A high-resolution compound vulnerability function for severe convective storm losses

Severe convective storms (SCS) are extreme weather events that can produce hazardous hail, precipitation, wind, and lightning, either alone or combined. In recent decades, growing exposure and climate change have rapidly increased SCS-related economic losses in the United States and Europe. Consequently, it has become more important for risk management to accurately predict their potential losses. Vulnerability functions form the foundation for these loss predictions, as they describe the relationship between a natural hazard’s intensity and damage to an asset. We construct a novel vulnerability function for residential buildings that can account for SCS damage from hail, precipitation and from their combination. This compound vulnerability function is estimated on a large sample of object-level insurance claims and on ultra-high-resolution meteorological observations, using truncated beta regressions. Comparing our vulnerability function with a conventional specification, which only accounts for hail damage, reveals that the latter overestimates damage by 12% and underestimates damage by [10%] {15%} for SCS with 24-hour cumulative precipitation levels of 0 mm [100 mm] {150 mm}. Additionally, our compound vulnerability function is better equipped to predict SCS losses for climate change scenarios because it can accommodate the anticipated changes in the joint distribution of damaging hail and precipitation.