Probabilistic rainfall thresholds for shallow landslide triggering

We present probabilistic rainfall thresholds for shallow landslides triggering developed by using the Iverson’s (2000) physically based model. These are applied to a 180 km2 area in Northern Italy. We apply a spatially-distributed version of the Iverson’s (2000) model, using a 5x5 m DEM. To account for the uncertainty about the model parameters, we use a Monte Carlo approach, and we obtain probabilities of failure associated to different combinations of rainfall intensity, rainfall duration, and potentially unstable area. We prove that the physically-based model provides stochastic thresholds that express the probability of landslides triggering, and an estimation of the percentage of unstable areas that can be triggered with a certain probability of failure.