Wind Storm Risk Management.

Summary Models and forecasts of damage from wind storms are a major issue for insurance companies. In this article, we focus on the calculation sensitivity of return periods for extreme events. Numerous elements come into play, such as data quality (location of insured buildings, weather report homogeneity), missing updates (history of insurance portfolios, change of ground roughness, climate change), the evolution of the model after an unprecedented event such as Lothar in Europe and temporal aggregation (events defined through blocks of 2 or 3 days or blocks of one week). Another important aspect concerns storm trajectories, which could change due to global warming or sweep larger areas. We here partition the French territory into 6 storm zones depending on extreme wind correlation to test several scenarios. We use a storm index defined in \cite{Ma} to show the difficulties met to obtain reliable results on extreme events.