GNGTS 2014 - Atti del 33° Convegno Nazionale

52 GNGTS 2014 S essione 3.1 The endof theMSCevent probablycoincidedwith themarine re-floodingof theMediterranean basin at the Miocene/Pliocene boundary, 5.33 Ma ago (Krijgsman et al. , 1999; CIESM, 2008). This is testified by pelagic sediments draping the cored margins (Ryan et al. , 1973; Rehault et al. , 1984) and extending throughout the entire north-western basin. The Lower Pliocene unit in the north-western Mediterranean basin has a typical semi-transparent acoustic facies. It lies beneath a more reflective facies (Rehault et al. , 1984; Sage et al. , 2005; Geletti et al. , 2014) representing the Upper Pliocene-Quaternary turbiditic bodies of coarser clastic sediments transported by the Rhône (Rehault et al. , 1984; Aslanian et al. , 2012). On the contrary, the unit is often absent or very thin on the Sardinian slope where sediment supply has been minor: the PQ units are 1200-1600 ms TWT thinner at the foot of the Western Sardinian slope than that in the north-western sector of the deep basin (Geletti et al. , 2014). Reprocessing in time domain. In general, the available seismic dataset is affected by several problems, that can be summarized as follows: - structural complexity characterizing part or all the sedimentary cover of the Messinian salt, due to the halokinesis generating domes, diapirs, faults, etc; this produces intense lateral lithological changes, implying strong lateral velocity variations for the seismic wavelets that propagate through the subsurface. - Pre-salt geological features are mostly unidentifiable, except on some of the WMR and CROP data, due to salt presence. - Salt-related pull up events. - Sea bed multiple (at about 7 s TWT) that may overlap primary reflections. - Over-migration of deep seismic reflectors. - Out of plane effects, which are commonly related to salt diapirism. Only accurate migration of a 3D seismic dataset will solve this geophysical problem (Yilmaz, 2001). The reprocessed part of the CROP C88M1 line, using the Focus (Paradigm Inc., 2007) software, is named C88M1 REPRO (location in Fig. 1A). It is 56.6 km long and it extends from the Western Sardinian slope to the north-western part of the eastern Sardo-Provençal deep basin. It was selected because of the investigation depth reached by the CROP survey itself and because it represents one the best examples of the geological and structural setting within the explored sedimentary basin, as identified by a first stage of interpretation. It contains also several different shaped salt diapirs, widely disrupting the MU overburden and causing high lateral velocity variation. A wide salt diapir is also present reaching the sea bed and generating a suspected (considering the first seismic interpretation) pull up event of about 0.6 s TWT for the underlying reflectors. Starting from raw data, the processing flow comprises: trace editing, multiple attenuation, normal-move out-correction, gain, spike deconvolution, stack, migration. Because of strong lateral velocity variability, the velocity function (named VELGEOD and executed using the GeoDepth software -Paradigm Inc., 2008) was determined by iterative analyses on semblance panels derived from CDP gathers chosen alternately within a small sedimentary basin and in correspondence to a salt structure. The velocity picking was overall based on the amplitude response displayed on semblance panels, carefully evaluated beside the correspondent best dynamic correction (normal-move-out) of mayor reflection hyperbolas displayed on each CDP gather. The preliminary seismic interpretation was consulted in case of ambiguity (e.g., doubtful velocity inversions, primary reflections) in order to respect geological and geophysical constraints. However, the computation was hampered due to the lack of calibrations by well data. After the reprocessing in time domain a geological interpretation was done. Adetail is shown Fig. 2A, where the wide salt diapir was newly interpreted as connected to the mother layer, since the reflector (affected by pull up) that was present inside the structure and that was interpreted as the top of the UU, is not visible in the new outcome. Somewhere in the reprocessed data, the thickness of the UU resulted higher than what was preliminarily interpreted, this is due the