Comment on “Seismic Hazard Assessment (2003–2009) for the Italian Building Code” by Massimiliano Stucchi, Carlo Meletti, Valentina Montaldo, Helen Crowley, Gian Michele Calvi, and Enzo Boschi

A new reference probabilistic seismic hazard map of Italy was recently presented by Stucchi et al. (2011). The map represents the most recent development of the “Cornell‐McGuire” approach adopted in Italy in the last 20 years for hazard assessment and implements current models for epistemic uncertainty treatment. This comment does not address the possible limitations of the considered approach itself, but some important (and possibly misleading) consequences of the representation chosen for hazard outcomes, selection of alternative procedures to be included in the logic‐tree considered to manage the epistemic uncertainty, and some statements about the possibility of using recent seismicity to validate these outcomes. This last point is of major importance since any future development of hazard mapping in Italy strongly depends on a careful evaluation of possible drawbacks inherent in the present formulations that should be pointed out by comparing theoretical outcomes and observations. The first aspect is related to one main result provided in the new hazard map that made it possible to base the design spectra for a building on point hazard data instead of on four intervals of peak ground acceleration (PGA), as was the case for the previous code. The authors claim the advantage of the new map is that it provides unprecedented detail of reference ground motion values associated with a fixed return time. In particular, officially released values (INGV, 2008) provide a PGA with four significant digits at more than 16,000 points. This appears overwhelming when compared to the mere four values associated with the zones of the previous map, with PGA set at 0, 0.15, 0.25, and 0.35 g . However, a closer look at the starting data may cast some doubts …