Comment on “Layered Seismogenic Source Model and Probabilistic Seismic-Hazard Analyses in Central Italy” by Bruno Pace, Laura Peruzza, Giusy Lavecchia, and Paolo Boncio

Pace et al. (2006) define a seismogenic model aimed to provide a (time-dependent) seismic-hazard assessment for Central Italy. One of the main novelties of the model is that it introduces seismogenic structures (faults) in seismic- hazard calculations, and it imposes a time behavior on these structures, leading to a time-dependent seismic-hazard assessment. In my opinion, the approach used in this article (1) does not explain the “observed” time distribution of historical earthquakes (one of the few empirical evidence that can be used to check the reliability of any seismic-hazard or earthquake-forecasting model), and (2) uses values of parameters of the statistical distribution that are clearly unreliable. These problems introduce significant biases that potentially make the hazard assessment made by the authors less reliable that the one obtained by more traditional approaches. In the following, I discuss in detail these points. I focus the attention only on earthquakes with M 5.5+ that are considered “destructive” for the Italian territory. Sometimes, for the sake of conciseness, I drop the specification of the magnitude of events, and then it should always be considered equal to or larger than 5.5. Traditionally, seismic hazard and earthquake forecasting in Italy are based on regular spatial grids or on the definition of seismogenic provinces (Slejko et al. , 1998; Albarello et al. , 2000; Meletti et al. , 2000; Scandone and Stucchi, 2000; Gruppo di Lavoro, 1999, 2004; Romeo et al. , 2000; Lucantoni et al. , 2001; Faenza et al. , 2003; Cinti et al. 2004). This choice has two main rationales; first, despite the common acceptance that almost all earthquakes occur on faults that …