A Methodology for the Spatiotemporal Identification of Compound Hazards: Wind and Precipitation Extremes in Great Britain (1979–2019)

Abstract. Compound hazards are two different natural hazards that impact the same time period and spatial area. Compound hazards can have a footprint that can operates on different spatial and temporal scales than their component single hazards. This article proposes a definition of compound hazards in space and time and presents a methodology for the Spatiotemporal Identification of Compound Hazards (SI–CH). The approach is applied to the analysis of compound precipitation and wind extremes in Great Britain from which we create a database. Hourly precipitation and wind gust values for 1979–2019 are extracted from climate reanalysis (ERA5) within a region including Great Britain and the British channel. Extreme values (above the 99 % quantile) of precipitation and wind gust are clustered with the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, creating clusters for precipitation and for wind gusts. Compound hazard clusters that correspond to the spatial overlap of single hazard clusters during the aggregated duration of the two hazards are then identified. Our ERA5 Hazard Clusters Database (given as a supplement) consists of 18,086 precipitation clusters, 6190 wind clusters, and 4555 compound hazard clusters. The methodology’s ability to identify extreme precipitation and wind events is assessed with a catalogue of 157 significant events (96 extreme precipitation and 61 extreme wind events) that occurred in Great Britain over the period 1979–2019 (also given as a supplement). We find a good agreement between the SI–CH outputs and the catalogue with an overall hit rate (ratio between the number of joint events and the total number of events) of 93.7 %. The spatial variation of hazard intensity within wind, precipitation and compound hazard clusters are then visualised and analysed. The study finds that the SI–CH approach (given as R code in supplement) can accurately identify single and compound hazard events and represent spatial and temporal properties of compound hazard events. We find that compound wind and precipitation extremes, despite occurring on smaller scales than single extremes, can occur on large scales in Great Britain with a decreasing spatial scale when the combined intensity of the hazards increases.