GNGTS 2014 - Atti del 33° Convegno Nazionale

82 GNGTS 2014 S essione 2.1 has been modelled according to the Gutenberg – Richter (G-R) distribution and the maximum magnitude for each seismogenic source has been calculated by the Kijko and Graham (1998) statistical algorithm. The ground motion prediction equation (GMPE) by Cauzzi and Faccioli (2008) (C&F, hereafter) has been selected for modelling the attenuation because it is claimed to be robust and it is defined for application in different tectonic environments. Figs. 1b and 1c show the expected ground motion, in terms of horizontal peak ground acceleration (PGA) obtained for a 475-year return period. It can be seen that the use of a GMPE that considers the specific tectonic style of each SZ produces some differences in the expected ground motions with a general decrease of PGA. More precisely, the maximum values are more concentrated and refer only to the SZ Romagna Apennines-Po Plain Margin. To introduce, in a very preliminary way, the correct 3D geometry suggested by the tectonic setting within the proposed zones, the SZs are no more considered as plain surfaces but as inclined planes (seismogenic planes, SPs hereafter) which occupy the whole area of the boxes of Fig. 1a. It is clear that this geometry is extremely rough because it simplifies in a single inclined element the totality of faults present in the boxes (see Tab. 1). Moreover, assuming for the SPs the dominant dip in each SZs and extending the SPs to the whole SZ, some unrealistic depths are reached. It must be clarified that this is only a preliminary test to evaluate the influence of inclined rather than subhorizontal planes as sources in PSHA. The obtained expected ground motion is further modified with respect to the previous maps (Figs. 1b and 1c) and should, in principle, better depict the shaking in the proximity of the reference fault trace on the surface. The C&F GMPE with unclassified tectonic style has been applied in the first elaboration (Fig. 2a) and it can be seen that the computed shaking is notably lower than that calculated with superficial planes (Fig. 1b) and the maximum PGA is concentrated in the SZ Romagna Apennines. The dominant tectonic style of the different SPs has been considered in the following elaboration by applying differentiated C&F GMPEs, simulating, then, the complete geometrical and kinematic behaviour of the SZs. A further decrease of PGA has been obtained (Fig. 2b) with respect to the previous map with inclined SPs not differentiated for the tectonic style especially in the normal SPs (Fig. 2a). The comparison with the map calculated with plain sources of differentiated tectonic style (Fig. 1c) put in evidence the very different role played by the sources. It must pointed out, anyway, that the transcurrent SPs have been modelled as vertical planes of very limited width (3 km). A comparison with the values of the seismic hazard map based on the ZS9 zonation, that is presently at the basis of the official estimates for Italy (Working Group MPS04, 2004; Stucchi et al. , 2011), is possible only in terms of plain sources with differentiated fault kinematics (Fig. 1c). Actually, the seismic hazard map considered here is a modification of the official one because the recent C&F GMPE has been applied (Fig. 2c) in substitution of the original suite of GMPEs (see Working Group MPS04, 2004; Stucchi et al. , 2011). The expected ground motion computed with the new seismogenic zonation (Martelli et al. , 2014) is much more articulated than that of the Italian map because the new zonation is much more detailed than that used for the Italian seismic hazard map, and the new SZs are narrower, thus concentrating, sometimes, the shaking. In particular, the largest PGA is now higher and concentrated in the SZ Romagna Apennines-Po Plain Margin, and two additional areas of large PGA appear in the SZs Emilia Apennines-Po Plain Margin e Garfagnana. These pieces of evidence are not supported, conversely, by the map computed with inclined SPs with differentiated GMPEs (Fig. 2c). Seismic hazard with dipping planes. The previous simplification of the tectonic setting with a single SP is overtly not satisfactory, because each SZ is characterized by the presence of at least one system of faults, sometimes with different vergence, and a simple dipping plane, often extrapolated to an unrealistic depth for maintaining the proposed dip angle (see Tab. 1), does not reproduce the actual tectonic framework. As a full geometric description has been given for the different SZs, a more sophisticated approach consists in defining inside each SZ more dipping planes adequately located, that can better represent the tectonic scenario (Tab. 1).