06 — Bleaching Tracks Accumulated Heat (`bleaching-degree-heating`)

Supplied working report; original verdicts not validated.
Review status: Needs review. This is an archival source note, not a passed proof.
REPORT.md

# 06 — Bleaching Tracks Accumulated Heat (`bleaching-degree-heating`)

- Source: Incoprea Theory, Volume XXIX (`bleaching-degree-heating`)
- Claim: Coral bleaching tracks degree-heating weeks better than peak temperature, because thermal stress accumulates.
- Prediction: Bleaching should track the integral (accumulated heat, DHW).
- Refutation: Bleaching tracking peak temperature would remove accumulation.

## Test performed

Two prongs. **(a) Literature:** the NOAA Coral Reef Watch degree-heating-week (DHW)
operational index and its validation studies were summarized (Kayanne 2017; Liu et al.
2003, 2013; Heron et al. 2016; Hughes et al. 2017; van Woesik et al. 2022; a 2026
re-analysis of DHW skill). **(b) New analysis (evidence class a):** we downloaded the
published Reef Check × CoRTAD merged dataset (Sully et al. 2019, n = 9,215 site-surveys
1998–2017, hosted on GitHub) and compared how well survey bleaching is tracked by
*accumulated heat* (12-week DHW at survey time) versus *instantaneous temperature*
(SST and SST anomaly at survey time). Note: the `*_Max` CoRTAD columns in this dataset
are 40-year climatological maxima (per-pixel constants) per the GCBD data dictionary
(van Woesik et al. 2022) and were therefore excluded.

## Data (sources, URLs, sample sizes)

- Reef Check bleaching surveys merged with CoRTAD satellite variables, Sully et al. 2019,
  [GitHub: InstituteForGlobalEcology/Coral-bleaching-a-global-analysis-of-the-past-two-decades](https://github.com/InstituteForGlobalEcology/Coral-bleaching-a-global-analysis-of-the-past-two-decades)
  (file `Reef_Check_with_cortad_variables_with_annual_rate_of_SST_change.csv`);
  n = 9,215 rows; tropical subset (|lat| ≤ 30°) n = 9,168; 3,152 surveys with bleaching > 0.
- Variable definitions: van Woesik et al. 2022, *Sci. Data* 9:20,
  [Global Coral Bleaching Database](https://www.nature.com/articles/s41597-022-01121-y).
- Derived outputs: `data/test06_pointbiserial.csv`, `data/test06_event_severity.csv`,
  `figures/test06_prevalence_by_heat_metric.png` (code: `code/analyze_bleaching.py`).

## Result (key numbers; reference figure files)

Literature: DHW is NOAA CRW's operational bleaching index; DHW ≥ 4 °C-weeks = significant
bleaching, ≥ 8 = widespread bleaching and significant mortality (Heron et al. 2016).
Kayanne (2017, Coral Reefs 36:63–70) validated DHW against 1982–2015 bleaching records in
the NW Pacific, and the 2016 event severity matched DHW closely (Ishigaki 10.7 °C-weeks,
Okinawa 10.5, Amami 8.2, Kochi 6.2, Kushimoto 7.8, Ogasawara 4.8, Guam 1.4, Palau 2.3).
Heron et al. (2016) found DHW correlated with % bleaching at r = 0.64 in CNMI and that
"bleaching increased with increasing DHW". Hughes et al. (2017) showed the 1998/2002/2016
GBR bleaching footprints were "determined by the spatial pattern of sea temperatures in
each year".

Our new analysis (9,168 tropical site-surveys):

| predictor (year-specific) | r (point-biserial, any bleaching) | ρ (severity, events only) |
|---|---|---|
| 12-wk accumulated DHW (hot-spot based, TSA_DHW) | **+0.075** (p < 1e-12) | **+0.164** (p < 1e-20) |
| 12-wk accumulated DHW (SSTA-based) | +0.053 (p < 1e-6) | +0.134 (p < 1e-13) |
| weeks ≥ 1 °C in previous 52 wks (also accumulation) | −0.005 | +0.189 |
| SST anomaly at survey | +0.024 (p = 0.02) | +0.075 |
| hot-spot anomaly at survey | +0.024 (p = 0.02) | +0.147 |
| SST at survey (K) | +0.018 (p = 0.08, n.s.) | +0.115 |

In a multivariate logistic regression (standardized coefficients): DHW +0.149, SSTA
+0.015, SST +0.004 — accumulated heat carries essentially all of the heat signal.
Bleaching prevalence is flat across bins of instantaneous anomaly/SST but climbs from
0.33 (DHW = 0) to 0.64–0.70 at DHW 20–30 °C-weeks
(`figures/test06_prevalence_by_heat_metric.png`).

Caveat from the 2026 re-analysis (Front. Mar. Sci.): global DHW skill is imperfect and
declines as climatological SST rises above ~28.5 °C (AUC 0.705 → 0.568) — DHW is the best
heat index but not a perfect predictor.

## Verdict

**Support** — in both the published validation literature and our fresh analysis of 9,168
tropical reef surveys, bleaching tracks *accumulated* heat (12-week DHW) clearly more
strongly than instantaneous temperature or its anomaly; the integral beats the peak.

## Limitations

Our data prong compares 12-week accumulation versus instantaneous at-survey temperature,
not the annual maximum SST per event (the `*_Max` CoRTAD columns are 40-year constants).
The Reef Check survey data are opportunistic, with timing/scale noise; all associations
are weak in absolute terms (max r ≈ 0.08–0.16). Association is not causation; survey
depth, taxa, and adaptation (e.g., local SST-variability effects, Sully et al. 2019) are
not controlled here.

## References

- Kayanne H. (2017) Validation of degree heating weeks as a coral bleaching index in the northwestern Pacific. *Coral Reefs* 36:63–70. https://doi.org/10.1007/s00338-016-1524-y ; summary: [Galaxea 19:17–18](https://www.jstage.jst.go.jp/article/galaxea/19/1/19_17/_pdf)
- Liu G., Skirving W.J., Strong A.E. (2003) Remote sensing of sea surface temperatures during 2002 Barrier Reef coral bleaching. *Eos* 84:137–144. https://doi.org/10.1029/2003EO150001
- Liu G. et al. (2013) NOAA Coral Reef Watch 50 km satellite SST-based decision support system for coral bleaching management. NOAA Tech. Report NESDIS 143.
- Heron S.F. et al. (2016) Validation of reef-scale thermal stress satellite products for coral bleaching monitoring. *Remote Sensing* 8:59. https://doi.org/10.3390/rs8010059
- Hughes T.P. et al. (2017) Global warming and recurrent mass bleaching of corals. *Nature* 543:373–377. https://doi.org/10.1038/nature21707
- Sully S., Burkepile D.E., Donovan M.K., Hodgson G., van Woesik R. (2019) A global analysis of coral bleaching over the past two decades. *Nat. Commun.* 10:1264. https://doi.org/10.1038/s41467-019-09238-2 (data: [GitHub](https://github.com/InstituteForGlobalEcology/Coral-bleaching-a-global-analysis-of-the-past-two-decades))
- van Woesik R. et al. (2022) A global coral-bleaching database, 1980–2020. *Sci. Data* 9:20. https://doi.org/10.1038/s41597-022-01121-y
- Frontiers in Marine Science (2026) re-analysis of DHW predictive performance. https://doi.org/10.3389/fmars.2026.1835611
Included code and results

tests/06-bleaching-degree-heating/code/analyze_bleaching.py