European seas, a study attributes this summer's record temperatures to human-caused warming

A story the morning edition did not carry. On August 19, 2026, the collective of researchers World Weather Attribution published a study attributing to human-caused climate warming the record sea temperatures measured this summer in European seas, from the Atlantic off Ireland to the Mediterranean (Reporterre, Reuters via Internazionale).

The figures underpinning the work are three, and it is worth separating them.

The first is an average: in July 2026 the average temperature of the Mediterranean was 27 degrees Celsius, the highest ever recorded for a month of July. An average across an entire basin is the most robust of the three figures, because it smooths out local anomalies.

The second is a single peak: a buoy off Majorca recorded 33 degrees Celsius in early August 2026. This is a single measurement, tied to one instrument and one point, and as such it does not describe the state of the basin: it describes how high the peak reached.

The third is a deviation from the norm: in June 2026 some areas of the Mediterranean exceeded seasonal averages by five degrees. Here too, the indication concerns “some areas,” not the entire sea.

The most significant part of the study, however, lies not in the temperatures but in the frequency. According to the authors, sea surface temperature conditions that statistically occurred once every 350 or every 1,200 years — depending on the area considered — now have a return period of five years or even one year. This is the kind of statement that attribution studies produce: they do not say that a single event was caused by climate change, they say how much its probability has changed in a warmed climate compared with one that was not.

The fourth figure looks ahead: continuing along the current warming trajectory, events like those observed in 2026 would be 1.3-1.9 degrees warmer. This is a projection conditional on a scenario, not a forecast, and should be read together with the range that accompanies it.

Regarding the limits of the analysis, one must be explicit. The available sources report the study’s conclusions and statements by one of the co-authors, University of Bern researcher Thomas Frölicher, according to whom the rise in temperatures in European seas — from the Irish Atlantic to the Mediterranean — is of alarming scale; they do not indicate whether the work has completed a peer-review process. Rapid attribution studies are designed to be published close to the event, and this timeliness is at once their usefulness and their methodological limit.

The European dimension of the data lies in its geographic extent: the study does not concern a single sea, but a continuous stretch running from the Irish Atlantic to the Mediterranean, that is, basins with different depths, currents and thermal regimes from one another. The available sources do not specify what consequences the authors associate with these temperatures, and this piece does not attribute any: the verifiable content of the study concerns the measured values, their return period, and the projection to 2026 under the current warming scenario.

The most concrete figure to retain is that of the return period. When an event estimated to occur once every 350 years becomes a five-yearly occurrence, the statistical parameters on which the historical reference series were calculated no longer describe the situation observed: it is the frequency, more than the single record, that is the figure changing the basis of calculation.