Effect of the input parameters on the spatial variability of landslide susceptibility maps derived by statistical methods

This study is aimed at assessing different spatial patterns of predicted values of landslide susceptibility maps with almost similar success and prediction rate curves. Our approach is applied to an alpine environment (Italian Central Alps) where debris flows represent a frequent damaging phenomenon. The Weights of Evidence modelling technique (a data driven Bayesian method) was applied using ArcSDM (Arc Spatial Data Modeler) an ArcGIS extension. The output prediction maps were reclassified in the same way to compare the predicted results: a relative classification, based on the proportion of the area classified as susceptible, was made. The thresholds among different susceptibility classes were put at each 10 % of the study area, classified decreasingly from the highest to the lowest susceptibility values. After applying Kappa Statistic, Cluster Analysis, and Principal Component Analysis (PCA), we analysed the spatial variability of the predicted maps. The results have shown great differences within the output spatial patterns of the predicted maps, and also within the highest susceptibility classes.