Maximum Hail Size Prediction

Abstract

We examine the possibility of building a meteorological prediction tool using data from the Greek National Hail Suppression Program. More specifically, we focus on maximum hail size prediction from operational meteorological radar and/or sounding data. Factor analysis and linear regression are applied in order to identify the optimum number of independent variables and the sequence to build the corresponding meteorological tool. A significant linear relationship is discovered for non-seeded storms relating hail size to various radar parameters, such as the Reflectivity or the group of Vertically integrated liquid density and Cloud top. A relationship for predicting hail size using radar parameters for seeded storms failed to be statistically significant.