Update operational workflow to work with GSI develop branch #3068
Conversation
…bal-workflow into dev/gfs.v16.GSIdev
…bal-workflow into dev/gfs.v16.GSIdev
The order of the upgrade to obsproc/v1.3 and the new AFWA global file was moved back and thus the release branch needed to be remade. This PR includes all updates for the GFS upgrade to obsproc/v1.3 and the new global AFWA data file: - new UFS_UTILS tag (`ops-gfsv16.3.20`) to support new global AFWA file - new GSI tag (`gfsda.v16.3.22`) to support obsproc/v1.3 and saildrone - workflow changes to support obsproc/v1.3 (and prepobs/v1.2 internal to obsproc) and AFWA global file - also workflow changes to allow developers to run the emcsfc_sfc_prep job within the rocoto mesh via `DO_SFCPREP=YES` in config.base This PR is a repeat of prior ones to get this package in place. Refs NOAA-EMC#2913
NCO provided increase in ncpus to avoid ncpu burst warnings. Refs NOAA-EMC#2913
…3.21) (NOAA-EMC#3200) Update the `sorc/syndat_getjtbul.fd/getjtbul.f` code to resolve a problem parsing 3-character storm names. This update is being included in the GFSv16.3.21 package (with approval by NCO). From the developer @Qingfu-Liu: > The `gfs_atmos_tropcy_qc_reloc` job did not properly parse the vitals info from a raw bulletin with a tab in the storm name field for storm "SIX". The code sorc/syndat_getjtbul.fd/getjtbul.f converting a tab to a space worked well for storms that have at least 4 characters in their names, but does not work as well for storms with 3-character names. This code change will fix the parsing problem for 3-character storms. Refs NOAA-EMC#2591
…release/gfs.v16.3.22 * origin/release/gfs.v16.3.21: Fix for syndat_getjtbul code parsing 3-character storm names (GFSv16.3.21) (NOAA-EMC#3200)
The WAFS is now a separate package in operations and is no longer a part of the GFS. Also included in this package are the following unrelated updates: 1. improvements to the enkfgdas_update job resource and environment settings 2. a fix for the syndat_getjtbul code when parsing 3-character storm names Refs NOAA-EMC#2591
The `ww3_outp` is optimized to read per time step binary output and create per point spectral and bull/cbull output. Refs NOAA-EMC#2913
…6.3.22 * origin/dev/gfs.v16: GFSv16.3.21 - WAFS removal from GFS (NOAA-EMC#3281)
Update the name of the "Gulf of Mexico" to the "Gulf of America". This includes fix/parm file updates and a new UWM tag for WW3 rename updates. Refs NOAA-EMC#3326
Make additional "Gulf of Mexico" to "Gulf of America" renames. The bufr station list is updated and the EMC_verif-global tag is updated to a version with the same name change. Refs NOAA-EMC#3326
Introduces bufr station updates into the GFSv16.3.23 operational package. There are three updated stations and 121 new stations added to the bufr station parm files. Changes provided by @BoCui-NOAA . Refs NOAA-EMC#2913
…MC#3387) Adjust the package to the new v16.3.23 and incorporate updates from the new v16.3.22 package (now ahead of this update). Additionally, the obsproc (and prepobs) updates are backed out for a later upgrade. Now use `obsproc/v1.2.3`. Refs NOAA-EMC#2913
…se/gfs.v16.3.23 * origin/release/gfs.v16.3.22: Update release notes Additional Gulf of Mexico name changes in GFS (NOAA-EMC#3353) Rename Gulf of Mexico to Gulf of America (NOAA-EMC#3352)
- The radmon_diag_ck.sh and rstprod.sh scripts were missing from the link script step and were causing errors in ops. - Add the two missing scripts to the link script and the gitignore file. Refs NOAA-EMC#2913
|
Code has been updated to release/gfs.v16.3.23. The workflow before these changes successfully ran for seven cycles. I still need to do a longer test with the latest version, but following conversations with @aerorahul, I think this is ready to take out of draft. Note that CRTMv3 is not longer part of this upgrade. |
|
Actually I have one question. Externals.cfg has the following lines: What does this do? I am checking out the develop version of GSI right now. |
@ADCollard We aren't using |
Awesome, feel free to move this out of draft. Note, I need to make a GFSv16.3.24 package that will have the obsproc/prepobs version updates. I'll let you know when I've done that so you can sync that into the v16.4 release branch or I can do the sync if it happens after this goes into the release branch. Just a heads up.
Noted, thanks! |
|
@KateFriedman-NOAA Just to confirm that we will re-tag FV3GFS to 16.4, even though there will not be changes? |
The ufs-weather-model was tagged for v16.4 because some WW3 updates were made and requested to be included in v16.4. So the g-w release branch will include the DA updates and a UWM tag update. |
|
@aerorahul Is this enough to kick off phases 1 and 2 that we discussed yesterday. As I said then, I am on leave from tonight until 3/14 without WCOSS access, so cannot kick off a parallel before then. Hopefully we could still do a code handoff around May 1st. Thanks. |
There was a problem hiding this comment.
Thanks @ADCollard ! I am going to merge, sync it with v16.3.24 changes, and then review further.
Update operational workflow to work with GSI develop branch. This mostly involves changing the links to elements that have been moved out of the GSI repository since GFSv16.3. The ultimate aim is to be able to cycle with v16 workflow but using a version of GSI close to develop. This will not reproduce operations.
Resolves #3045
Type of change
Change characteristics
How has this been tested?
Testing is limited to WCOSS2. Initialised full resolution cycling experiment. Will compare to operational performance.
Checklist