GNGTS 2013 - Atti del 32° Convegno Nazionale

28-38 m, Fig. 3a). These are layered over the high reflective unit and can be correlated with the ochre sands mapped on the nearby outcrop (Fig. 3b). The combined analysis of the 3D GPR volume on vertical sections and depth slices provide an optimal visualization and imaging of the structures, highlighting a sector of the volume between the interval 2.5-5.5 m in the NW- SE direction (Y axes), where the geological structures are clearly defined (limited by a blue dashed line in fig. 3). In the area ranging from 0-3 m, with the exception of the shallower layers (0-40 ns), the structures are not well defined in depth due to a strong attenuation of the GPR signal. The boundary between the two zones in the 3D volume is sharp and easily detectable, and it separates two units showing very different electric properties. The North sector of the volume, located close to the scarp in field, highlights the discontinuity between the reflectors interpretable with the fault zone at about 10 m along the profiles (along the SW-NE direction, from SW to NE). In addition to this discontinuity, some diffractions already described in the 2D profile, represent features defining a characteristic geophysical signature of the fault zone. Fig. 3c reports a close-up obtained trough a depth-slice at about 2 m. Though the management and the interpretation of the whole volume at different depth, the fault zone visible also in the nearby outcrop was localized with high-resolution, extending in depth the morphological and geological surface data and providing new data useful for further paleoseismological analysis. Some quantitative and geometrical characteristics of the fault zone and others considerations about the sedimentary units can be done: a mean dip-direction has been estimated in about 235°, whilst the dip angle was close to 70°. These values are in agree with ones provided after the further ground-truth done on the site by the INGV research group of paleo-seismologists (UR Cinti). The new trench was focused across the two areas showing different GPR signature, intercepting both the fault and the sharp boundary highlighted by the 3D GPR volume, ensuring a successful data validation about the geometric characteristic of the fault. The blue dashed boundary therefore actually represents the separation between the natural subsoil nearby the scarp (fan deposits) and progressively thicker backfill layers located above the road on the survey site. The latter resulted stratified above a colluvial-fluvial unit originated by the erosion and re-sedimentation of a NE-SW stream, which generated the small NE-SW valley where the 3D data were acquired. Discussion and conclusions. The 2D/3D GPR “Grotta Carbone” and “Castrovillari” datasets highlighted characteristic geophysical signature of fault zones, efficiently defining the tectonic structure through some features like lateral truncations of layers, different dip of the reflectors on the two sides of the fault, presence of diffraction hyperbolas (un-migrated data) and a narrow strong attenuated zone. Some others considerations can be done on the areas surrounding the fault zone. The survey site shows reflectors having a W-dipping direction trend, excepted for the NE sector of the radargrams, where the layers are gently E-dipping. The 3D GPR analysis highlights how the S-SE sector of the acquisition is characterized by electrical properties of the investigated materials, which prevent to image geological structures in the S-SE side due to strong signal attenuation. Close to the outcrop (scarp), already from about half of the survey grid, both the fault zone and the sedimentary layers are well imaged by the data. The subsoil between the scarp and about 3 m towards South, is characterized in the first 3 m of depth by strong E-dipping reflectors. Some vertical sections and the depth-slices illustrate this unit is laterally extending in depth but then abruptly interrupted, probably due to the erosion / re-sedimentation operated by a pre-existent stream. In the shallow part a boundary with some backfill layers filling the valley, probably produce high attenuation and low amplitude reflections, reducing the investigation depth and complicating the visualization/interpretation within this sector. The structures interpreted in the 2D GPR sections and 3D GPR volume, like the fault zone, some important layers and stratigraphic units have been successively validated by the trench excavation. The trench revealed a correct geophysical interpretation not only of the geometrical characteristic of the fault (strike, dip direction, dip angle), but also of the lateral boundary visible in the depth-slices (backfill, alluvial deposits…etc). The survey represents an 174 GNGTS 2013 S essione 1.2