From Observation to Simulation: Building a Continental-Scale Hail Model for Re/Insurance Applications

In recent decades, insured losses from hailstorms have shown a marked upward trend across Europe. Notable events such as Qiara and Maya in France (2022), and Unai in northern Italy (2023), underscore the growing potential of severe convective storms to cause catastrophic damage. These events highlight the increasing relevance of hail as a key peril for the insurance and reinsurance sectors, necessitating a modernised and data-driven approach to risk assessment.

To address this evolving risk landscape, Swiss Re has developed a new severe convective storm model that incorporates a pan-European probabilistic event set. This model aims to realistically capture the insurance-relevant characteristics of hail events at a continental scale, including their frequency, severity, spatial extent, and clustering behaviour. Key parameters such as the geometry of hail streaks, the distribution of hailstone sizes within these streaks, and the spatial correlation of impacts are explicitly modelled to reflect the complexity of real-world events.

One of the primary challenges in constructing such a model lies in the scarcity and heterogeneity of observational hail data across Europe. To overcome this, we employed statistical learning techniques to infer the daily probability of hail occurrence from convective environments using reanalysis datasets. This approach enabled the creation of a robust, multi-decadal hail climatology, which serves as the foundation for geographically differentiated risk estimates and realistic event footprints.

Building on this climatology and supplemented by literature-based insights into hail intensity distributions, we generated a stochastic event set simulating tens of thousands of years of hail activity. This synthetic catalogue supports underwriting applications by enabling consistent risk evaluation at both local and portfolio scales. The model’s outputs have been validated against existing climatologies to ensure their suitability for operational use in insurance pricing and risk management.

This contribution presents the methodology, validation, and implications of Swiss Re’s new hail model, offering insights into how advanced statistical and climatological techniques can enhance the industry's understanding and quantification of hail risk in Europe.