New GSI fix set for GFS v16.3.12

[ADCollard]

Target fix directory

gsi

Type of change

• Update existing files
• Remove files
• Add new files to existing directory
• Add new fix directory

Any workflow changes needed?

No change needed

Related issues

NOAA-EMC/GSI-fix#13

Pickup location

/scratch1/NCEPDEV/da/Andrew.Collard/git/GSI_gfsda_v16/fix/Rcov_crisn21 & GSI-fix repo clone

Size

Full new copy of GSI fix subset

[KateFriedman-NOAA]

@ADCollard I can certainly add this new file into the current fix/gsi/20230911 timestamp version within the developmental fix set. Are the updated GFSv16 global_anavinfo.l127.txt, global_convinfo.txt, and global_satinfo.txt files coming later? I would create a new GSI timestamp set for them when they do.

[KateFriedman-NOAA]

Is this a duplicate of issue #1919?

[RussTreadon-NOAA]

@KateFriedman-NOAA , you are right. #1919 is no longer needed.

[KateFriedman-NOAA]

I have copied the new Rcov_crisn21 file into the GSI 20230911 timestamp subfolder:

[role.glopara@hfe05 20230911]$ pwd
/scratch1/NCEPDEV/global/glopara/fix/gsi/20230911
[role.glopara@hfe05 20230911]$ ll Rcov_crisn21
-rw-r--r-- 1 role.glopara global 40436 Feb  1 16:33 Rcov_crisn21

[ADCollard]

@KateFriedman-NOAA Thanks!

The other files are included in the GSI-Fix PR13. Do I need to explicitly add then to this issue?

(Apologies, I cannot assign any one in the GSI-Fix repositorty, but I will go ahead and tag people in the commentary)

[KateFriedman-NOAA]

@ADCollard Previously, a DA team member provided me with a copy of the newest GSI-fix on Hera and I copied it into a new timestamp subfolder of fix/gsi within the developmental fix set (and synced to the other platform copies). Once NOAA-EMC/GSI-fix#13 is in, I would expect to get a request to add a new timestamp copy and populate it with the updated GSI-fix hash. Then I would update gsi_ver in versions/fix.ver in global-workflow in the same branch that would collect the related script updates that go with the fix updates (issue #1918). Since I know the global_satinfo.txt and two other global_*.txt files are changing compared to their current versions in the latest GSI fix timestamp, I need to make a whole new timestamp copy to preserve the existing set. I was able to add that new Rcov_crisn21 fix file to the current timestamp copy since it didn't already exist. Hope that makes sense. Would handing off the last GSI-fix hash after that PR goes in work for you for the fix updates?

Also, please see the checklist in issue #1918 for the updates I am expecting to make in global-workflow develop to bring in recent GFSv16.3.* changes. Let me know in that issue if any of those checklist items won't be happening in that effort. Thanks!

[ADCollard]

@KateFriedman-NOAA NOAA-EMC/GSI-fix#13 has now been merged into develop. Can we no go head and produce a new timestamp copy?

Thanks,
Andrew.

[KateFriedman-NOAA]

@ADCollard Certainly! I also see that NOAA-EMC/GSI-fix#14 was merged into GSI-fix develop. Do you want that update included (new hash 298bdc0) or just the state of GSI-fix after NOAA-EMC/GSI-fix#13 went in (new hash NOAA-EMC/GSI-fix@cd0a90c)?

[ADCollard]

@KateFriedman-NOAA Yes, please proceed with both updates.

[KateFriedman-NOAA]

@ADCollard I have cloned the HEAD of GSI-fix into a new 20240208 timestamp subfolder. See here on Hera: /scratch1/NCEPDEV/global/glopara/fix/gsi/20240208. It looks like it's missing some files, in particular Rcov_crisn21 and other Rcov_* files. How do you suggest we collect this new timestamp copy with all files? Previously I think someone provided me with a staged complete copy to pull in. Perhaps that would be best this time?

[ADCollard]

@KateFriedman-NOAA Thanks for doing this. I think the cleanest way is to copy the binary files from the previous timestamp folder.

[KateFriedman-NOAA]

@ADCollard Alrighty...I removed the 20240208 copy I'd cloned, copied the 20230911 folder as 20240208, and then did a git pull in the 20240208 folder to pull in the GSI-fix repo updates. Here are the diffs that got pulled in:

[role.glopara@hfe08 20240208]$ git pull
remote: Enumerating objects: 9, done.
remote: Counting objects: 100% (9/9), done.
remote: Compressing objects: 100% (8/8), done.
remote: Total 9 (delta 3), reused 5 (delta 1), pack-reused 0
Unpacking objects: 100% (9/9), done.
From https://github.com/NOAA-EMC/GSI-fix
   5722cd4..298bdc0  develop    -> origin/develop
Updating 5722cd4..298bdc0
Fast-forward
 global_anavinfo.l127.txt          |   6 +-
 global_anavinfo_soilanal.l127.txt | 137 ++++++++++++++++++++++++++++
 global_convinfo.txt               |   2 +-
 global_convinfo_2mObs.txt         | 262 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 global_satinfo.txt                | 470 +++++++++++++++++++++++++++++++++++++++++++++++------------------------------------------------
 5 files changed, 637 insertions(+), 240 deletions(-)
 create mode 100644 global_anavinfo_soilanal.l127.txt
 create mode 100755 global_convinfo_2mObs.txt

Please take a look at this new /scratch1/NCEPDEV/global/glopara/fix/gsi/20240208 GSI fix subfolder on Hera and let me know if it's good-to-go. I'm guessing that there may be files in the 20240208 copy that no longer exist and could be removed....please let me know if so and which files. I'd like to keep the 20240204 set as clean as possible. Thanks!

[ADCollard]

@KateFriedman-NOAA The contents of /scratch1/NCEPDEV/global/glopara/fix/gsi/20240208 are as expected. I do not think there are any files that do not exist, but there is probably scope to remove obsolete files. But that is not something that I would be comfortable doing unilaterally - we probably need the DA team as a whole to review.

[KateFriedman-NOAA]

@ADCollard Gotcha. I will proceed with syncing the new fix/gsi/20240208 version to the other supported platforms and then close this issue as complete. Make sure to update gsi_ver=20240208 in your branch for issue #1918. I will add another checklist item to that issue as a reminder for us.

[KateFriedman-NOAA]

@WalterKolczynski-NOAA If you haven't already since yesterday, please sync the fix set on Orion to pull in the new 20240208 GSI subset. Thanks!

[KateFriedman-NOAA]

New GSI fix set (20240208) has now been synced to all supported platforms. Closing issue as complete.

The gsi_ver in versions/fix.ver will be taken care of in issue #1918.