Document the September 2026 OM4 data release

[alxmrs]

Summary

• mark v2026-09 as the canonical processed OM4 release and document all 12 averaged, filtered, and snapshot datasets
• explain freshwater forcing and snapshot timing semantics, plus the intended filtered/unfiltered multi-scale pairing
  • additional human context: The documented note on the intention for what data to use for multi-scale modeling is useful to review (it was a recommendation from Laure and Alistair).
• update the shipped OM4 demo, generated Colab quickstart, SkyPilot defaults, configuration docs, and normalization example to use the September release
• point clone_data.py at the canonical Samudra namespace and fix its existing missing .zarr suffix
• document all filtered preprocessing and normalization commands

Observation data remains pinned to v2026-07/obs; this PR changes the canonical OM4 release only.

Verification

• confirmed public HTTP 200 access to .zmetadata for all 36 stores (12 datasets × data/means/stds)
• uvx pre-commit run --all-files
• python -m pytest tests/test_clone_data.py tests/test_quickstart_config.py -q
• python scripts/build_quickstart_notebook.py --check

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e80f3f99c2

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[alxmrs]

I reran each documented xr.open_zarr(...) call against the live September stores and updated all four displayed Xarray representations verbatim. The refreshed output now shows the current sizes, wfo, 100 data variables, 17 attributes, and current provenance fields. I also verified programmatically that each live repr(ds) appears exactly in docs/data.md. 🤖

[alxmrs]

quick review note.

[alxmrs]

Let's omit this file.

[alxmrs]

Removed in 6673c01. The production path correction remains, and the remaining focused checks and pre-commit suite pass. 🤖

[jder]

LGTM, thanks so much for the details docs!