Source signature and static shifts estimations for multi-component ocean bottom data

Introduction

Many migration algorithms, such as shot-profile, plane wave, and reverse time migration, require the input of source wavelets as part of the calculation. Ocean bottom data have a distinct advantage in the estimation of the source wavelet, because an ocean bottom seismometer and hydrophone can pick up the source signature as direct arrivals through a relatively homogeneous medium. As shown in figure 1, the source wave path only needs to go through the water once to reach the receivers.

DirectSource
Figure 1. For ocean bottom data acquisition, the source wavelet is picked up by the down-going direct arrival data.[NR]

Source signatures are offset-dependent because each shot is generated from an array of airguns and the obliquity of the source wave path directly affects the shape of the wavelet. A crude way to extract the source signature from a direct arrival is to time-window the pressure signal at the receiver. However, this technique is often inaccurate because the source bubble often overlaps with the recorded primary reflection. A better way to extract the source signature is to first separate the total data into up-going and down-going wavefields, and then perform time windowing only on the down-going component. We can do that because the direct arrival is strictly down-going, as illustrated in figure 1. For multi-component ocean bottom data, we can perform up-down separation using the pressure (P) and vertical velocity (Z) wavefields. This method is called PZ summation.

There is significant literature on the method of PZ summation to decompose data into up-going and down-going wavefields (Sonneland et al., 1986; Amundsen, 1993). Some methods involve separation in the $\tau - p$ or $\omega - k$ (Fourier) domain. However, transformation into other domains poses a problem when the data are aliased or sparsely sampled. In this study, we perform PZ summation in the physical (t-x) space to avoid this difficulty.

In the next section, we first present the field dataset and discuss some of the challenges. Futhermore, we estimate the source statics in a common receiver gather. Variations in shot deployment depth and the water column cause time anomalies that can be approximated as surface-consistent static time shifts. We compare two methods of static time shift estimation. One method uses the maximum pulse of the source while the other use cross-correlation. Finally, we discuss and show the result of the source signature extraction on the field dataset.

Source signature and static shifts estimations for multi-component ocean bottom data