Publications

Broadband acquisition aims to improve the bandwidth of seismic data, which in practice means extending the low-frequency end of the spectrum without limiting the high-frequencies beyond the natural Earth response. These techniques focus on the receiver-side ghost, and commonly used are co-located velocity and pressure sensors and dual-depth or variable depth hydrophones. All these methods rely on processing to achieve the final receiver side de-ghosted data, as the “dumb sum” of the measurements will lead to poor results. With sufficient signal-to-noise in the data it is possible to de-ghost the receivers towed at a moderate single depth by tuning the acquisition design, with consideration of the source emission response in combination with the streamer reception response. For the source emission response we consider the bubble in providing the low-frequencies which are reduced as the hydrostatic pressure increases. The source ghost frequency increases with higher emission angles giving notch diversity. On the receiver side we need to tow deep to minimize the noise from swell and low-frequency zero notch. Finally we want the streamer ghost notch, or weakest signal reception, to be where the source emission is strongest. Putting the source and receiver issues together, we can conclude that the optimum would be to tow the source at 5-7 m, and the streamer at multiples of 2-3 of the source (15-25 m). A test line was acquired that shows the equivalency of slant streamer and flat depth streamers in terms of post-stack amplitude spectra, showing that the acquisition design and pre-stack de-ghosting processing methodology is effective in providing roadband data, and low-frequencies are harder to recover through processing.