Clarify rh_from_tdew variable naming, comments and references

[gaoflow]

• Closes Inverted vapor pressure and saturation vapor pressure #2734
• I am familiar with the contributing guidelines
• I attest that all AI-generated material has been vetted for accuracy and is in compliance with the pvlib license
• Tests added
• Updates entries in docs/sphinx/source/reference for API changes.
• Adds description and name entries in the appropriate "what's new" file in docs/sphinx/source/whatsnew for all changes. Includes link to the GitHub Issue with :issue:`num` or this Pull Request with :pull:`num`. Includes contributor name and/or GitHub username (link with :ghuser:`user`).
• New code is fully documented. Includes numpydoc compliant docstrings, examples, and comments where necessary.
• Pull request is nearly complete and ready for detailed review.
• Maintainer: Appropriate GitHub Labels (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.

This is a clarity change with no change to the returned values. As noted in #2734 and independently confirmed by @echedey-ls (cross-checked against an online calculator and NREL's pvdeg), rh_from_tdew already returns correct relative humidity — e.g. rh_from_tdew(20, 10) == 52.56 %, rh_from_tdew(25, 25) == 100 %, and it round-trips with tdew_from_rh. What confused the report is that the two intermediate variables were named backwards: e was computed from temp_air (so it was really the saturation pressure) and es from temp_dew (the actual pressure), and 100 * es/e with those swapped names happens to equal the conventional 100 * e/es.

Changes:

• Compute e from the dew point (actual vapor pressure) and es from the air temperature (saturation) so the source matches the usual convention, and write 100 * e/es. Numerically identical — the existing value-pinning tests (test_rh_from_tdew, test_tdew_from_rh, test_tdew_to_rh_to_tdew) pass unchanged.
• Fix the now-inaccurate derivation comment in tdew_from_rh (it referenced es/e).
• Add a short Notes section and the Alduchov & Eskridge (1996) reference for the Magnus form, since the WMO guide alone doesn't name it (per @echedey-ls).
• Add a physical-bounds regression test: saturation when dew point == air temperature, RH below 100 % otherwise and monotonic in the dew point, so the direction can't silently invert.

I left the exp-division micro-optimisation @echedey-ls mentioned out of this first pass since he flagged maintainers may differ on it — happy to add it (or expand the applicability notes with a max-error / valid-range figure) if wanted.

[kandersolar]

@gaoflow in the future please include the PR template checklist when making PRs. I've added it back to this one.

[gaoflow]

Thanks, and apologies for omitting the checklist - I will include the PR template on future PRs.

[cwhanse]

Suggested change

	:py:func:`pvlib.atmosphere.tdew_from_rh`) so the actual and saturation vapor
	pressures are no longer swapped in the source. The returned values are
	:py:func:`pvlib.atmosphere.tdew_from_rh`) related to the actual
	and saturation vapor pressures. Function output is

[cwhanse]

Suggested change

	Relative humidity is computed as ``100 * e / es``, where the actual vapor
	pressure ``e`` is the saturation vapor pressure at the dew point and the
	Relative humidity is computed as ``100 * e / es``, where ``e`` is the
	saturation vapor pressure at the dew point temperature, and ``es``

[cwhanse]

Suggested change

	saturation vapor pressure ``es`` is evaluated at the air temperature, both
	from the Magnus equation ``A * exp(B * T / (C + T))``. The default
	is the saturation vapor pressure at the air temperature, both
	from the Magnus equation ``A * exp(B * T / (C + T))``. The default

[cwhanse]

@gaoflow can you tick off the checklist items that are complete?

[gaoflow]

Done, checklist ticked. I left only the docs/sphinx/source/reference API-changes box unchecked since this is a clarity change with no API change (the whatsnew entry is in v0.15.2.rst). Tests and the new physical-bounds regression are in tests/test_atmosphere.py.

[cwhanse]

@gaflow there are three editorial suggestions, if you can accept or indicate otherwise. Then I think this is ready to go.