feat(weather): add focused case diagnostics

[stacknil]

Summary

• Treat the focused local PDF as the primary technical source, without preserving its identity fields, original prose, figures, or datasets.
• Add dry pressure-coordinate temperature-tendency decomposition: zonal, meridional, vertical, adiabatic, and summed dry-dynamic terms.
• Add horizontal-advection component outputs and ridge alpha-grid metric review for the physics-informed baseline modeling arc.
• Add focused-case abstraction docs and update reviewer, methodology, calculation, diagnostic, example, config, and sanitization surfaces.

Validation

• python -m ruff check .
• python -m pytest -> 35 passed
• python -m compileall src scripts
• CLI help smoke checks for thermodynamic, dynamics, precipitation, climate-statistics, and ensemble scripts

Privacy and Data Boundary

• Candidate-file scan returned 0 restricted artifact files, 0 text privacy matches, and 0 non-ASCII text files.
• No raw weather datasets, original PDF content, local path, identity fields, course materials, generated binary artifacts, or report figures are included.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a27b695671

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Use baseline split default in ridge alpha-grid helper

ridge_alpha_grid is documented and positioned as using the same time-ordered split as the baseline ridge helper, but its default train_fraction is 0.7 while ridge_regression_fit_predict defaults to 0.8. Calling both with defaults therefore evaluates different train/test windows, which can change RMSE ordering and make alpha-grid diagnostics non-comparable to the baseline results users are likely to contrast against.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a27b695671

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Guard against single-level pressure inputs

Add an explicit check that pressure_hpa has at least two levels before computing np.gradient; with a single pressure level (a common case when users subset to one isobaric surface), np.gradient(..., axis=0) raises at runtime and aborts the entire diagnostic instead of returning a controlled validation error. This makes temperature_tendency_terms fail unexpectedly for valid-looking inputs and can break downstream experiment pipelines.

Useful? React with 👍 / 👎.