Defining compound extreme events on objective spatiotemporal scales

Compound extreme events describe the simultaneous occurrence of two or more individual extreme weather or climate events that often have a significant impact on environment, society or economy. Many studies have investigated such events, often using different spatiotemporal scales for the same event, depending on e.g., the country or region of interest. Although appropriate from an impact point of view, this practice might lead to conflicting or inconsistent results. It is therefore necessary to find objective definitions of extreme events for attribution studies or to investigate how likelihoods of certain extreme events change over time.

Building on previous work for single extreme events, we propose a roadmap for obtaining objective compound event definitions, especially with regards to their spatiotemporal characteristics, by estimating multivariate probability distributions via copulas and then maximizing the rarity of the event across several scales. We present applications to past compound extreme events with considerable impact on e.g., human health and agriculture, such as the European heat wave/high ozone event in summer 2003, and also investigate how probabilities of these events change under different emission scenarios.