Figure 4. (a) An example of raw receiver gather at station 487 showing notable backscattering of surface-waves marked by black arrows. (b-d) The same receiver gather as (a) but after different processing stages, (b) after the application of refraction static corrections, (c) after spectral equalization, deconvolution, time variant filter and median filter, and (d) after surface-consistent residual static corrections and top mute.
Landstreamer and wireless data were processed separately and merged after stack using a similar procedure as described by Malehmir et al. (2022). Landstreamer data were handled with a straightforward processing scheme with the aim of imaging the top of the bedrock. The streamer processing is largely based on building a velocity model for the NMO corrections and maintaining higher frequencies. In contrast, the wireless recorder data processing required carefully selected processing steps and parameters due to deeper targets and a lower S/N ratio.
Four seconds of wireless recorder data were processed. First-break picking was the most time-consuming step, however rewarding as the refraction static correction step was crucial for improved imaging. Other important steps were trace editing, median filters, and velocity analysis coupled with surface-consistent residual static corrections. For stacking purposes, a slalom CMP line was positioned following the centers of midpoint distributions; an identical CMP geometry was also used for the landstreamer data to enable merging of the two data sets. After stack, phase-shift migration was applied and data were time-to-depth converted using a constant velocity of 6000 m/s. Table 2 shows the key reflection seismic processing steps applied to the data.