Incorporating Spatial Variability of Source Wavelets in a Robust CL-SRME Framework for Offshore Wind Seismic Data

Multiple reflection attenuation is essential for offshore wind farm site selection, but conventional closed-loop surface-related multiple elimination (CL-SRME) struggles with complex geological conditions due to its assumption of spatially uniform wavelets. To break this assumption, we propose an enhanced CL-SRME framework that incorporates a multichannel surface correlation operator, explicitly designed to conclude the spatial variability of source wavelets. The framework utilizes wavelet continuity to stabilize and optimize this ill inversion by integrating a low-rank approximation-based denoising engine within a regularization by denoising (RED) constrained inversion process. Then we introduce an improved alternating direction method of multipliers (ADMM) algorithm, ensuring computational efficiency even under complex geological and noise conditions. The effectiveness of our framework is demonstrated through synthetic, field, and passive seismic data, showcasing significant improvements in seismic data processing. The proposed technology can significantly improve the identification of geological features and holds great potential for enhancing the efficiency of marine seabed surveys.