Projecting the effects of a warming climate on the hurricane hazard and insured losses: Methodology and case study

Abstract

This paper summarizes research (conducted previously by the author) to incorporate (1) sea-surface temperature projections, (2) event-based hurricane simulation models, and (3) loss models to assess the impacts of the changing climate on the hurricane hazard and estimated insured losses. The focus of the work is on the US east coast and hence the formation and movement of hurricanes over the Atlantic Basin. As an illustrative case, the area around Charleston, South Carolina is considered. Specifically, this paper presents a methodology to probabilistically estimate regional hurricane loss considering both hurricane intensity and size as well as projected sea surface temperature change as a function of climate change.It is most common to characterize the hurricane wind hazard using a measure of intensity only. Recent studies have demonstrated the importance of accounting for storm size when characterizing the hurricane hazard in order to properly estimate/predict the spatial extent of damage. For purposes of regional loss estimation, therefore, considering maximum wind speed only may not be adequate. A storm with larger size but lower intensity might result in more damage and hence greater total loss than a smaller, more intense storm. Here, the maximum wind speed (intensity) and the radius of maximum winds (spatial extent) are selected as the two dominant indicators (parameters) characterizing the hurricane (event) hazard. The resulting bivariate hazard model is then used as input to a regional loss estimation model to illustrate its potential application. Only hurricane wind damage (and associated losses from both wind damage and water intrusion) is considered in this study. Collateral damage due to flood or storm surge is not addressed here.Simulated events impacting this region are extracted from a synthetic hurricane database (developed by the author and his colleagues) and the joint distribution of intensity and size is established to characterize the hurricane hazard. Regional loss for each simulated hurricane event is then estimated using the loss estimation module in HAZUS_MH. Finally, distributions of regional loss at different (event) hazard levels are generated.A future climate scenario (increasing sea surface temperature) is next considered and the joint distribution of hurricane intensity and size and the resulting loss distribution curve is recomputed. The results are then able to be compared to those under the current climate scenario. Such comparisons can help to inform codes and standards committees, for example, as they consider whether and how to take potential climate change impacts (in the design lifetime) into account. They also can inform ongoing public discourse around mitigation and adaptation, whether hardening of buildings and other infrastructure systems, community resilience planning, construction practices, or related policies.