Shear Noise Suppression for Seafloor Vertical Geophone Using Robust Reduced-Rank Filtering

The present study proposes a data processing application for shear noise suppression of seafloor vertical geophone seismic data. Typically, seafloor seismic acquisition systems record up to four components, namely, one hydrophone and three geophone (one vertical and two horizontals) components. A key advantage of multicomponent data acquisition using ocean bottom sensors, either nodes or cables, is the ability to separate the full wavefield into up- and downgoing wavefields by utilizing the hydrophone and vertical geophone. Unlike the hydrophone, the vertical geophone data recorded at the seafloor are usually contaminated by a number of noise categories, such as Scholte waves, backscattering noise, and various other types of noise. In particular, shear noise recorded by the vertical geophone causes a number of issues and degrades the effectiveness of subsequent data processing steps. Therefore, we need to take a shear noise suppression procedure into consideration prior to full data processing workflows. In this study, we suggest a denoising method for seafloor vertical geophones by robust rank-reduced filtering. We mainly aim to remove shear noise that exhibits random behavior in the common shot domain by applying a 3-D singular spectrum analysis via a joint sparse and low rank approximation. The proposed denoising application for seafloor vertical geophones successfully attenuates the shear noise and will significantly contribute to the subsequent data processing including up- and downgoing wavefield separation.