GNGTS 2014 - Atti del 33° Convegno Nazionale

72 GNGTS 2014 S essione 2.1 seismic rate computation [“higher not highest”, see Slejko et al. (1998)], one approach for the maximum magnitude determination [“one step beyond”, see again Slejko et al. (1998)], and three GMPEs (Ambraseys and Simpson, 1996; Cauzzi and Faccioli, 2008; Bindi et al. , 2011). The Cauzzi and Faccioli (2008) GMPE is defined for distances larger than 20 km: its application in the near field is, then, questionable. The usual approach followed in PSHA consists either in extrapolating the functional form also outside its distance range of definition or in introducing an asymptotic upper bound for the near field. We have considered both the two options: its simple extrapolation to short distances and a saturation effect in the near field (PGA does not increase for distances less than 20 km). To identify the near-field effects, the high level seimicity has been associated to linear sources according to the fault parametrization in the Italian fault database DISS (DISS Working Group, 2010) considering the characteristic earthquake model for their activity. This latter model was introduced by Schwartz and Coppersmith (1984) for the seismicity in California and states that each fault can produce events of a specific narrow range of magnitude (+/- 0.5) in agreement Fig. 1 – Seismicity and seismic hazard for rocky sites in NE Italy: a) epicentres of the earthquakes with a magnitude larger than 3.0 and location of the E-W oriented transect (red line), the square indicates the study area; b) vertical PGA with a 475-year return period considering an extrapolation of the Cauzzi and Faccioli (2008) GMPE; c) vertical PGA with a 475-year return period considering a saturation effect in the near field for the Cauzzi and Faccioli (2008) GMPE. The blue polygons in (a) represent the FRI zonation (Slejko et al. , 2011), the black rectangles show the DISS sources.