GNGTS 2017 - 36° Convegno Nazionale

694 GNGTS 2017 S essione 3.3 For the present work, together with the group-velocity spectra of the Z and R components and the RPM and RVSR curves, we also considered the transversal (T) component (i.e. Love waves) and the Horizontal-to-Vertical Spectral Ratio (HVSR). The adopted nomenclature (ZVF, RVF and THF - Dal Moro, 2014) refers to the vertical and radial components obtained while adopting a Vertical Force (VF) and to the transversal component obtained by means of a Horizontal Force (HF) [wooden beam]. The HS approach represents an improvement of the classical MFA (Multiple Filter Analysis - Dziewonski et al. , 1969) or FTAN (Frequency-Time ANalaysis - Natale et al. , 2004) methods (which, incidentally, are fundamentally identical). Differently than the MFA/FTAN approach, HS relies on the analysis of all the components, which is accomplished not through the interpretation of the modal dispersion curves but by adopting the FVS ( Full Velocity Spectrum - Dal Moro, 2014) approach, which does not require the interpretation of the velocity spectra in terms of modal dispersion curves. Furthermore, in addition to the velocity spectra, the HS method also considers the information regarding the ratio between the amplitude spectra of the radial and vertical components (RVSR) and the actual particle motion (RPM frequency curve) (Dal Moro, 2017). The single-offset active data recorded at the SED site were recorded according to the acquisition parameters reported in Tab. 1 (the HVSR was computed by considering a passive dataset - record time: 2 h; sampling frequency: 200 Hz). The site is dominated by mudstones and fine-grained sandstones typical of the Swiss foreland basin (see also Dal Moro et al ., 2015). The results of the accomplished joint inversion are summarized in Figs. 2 and 3 (the minimum-distance model represents the model having the minimum geometrical distance from the utopia point - for details see Dal Moro, 2014, 2017). In spite of the very limited equipment (a single 3C geophone) and consequent limited field effort, the shear-wave velocity (V S ) profile obtained by following the described procedure is in good agreement with the results obtained through the analysis of classical multi-component and multi-offset data and with the available borehole data (Vertical Seismic Profiling - VSP) (see Dal Moro et al. , 2015). Tab. 1 - Acquisition parameters of the active 3C dataset (see Fig. 1). Recording time 2 s (then reduced to 0.75 s) sampling frequency 1000 Hz (1 ms) offset 70 m stack 5 source 8 kg sledgehammer Some general conclusions can be drawn: - a joint inversion is necessarily a sort of compromise between the misfit values of all the considered objective functions and, as a consequence, it is not possible to obtain a perfect match for each single considered object (see also Dal Moro and Puzzilli, 2017); - the active data recorded by a single 3C geophone can be used for determining up to five objects to exploit for a very-well constrained joint inversion procedure: the group-velocity spectra of the Z, R and T components, the RVSR and the RPM frequency curve; - the 3C geophone used to record the active data, can also be used to record the passive data useful to compute the HVSR which can be added to the joint inversion, thus further constraining the inversion procedure and increasing the investigated depth otherwise limited to about two thirds or half of the offset adopted for the active acquisition; - compared to the RVSR, the RPM frequency curve appears less sensitive to subtle stratigraphic details but more robust (less sensitive to data pre-processing);