GNGTS 2019 - Atti del 38° Convegno Nazionale

382 GNGTS 2019 S essione 2.2 As an example, we report the case of the station CLF (Colfiorito), NTC18-D soil class, located in central Italy, very well sampled, having recorded more than 50 earthquakes (Fig. 2). CLF exhibits a clear peak at 1 Hz and other lower peaks at higher frequencies (3.3, 6.7, 14 Hz), similarly to the results obtained by other experimental analyses based on earthquake and noise measurement (http://itaca.mi.ingv.it ; Pacor et al. , 2011). In order to obtain robust results, the comparison with the NTC18 amplification factors was carried out considering stations with at least 5 records. Consequently, only 138 stations were selected for this purpose. Specifically, the verification of the NTC18 site amplification factors was performed by comparing the δS 2 S A median values, computed at short period range (0.07-0.2 s) and intermediate one (0.7-2.0 s), using the MIX model. Since we are not able to separate the stratigraphic and topographic amplifications from δS 2 S A , we also multiply the amplification factor to the period independent topography amplification factor, following the NTC18 prescriptions. Fig. 3 – Comparison between the NTC18 amplification factors and δS 2 S A of stations with at least 5 records for each site class. The little square indicates the median value of δS 2 S A for a station in a range period, while the full dot indicates the median value of δS 2 S A of all stations. We refer at short periods with blue color, while at intermediate periods with red color. As shown in Fig. 3, a good agreement between NTC18 and δS 2 S A median values for the 138 Italian accelerometric stations was observed, although there is a great variability among the experimental data; moreover, for class E sites, there is an overestimation of NTC 2018 amplification factor with respect to that obtained by δS 2 S A at intermediate periods. Discussion and Conclusions. In this study, to test and validate the amplification factors proposed by the seismic standards, an experimental approach is proposed. Our method exploits data and metadata contained in flatfiles of strong motion dataset in order to classify the seismic stations according to the standards and provide an estimate of the site amplifications. We used the ESM flatfile and the information contained in the ITACA (http://itaca.mi.ingv.it ) and ESM (http://esm.mi.ingv.it) databases to evaluate the site-to-site term of 138 Italian stations well characterized. The median values of the site terms, computed with respect a reference ground motion level, are compared with the amplification factors