In Switzerland, hailstorms are one of the costliest damaging natural catastrophes. Furthermore, the complex topography, the available high quality radar measurements of MeteoSwiss, the location in central Europe and the previous research done in this field makes this location an ideal laboratory for hail research. With today’s data and recently developed modeling methods, there is great potential in improving the hail forecast in Switzerland. This work aims to improve the forecast, so that the possibility of preventive measures is increased and the impact of hail is reduced.

To get familiar with the subject and data, I am presently focusing on the verification of radar hail detection algorithms using more than 50'000 crowd-sourced hail size reports from the App of MeteoSwiss. The focus of the second part of this PhD Project has yet to be set in stone: One possibility is to improve the MeteoSwiss forecasting model itself by testing the 2-moment-scheme or for example by incorporating the diagnostic algorithm HAILCAST into the MeteoSwiss COSMO model. The other option is to focus on regional hail producing environments and using methods such as multidimensional EOFs and/or other machine learning methods to define new hail predictors for the model.