An evolving network approach to assess compounding heat and dry extremes in Europe.

Abstract

The probability of incidence of compound events is increasing due to human-induced climate change: in particular, there is high confidence that concurrent heatwaves and droughts will become more frequent with increased global warming1. Hereby, understanding the aggregated impact of multiple and synchronized compound hot and dry events at different spatial regions is a pressing issue, especially when it comes to predicting these extremes. In order to assess the evolution of these climate hazards, it is crucial to identify the synchronization structures of compound hot and dry events. To achieve this goal,  we highlight the hotspot regions where extremes are increasing and analyse the atmospheric precursors driving these anomalous conditions. Complex networks represent a promising tool in this perspective. In this work, we present an evolving network approach to assess the time evolution of synchronized compound hot and dry extremes due to global warming in continental Europe. Under this framework, we identify those regions where the frequency of these events has increased in the past 80 years and we describe their atmospheric drivers. Using ERA5 reanalysis data2 and focusing on the extended summer seasons (from April to September) of the period 1941-2020, we construct an evolving network constituted by 51 consecutive layers. Each layer models the synchronization structure in space of compound hot and dry events for a certain time window. Once the evolving network is established, the 51 layers are analysed to highlight the main changes in the graph structure. In particular, by looking at different centrality and clustering metrics and their evolution, we identify hotspot regions, and consequently we describe the atmospheric conditions which drive the compound events at these key locations. Climate complex networks prove to be a powerful tool to reveal hidden features of climate processes; this approach indeed brings out key aspects concerning the spatial dynamics of hot and dry events, laying the foundations to build a forecasting method for these extremes.