Models for landslides induced by precipitation in Norway

This work summarizes the preliminary results of a post doc project between University of Calabria and Norwegian Geotechnical Institute (NGI). The topic is the development of an integrated modeling to forecast rainfall fields and triggering mechanisms of landslides. The activities are organized in three research lines: (1) analysis and modeling of rainfall fields; (2) development of a modeling for landslide initiation; (3) integration between rainfall forecasting and landslide initiation models. The preliminary results reported in this paper are related to the second research line, which comprises: a) development of a physically and DEM based model at regional scale for triggering of landslides; b) integration with empirical/hydrological models for early warning purposes. Concerning the former activity, the study area, formed or not by different soil layers, has to be represented by a threedimensional grid of cells; the study of the effects induced by precipitation on slope stability has to be performed by using models of infiltration and runoff generation, coupled with a global stability analysis. The basic idea of this second activity is to test the feasibility of establishing relations between the results of empirical approaches, and the output of physically based models. The advantage of this procedure consists in the use of the physically-based model only for the calibration of the aforementioned relations, while for early warning purposes it could be possible to use only the simplest model (i.e., the empirical) with a significant reduction in computational costs. The work is outlined as follows. Section 2 describes the development of an infiltration module for unsaturated layered soils and transient flow, which is a component of the physically- and DEM-based model (in preparation), comprising also modules for runoff generation and slope stability. Section 3 illustrates the results of a previous NGI project, related to the probabilistic estimation of thresholds for triggering of rapid soil slides in Norway; starting from these results, we emphasized the necessity of coupling with other data source, like radar, in order to improve the forecasting and the performance of models for landslide triggering.