Local scale multiple quantitative risk assessment and uncertainty evaluation in a densely urbanised area (Brescia, Italy)

The study of the interactions between natural and anthropogenic risks is necessary for quantitative risk assessment in areas affected by active natural processes, high population density and strong economic activities. We present a multiple quantitative risk assessment on a 420 km2 high risk area (Brescia and surroundings, Lombardy, Northern Italy), for flood, seismic and industrial accident scenarios. Expected economic annual losses are quantified for each scenario and annual exceedance probabilityloss curves are calculated. Uncertainty on the input variables is propagated by means of three different methodologies: Monte-Carlo-Simulation, First Order Second Moment, and point estimate. Expected losses calculated by means of the three approaches show similar values for the whole study area, about 64 000 000 € for earthquakes, about 10 000 000 € for floods, and about 3000 € for industrial accidents. Locally, expected losses assume quite different values if calculated with the three different approaches, with differences up to 19%. The uncertainties on the expected losses and their propagation, performed with the three methods, are compared and discussed in the paper. In some cases, uncertainty reaches significant values (up to almost 50% of the expected loss). This underlines the necessity of including uncertainty in quantitative risk assessment, especially when it is used as a support for territorial planning and decision making. The method is developed thinking at a possible application at a regional-national scale, on the basis of data available in Italy over the national territory.