Shifting (ir)regularity regimes in extreme climate impacts under global warming

The spatio-temporal patterns of extreme climate impacts have been studied in various ways, yet two questions lack a profound analysis: Do such events occur regularly and if so, how do existing regularity patterns change under global warming? Here, we address these questions with a new statistical method for detecting dominant periods in noisy spatio-temporal data. Under pre-industrial conditions we find regularity patterns in close to 57% of cropland affected by crop failure, likely related to the influence of climatic oscillations such as the El Niño-Southern Oscillation, while extreme heatwaves occur without significant regularity. Wildfires exhibit regularity in 90% of all affected areas, yet linked to the specific regrowth dynamics of different biome types. Under climate change, the warming trend determines (ir)regularity, leading to a decrease of regularity in crop failure and wildfire by additionally 4% and 2%, respectively, and the emergence of regularity in 10% of the areas affected by heatwaves at the transition from the pre-industrial era to the anthropocene. In the anthropocene, crop failure, wildfire and heatwave occurrences increase and undergo a shift towards monotonic growth in all impacts replacing pre-existing natural regularity patterns. Moreover, de-trended impact projections indicate an additional shift towards higher frequencies of extreme event patterns due to climate change. The observed shifts in extreme event regularity are important for the planning of adaptation strategies and our method offers a new perspective on studying extreme climate impacts.