Distribution, statistics and control factors of landslide processes in the mountain landscape of southern Apennines, Italy

Recent research trends in landslide science highlight a growing diffusion of automated techniques for the detection and mapping of landslides at different spatial and temporal scales. Although emerging techniques based on AI algorithms and remote sensing techniques can facilitate the creation of landslide inventory, conventional geomorphological methods of production of landslide maps still play a central role in the compilation of reliable census of landslide processes at a regional scale. After a synoptic view of the limitations and advantages of the new techniques of landslide mapping, this work focused on the statistical analysis of a 1:10,000 scale landslide inventory map of a large sector of the southern Apennine belt, which has been created by extensive visual interpretation of stereoscopic aerial photography, supported by field surveys. GIS-based statistical analysis of the landslide inventory map provided a clear picture of the main predisposing factors that controlled the distribution, size and pattern of landslide processes within the different morpho-structural units of the chain. The non-random spatial distribution of landslide processes is strongly controlled by lithological and morpho-structural factors and the resulting zonation represents an effective basis for landscape planning purposes and a key tool for more advanced analyses based on more innovative techniques such as InSAR monitoring, slope stability models or definition of rainfall thresholds. More specifically, landslide-dominated landscapes prevail in sectors with a relevant tectonic activity and Quaternary relief growth. Finally, the work explored several case studies where the integration of conventional and innovative methods provided relevant results on the surface and subsurface characterization of mass movements and the estimation of displacement fields and mobilized volumes.