Susceptibility assessment to different types of landslides in the coastal cliffs of Lourinhم (Central Portugal)

The coastal zone of Lourinhã (located in Central Portugal) is characterized by beach–cliff systems, where beaches are narrow and cliffs have notorious slope instability. These cliffs evolve by different types of landslides, which are one of the main sources of natural hazard and risk in this coastal region. s work, aerial photo interpretation and a systematic field survey were performed in order to obtain an inventory of landslides of the following types: rotational slides, translational slides and debris flows. The entire coast was then split into 261 terrain mapping units. For each unit, landslide predisposing factors were derived and classified: cliff elevation, slope angle (maximum, mean and standard deviation), potential solar radiation, slope curvature (profile and plan), lithological units and geologic structure. The predictive susceptibility models were computed for each type of landslide using a bi-variate statistical method — the Information Value Method. The degree of fit and the predictive capacity of the models were assessed using the Effectiveness Ratio, the standard Receiver Operator Characteristic curves and the respective Area Under Curve. s show that each landslide typology occurs in particular terrain conditions. Individual susceptibility models evidence better predictive capacity than susceptibility model for total landslides.