GNGTS 2015 - Atti del 34° Convegno Nazionale

Tab. 1 - Number of EC8 site class classification before and after the S2 DPC-INGV project (2014-2015). # EC8 site class Before S2 After S2 A 1 9 A* 8 2 B 1 14 B* 25 3 C 0 18 C* 14 2 E 0 1 the large V S,30 interval characterizing this soil class (360-800 m/s) and to the difficulty to detect rock sites from the geological map. On the contrary, class C* is more stable, as deep sediments are easily identifiable on the basis of the surface geology. Tab. 1 quantify the upgrade in terms of EC8 class classification before and after the S2 DPC- INGV project (2014-2015). In general, the V S,30 measurements provides a reliable estimation of the site amplification when the geological settings are close to one-dimension approximation. When the sites are located in complex geological settings, the V S,30 in not a good proxy of site response. As an example, we present the case of the Malcesine station (code MLC). This station is located on one of the flanks of the Garda Lake and was formerly classified as A* on the base of the geological information (Di Capua et al. , 2011). After the geophysical surveys carried-out in the frame of INGV-DPC S2 Project (Barnaba et al. , 2015), the station moved tosoil category B since the measured average shear wave velocity is V S,30 = 580 m/s. The variability of the ground-motion observed at this station were investigated by Lanzano et al. (2015) through the residual analysis according to Rodriguez-Marek et al. (2011) and Luzi et al. (2014). This analysis was carried out on a 2015 updated data set of records (retrieved by ITACA 2.0 and the European Strong Motion database (ESM http://esm.mi.ingv.it) ), using the Bindi et al. (2011) as a reference Ground Motion Prediction Equations. The residuals have been decomposed to retrieve the event, site and record-to-record errors. In particular, the site-term could be used to evaluate if the site response is coherent with the median predictions for the soil categoryassigned to the considered site. The MLC station recorded 18 earthquakes, most of them belonging to the 2012 Emilia seismic sequence. Independently on the soil classification adopted to MLC, the site term estimated trough the residual analysis always shows a remarkable amplification around T=0.1 s; this behavior is also confirmed by the noise measurements. This high frequency amplification is not predicted for EC8-classes A and B and could be ascribed to the specific geological setting that are notrepresented by a simple 1D scheme. Conclusion. The results of geophysical and geomorphological surveys performed in the framework of different scientific projects, collected in the ITACA 2.0 release, allow us to: i) update site-information at strong-motion Italian recording stations; ii) better individuate peculiar features that can be ascribed to site effects or soil-structure interaction; iii) improve the performance of empirical model for ground motion prediction. Site classification inferred only from geological information can be misleading and surveys of site characterization of the recording stations should be promoted, in the future, in order to improve the current knowledge on strong motion seismology (site effects, Ground Motion Prediction Equations, hazard map validation). GNGTS 2015 S essione 2.2 127