Field Data Test

The proposed methods are tested on a 2D marine dataset. There are 496 shots with a shot interval of 37.5 m. Each shot is recorded by a 6 km long cable with 480 receivers at an interval of 12.5 m. The nearest offset is 198 m. Each trace is multiplied by $\sqrt{i/\omega}$ in the frequency domain and then scaled by $\sqrt{t}$ in the time domain to correct for 3D geometrical spreading (Zhou et al., 1997). Then, the CSGs are filtered with a Wiener filter to transform the original wavelet to a Ricker wavelet with a 25 Hz peak frequency. The original wavelet is estimated by stacking traces with a strong water bottom reflection, and windowing the water-bottom reflection event. Figure 3.13 shows a shot gather with source at $x=11.3~km$ after preprocessing. For convenience, the trace length is set to be 10 secs, so that the frequency sampling is 0.1 Hz. For the frequency band of 0-62.5 Hz, there are 625 frequency channels to accomondate up to 625 shots.