Demonstrating the plausibility of worst-case month-long heatwave storylines in Western Europe

What are the worst-case heatwaves that are plausible in the present or near-future climate? Model-based experiments using ensemble boosting, a computationally efficient method to simulate unprecedented extremes, suggest that month-long heatwaves that break previous records by more than 5 K across Germany and France are possible in the near future. But how can we assess the plausibility of these heatwaves unprecedented in the observational record? We here test whether the most extreme simulated month-long heatwaves in Germany and France are consistent with current process understanding and with historical heatwaves.
We show that despite their extreme record-breaking characteristics both events cannot be ruled out as implausible. To demonstrate this, we compare these two worst-case events with historical heatwaves in the reanalysis record. To this end, we calculate standardized anomalies relative to a time-evolving climatology of relevant physical variables such as temperature, 500 hPa geopotential, surface solar radiation, and soil moisture. We focus on two different worst-case events — one in Germany and one in France — which exhibit distinct characteristics and physical drivers. The event in Germany features extreme anomalies in most physical drivers, particularly those associated with land-atmosphere feedbacks, and features three short heatwaves in quick succession. In contrast, the event in France mostly features less extreme anomalies in these drivers and consists of one less intense but very persistent heatwave caused by anomalously weak zonal flow combined with above-average southerly winds. Using a multilinear statistical model and comparing with historical analogues, we show that the characteristics and physical drivers of both events are consistent with current process understanding and with historical events.