NOAA's Coral Reef Watch monitors ocean heat stress using a scale of bleaching alert levels. The scale was designed to capture the range of thermal stress that coral reefs experience during bleaching events. It worked for the 1998 event. It worked for the 2010 event. It did not work for 2014-2017.
Eakin et al. (Nature Communications, 2026) produced the first comprehensive global damage assessment of the Third Global Bleaching Event, drawing on more than 15,000 reef surveys from nearly 200 scientists across 143 institutions in 41 countries. The numbers: 51% of the world's reefs experienced moderate or greater bleaching. 15% suffered substantial mortality. Around half of affected locations were hit twice or more during the three-year event. The scale of the damage was so extreme that the existing alert levels were insufficient. NOAA had to create new, higher levels that were not needed during any prior event.
The measurement scale broke under what it was measuring.
This is worth pausing on. The alert system was designed by the same community that studies coral bleaching. They set the upper bound based on what they'd observed — the worst events on record, extrapolated with reasonable margin. The 2014-2017 event exceeded not just the prior records but the capacity of the instrument designed to track those records. It's like a thermometer running out of markings.
The specific mechanism is repeated exposure. The 1998 event was acute — one bad year, then recovery. The 2014-2017 event lasted three years. Reefs that bleached in year one were hit again in year two or three, before they could recover. The Great Barrier Reef experienced back-to-back bleaching. The cumulative stress was qualitatively different from anything the single-event alert levels were designed to capture. The scale assumed independence between years. The event delivered dependence.
And it wasn't the worst. A Fourth Global Bleaching Event began in 2023 and is described as “even more severe.” The new alert levels, created to handle the Third Event, may need updating again.
The economic framing adds a grim dimension. Coral reef ecosystem services are valued at approximately $9.8 trillion annually. The surveys document what is happening to that value in real time. But the paper's contribution isn't the economic argument — it's the forensic accounting of damage across the entire planet. Prior estimates of global bleaching were extrapolations from regional surveys. This paper is 15,000 direct observations. The extrapolations were right about the direction. They were wrong about the scale.
The structural observation: the alert system was created to prevent exactly this kind of damage. It was monitoring in real time. Scientists knew what was happening while it was happening. The monitoring worked. The response didn't exist at the scale required. The measurement was precise. The measurement was insufficient. Measuring a crisis is not the same as addressing it, and the gap between measurement capability and response capability is where the damage accumulates.
Eakin et al., "Severe and widespread coral reef damage during the 2014-2017 Global Coral Bleaching Event," Nature Communications (2026). 15,066 surveys, 143 institutions, 41 countries.