GNGTS 2019 - Atti del 38° Convegno Nazionale

GNGTS 2019 S essione 3.1 569 with maximum static allowed per iteration equal to 10, 15, 20, 30 ms respectively and a final maximum static of +/- 100 ms. We needed two runs of residual static corrections where the latter was used as final refinement. The first Frequency-Space deconvolutions were applied to further reduce the noise and to improve the coherency of the reflected signals. The first 2D deconvolution is applied in the shot domain. The second 3D deconvolution was applied on 14 common offset sub-volumes in a frequency range of 10-45 Hz. The prediction considered distances equivalent to 3 Inlines and 3 Xlines. Finally, in the pre-stack domain, we evaluated the spatial correlation between adjacent samples by means of the crosscorrelation and we assigned a weight to each sample on the base of it. In this way, the amplitudes of the uncorrelated samples are reduced by small weights, while the amplitudes of spatially-correlated samples were enhanced. The step was used to improve the quality of the stack section as well as the reliability of the events observed on the stack volume. Fig. 1b shows a shot gather after the processing discussed in this section, while Fig. 3 shows the comparison between a portion of the stack volume before (Fig. 3a) and after (Fig. 3b) processing. Fig. 2 - Processing sequence applied to the data. The colour indicates the steps particularly pertinent to the denoising (blue), to the static corrections (grey), to the coherency enhancement (green) and to weighting of the stack (orange). The white colour is used for other standard processing steps. Fig. 3 - Comparison between the stack volume before and after the processing sequence illustrated in Fig. 2.