The application of predictive modeling techniques to landslides induced by earthquakes: the case study of the 26 September 1997 Umbria-Marche earthquake (Italy)

During September and October 1997, an earthquake sequence of moderate magnitude struck the Umbria and Marche regions, central Italy. As a consequence of the most severe shocks, several landslides were triggered on bedrock, calcareous debris and sandy clay deposits. The landslide inventory, collected in the field and by aerial-photo interpretation, provided an opportunity to test the accuracy of existing predictive relations for landslide hazard zoning and to formulate new ones. Shallow soil slides and debris slides were selected to verify the prediction performance of the methods, which are based on a simplified model [Geotechnique 15 (1965) 139] that calculate landslide displacements during Cyclic seismic loading. Rock falls were modeled using a different approach to produce an empirical predictive rule for determining the probability of detachment of blocks. When landslides occur along pre-existing or newly formed failure surfaces, the Arias intensity and destructiveness potential are the strong ground-motion parameters that best relate to displacements. The predictive relation proposed for rock falls evaluates the probability of occurrence as a function of the slope angle and the peak ground acceleration (PGA). (C) 2003 Elsevier Science B.V. All rights reserved.