Various wave updates: Reformat WW3 inputs; enable PIO for wave restarts; enable WW3 cycling; synchronize UFS restart writes; add uglo_15km WW3 grid #3190
Conversation
Conflicts: .gitmodules
…reading if desired
…combineforww3testing
Co-authored-by: Rahul Mahajan <[email protected]>
Co-authored-by: Rahul Mahajan <[email protected]>
Co-authored-by: Rahul Mahajan <[email protected]>
Co-authored-by: Rahul Mahajan <[email protected]>
fix files are still needed and I'm working to stage high-res ICs. Running a last test or two before I'm ready to pass off the fix files. Also - as soon as #3188 is merged, there will be conflicts to resolve. |
| @@ -187,7 +187,7 @@ def get_task_names(self): | |||
| if options['do_jedisnowda']: | |||
| task_names[run] += ['snowanl'] | |||
|
|
|||
| wave_prep_tasks = ['waveinit', 'waveprep'] | |||
| wave_prep_tasks = ['waveinit'] | |||
There was a problem hiding this comment.
If waveprep is no longer needed, should we be removing all the related scripts?
There was a problem hiding this comment.
I'd feel better if we waited a little longer to completely remove waveprep to be honest. But likely, yes we will be removing these scripts soon.
| @@ -931,6 +931,11 @@ def _fcst_cycled(self): | |||
| dep = rocoto.add_dependency(dep_dict) | |||
| dependencies = rocoto.create_dependency(dep=dep) | |||
|
|
|||
| if self.options['do_wave']: | |||
| wave_job = 'waveprep' if self.options['app'] in ['ATMW'] else 'waveinit' | |||
There was a problem hiding this comment.
If waveprep is still needed for ATMW (should we ever get that to work again), shouldn't it still be in the task list above for ATMW? Right now there will never be a waveprep task.
There was a problem hiding this comment.
Wave prep could still be used for ATMW, but i do not have the bandwidth to get that running again and am unsure of the immediate use case, other than it would have sped up a lot of the testing I've done here.
There was a problem hiding this comment.
I think we can handle wave_prep in another PR.
We could open an issue and add it to the backlog. Realistically, defer it to @JessicaMeixner-NOAA to prioritize ATMW a working option and have a test to keep it unbroken. The wave development team will benefit from such an option if there are more updates to the wave post-processing expected.
Either way, suggestion here is to park the wave_prep issue for now and let this PR continue through. Open to other suggestions.
There was a problem hiding this comment.
I'd appreciate parking the waveprep issue.
I think this PR is almost ready to go.... One last minor tweak (i hope) and should be ready to share the new fix files.
Co-authored-by: Walter Kolczynski - NOAA <[email protected]>
with new unstructured grids
JessicaMeixner-NOAA
dismissed stale reviews from WalterKolczynski-NOAA and aerorahul
via
4698739
|
Fix file update: #3218 |
|
A recent cycled w/wave test output is here: /scratch1/NCEPDEV/climate/Jessica.Meixner/WaveUpdatesHR5/cyclewwavet03 The fix file version has been updated. Once it's syn'd to other machines this should be ready for CI testing. |
Description
This PR adds the following:
Notes on ufs-weather-model updates:
Note this PR requires the following:
Co-author: @sbanihash
Related Issues:
Type of change
Change characteristics
How has this been tested?
Tests were done by adding waves to C48mx500 3DVar tests for cycling.
High resolution tests cases were also added.
This likely needs a new round of testing before CI and after staged ICs & new fix files are added.
Checklist