Comparing 600 years of extremely hot Central European summers to future projections

We use a unique combination of (paleo-)reanalysis (ModE-RA, ERA5) and model simulations (ModE-Sim) to examine extreme summer heat events in Central Europe over the past 600 years (1421-2008) and compare their occurrences to CMIP6 future climate projections. Using a common, fixed climatology, 2003 was identified as the hottest summer of the past 600 years. However, using a moving climatology approach, we identify 1540 (April–September, +2.2 $$^\circ$$C) and 1590 (June–August, +2.8 $$^\circ$$C) as the most extreme summer anomalies. The two summers differ in their temporal development, their impacts, and partly their atmospheric circulation patterns. We then analyze similarly anomalous summers in two sets of CMIP6 projections and the ModE-Sim ensemble and connect their occurrence to sea surface temperature anomalies over the North Atlantic. ModE-Sim, which comprises 11,760 model years, produces 0.14% (April-September) and 0.24% (June-August) Central European summers with temperature anomalies from a moving climatology surpassing 1540 and 1590. In CMIP6 projections such anomalies are more frequent, but none reaches the magnitude of the largest June-August anomaly in ModE-Sim which exceeds 4 $$^\circ$$ $$^\circ$$C. A similarly extreme anomaly in a future climate could dramatically affect ecosystems and societies. Overall, the combination of reanalysis and model simulations provides a unique and comprehensive framework for understanding the drivers of past and future summer heat extremes.