Skip to content

Environment: cloud cover vs catches, moonlight and nursery water - #105

Merged
Luen merged 5 commits into
mainfrom
cloud-cover
Nov 3, 2026
Merged

Luen merged 5 commits into
mainfrom
cloud-cover

Conversation

@Luen

@Luen Luen commented Nov 3, 2026

Copy link
Copy Markdown
Collaborator

Closes #104. Stacked on #103 (the Moon chart cites the lunar-cycle paper). Merge #103 first; this diff then shrinks to the cloud work.

What the data shows (computed live)

  • Crew score vs ERA5. In the 2020–22 set logs the crew scored cloud from 1 to 4. ERA5 cloud cover during the same sets rises with every step of that score (ρ = 0.56), so the reanalysis is a fair stand-in on nights nobody scored.
  • Cloud and moonlight. Bright-moon nights average the same catch rate under a clear sky and under cloud (Welch p = 0.52). A moonlight index (illumination × moon height × clear sky) shows no link with catch rate (ρ = 0.04, 1,185 sets). Accounting for cloud does not reveal a moon effect, which agrees with Debaere et al. 2024.
  • Cloud and nursery water. Compared with clear days in the same month, overcast days peak about 1 °C cooler on the nursery sand, with a day–night swing about 1 °C smaller. The raw correlation is ~0 because the cloudy wet season is also the warm season, so the comparison is made within months.

Changes

  • analytics/cloud.ts: moonlight(), crewVsEra5, moonCloud, cloudWater.
  • Weather: "The crew's cloud score vs the reanalysis".
  • Moon: "Does cloud hide the moon?", with a link to Debaere et al. 2024.
  • Water quality: "Clouds and the nursery's afternoon", with a link to Trujillo et al. 2025.
  • Drivers: "Moonlight reaching the water" and "Cloud cover noted by the crew (1–4)".
  • Finding: "Clouds take the edge off nursery afternoons".
  • Tests for the new calculations; AGENTS.md notes.

Checks

lint, check, unit tests (cloud tests included), build, e2e (29 passed, axe light and dark). Checked the numbers against the real data and the charts by hand.

🤖 Generated with Claude Code

Luen and others added 5 commits November 3, 2026 12:58
- publications.ts: ids, rummerlab.com PDF file names and 15 more Mo'orea
  papers (lunar cycle, umbilicus age, home range, nursery criteria,
  thermal tolerance, escape performance, heatwave stress, aggregation,
  photo-ID, injuries, genetics, juvenile density, oxygen supply).
- ChartCard papers prop: "Paper: Authors (year), Journal (PDF)" under
  the chart, linking the PDF on rummerlab.com or the DOI. 51 charts on
  17 pages cite the study that analysed the same question; mappings were
  checked against each abstract.
- About lists PDF links; AGENTS.md documents the convention.
- e2e: exact match for the Biology menu entry (journal names contain it).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- analytics/cloud.ts: crew cloud score vs ERA5, catch rate on dark nights
  and bright-moon nights under clear or overcast skies, and cloud vs the
  in-situ daily high and swing compared within the same month.
- Weather: "The crew's cloud score vs the reanalysis".
- Moon: "Does cloud hide the moon?"
- Water quality: "Clouds and the nursery's afternoon".
- Drivers: "Moonlight reaching the water" and the crew's cloud score.
- Finding: overcast afternoons are cooler on the nursery sand.

Closes #104

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@Luen
Luen merged commit 6b43b2a into main Nov 3, 2026
5 checks passed
@Luen
Luen deleted the cloud-cover branch November 3, 2026 03:19
@Luen Luen mentioned this pull request Nov 3, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Environment: cloud cover vs catches, moonlight and nursery water temperature

1 participant