GNGTS 2021 - Atti del 39° Convegno Nazionale

431 GNGTS 2021 S essione 3.2 The tomography was performed using CAT3D software (Böhm, 2014) and the results are shown in Figure 2b, together with the computed V S30 along the whole line. After a very shallow, more fractured layer showing low velocities, the velocity increases quite rapidly. In fact, the values of V S30 are above 800m/s along the whole line, making the soil classifiable as A-class. The importance of acquiring an S-wave seismogram is evident in the case of Kastela, as no dispersive event could be identified to perform surface-wave analysis, as can be seen in the frequency-wavenumber spectrum shown in Figure 2c. Furthermore, the importance of acquiring all three wavefields (P,SH and SH), in the case of such a complex geology as that of the dalmatian flysch is evident in the fact that it allows not only to detect the anisotropy (V SV >V SH , Figure 2d), but also to estimate the bedding angle (Böhm et al., 2020). In Figure 3a we show the results of the travel-time tomography relative to the SV seismogram acquired in Bibione. We can identify loose sands in the shallowest southern part, while in the north the sand is more compacted already on the surface. We can also identify an increase in the velocity at approximately 17 m depth and a sharper one at 27 m depth. The former is probably due to the presence of more compacted sand, while the latter is probably due to an increase in the clay content of the sediments. As for the surface-wave analysis, we extracted a dispersion curve and inverted it using the neighbourhood algorithm of the GeoPsy free software (Wathelet, 2008) and we show in Figure 3b the 1000 best fitting profiles. These results are quite consistent with those of the tomography. To be noted that, due to the lack of a-priori information, the use of both methods allowed us to better constrain the inversions and obtain more reliable results. Conclusions We showed two case studies, from two very different environments, both aiming at characterizing the shallow subsurface. In the first case study, we acquired data along a road in a village, where the geology is known to be complex, consisting mainly of thrusted hard- rock flysch. In the second, the seismic data was acquired on a sandy beach. Our experience highlights in both cases the importance of integration of both first-break tomography on S-wave seismic data and surface-wave analysis. In fact, in the first case study, no dispersive event could be retrieved and therefore essential engineering parameters such as the V S30 could be computed only from the travel-time tomography. In the second case study, the lack of a-priori information made the integration of both methods a valuable tool to constrain both the travel-time tomography and the surface-wave inversion. In conclusion, while surface-wave analysis remains the most time-effective tool to characterise the shallow subsurface, in many circumstances, especially when very little a priori information is present, its integration with S-wave seismic data can be essential. Figure 3. a) Travel-time tomography of the SV data. b) 1000 best fitting profiles from the multichannel analysis of surface-waves.