Constructing statistical models for hailstorm occurrence in US and Europe.

Hailstorms are among the costliest extreme weather events, causing major damage to agriculture and infrastructure, and leading to substantial losses in the insurance sector.  

Hailstorm modeling is extremely challenging and requires computationally expensive high resolution physical modeling.   Machine Learning approaches have recently emerged as a way to bypass some limitations of physical models, combining hail reports as target data with meteorological predictors. Limitations to this approach appear from the scarcity of consistent observational data in most regions. 

In this study, we compare domain-shift adaptation methodologies to propose an optimal approach to produce hailstorm models in data scarce regions. Results show that models combining data from data-rich and data-scarce regions offer the best balance between regional skill and cross-domain generalization. 

Furthermore, we introduce a probability calibration methodology to improve interpretability of the model inferences and demonstrate how these models can be used to construct hailstorm hazard maps, providing valuable tools for stakeholders. 

In addition to historical climatology, we present results for hail climate projections. 

This work has been partially funded by the EDF Project KOIOS GA 101103770