Validation and extension of a statistical usability model for unreinforced masonry buildings with different ground motion intensity measures

Predicting the usability of a building, i.e. its condition of being occupiable after a seismic event, is relevant both in a post-emergency situation and within a risk-reduction policy. In the past an empirical model was proposed, involving the computation of a usability index based on macroseismic intensity and on seven building parameters, combined by means of regression coefficients and weights. The statistical model was calibrated on data of about 60,000 buildings affected by the 2009 L’Aquila earthquake in Italy. Therefore, it is useful to validate the model against data from the 2002 Molise earthquake in Italy. Good agreement between predicted and observed usability is shown, despite the fact that in 2002, macroseismic intensity was attributed to an entire municipality instead of a more limited area. Moreover, given the current availability of the shakemaps for the 2009 event, a novel model replacing conventional macroseismic intensity by an instrumental intensity measure is proposed. Three ground motion parameters are considered here: peak ground acceleration, peak ground velocity, and spectral pseudoacceleration at a period of vibration of 0.3 s. The model has been streamlined by reducing the building parameters from seven to five: building position within the structural aggregate, roof type, construction timespan, structural class, and pre-existing damage to structural elements. Peak ground acceleration and spectral pseudoacceleration are shown to be less effective than peak ground velocity in predicting observed usability. Therefore, usability probability matrices are computed in terms of peak ground velocity; the model is presented with all necessary coefficients and weights, and a worked-out example shows how to apply the procedure.