The potential cost associated with damage caused by severe hail storms is becoming an increasingly considered aspect of natural catastrophe modelling in the insurance industry. To calculate this potential cost we need to understand both the damage an event would cause, and the probability of said event. This thesis focuses on the latter and uses recently released meteorological data to develop a regression model for the probability of severe hail. We first discuss the meteorological theory of the development of thunderstorms and how severe hail grows, followed by the necessary statistical theory to analyze and fit regression models to our data. We then preprocess and perform a statistical analysis on our data. A generalized additive model and logistic model are fit for the probability of a thunder-storm and severe hail given that there is a thunderstorm respectively. These two models can then be multiplied to calculate the probability of severe hail. We compare our two separate models to those from another study to verify our results.      «

The potential cost associated with damage caused by severe hail storms is becoming an increasingly considered aspect of natural catastrophe modelling in the insurance industry. To calculate this potential cost we need to understand both the damage an event would cause, and the probability of said event. This thesis focuses on the latter and uses recently released meteorological data to develop a regression model for the probability of severe hail. We first discuss the meteorological theory of th...     »