The signal, without the headline overreach
El Nino still changes where and when marine heat accumulates. The new result is that it no longer supplies the larger contribution to estimated bleaching heat stress across most of the coral regions in which both signals can be resolved. Pershing, Bruno, Giguere and Khrizman compare the warming contribution with a strong El Nino, defined by an Oceanic Nino Index of 2, across 74 regions where both terms are statistically significant. At the study's 2025 global-temperature level, the warming contribution is larger in 71 of 74 regions, or about 96%.
The three exceptions are Cocos Islands, Northern Galapagos and Line Islands. They show why the careful statement is warming dominates across most of the analysed dual-signal regions, not that ENSO is irrelevant or that every reef on Earth has crossed the same threshold.
| Reference point | Warming exceeds strong El Nino | Evidence status |
|---|---|---|
| 1989, GMT15 about 0.50°C | 50% of dual-signal regions | Historical crossover reported by the study |
| 1998 global event | 61 of 74, 82% | Historical attribution |
| 2025 study endpoint, GMT15 1.35°C | 71 of 74, 96% | Historical attribution endpoint |
| 2028 | 74 of 74, 100% | Projection under SSP3-7.0 |
The 2028 endpoint belongs to the paper's SSP3-7.0 pathway. It is not a forecast common to every emissions scenario, and it does not mean every coral region on the planet becomes ENSO-independent in 2028. It means the modelled warming contribution exceeds a strong El Nino across the paper's 74-region dual-signal subset under that pathway.
The fourth event disappears in the counterfactual
The sharper attribution test reconstructs a climate in which the anthropogenic warming contribution is removed while the observed weather pattern is retained probabilistically. The authors then run both observed and counterfactual temperatures through the NOAA Coral Reef Watch degree-heating-week framework.
70 of 71
Reported coral regions in the 2018 to 2025 event that fall below moderate bleaching risk when the study removes the anthropogenic warming contribution.
Spady and colleagues reported bleaching across 71 regions during the 2018 to 2025 study window. With anthropogenic warming removed, 70 fall below the moderate-risk threshold of 4°C-weeks. Only Central Kuroshio remains at moderate risk. Its observed maximum exceeded 17.4°C-weeks, well into the paper's extreme-risk category.
This supports the authors' conclusion that the recent event requires anthropogenic warming in the attribution framework. It does not mean those 71 reefs were measured twice. The observed record comes from NOAA sea-surface temperatures and bleaching reports; the climate without anthropogenic warming is modelled.
How the attribution was built
The analysis represents 100 Marine Ecoregions of the World coral regions using one-degree ocean cells. Daily sea-surface temperature comes from NOAA's quarter-degree Optimum Interpolation Sea Surface Temperature record. Global warming is represented by GMT15, a centred 15-year mean of HadCRUT5 global temperature relative to 1850 to 1900.
For each local cell and temperature state, the authors construct daily sea-surface-temperature distributions from 13 CMIP6 simulations and two empirical scale-factor methods. A counterfactual day is selected by matching its exceedance probability to the equivalent distribution at GMT15 equal to zero. Observed and counterfactual daily series then enter the same maximum degree-heating-week calculation.
| Layer | Input or method | What it contributes |
|---|---|---|
| Observed ocean state | NOAA OISST, daily, 0.25 degree | The historical local sea-surface-temperature sequence |
| Global warming index | HadCRUT5 GMT15 | The smoothed anthropogenic warming coordinate |
| Local attribution | 13 CMIP6 simulations plus two empirical methods | How each local daily distribution changes with GMT15 |
| Risk translation | NOAA Coral Reef Watch DHW framework | Accumulated heat stress over rolling 12-week windows |
| Future crossover | SSP3-7.0 after the observed period | Pathway-dependent dates for warming dominance |
The design is best described as observation-led attribution using modelled counterfactuals. That is stronger and more precise than calling the result either a direct measurement of causation or a free-running climate-model forecast.
What changed between 1998 and 2025
In 1998, the first recognised global mass bleaching event and an archetypal strong-El-Nino year, the warming contribution was already larger than the strong-El-Nino contribution in 61 of the 74 dual-signal regions. The paper marks the event's starting GMT15 at 0.72°C. By the 2025 endpoint GMT15 reaches 1.35°C and the count rises to 71.
The paper also compares warming with local natural variability. At the current temperature level it reports 90 of 97 regions, about 93%, above that local variability benchmark, with all expected to cross by about 2040 under its projection. The article text contains slightly different regional totals in an adjacent sentence, so Solar Analytica retains the explicit 90-of-97 result rather than silently reconciling the authors' denominator.
The latest NOAA operating context
NOAA Coral Reef Watch tracked the fourth global coral bleaching event from 1 January 2023 to 30 September 2025 and estimates that bleaching-level heat stress affected about 84.4% of the world's coral reef area across at least 83 countries and territories. In June 2026, NOAA said the event likely concluded in mid-2025. That update matters: the event is now a closed attribution window for this report, not an ongoing bulletin.
The NOAA global total and the paper's regional analysis are different denominators. NOAA describes the observed footprint of a global event. Pershing et al. analyse a defined set of regional cells and compare observed heat stress with a counterfactual. The two records are complementary, not interchangeable.
Critical assessment
| Strength | Limitation carried with it |
|---|---|
| Daily observed SST anchors the historical sequence. | The no-anthropogenic-warming sequence is inferred, not observed. |
| Multiple climate models and empirical methods inform local scaling. | Results still depend on the attribution model, GMT15 construction and distributional assumptions. |
| The same DHW risk model is applied to observed and counterfactual series. | DHW is a heat-stress proxy; bleaching also varies with depth, species, acclimatisation and local conditions. |
| Regional cells provide broad global reef coverage. | One-degree cells cannot resolve every reef microclimate or within-region exposure. |
| Historical attribution and future projection are connected in one framework. | The 2028 and 2040 crossover dates are pathway and method dependent. |
The defensible conclusion is substantial but bounded: anthropogenic ocean warming is now the dominant estimated driver of accumulated bleaching heat stress across almost all regions where the study can separately identify warming and ENSO, and the recent global event largely disappears in the modelled counterfactual. The paper does not demonstrate that ENSO no longer matters, that every reef responds identically, or that one emissions pathway supplies an inevitable calendar date.
Why Solar Analytica carries this signal
Coral bleaching is not a PV-output variable and does not enter Solar Analytica's solar-resource, grid-mix or yield tables. It belongs in the wider climate evidence record because it tests whether cumulative ocean warming is now larger than familiar natural variability in a globally distributed physical system.
The connection is methodological rather than rhetorical. The Pacific low-cloud assessment separates mechanism evidence from a global climate-sensitivity claim. The State of the Solar Climate separates measured operating conditions from future scenarios. This report applies the same rule: observed heat stress, attributed counterfactual and pathway projection remain visibly different layers.