[cheginit]

Closing document for delivery. doc/contributions_2026.md summarizes the features, performance
work, and fixes merged into development since May 2026, with figures and tables, and lists the
open items a deployment needs to know about. The PR also commits the benchmark scripts that
benchmark/README.md and benchmark/RESULTS.md document but that never reached
development, names the NGWPC Hydrofabric consistently, and bumps the pre-commit hook versions.

Additions

• doc/contributions_2026.md: an executive summary, then sections on benchmarking and
  performance, the area-scaled streamflow DA, the Old River water transfer, reservoir routing
  and DA, NextGen integration, fixes, tests and tooling, and open items at delivery. Its 15
  figures and 6 tables are numbered, and each is cross-referenced in the text.
• doc/images/contributions/: the 10 figures the document embeds from demonstrations outside
  this repository. The five benchmark figures are referenced from benchmark/figures/.
• benchmark/scripts/ (18 files): the tier drivers, data preparation, regression check, and
  figure generators. The reorganization that moved them out of benchmark/ into
  benchmark/scripts/ never committed the new directory. The 16 scripts that existed before
  carry only the path updates from that move, plus new override options in
  sweep_max_loop_size.py; paths.py and prep_vpu.py are new.

Removals

• None.

Changes

• benchmark/README.md, test/nhf/README.md, and three benchmark scripts expand NHF as NGWPC
  Hydrofabric and link nhf-builds on first mention.
• .pre-commit-config.yaml: pre-commit-hooks v4.6.0 to v6.0.0 and ruff-pre-commit v0.11.5 to
  v0.16.9. Both hooks still run only on src/troute-rnr/.

Testing

1. Each number in the document was checked against its pull request body, commit message,
   benchmark/RESULTS.md, or the notebook output it comes from.
2. Every figure and table is referenced in the text, and every reference resolves to an
   anchor.
3. The benchmark scripts compile (python -m py_compile), both shell scripts parse
   (bash -n), and the paths.py self-check passes. The suite was not run end to end: the
   local benchmark datasets predate the lake columns the NHF reader now requires (open item 7
   in the document).
4. No code under src/ changes.

Notes

• The performance section reproduces the historical study in benchmark/RESULTS.md, measured
  on Python 3.9 and 3.11 before the move to 3.12, as delivered.
• Section 9 of the document lists the open items. The first two affect the first run after
  deployment: state files written before Route run-of-river dams and NID dam geometry; harden the level pool, restarts and reservoir DA #129 are refused until regenerated, and the NHF
  reader requires the lake, flowpath, and gage columns it names.

Checklist

• PR has an informative and human-readable title
• Changes are limited to a single goal (no scope creep)
• Code can be automatically merged (no conflicts)
• Code follows project standards (link if applicable)
• Project documentation has been updated (including the "Unreleased" section of the CHANGELOG)
• Reviewers requested with the Reviewers tool ➡️