GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.2 GNGTS 2023 Estimation of the seismic microzonation amplification factors from the GMMs between-station residuals G. Brunelli, G. Lanzano, F. Pacor, C. Felicetta, S. Sgobba & PRIN SERENA WP06 WG Istituto Nazionale di Geofisica e Vulcanologia (INGV), Milano, Italy Introduction Seismic Microzonation (SM) is a key tool for planning strategies for risk mitigation (Moscatelli et al. 2020) and consists of the assessment of local seismic hazard by identifying and mapping zones with homogeneous seismic behavior, commonly at municipality scale. SM studies are complementary to the regional seismic hazard studies, which calculate with probabilistic or deterministic methods the ground motion parameters at a site. At the most detailed levels (Levels 2 and 3), the results of the seismic microzonation are amplification factors (AFs) for each zone prone to local amplification. These AFs are generally calculated by numerical simulations, starting from in-situ measurements of the shear wave velocity profile with depth and building a geological model underneath the site. This approach was extensively applied to the entire Italian territory by Falcone et al. (2021) which perform a huge number of simulations (about thirty million 1D local seismic site response analyses) to characterize homogeneous areas (in terms of morphogeological and lithological properties) based on lithology distribution and Vs profile with depth (about 16 ′ 000 profiles of shear wave velocity from geophysical surveys and 44 ′ 000 logs from continuous coring boreholes). Alternatively, when multiple seismic recordings at a single site are available, they can be exploited to estimate AFs empirically through many methods, such as: a) spectral ratios between the horizontal and vertical components (H/V), both for acceleration response and Fourier spectra; b) Standard Spectral Ratio (SSR); c) amplification functions obtained from parametric or non-parametric generalized inversion techniques (GIT); and d) repeatable site terms ( δ S2S) obtained from decomposition of residuals of a nonergodic Ground Motion Model (Rodriguez-Marek et al., 2011; Priolo et al., 2020, among the others). Given the same empirical data, the major differences between the methods are related to how the reference conditions are set in the amplification computation: for SSRs and GIT functions, it is necessary to define one or more "local" reference sites; for H/V, the reference is the vertical motion that is assumed to be unaffected by site effects; for δ S2S, the reference is the median prediction of a ground motion model calibrated on rock sites.