GNGTS 2014 - Atti del 33° Convegno Nazionale

GNGTS 2014 S essione 3.1 117 overlapped on an orthophoto of Innerhytta pingo (c). The source and the receiver points were positioned with DGPS. We acquired three lines for reflection/refraction seismic in a tomographic inversion perspective. We used five 24-channels Geode seismographs, sampling rate 0.250 ms, for a total recording window of 2 s. The vertical 14 Hz geophones were listening, together with the snow streamer. Seisgun, sledgehammer and firecrackers were used as sources. To increase the coverage of the tomographic experiment, while acquiring the 2D lines, we recorded also cross- line, using both the snow-streamer and the geophones, or shooting offline. The design of the seismic lines was done taking into account the needs of illuminating the different parts of the pingo area, reducing the possible effects of the rough topography and optimizing acquisition costs and duration. Line L100 is traced at the foothill of the pingo, on the bottom of the above the Adventelva (Advent River) floodplain. Line L200 is traced from L100 across the NE corner of the pingo, while L300 starts from about half of L200 toward a saddle to the East of the pingo’s apex. This pattern implied a high coverage above the pingo apex and the areas where the highest lateral variability in depth is expected. Summarizing (see Fig. 2 c): L100 – 91 channels: 67 channels inline and 24 channels cross-line (snowstreamer S1), 21 shot points; L200 – 120 channels: 72 channels inline (48 geophone + 24 channels S1 snowstreamer) and 48 channels cross-line (L300), 18 shot points; L300 – 120 channels: 48 channels inline and 72 channels cross-line (L200: 48 geophone + 24 channels S2 snowstreamer), 24 shot points. In addition to this, two experiments of surface-wave recording were done, one along L100 and the second one along L200 using two 24-channels Geode seismographs, sampling rate 0.250 ms, for total recording window of 2 s. The sensors were vertical 4.5 Hz geophones, and the sources were sledge-hammer in the case of L100 and the two different weight drops on L200. The data and data analysis. Since one of the goals of the IMPERVIA spring 2014 experiment was to test the capability of the seismic method for near-surface, high-resolution targets in extreme conditions as Svalbard, analyses were carried out both on the field and after the survey, in order to state the efficiency of the different sources for our goals. In Fig. 3 a comparison between a shot gather when the source is the seisgun (a) and the firecrackers (b), respectively. It is evident the higher resolution due to the latter source, more impulsive and with stronger energy with respect to the seisgun. The sledge-hammer did not provide comparable results. A better quality could be obtained through several shots repeated and stacked, what would, however, imply too long times and more problems in ground coupling, with respect to the other two sources. As regards as the surface wave experiments, the preliminary data analysis state that the best results were provided using the weight-drops (the heaviest one) with respect to the sledge- hammer, seisgun and firecrackers. The data processing and data analysis are on-going: it encompasses both standard processing of the reflection data, tomographic inversion of reflected and refracted arrivals, and surface wave inversion. Due to the strong velocity variations and the presence of negative gradients conventional reflection processing can be difficult. The velocity field provided by a tomographic inversion, exploiting all the arrivals generated throughout the experiment (e.g. Vesnaver et al. , 1999) can be the input for a correct move-out data correction and successive stack, or as input for a depth data migration. In particular, we started the tomographic analysis of the first line L100, considering both the inline, as well as the cross-line snow-streamer traces. We use the method described in detail by Vesnaver et al. (1999) and Rossi et al. (2007), with a minimum time ray-tracing modified from Böhm et al. (1999) and the iterative inversion SIRT process (van der Sluis and van der