GNGTS 2018 - 37° Convegno Nazionale

688 GNGTS 2018 S essione 3.2 ERT lines with floating cables, having equi-spaced electrode locations, were stretched across the pond. Polyethylene foam floaters were used to maintain the cable at the water surface, but ensuring the cable takeout (directly used as electrodes) to be submerged. At the ERT line edges, the cable takeouts were connected to stainless steel electrodes fixed in the ground of the pond shore. A total of 9 ERT profiles were acquired on the pond, with 24 electrodes at 3-m spacing. An additional line, with 48 electrodes at 1-m spacing, was additionally acquired close to the southern shore of the pond (Fig. 2). The first and last electrodes of each line were geo-referenced using a Garmin GPS 60 to retrieve the position of each survey line. AWenner-Schulumberger acquisition sequence was adopted, using a multichannel resistivity system (Syscal Pro - Iris Instruments). Before each current injection, the natural Spontaneous Potential (SP) between each current dipole was additionally recorded. Data processing. GPR data processing consisted of i) trace regularization, on the basis of the effective profile length and number of traces; ii) time cut, to reduce the trace length after a check of the deterioration of the S/N ratio; iii) move start time, to remove the delay introduced by the system; iv) dewow, to reduce very low frequency components; v) further time cut, to equalize the sample number within the traces of the two surveys; vi) divergence compensation, to increase deep echoes strength; vii) subtracting average, to remove horizontal bands and viii) muting above the pond-bed picking. No clear velocity domains were found to migrate the many diffraction hyperbola present in the radargrams. The manual picking of the pond bottom on the radargrams was further used to reconstruct the pond bathymetry. Given the density of the GPR traces, retrieved water depths were then interpolated over the surveyed area to obtain a reference bathymetry map (Fig. 2). Electrical profiles were processed with both 2-D and 3-D inversion procedures. Raw apparent resistivity measurements were filtered basing on their related standard deviation (each measurement is the average of 10 subsequent measures) to improve the overall data quality. Measurements with standard deviation higher than 1% were filtered out from the dataset. Two different 2-D robust inversions were performed on each ERT profile (Res2DInv, Geotomo): the first not introducing a-priori information on the water layer; the second constraining the bathymetry of the pond (from GPR data) and the water resistivity (fixed to 125 Ohm m, as obtained from independent in-situ measurements with a portable conductimeter). The 3-D inversionwas carried out without a-priori constraints (ERTLab, Geostudi). The SPmeasurements related to the dipoles with lower spacing were also extracted and mapped over the pond. Each SP measure was plotted in the centre of the measuring dipole. Minimum distance between two current electrodes (AB) was 9 m, consequently raw SP data are approximately related to a depth of investigation of around 4.5 m (AB/2). This depth corresponds to the first 2-3 m below the pond bottom and may be diagnostic of recharge (positive SP anomaly) or seepage (negative SP anomaly) hydrodynamic processes occurring at the pond bottom. Results and discussions. The bathymetry map obtained from GPR data processing (Fig. 2 and 3) highlighted tiny water depths over the investigated area, with a maximum of approximately 2 m in the centre of the pond. The majority of the diffraction hyperbolas observed in the radargrams were due to the widespread presence of schist slabs and decimetric blocks laying on the pond bottom. These hyperbolas showed asymptote slopes tending to 0.03 m/ns, i.e. the radar propagation velocity in water. Nevertheless, other diffraction hyperbolas showing higher velocities (around 0.05 m/s) were also noticed and not so clearly interpretable. Only in restricted parts of the pond bottom, thin decimetric layers of lacustrine sediments were found. As a consequence, it is unlikely that the aforementioned anomalous diffractions were due to blocks or slabs within the sediments. Apart from the blocks sliding within the pond from the lateral debris covers, the sedimentation of finer materials within the basin is almost negligible and the pond seems to directly lie on a rocky bedrock or on very coarse debris bodies. In this respect, ERT results (Fig. 3) highlighted a submerged high-resistivity promontory, elongating from the SW to the NE of the pond. The morphology of this body is highlighted with the