Speed up tif data access

[konstntokas]

• Use chunks="auto" by default in rioxarray.open_rasterio within
  xcube.core.store.fs.impl.rasterio when reading GeoTIFF and JPEG2000 files.
  This enables efficient, storage-aware data access without forcing explicit
  rechunking.
• Added the optional argument band_as_variable, which allows to preserve the
  original dataset structure as returned by rioxarray, rather than splitting
  raster bands into separate data variables. This improves data access patterns and
  avoids unnecessary transformations. Defaults are set to True.

Note: I needed to update the stac jsons, because the htmlRepr of the datasets have changed (probably due to an update in xarray).

Here is an small example of a GeoTIFF access as it was implemented before:

Here is an small example of the same GeoTIFF access with chunks="auto":

Checklist:

• Add unit tests and/or doctests in docstrings
• Add docstrings and API docs for any new/modified user-facing classes and functions
• New/modified features documented in docs/source/*
• Changes documented in CHANGES.md
• GitHub CI passes
• AppVeyor CI passes
• Test coverage remains or increases (target 100%)

[konstntokas]

Example Notebooks showing local data access with different opening parameters

access_tif_local0.ipynb
access_tif_local1.ipynb
access_tif_local2.ipynb

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.55%. Comparing base (3029575) to head (fac81aa).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1192   +/-   ##
=======================================
  Coverage   89.55%   89.55%           
=======================================
  Files         280      280           
  Lines       21646    21653    +7     
=======================================
+ Hits        19384    19391    +7     
  Misses       2262     2262           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[b-yogesh]

LGTM. I did not test it but I believe the screenshots on the speedup. Should we add any tests that show that using chunks is faster than any other tile size? Also, please have a look at my comments.

[b-yogesh]

Here it says the defaults are set to True but in the PR description it says it is set to False.

[konstntokas]

Defaults are True to ensure backward compatibility.

[b-yogesh]

Here as well

[konstntokas]

Defaults are True to ensure backward compatibility.

[b-yogesh]

Here as well.

[konstntokas]

Defaults are True to ensure backward compatibility.

[b-yogesh]

Here and other places too. If true is correct, change the PR description

[konstntokas]

Defaults are True to ensure backward compatibility.

[b-yogesh]

Here it is set to False

[konstntokas]

This is changed to True.