TEMPO code cleanup and refactor for v3.0.0 and MPAS release

[AndersJensen-NOAA]

• removed unused configuration variable
• added graupel volume to init_atmosphere registry (scalars are inconsistent between init_atmosphere and atmosphere #167)
• formatted code

The title above should be a 1 line short summary of the pull request (i.e. what the project the PR represents is intended to do).

Enter a description of this PR. This should include why this PR was created, and what it does.

Testing and relations to other Pull Requests should be added as subsequent comments.

See the below examples for more information.
MPAS-Dev/MPAS#930
MPAS-Dev/MPAS#931

Priority Reviewers

• Please list the developers/collaborators you'd like to prioritize for review
• Example:
  • @clark-evans

[guoqing-noaa]

Any updates on the timeline of this PR? Thanks!

[clark-evans]

Any updates on the timeline of this PR? Thanks!

It's currently being tested in dev1. The output so far looks fairly reasonable. If that holds with formal verification, I expect we'll be ready to merge next week or shortly thereafter.

[guoqing-noaa]

Any updates on the timeline of this PR? Thanks!

It's currently being tested in dev1. The output so far looks fairly reasonable. If that holds with formal verification, I expect we'll be ready to merge next week or shortly thereafter.

That's great! Thanks for the good news!

[joeolson42]

I suppose... ;-)

[AndersJensen-NOAA]

@dustinswales @clark-evans Three tests are failing. I though it was an issue with the lookup tables, so I made a few commits to try and modify, but it could be a namelist issue also with the update to tempo v3. Suggestions for easy ways to debug would be helpful.

[clark-evans]

@dustinswales @clark-evans Three tests are failing. I though it was an issue with the lookup tables, so I made a few commits to try and modify, but it could be a namelist issue also with the update to tempo v3. Suggestions for easy ways to debug would be helpful.

@dustinswales are we still pulling the namelists from barlage/mpas_testcase? That's what it looks like to me in the workflow. I'm guessing the feature tests are failing because Anders moved the TEMPO namelist parameters to a new namelist block that isn't represented in the testcase namelists.

If this is the case, then the tests would fail until that repo is updated - but we don't necessarily want to update that repo until after this PR is merged. Is there a hacky way of getting around this to test the updated code?

[dustinswales]

@clark-evans You are correct about the nml.
@AndersJensen-NOAA just stopped by and we chatted. The plan is for Anders to update his fork of the MPAS case configuration repo, and make any necessary changes for the TEMPO namelist updates.
Also, we should update the CI scripts to use Anders repo rather than updating the one owned by @barlage (We should've handed this off to someone at GSL when Mike left, so it's a good time for Anders to adopt it :))

[AndersJensen-NOAA]

@clark-evans You are correct about the nml. @AndersJensen-NOAA just stopped by and we chatted. The plan is for Anders to update his fork of the MPAS case configuration repo, and make any necessary changes for the TEMPO namelist updates. Also, we should update the CI scripts to use Anders repo rather than updating the one owned by @barlage (We should've handed this off to someone at GSL when Mike left, so it's a good time for Anders to adopt it :))

Thanks -- I may have to iterate a bit but I think I can get this working with running the base with old tables and old namelists and the feature with new tables and new namelists. It may take a few tries. If I get stuck I'll reach out again.

[AndersJensen-NOAA]

@dustinswales -- I have things working. Can you take a quick peek at my changes to run_mpas.yaml and run_mpas_hrrr.yaml to make sure that things look good?

@clark-evans ready to merge once Dustin and you okays this.

[clark-evans]

The comparison script output from the GSL hrrrv5 tests includes the following line for all three tests:

"DIFFER : Failed to find variable ng in file "/home/runner/work/run_bl/history.2024-02-02_18.00.00.nc"."

This also appears for all subsequent forecast hours. Since it's in the run_bl directory, I'm guessing Ng is not in the baseline output but is in the feature output. Is that to be expected, @AndersJensen-NOAA?

[dustinswales]

@AndersJensen-NOAA Looks good to me.

I will say that sleuthing through the CI logs to see that this is answer changing is a big pain.
We could add a plotting step to the CI script that plots any differences and stores them as an artifacts.
(This we the diffs would be right in your face in the PR, and we could double-check the developers "blessing of the changes")

[AndersJensen-NOAA]

The comparison script output from the GSL hrrrv5 tests includes the following line for all three tests:

"DIFFER : Failed to find variable ng in file "/home/runner/work/run_bl/history.2024-02-02_18.00.00.nc"."

This also appears for all subsequent forecast hours. Since it's in the run_bl directory, I'm guessing Ng is not in the baseline output but is in the feature output. Is that to be expected, @AndersJensen-NOAA?

great catch @clark-evans -- I reverted the changes to the namelist for the baseline run only, in an attempt to not change the baselines as I had inadvertently turned off the hail aware option when updating namelists.

New tests are running now