Revisiting recent changes in European summer temperature distributions and assessing their role for extreme summer temperatures

Following the extreme European summer heatwave of 2003, it has been suggested that the event might have been associated with changes in the distribution of summer temperatures. Here we revisit this hypothesis and investigate observed European and Swiss summer temperatures for the period 1864-2024.

The pronounced increase in skewness has a number of important implications: (1) It implies that extreme hot summers have become more frequent than expected from the median warming. In particular, the increase in skewness strongly affects estimates of the probability of extreme summer heatwaves such as 2003 and 2018. (2) It is demonstrated that the increase in skewness can partly be explained by the accelerating warming around 1980. It is thus not clear whether the high values in skewness will persist into the future. (3) There is a statistically significant difference in the trends of median and mean warming, with mean temperatures warming stronger than the median. (4) These different warming rates explain a non-negligible fraction of the so-called mismatch (i.e., summer temperatures in observations have warmed stronger than in CMIP and CORDEX scenarios). (5) It is demonstrated that understanding this mismatch requires an assessment of extreme summer temperatures, beyond the more commonly used mean summer temperature trends.

We will also provide estimates of the frequency of 2003-like summer heatwaves for the current and future climate, making different assuptions about the persistence of the aformentioned changes in skewness.