85 3. Seismic tomography This is due to: • the pattern radiation of the source having a strong S lobe perpendicular to the borehole; • the acquisition geometry with large angles of incidence, favouring the shearwave conversions. The acquired data were processed with tube-wave filter removal, down-wave field separation and band-pass filters (40-60-300-450) Hz. Figure 3.18 (b) shows a subset of processed data. Figure 3.19 shows the raw geophone data after rotation from (H1, H2) directions to (X, Y) directions, where X is the horizontal axis in the acquisition plane and Y is the transverse (out-of-plane) direction. (a) (b) Figure 3.19 Geophone data: Z-component (upper) and X-component (lower). (a) Raw data after reorientation (X, Y). (b) After down-wave field separation and used as input data for the tomography. Adapted from Beydoun et al. (1989). Given that seismic wave amplitude and travel time information are both used in the inversion algorithm, for a successful solution it is fundamental that careful data pre-processing is carried out to preserve both amplitude and travel time parameters. The pre-processing steps for both datasets were similar: • least-squares approach in the frequency and depth domains to estimate simultaneously the up and down-going tube waves by minimizing the separation residual; • residual waves were filtered by similar processing to eliminate only down-going P and S-waves; • only upgoing S-S and S-P reflected events are used for the tomography;
RkJQdWJsaXNoZXIy NjA3NzQ=