Comparison between seismic amplification factors from building code and those obtained by data from Italian seismic microzonation studies

Given their ease of use, the simplified approaches based on the use of some seismic “proxies” and contained in the most common building code provisions are the most widespread methods in the professional activity to estimate the ground-motion amplification due to the local seismo-stratigraphical features. This kind of approach is also present in Italy in the actual national building code. As is now well known, this procedure assigns an acceleration response spectrum through the estimation of the soil class of the investigated site, which is identified by the values of two proxies, that is the seismic bedrock depth H (i.e., the layer with Vs ≥ 800 m/s) and the time average Vs down to H (VS_H) or to 30 m depth (VS₃₀), if H is lower or higher than 30 m respectively. The objects of this study are the seismic Amplification Factors (AFs) refer to each soil class, defined as the ratio between the integral of the acceleration response spectrum of soil classes B, C, D, E respectively and the integral of the response spectrum of the soil class A. These values were computed for three period intervals (0.1–0.5 s, 0.4–0.8 s and 0.7–1.1 s) and are determined for all the Italian municipalities considering the response spectra referred to the Italian Seismic Hazard Map relative to the ground motion expected to be exceeded with a probability equal to 10% in 50 years. The aim of this work is to test the effectiveness of these AF estimates verifying if these values are enough conservative with respect to those obtained by 1D numerical simulations carried out in the same sites. To perform this analysis, we evaluate if the number of excesses from these estimates is significant from the statistically point of view or it can be considered as a random fluctuation. On this purpose, outcomes from the “Italian Map of Expected Values of Amplification Factors” obtained in the framework of a national project in 2019-2021 were considered. In particular, seismo-stratigraphical and geotechnical data from sesimic microzonation studies of 1689 Italian municipalities were used to perform 1380600 1D equivalent linear numerical simulations, from which it was possible to compute the AFs value from the ratio between the integrals of the relevant output and input response spectra. For each simulated profile, AF value from building code is then retrieved estimating the soil class and considering the municipality of the data origin. Finally, the two AF datasets were compared following the statistical approach mentioned above.