Early warning systems are nowadays the most implemented strategy to manage risk coming from natural hazards induced by extreme weather conditions. With respect to shallow landslides triggered by rainfall events, the Italian SANF (National Early Warning System for Rainfall Induced Landslides) system has been developed by CNR-IRPI research group (Rossi et al. 2012). It is aimed at forecasting rainfall-induced landslides based on the comparison between the rainfall measurements from rain gauges and the empirical rainfall thresholds traced through the method by Brunetti et al. (2010). Recently, a new automated method by Vessia and Parise (2012) has been proposed to trace Cumulate (E) vs Duration (D) rainfall empirical thresholds within SANF system. It is applied to a sample of 100 rainfall events triggering shallow landslides occurred on Italian territory. This study investigates the statistical properties of the two samples of ED couples from the two methods that provide two power laws as mean thresholds (Fig. 1). According to this study, the two samples belong to the same probability distributions, that are the Lognormal for E and the Weibull/Lognormal for D. These results support the use of both methods to forecast possible shallow landslides within SANF system.
File in questo prodotto:
Non ci sono file associati a questo prodotto.