Landslide hazard evaluation in a sample area of the Serchio River basin (Tuscany, Italy) using geological, geomorphological and lithotechnical analysis, statistical approach and GIS technique

The study area is located in the middle part of the Serchio River basin. This area is characterized by high landslide susceptibility, because of the geological, geomorphological and climatic features of the basin, among the rainiest areas in Italy. In fact, it is frequently hit by severe rainstorms which often cause many landslides, involving villages, infrastructures and crops. The proposed experimental study is aimed at contributing to the landslide hazard evaluation using a multidisciplinary approach: geological, geomorphological and litotechnic survey, together with multivariate statistical analysis and GIS technique. It is included in a wider research project, promoted by the Tuscany Region Administration and APAT - Italian Geological Survey. This project is aimed at defining the landslide hazard in the area of the Sheet 250 "Castelnuovo di Garfagnana" (Topographic Map of Italy at 1:50,000 scale). The method applied in ranking slope instability proceeded step by step, as follows: - geomorphologic investigation, in order to realize a landslide inventory map and understand the relationship between geomorphological processes and instability; - lithotechnical survey and characterization with in situ and laboratory tests (compression strenght and grain size analysis), in order to analyze the physical and mechanical properties of rock and soil and link them to the instability phenomena; - organization of the spatial distribution of the considered factors in different layers, each related to a specific factor. The spatial overlay of the layers and their matching with the landslide distribution lead to assume connections between different types of instability factors and landslide occurrence; - finally, GIS supported statistical analysis (spatial analysis, conditional and multivariate analysis) allowed to supply the hypothetical connections with an objective and quantitative response. In short, the outcome of this study is a classification of the land surface in some different hazard domains. It evolved towards two distinct statistical methods, both indirect and quantitative. At first, a conditional analysis was applied: the instability factors were individually analyzed, matching them with the landslide distribution. Subsequently, a multivariate analysis was applied: every factor of instability was compared with the others in relation to the landslide occurrence; thus, a more objective classification of the map units was done. The results of this study were compared with the actual geomorphological and stability characteristics of the considered area and a good fitting was verified. On the other hand, the research is still in progress and needs further investigations to understand advantages and limitations of this methodological approach.