Fixes to search_data_catalogs

[RondeauG]

Pull Request Checklist:

• pre-commit hooks are installed/active in my local clone ($ pre-commit install)
• This PR addresses an already opened issue (for bug fixes / features)
  • This PR fixes search_data_catalog crashes if the result is empty #42
  • This PR fixes Manage fixed fields in both historical and future periods #81
• (If applicable) Documentation has been added / updated (for bug fixes / features)
• If a merge request has been made in parallel to this PR in xscen-notebooks, it is merged and the submodules have been updated.
• HISTORY.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?

• search_data_catalogs no longer crashes if it finds nothing. It now returns an empty dictionary and a warning.
• _dispatch_historical_to_future no longer duplicates fixed fields if they exist in both periods.
• If a fixed field is missing from the search, it is looked for in other experiments or members.

Does this PR introduce a breaking change?

Other information:

[RondeauG]

Question:
I have 3 members for the same model: 'r1i1p1f1_NorESM2-MM_gn_NCC', 'r2i1p1f1_NorESM2-MM_gn_NCC', and 'r3i1p1f1_NorESM2-MM_gn_NCC'. For some reason, r1, r2, and r3 all have orog and areacella, but only r2 has sftlf.

Should we provide an option to broadcast fixed fields to other members or is that too dangerous?

[Zeitsperre]

Should we provide an option to broadcast fixed fields to other members or is that too dangerous?

I can say that for CORDEX-related datasets, this is how the fixed field data is presented between experiment runs. If you feel it's dangerous, you could add warnings when this is being assumed.

[aulemahal]

I have a few changes I want to make to _dispatch_historical_future from an issue encountered by Chris. I could do them here, but it can also wait for another PR.