GNGTS 2013 - Atti del 32° Convegno Nazionale

Seismic data were collected on the Massalezza lobe only (lobe A in Fig. 1) using a DMT Summit modular system with more than 200 double channel A/D conversion units. Each receiving station was equipped with a three component 10-Hz geophone to detect the incoming signal. The three component sensors were laid out along four lines (L100, L200, L300, L400). The average station spacing was 10.0 m for a total of 276*3 live channels. Elastic waves were generated and propagated into the ground using a medium size vibrator (both in P-wave and S-wave modes). The source stations were located along the roads bounding and crossing the Massalezza lobe. Recorded seismic data were generally of good quality; first breaks, in P-wave mode, were sharp and easy to pick even at offsets larger than 500 m while in S-wave mode the signal was slightly lower amplitude. A total of about 80000 first arrivals were picked and pre-processed and lately inverted using a 3D traveltime tomography. Inversion was carried out with the CAT3D proprietary software. Data resolution was improved using staggered grids (Vesnaver and Bohm, 2000). Resistivity data were collected on the two landslide lobes (A and B in Fig. 1) using a 48-electrode Syscal R1 system, a 96-electrode Syscal Pro system and an experimental 36-electrode wireless resistivity system manufactured by MultiPhase Technologies LLC. A total of six ERT profiles, in Wenner and in Wenner-Schlumberger configuration, and two ERT volumes, in pole-dipole configuration, were collected on the Massalezza lobe. Three additional ERT profiles and a ERT volume were collected on the eastern lobe. Data were collected in separate sessions during early spring and middle autumn after several days of heavy rain to improve the coupling. The high permeability of the landslide mass allowed for assuming the subsurface layers as homogeneously wet. Recorded resistivity data also resulted of good quality and just few points needed to be removed from the dataset prior to run the inversions. The inversion was carried out using the package ERTLAB+ that is based on a sophisticated reweight (Morelli and LaBreque, 1996) of the inversion parameters at each iteration. Results and discussion. Unit a’’ in the reference section appears to be low both in resistivity (0.5-1.0 KΩ·m) and P-wave velocity (2200-2700 m/s). Unit b , is very thin and, due to the large sensor spacing, is outside the resolution capability of both the two techniques. Units c , d and e in the middle part of the section are moderately compact limestones and appear show a similar geophysical response. In these units the resistivity is fairly high (2.5-4.5 KΩ·m) as well as the P-wave velocity (3400-3800 m/s). At the top of the rock wall, where unit f outcrops, there is a sudden lowering of both the resistivity (0.5-1.0 KΩ·m) and the velocity (2300-2800 m/s). The terrain resistivity in the landslide accumulation exhibit a large degree of fluctuation. The majority of the values range from 0.15-0.20 KΩ·m to 3.50-4.00 KΩ·m. The resistivity distribution along the profile ERT1 (Fig. 2) appears to be quite complex and the Average values are slightly lower as compared to the reference profile. This is somewhat related to the fracturing and to the severe changes occurred in the geological layers involved in the landslide. Unfortunately along profile ERT1 there are no borehole data available to constrain the interpretation. The deep conductive unit a’’ was utilized to define the initial geological layout along the profile. The prominent structure is a narrow syncline fold located in the middle portion of the profile. The axis is probably elongated along the old Massalezza valley. The bottom layers are probably belonging to the a’’ unit while in the top layers there are the typical lithologies of the c unit. Some resistive bodies could be associated with a partly undifferentiated c - d - e lithological sequence while the conductive unit, visible in the distance interval 50-200 m, according to Rossi and Semenza (1965) still belongs to units c, d and e. Possible explanation of this anomaly are the high degree of fracturing in the landslide or the occurrence of lateral changes of lithology moving from the right to the left side of the Vajont valley. These folding with an apparent N-S trend of the lithological units is also visible on the sliding surface (Massironi et al. , 2013). The preservation of the structural settings, before and 193 GNGTS 2013 S essione 3.3