GNGTS 2015 - Atti del 34° Convegno Nazionale

used for the dispersion curve interpretation is 1D. Hence, every time the velocity distribution is laterally heterogeneous, and the model errors can have significant consequences on the reliability of resulting shear-wave velocity distributions. For this case studies we the possibility to enhance lateral heterogeneities by an innovative 2D surface wave dispersion analyses inversions with a multi-offset phase analysis (MOPA) able to detect and locate the lateral discontinuities. In the frequency domain, the surface wave response can be described by the superposition of modal contributions (Aki and Richards, 2002): s (ω, x ) = � s m (ω, x ) = � B m (ω, x ) e j (ω t-k m (ω) x + ϕ 0 (ω)) (1)       m       m where: j is the complex unit, x is the source-receiver distance, ω the angular frequency, and k m ( ω ) the wavenumber of the m -th mode. If we disregard the amplitude B m ( ω, x ) and consider only the modal phase: φ (ω, x ) = – k m (ω) · x + ϕ 0 (ω). (2) The experimental phase can be extracted from time-to-frequency Fourier transform applied on the collected traces. Thus, under the assumption that the m -th mode carries the largest part of the energy and so the summation in Eq. 1 can be approximated to consist of only its m -th term, the wavenumber k m ( ω ) can be recovered by using a simple linear regression (Strobbia and Foti, 2006). Eq. 2 means that phase-offset points lie on a straight line with slope coefficient proportional to k m ( ω ). Hence, lateral velocity variations reveal themselves as a slope changes in the presence of lateral heterogeneities. For different subsets of contiguous geophone locations ( x i ), different values of k m (ω, x i ) are obtained for the same frequency ω. Thus, the location of slope changes in the phase-offset piecewise linear function can highlight the locations of lateral discontinuities (Vignoli and Cassiani, 2009; Vignoli et al., 2015). If slope changes occur consistently at the same location for a large frequency range, we can confidently guess that, at that location, there is a lateral velocity variation. In the Ex-Esso area we adopted the MOPA technique detecting lateral variation using N. 48 4.5 Hz geophones and several sources (vibroseis, hammer, seismic gun). Geological investigation. A hydrological study and monitoring of groundwater was carried out for a period of 8 months in 5 piezometric stations, to assess the development of the water table. Pluviometric measurements used are those measured by the regional Civil Protection. During the project were carried out 12 drilled boreholes (E1 - E12), to reconstruct the stratigraphy of the area and determine the geotechnical parameters. The investigation depth Fig. 1 – Electrical Resistivity Tomography (ERT) tridimensional 2D pseudosection resistivity model of the Ex-Esso industrial site (Trieste, Italy). 72 GNGTS 2015 S essione 3.2