Pseudo-3D Receiver Deghosting of Seismic Streamer Data Based on l1 Norm Sparse Inversion

Abstract

The ghost effect in marine seismic data causes low-frequency suppression and frequency notch, resulting in incomplete frequency information for seismic records, which poses challenges for high-resolution imaging. The deghosting effect depends on the approximation of the ghost delay operator. Due to the strict requirements of dense sampling, the 2D deghosting method for a densely sampled inline dataset is still the mainstream method for marine data processing. As the trade-off, inversion-based methods are widely used in the industry to reduce the influence of the inaccurate ghost delay operator. In order to overcome the sampling limits and improve the 2D deghosting effect, we propose a pseudo-3D deghosting framework based on an l1 norm sparse inversion. In the framework, the inversion problem is divided into two subproblems, i.e., pseudo-3D operator building and optimization inversion. Considering the data coherence along the shot direction, we derive a pseudo-3D ghost delay operator to deghost simultaneously for multi-shot gathers. We then introduce a sparse inversion method in the pseudo-3D radon domain (multi-shot gathers) to further improve the inversion accuracy. The proposed framework is easy to implement, is not sensitive to noise, and shows superior performance in synthetic and field examples.