Validation of insensity attenuation relationships

A new approach is proposed for the empirical validation of intensity attenuation relationships to be implemented in standard procedure devoted to probabilistic seismic hazard assessment (PSHA). To this purpose, the overall number of documented intensities above a given threshold in the area of interest is compared with the one expected on the basis of the considered attenuation relationship. By using this methodology, the reliability of empirical relationships can be analyzed by also taking into account the uncertainty affecting ill-defined intensity attributions both at the epicenter and at the site. To assess the feasibility of this procedure, an attenuation relationship recently proposed for the Italian region has been evaluated by considering felt intensities documented in the area during the past two centuries. Although the macroseismic database considered for validation is the same used for the parameterization of the investigated relationship, important discrepancies have been detected between observed and computed intensities in the range of values significant for seismic hazard. This result indicates that a careful parameterization of attenuation relationships in their complete probabilistic form is mandatory when such relationships have to be implemented in PSHA procedures.