits_011818.nc and ufs.topo_edits_011818.nc.

The first of these files contains 103 values (nEdits = 103). It is the file which comes directly from GFDL, but which leaves an open boundary along the southern most row.

In cpld_gridgen, we read the original topo_edits_011818.nc file and add 40 "edit" points to close (add land) to the southernmost row. This creates the ufs.topo_edits_011818.nc. This file holds nEdits=143.

UWM 1-deg MOM6 should only use the ufs.topo_edits_011818.nc file.

In the 20240416/100 directory, there is an incorrect link:

lrwxrwxrwx 1 role.glopara global       24 Aug  9 15:38 topo_edits_011818.nc -> ufs.topo_edits_011818.nc

This means that when cpld_gridgen reads the topo_edits_011818.nc, it creates a file with duplicated edits (nEdits = 183) which closes the already closed boundary. This breaks the cpld_gridgen RT.

This apparently was done in NOAA-EMC/global-workflow #2812

What should have happened?

The files should be unlinked; the original 103 file (topo_edits_011818.nc ) needs to be placed in the fix directory instead of linking it to the ufs file. The original file can be recovered from 20220805/100/topo_edits_011818.nc.

Note that the 20231219/100/topo_edits_011818.nc is NOT correct. It is in fact the ufs.topo_edits_011818.nc file.

What machines are impacted?

All or N/A

What global-workflow hash are you using?

4ab8cf8

Steps to reproduce

1. Check the nEdits size in topo_edits_011818.nc and ufs.topo_edits_011818.nc. The first should contain a dimension of nEdits = 103 and the second should contain nEdits = 143

Additional information

No response

Do you have a proposed solution?

No response

[KateFriedman-NOAA]

@DeniseWorthen I created a new fix/mom6/20250128 timestamp from a copy of the 20240416 one and then did the following based on your description above:

[role.glopara@hfe07 100]$ pwd
/scratch1/NCEPDEV/global/glopara/fix/mom6/20250128/100
[role.glopara@hfe07 100]$ ll topo_edits_011818.nc
lrwxrwxrwx 1 role.glopara global       24 Aug  9 15:38 topo_edits_011818.nc -> ufs.topo_edits_011818.nc
[role.glopara@hfe07 100]$ unlink topo_edits_011818.nc 
[role.glopara@hfe07 100]$ cp ../../20220805/100/topo_edits_011818.nc .
[role.glopara@hfe07 100]$ ll topo_edits_011818.nc
-rw-r--r-- 1 role.glopara global 1476 Sep  3  2020 topo_edits_011818.nc

Does that correct things for the new timestamp? Are other resolutions impacted as well and need fixing?

FYI, I will leave the older timestamps as is. The new timestamp should get set in develop versions/fix.ver, do you have plans to update it in a companion PR or shall we submit a PR just to update the fix version variable for mom6? Let us know, thanks!

[DeniseWorthen]

Thanks Kate. Those files look correct; let me run my cpld_gridgen RT w/ the new fix.ver.

The only file impacted in the MOM6-fix directories would be for the 1deg (100) topo-edits files. All the other resolutions have no additional topo-edits.

In terms of further changes, I'm in the midst of generating new files for the 9-deg ocean and ice. That will mean new fix files for both CICE and MOM6 (in the post). So it probably makes more sense for me to get all those new files ready and then make a PR for a new fix.ver for both CICE and MOM6. Does that sound right?

[KateFriedman-NOAA]

@DeniseWorthen Sounds good to me! If you're cool with it, I'll sync the new timestamp with the corrected topo-edits file to the other supported platforms and then close this issue as complete. When you have the new 9-deg ocean/ice files ready, please open a new issue to document adding them to the fix set. We will then add them to the new 20250128 timestamp and sync to the other platforms again. Then you can submit a PR that includes the fix variable update. Let me know if that sounds good, thanks!

[DeniseWorthen]

Yes, please sync the corrected topo-edits file.

I'll follow-up as you suggest w/ the new 9deg files.

Thanks.

[KateFriedman-NOAA]

Thanks @DeniseWorthen ! I'll begin syncing the new timestamp to the other platforms. @WalterKolczynski-NOAA please sync the new fix/mom6/20250128 fix to MSU, thanks! @DavidBurrows-NCO please sync the Hera fix to Gaea, thanks!

[KateFriedman-NOAA]

Synced mom6/20250128 to WCOSS2-Dogwood, Jet, AWS, and HPSS. Will sync to WCOSS2-Cactus when outage concludes.

[KateFriedman-NOAA]

I'll follow-up as you suggest w/ the new 9deg files.

@DeniseWorthen Would #3140 be the other issue we discussed? It's currently on hold awaiting files.

[DeniseWorthen]

Yes, that is the relevant issue. I needed first the C12 and C24 mosaic files in place (this has been done) and now that the topo-edits are fixed, I can open the PR to add the capability to cpld_gridgen. Once I create the new files (for cice, post), we can follow-up w/ updating the fix file directories.