Spatio-temporal modelling for avalanche risk assessment in the North of Italy

The main objective of this work is to evaluate the avalanche activity in a given location and at a given time taking into account a number of variables including the stratigraphy of snow cover, temperature, direction and wind speed, altititude, etc. To this end we propose a space-time point model where the intensity function indicates the limiting expected rate of occurrence of snow avalanches of a given size occurring on a certain day at location (x, y), conditioned on the historical information available prior to time t. Some meteorological and environmental data may be considered as the covariates of the model. To show the ability of the model in assessing the risk avalanche, data from digitalized Avalanche Database of the Trentino Region (North of Italy) is considered. Since not all locations in the Alpine zone are equally likely subject to snow avalanche, the model will be flexible enough for including a spatially-varying background rate of avalanche which may be estimate by kernel smoothing the observed avalanches.