Re-processing of the CROP-04 seismic data

We present the results of a re-processing of the CROP-04 NVR profile. This re-processing was carried out to improve the interpretability of the data that, due to a number of causes, was deemed as insufficient after the first processing. A complete discussion on the causes responsible for the low data quality can be found in another note in this volume (MAZZOTTI A., STUCCHI E., FRADELIZIO G.L. & ZANZI L., Analysis of the Crop-04 seismic data). In the new processing sequence, we have paid attention from the very start of the operations at the removal of the large incoherent noise component and at the solution of some additional problems related to the structural complexity of the area, to the crooked line acquisition and to the rough topography. The re-processing sequence consists of the following principal steps: semi-automatic killing of the data that are below a given S/N threshold, static corrections computed with two different refraction methods, common offset sorting and FX deconvolution, predictive deconvolution (in two time windows), CMP crooked line sorting (optimized by exploring three different criteria), velocity analyses by means of CVS and of velocity spectra, surface-consistent residual static corrections, reiteration of the velocity analyses, weighted stack based on S/N estimation (by means of Singular Value Decomposition). Part of the re-processing strategy was also driven by some a priori knowledge of the regional tectonic setting of the area and by models. Moreover, some steps required internally developed codes to overcome particular problems. The final result shows an increment of the signal quality up to 4-5 s, and, at least in some segments, the presence of deep events up to 7 s in the west and up to 9-10 s, in the eastern part of the line