The document discusses an automated method for classifying earth surfaces to assess landslide susceptibility using topographic data. It examines different geomorphometric classification approaches and parameters to distinguish landscape types related to landslides. The study tests supervised and unsupervised classification methods, compares results, and develops an integrated method using slope gradient, convexity and texture that identifies terrain features correlated with past landslide events. The automated integrated classification approach provides a useful tool for landslide susceptibility analysis at a territorial scale.