Fix handling of optional variables in Unified BGC datasets

[NoraLoose]

Summary

This PR fixes a bug in SurfaceForcing and ensures that optional variables in UnifiedBGCDataset and UnifiedBGCSurfaceDataset (used by InitialConditions, BoundaryForcing and SurfaceForcing) are handled correctly. Specifically, if an optional variable is present in the source data, it is now read from there rather than treated as missing and filled with a default value. This is especially relevant because Pierre has updated his "unified BGC dataset" so that it now includes not only required but also all optional variables (see #313).

Changes

• Corrected logic to correctly detect and use available optional variables in SurfaceForcing.

• Updated external test data to reflect a mix of present and missing optional variables for UnifiedBGCDataset and UnifiedBGCSurfaceDataset. The updated external data led to a necessary overwriting of the validation test data.

• Added tests for SurfaceForcing, InitialConditions, BoundaryForcing to verify:

  • Present optional variables in UnifiedBGCDataset and UnifiedBGCSurfaceDataset retain their original (non-constant) values.
  • Missing optional variables are filled with constant default values.

• Updated example notebooks to read from Pierre's extended unified BGC dataset, which now includes all optional variables.

Checklist

• Closes Integrate Pierre's updated unified BGC dataset #313
• Tests added
• Passes pre-commit run --all-files
• Changes are documented in docs/releases.md
• New functions/methods are listed in docs/api.rst
• New functionality has documentation

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ScottEilerman]

Only looked at the .py files, but those look fine to me!

[NoraLoose]

Thanks @ScottEilerman!