Statistical analyses of 104 natural hailpatterns recorded on the dense hailpad network of the Grossversuch IV experiment in Switzerland (1975–78) were carried out using standardized principal component analysis. Each hailpattern is considered to be a particular realization of a random phenomenon characterized by a set of variables, measured experimentally, in a multidimensional space. The dimension of the space is equal to the number of variables, with 13 variables derived from hail measurements at the ground and five others from meteorological parameters.
In the hail variable spaces, the hailfalls can be interpreted by a size effect (88%) and a shape or spreading effect (7%) due probably to the movement of the storm. In spite of a high level of variability, the hail phenomenon appears to be continuous from small to large events, at least in the region and for the hail variables considered. An additional independent sample of 60 hailfalls, recorded on the same network from 1979 to 1981, gives the same result. The advantage of this method is that it takes into account globally all the available information on hailfall without an a priori reduction in the number of variables.