Drop Python3.8, support Pandas v2.2, reduce Warnings

[Zeitsperre]

Pull Request Checklist:

• This PR addresses an already opened issue (for bug fixes / features)
  • This PR fixes Drop Python3.8 #1268
  • This PR fixes Adopt new pandas frequency codes #1534
  • This PR fixes standardized_index_fit_params with method="ML" raises TypeError in scipy #1617
• Tests for the changes have been added (for bug fixes / features)
  • (If applicable) Documentation has been added / updated (for bug fixes / features)
• CHANGES.rst has been updated (with summary of main changes)
  • Link to issue (:issue:number) and pull request (:pull:number) has been added

What kind of change does this PR introduce?

• Drops Python3.8 references and marks Python3.9 as base requirement.
• Drops older xarray and pandas support.
• Updates documentation and CI for dropped Python3.8.
• Raises required versions of scipy and numpy (1.9.0 and 1.20.0, respectively).
• Removes the winter_storm indice and indicator (deprecated in xclim v0.46.0).
• Adopts the new syntax for frequency codes:
  • "A" is no more
  • Y -> YE, Q -> QE, M -> ME
  • H -> h, T -> min, S -> s, L -> ms, U -> us, N -> ns
• Addresses several DeprecationWarning and RuntimeWarning messages related to numpy, xarray, and pint

Does this PR introduce a breaking change?

Yes, Python3.8 is no longer supported and base dependency versions have been augmented:

• numpy>=1.20.0
• pandas>=2.2.0
• scipy>=1.9.0
• xarray>=2023.11.0

Other information:

Python3.12 support is still not possible due to missing Python3.12 builds for numba. (This is being addressed in #1613) - This has been addressed.

A Pull Request has been opened to address DeprecationWarnings stemming from numpy calls in eofs (ajdawson/eofs#148; eofs has been removed due to licensing-related issues (#1621))

[github-actions]

Note
It appears that this Pull Request modifies the main.yml workflow.

On inspection, the XCLIM_TESTDATA_BRANCH environment variable is set to the most recent tag (v2023.12.14).

No further action is required.

[Zeitsperre]

@aulemahal I've left a few "FIXME" notices since changes will ultimately involve us needing to make some API/UI decisions.

[aulemahal]

Fixed! My solution also works with 2023.11.0.

[Zeitsperre]

Adding the approved label so that we can run more tests. This PR will still require a proper review.

[coveralls]

coverage: 90.169% (-0.1%) from 90.312%
when pulling 8bd6efc on drop-python38
into 8db48d8 on master.

[aulemahal]

Some minor comments.
But I hereby approve the lines I did not write.

[huard]

This PR mixes a lot of apparently unrelated changes (formatting, the pandas freq update, deprecations, etc). Makes it hard to review.

The code still accepts "A" as an offset, I assume it's because it's still supported upstream?

[aulemahal]

Sorry for the mammouth PR indeed. All those changes stem from the same two updates of python and pandas.

Xclim doesn't precribe what frequency code should be used, in most cases, it only passes them down to xarray. This PR changes all places where xclim was using the old codes (tests, checks, documentation). A user with pandas < 3 will stil be able to use "A" and should get a warning from pandas.

However, there a some instances where a frequency code is checked against the output of xr.infer_freq or parsed with pd.frequencies.to_offset. Changes there are breaking, which explains why we decided to pin pandas and xarray to the versions accepting the new codes.