GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 2.2 GNGTS 2024 All the available data underwent specific Quality Control (QC) in order to consider only reliable and state of the art informaCon. Data deriving from MASW tests (the most widely diffused technique for Vs profile determinaCon) underwent a specific QC control consisCng in checking: 1) the consistency of the dispersion curve, that should present a clearly visible and conCnuous fundamental mode in the frequency band of interest; 2) in case of presence of mulCple modes of vibraCon, they should be well separable, well disCnguishable and reliably interpretable independently; 3) the picking of the dispersion curve that should be reliable and fimng with the spectral maxima of the seismogram transform used for the analysis; 4) the inversion of the data should lead to a syntheCc dispersion curve very close to the experimental data i.e. good correspondence between the experimental data and the results of the inversion; 5) the depth of the Vs profile should be compaCble with the minimum frequencies observed in the analysis of the dispersion curve, i.e. invesCgaCon depth less than at least the maximum wavelength (preferably half the maximum wavelength); 6) the Vs profile should match the minimum parametrizaCon criterion, i.e. number of analyzed layers compaCble with the experimental informaCon. The results of the QC are about 1000 Vs profiles distributed over all the GDs (Figure 1), concentrated near the main urban seblements and most populated areas that are the main targets for such types of studies. The final step is the evaluaCon of specific Vs profile distribuCon within each GDs and their comparison among different GDs. Results and Discussion As an example, the results of the performed analysis over the GD 8 (Po plain fed by alpine rivers) are reported in Figure 2. In this GD 362 Vs profiles were available (Figure 2a) of which 66 reached the seismic bedrock (Figure 2b). The average Vs of the bedrock is 975 m/s and its depth is between 5 and 46 m. The Vs,z (harmonic average velocity) distribuCon of the non- bedrock layers was also computed for each profile (Figure 2c) together with the resulCng Vs,h according to NTC (Figure 2d). The Vs,z is indeed usually considered as a closer representaCon of the physics of the problem than the Vs layered profile (Comina et al., 2022). This allowed also to obtain a representaCve median Vs,z profile for the GD (together with its standard deviaCon). Median Vs,z profiles were then analysed in the different GDs (Figure 3). The distribuCon of the median Vs,z profiles show groups of GDs with very similar behaviours, reflecCng the similariCes in the properCes of the non-bedrock deposits. Following a global approach to the data analysis the median Vs,z profiles, eventually merged between similar GDs, will be the basis for the following randomizaCon and simulaCon steps. For this purpose specific randomizaCon approaches, based on the same Vs,z (Passeri et al., 2020) or on usually adopted randomizaCon criteria (e.g. Toro, 2022) will be evaluated. This proposed global approach allows to overcome the limitaCons inherited by the uncertainCes of the specific litho-straCgraphic semngs within each GD, due to the regionality scale of observaCons and the quality of the exisCng databases.