support runoff normalization for partial years with annual normalization data

[fneum]

Changes proposed in this Pull Request

It makes cutout.runoff(normalize_using_yearly=...) more lenient by supporting proportional scaling of partially represented years in the cutout, assuming even distribution of runoff throughout the year. In this case, a warning is printed due to the strong assumption.

Relevant for PyPSA/pypsa-eur#1613 (e.g. July-June meteorological years, in case the original cutout does not span Jan-Dec for each year included)

This also means we can include hydro in the CI test of PyPSA-Eur, which uses only a week of weather data.

Checklist

• Code changes are sufficiently documented; i.e. new functions contain docstrings and further explanations may be given in doc.
• [] Unit tests for new features were added (if applicable).
• [] Newly introduced dependencies are added to environment.yaml, environment_docs.yaml and setup.py (if applicable).
• A note for the release notes doc/release_notes.rst of the upcoming release is included.
• I consent to the release of this PR's code under the MIT license.

[Copilot]

Pull Request Overview

This PR enhances the runoff normalization function to support partial years by applying proportional scaling based on available data, while issuing a warning about the strong assumption of even runoff distribution.

• Improved handling of date ranges for runoff and normalization data
• Added partial year normalization logic with proportional scaling and a warning message
• Updated reindexing to use a dynamic coordinate based on the altered dimension

Files not reviewed (1)

• RELEASE_NOTES.rst: Language not supported

Comments suppressed due to low confidence (3)

atlite/convert.py:1026

• [nitpick] Ensure that reindexing using the dynamic dimension (dim) aligns with the intended coordinate, especially if the original 'countries' coordinate was explicitly required in other parts of the code.

).reindex({dim: result.coords[dim]})

atlite/convert.py:1010

• [nitpick] Verify that the proportional scaling logic correctly handles all edge cases, such as when the overlap period almost covers a full year, to ensure the computed fraction accurately reflects the intended partial data scaling.

fraction = (end - start).total_seconds() / ((year_end - year_start).total_seconds())

atlite/convert.py:965

• [nitpick] Consider renaming 'result_time' to a more descriptive name like 'time_index' to more clearly indicate that this variable represents a DatetimeIndex.

result_time = pd.DatetimeIndex(result.coords["time"].values)