SVD-based Hydrophone Driven Shear Noise Attenuation for Shallow Water OBS

Abstract

In ocean-bottom seismic (OBS) data processing, the vertical geophone component Z data is often corrupted by a high level of “shear” noise which is not recorded by the hydrophone P. This noise highly influences the quality of the wavefield separation and, consequently, the final up-and down-going wavefield imaging products. With the growing importance of large, dense OBS acquisitions, we need methods that are both capable of accurately attenuating the “shear” noise and applicable to large amounts of seismic data. In this paper, we propose a singular value decomposition (SVD)-based noise attenuation method that relies on the hydrophone to drive the algorithm, allowing a good signal preservation on the vertical geophone. The proposed method uses a cross-correlation matching pursuit to compare the hydrophone and geophone data, and is equipped with an optimized SVD kernel. We demonstrate the performance of the method on a dense wide-azimuth 3D OBS shallow water acquisition involving 56,000 4-component nodes. Results demonstrate the efficiency of the proposed method in removing “shear” noise on the vertical geophone component with no damage to the signal.