Fig. 1: Summer 2019 in Europe.

a European summer (JJA) land temperature anomalies (°C, with respect to 1981–2010) for 1500–2019 (vertical lines) and their probability density function (percentage, histogram). The five warmest (coldest) summers of 1500–2019 are highlighted in red (blue). Light (dark) grey shading shows a Gaussian fit of the distribution for 1960–1989 (1990–2019). These distributions are displayed for illustration purposes only, and they should not be considered a robust representation of the true distributions given the limited sample size. b Smoothed running decadal frequency of extreme summers (>95th percentile of 1500–2019), with dotted line showing the maximum decadal value that could be expected by random chance (p < 0.05). c Average TX (daily maximum temperature) anomalies for the 2019 and 2003 summers (calculated with respect to their corresponding previous 30-year climatological means), and their difference. d Same as (c) but for the highest TX anomalies registered in each summer and grid cell, regardless of the day of occurrence.