In her diploma thesis (working group "Atmospheric Risks"), Marie Kapsch investigated the long-term variability of large-scale weather types that are related to hailstorm occurrence over Germany. For this work, she used simulation results from different regional climate models between 1971 and 2050. By using additionally claims data of a building insurance company, Marie Kapsch was able identify specific weather types with an increased frequency of occurrence on hail days. According to linear trend analysis, the hail-related weather types show an increase in the probability of occurrence both in the past and future. During a research visit at the Statistical Department in Exeter, Marie Kapsch was involved in the development of a Bayesian model for the prediction of hail-damage days. With this novel and unique approach, it was possible to determine the annual probability of hail-damage days based on all weather types. The results of the Bayesian model suggest that the number of hail days will increase only slightly in the future (until 2050).
Awards ceremony on 23 May 2012