Hydraulic hazard mapping in alpine dam break prone areas: the Cancano dam case

Dam-break hazard assessment is of great importance in the Italian Alps, where a large number of medium and large reservoirs are present in valleys that are characterized by widespread urbanized zones on alluvial fans and along valley floors. Accordingly, there is the need to identify specific operative approaches in order to quantify hydraulic hazard which in mountain regions inevitably differ from the ones typically used in flat flood-prone areas. These approaches take advantage of: 1) specific numerical algorithms to pre-process the massive topographic information generally needed to describe very irregular bathymetries; 2) an appropriate mathematical model coupled with a robust numerical method which can deal in an effective way with variable geometries like the ones typical of natural alpine rivers; 3) suitable criteria for the hydraulic hazard assessment; 4) representative test cases to verify the accuracy of the overall procedure. This contribution presents some preliminary results obtained in the development of this complex toolkit, showing its application to the test case of the Cancano dam-break, for which the results from a physical model are available. This case was studied in 1943 by De Marchi, who investigated the consequences of the potential collapse of the Cancano dam in Northern Italy as a possible war target during the World War II. Although dated, the resulting report (De Marchi, 1945) is very interesting, since it mixes in a synergistic way theoretical, experimental and numerical considerations. In particular, the laboratory data set concerning the dam-break wave propagation along the valley between the Cancano dam and the village of Cepina provides an useful benchmark for testing the predictive effectiveness of mathematical and numerical models in mountain applications. Here we suggest an overall approach based on the 1D shallow water equations that proved particularly effective for studying dam-break wave propagation in alpine valleys, although this kind of problems is naturally subject to "substantial uncertainties and unavoidable arbitrarinesses" (translation from De Marchi, 1945). The equations are solved by means of a shock-capturing finite volume method involving the Pavia Flux Predictor (PFP) scheme proposed by Braschi and Gallati (1992). The comparison between numerical results and experimental data confirms that the mathematical model adopted is capable of capturing the main engineering aspects of the phenomenon modeled by De Marchi.