GNGTS 2019 - Atti del 38° Convegno Nazionale

GNGTS 2019 S essione 3.2 631 Data and methods. The main seismic parameters used during the acquisition are: 2 ms sampling rate, 2.5 m shot and 5 m receiver interval, maximum offset 6 km, for 596 channels. The energy source was a single Vibroseis, with a sweep ranging from 10 to 115 Hz, a sweep duration of 21 s, and an effective recording time of 6 s. The receivers were single geophones with a peak frequency of 10 Hz. Further details are reported by Vesnaver et al. (2016). The single-source single-receiver acquisition provided data less protected by ground roll, with respect to the conventional receiver array technology, but provided wider frequency band, which improves the Q factor estimation. Indeed, the adopted method of Quan and Harris (1997) exploits the differential frequency decay as a function of the propagation distance. In addition, the large available offsets and high fold made our estimations particularly robust, despite the significant noise present in the data. Further details about the Q factor inversion can be found in Rossi et al . (2007, 2011). For the characterization of the shallower layer, the first arrivals were picked in a semi- automatic way and a traveltime inversion was carried out, to estimate the P velocity with the staggered grid method (Vesnaver and Böhm, 2000). An initial 1D model with 30 layers was chosen with an increasing velocity gradient from 0.5 km/s at the surface to 2 km/s at a depth of 1.7 km. This first step provided the initial guess for a second, refined model composed of 3 blocks (pixels) in each layer, with a width of 6 km. We continued this multi-scale approach a few times, decreasing the pixel size (and so increasing their number) per layer, up to a minimum size of 63 m. This size is comparable to the maximum lateral resolution we may expect in the area, based on the estimated velocities and available frequency band at average offsets. Results. Fig. 2 displays the tomographic images for the P velocity and the Q factor, in the areas where the coverage per pixels exceeds 10 rays, to avoid interpreting areas biased by the null space and/or the initially guessed Earth model. First arrivals only were inverted as diving Fig. 3 - Superposition of P-wave velocity (upper image) and Q-factor values (lower image) to the pre-stack depth migrated seismic section.