A real-world application of Monte Carlo procedure for debris flow risk assessment

Abstract

A method for formal risk analysis in debris flow-prone areas is proposed. In this paper risk is defined as the mean annual probability for buildings located in hazardous areas to be damaged by a debris flow. As is well known, specific risk assessment involves the evaluation of both hazard and vulnerability. To quantify debris flow hazard, a Monte Carlo procedure is applied that randomly selects the input variables of mathematical models simulating triggering, propagation and stoppage of debris flows. This allows to estimate the probability density function of the output variables characterizing the destructive power of debris flow (for instance total force, sum of hydrostatic and hydrodynamic forces) at each point of the alluvial fan. Three different vulnerability functions are adopted: two of them are derived from assessments of different types of natural risk, such as snow avalanches and flood waves. The third vulnerability function is obtained from structural analysis of buildings damaged during a mudflow that occurred in Sarno, Italy. The proposed procedure is applied to assess specific risk on the alluvial fan of Ardenno, located in the Valtellina valley, Italian Alps, and the effect that the use of different vulnerability functions has on the risk maps is explored and discussed.