Identifying Hotspots of Large Hail Size in the Continental U.S.: A Bayesian Approach

Severe hailstorms represent a significant and costly hazard, often resulting in extensive damage to agriculture and infrastructure, leading to substantial economic and insured losses. In 2023 alone, losses for severe convective storms exceeded 34 billion USD insured, with the vast majority of this a result of hail. As both population density and building exposure continue to grow over time, it becomes imperative to estimate local-scale hail frequency for effective risk management planning, as well as mitigation strategies for design and choice of building materials. However, the estimation of hail size faces significant challenges due to inconsistent and heterogeneous hail size reporting. This study aims to overcome these limitations through the application of a Bayesian estimation technique associated with a Monte Carlo Markov Chain approach in evaluating the return level of hail size. Based on extreme value theory, the prior information from the previously available (1°×1°) hail size report is incorporated along with local scale information (0.25^° × 0.25^°) to minimize the uncertainty associated with hail size estimation. The analysis aims to identify the percentage of area of each state susceptible to more than 1.75 inches hail size for specific return periods. Based on the return levels and exposure, hotspots are identified based on hail severity. To demonstrate the accuracy of the Bayesian estimation technique, the outcomes are compared with a maximum likelihood method. This comprehensive analysis seeks to enhance our understanding of hailstorm risks and improve our ability to manage and mitigate their impacts.