Warnings / errors on the climatology applications page

[kafitzgerald]

See here: https://ncar.github.io/geocat-applications/applications/climatology.html

We have several warnings showing up on the climatology applications page. It'd be nice to clean these up to avoid the visual clutter and address any important issues.

In particular, these warnings from pythia-datasets have me a little worried. I did go ahead and open a PR to address this, but the review seems to have stalled out: ProjectPythia/pythia-datasets#203. Not sure if we should look into using something different.

/home/runner/micromamba/envs/geocat-applications/lib/python3.13/site-packages/pythia_datasets/__init__.py:4: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
  from pkg_resources import DistributionNotFound, get_distribution

These download notes aren't worrisome, but definitely add a good bit of clutter. It'd be nice to see if we can cleanly hide these.

Downloading file 'CESM2_sst_data.nc' from 'https://github.com/ProjectPythia/pythia-datasets/raw/main/data/CESM2_sst_data.nc' to '/home/runner/.cache/pythia-datasets'.

I suspect these serialization warnings have more to do with the source datafile. I suppose this kind of thing is the reality of working with "real" datasets, but it's a bit distracting IMO and something I'd like to clean up.

/home/runner/micromamba/envs/geocat-applications/lib/python3.13/site-packages/xarray/conventions.py:204: SerializationWarning: variable 'tos' has multiple fill values {np.float32(1e+20), np.float64(1e+20)} defined, decoding all values to NaN.
  var = coder.decode(var, name=name)

There's also stuff like this coming from Matplotlib that I think we can set a project level config to avoid or quiet with a semicolon if we want to go that route.

<matplotlib.collections.QuadMesh at 0x7fee01ccfe00>

I haven't looked in detail, but would be surprised if there weren't similar things elsewhere.

[kafitzgerald]

It sounds like (see here: ProjectPythia/pythia-datasets#203 (comment)) we probably should probably find an alternative to pythia-datasets.

[kafitzgerald]

This should probably be elevated in priority now that we're seeing errors as pointed out by @cyschneck.

I think pythia-datasets is working on a release (the code changes are in, but not released is the current status) since this is critical for Pythia as well (see: ProjectPythia/pythia-datasets#214), but really we should probably work to drop that dependency here.

[brian-rose]

I didn't realize the Pythia-datasets was being used as a dependency for other projects. It would be worth having a conversation about whether it makes collective sense to keep maintaining that project or not.