A quantitative approach for debris flow inception and propagation analysis in the lead up to risk management

This paper presents a quantitative analysis of debris flows in weathered gneiss through a methodology to identify key aspects in the lead up to risk management. The proposed methodology considers both the triggering and propagation stages of landslide using two physically based models “Transient Rainfall Infiltration and Grid-Based Regional Slope-Stability Analysis” (TRIGRS) and “Smoothed Particle Hydrodynamics” (SPH), respectively. The TRIGRS analyses provided information about the location and initial volume of potential landslides. The SPH model, adopting the initial triggering volumes as input data, allowed the back analysis of the propagation stage in terms of both main pathway and depositional area. Both models can be easily implemented over large areas for risk assessment and are able to provide interesting information for the design of risk mitigation structures. Clearly, the rigorous implementation of these models requires the use of geotechnical data obtained from in situ and laboratory tests. When these data are not available, literature data obtained for similar soils for genesis and stress history with those studied can be used. The applicability of the methodology has been tested on two debris flows which occurred in 2001 and 2005 in the province of Reggio Calabria causing extensive damage involving various lifelines. The model results have been validated with the real events, in terms of both triggering/inception areas and debris fans using two dimensionless indices ( I trig and I dep ). The indices compared, respectively, the real triggering/inception area with the simulated one and the real depositional area with the numerical one. For analysed phenomena, the values of I trig were higher than 90% while I dep assumes values higher than 70% which support the applicability of the proposed methodology.