[develop] Add MET/METplus to Jet

[EdwardSnyder-NOAA]

DESCRIPTION OF CHANGES:

Added MetPlus file paths to the Jet machine file in the develop branch. A similar PR #797 was approved for the release branch.

TESTS CONDUCTED:

Ran MET_verification and MET_ensemble_verification successfully on Jet.

DEPENDENCIES:

None.

DOCUMENTATION:

No documentation changes are needed.

ISSUE (optional):

CONTRIBUTORS (optional):

[willmayfield]

Did you check the MET outputs and logs? I had hoped to put this change in #779 but ran into errors from MET, even though the workflow tasks succeeded. I was thinking it needs an appropriate run_vx.local modulefile and/or updates to the miniconda environment.

[EdwardSnyder-NOAA]

Did you check the MET outputs and logs? I had hoped to put this change in #779 but ran into errors from MET, even though the workflow tasks succeeded. I was thinking it needs an appropriate run_vx.local modulefile and/or updates to the miniconda environment.

Hi @willmayfield - I didn't check the logs or outputs. What sort of errors did you receive? I'm noticing the ModuleNotFoundError: No module named 'dateutil' error is in most of the vx logs.

[willmayfield]

Hi @willmayfield - I didn't check the logs or outputs. What sort of errors did you receive? I'm noticing the ModuleNotFoundError: No module named 'dateutil' error is in most of the vx logs.

Yep, that's what I see too.

[EdwardSnyder-NOAA]

Added modulefiles/tasks/jet/run_vx.local script, which removed ModuleNotFoundError: No module named 'dateutil' errors and resulted in a successful run of both the MET_verification and MET_ensemble_verification tests.

[willmayfield]

Compared a few MET stat files to the same test on Hera. There were very small numerical differences, likely within reason between platforms, and the MET logs looked fine. Thanks Ed!

[EdwardSnyder-NOAA]

Thanks @willmayfield for doing the comparison!