Conventional streamer seismic acquisition suffers from the sea surface ghost and resulting notches. As a common broadband acquisition program, multi streamers and their combination are now in close pursuit. However, the traditional over/under "dephase and sum" algorithm might produce poor quality results in rough conditions because of the violation of calm sea surface assumption. In this abstract, we propose an adaptive over/under streamers data combination strategy. The new method adaptively estimates the amplitude spectrum of the ghost operator and reduces the equivalent notch effect to a theoretical minimum. Besides, by turning waste into treasure, we can make full use of the random receiver depth to improve resolution and signal-noise ratio (SNR) rather than to be interrupted by unfixed depth.
Conventional acquisition in marine seismic exploration can’t remove ghost effectively. Due to incidence angle of the upgoing wavefield and depth of the streamer cable, interference between upgoing and downgoing wavefields creates nulls or notches in the recorded spectrum. It is well known that a streamer towed at shallow depth is good at receiving high-frequency components but greatly attenuates the low-frequency components. This leads to high resolution imaging of shallow area but poor in frastructure. Although a deep-towed streamer is good at receiving low-frequency components and propagate robustly in deep,it produces notches within the general seismic frequency range and leads to narrow bandwidth. As illustrated in Figure1, the notches reduce the useful bandwidth especially at increasing streamer depths.
Over/under streamers and variable-depth streamers are common ones, but they are all limited by kinds of additional factors. First, in rough sea surface conditions with environmental noise, the streamers can’t be towed in a fixed depth, a streamer towed within an error range of one meter can be considered reasonably. Besides, as shown in Figure1, variable-depth streamers could obtain sufficient low-frequency components, but relatively lack of high-frequency energy.
In this work, we propose an adaptive over/under streamers data combination strategy and adaptively estimate the amplitude spectrum of the ghost operator corresponding to each trace. Rather than to be interrupted by receivers of random depths, our method can furthest benefit from the deep receivers to improve SNR and shallow ones to acquire more high-frequency components, respectively. In this case, variable-depth streamers could also be combined with a shallow streamer to strengthen its high-frequency energy. We applied this strategy to synthetic data examples from over/under rough streamers model, and new proposed over/variable-depth streamers model, and a real data case from China. We got good results for both cases.