Observed damage and empirical predictive model of the school building heritage of the Marche region

The 2016 seismic sequence that struck Central Italy, with its three mainshocks and numerous aftershocks, caused widespread damages to building heritage. Among the buildings that suffered damages, schools are considered relevant structures from multiple points of view. Educational institutes, indeed, host children and young people during their training years in facilities that are often old and realized without recent safety Standards becoming, therefore, non-resilient, nor robust structures towards multiple natural hazards. In this paper, a database containing information regarding the geolocation of individual buildings, no. of floors, the floor surface, the construction period, the structural type and the potential presence of landscape or hydrogeological constraints, has been realized with reference to the schools of the Marche region, by merging different sources of information. In addition, the database also collects information regarding observed damages and recorded seismic intensities to schools during the 2016 Central Italy earthquake. Once built up the database, an empirical predictive model, known in literature and already used for the evaluation of potential post-earthquake scenarios to cultural heritage, is applied to school buildings located into the seismic crater of the Marche region, by estimating the parameters of the model through inference techniques. The probabilistic model applied, relates a measure of the ground motion intensity to an overall damage index, continuous in the range [0,1]. Moreover, if one integrates the partial information related to damaged buildings often provided by the post-earthquake surveys, with the data related to the undamaged and/or collapsed structures, the model can provide a full risk assessment procedure. Finally, fragility curves are also derived, to make it possible the use of the research outcomes in risk frameworks based on a discrete description of the consequences.