Some issues with the CRUJRA2024b datasets

[ekluzek]

There's a couple issues I noticed with the CRUJRA2024b datasets, to discuss. We already noticed issues with NaN's earlier.

• NaN's in the _FillValue for time
• NaN's in fields -- many fields have NaN's
• time is not unlimited dimension (which means you can't use cprnc to compare it for example)
• More global attributes to describe the source of the data and the scripts run for each step
• Add Source, Host, comment, Refernces, title, institution, mpi_era
• Optional: Follow more of the CF conventions:
  • add axis to lat, lon, time
  • add coordinate attribute to variables
  • add cell_method attribute to variables
  • add standard_name to variables it can apply to
  • add latixy, longxy for the latitude/longitude grid in 2D
  • could add realtopology = linear attribute to time

We noticed the second one above in this comment: #3764 (comment)

If we do anything here, this will need to go into CDEPS. But, we can start the discussion here. And any resolution would need to be done first in the processing scripts for the dataset, which would be handled in CTSM.

Because, we don't need to do anything immediately I'm putting this in for a ctsm6.1 milestone as we should update scripts for the next TRENDY update.