Application of Sparsity-Enhanced Predictive Deconvolution in Offshore Wind Farm Site

Abstract The development of offshore wind power has become an integral part of the global energy transition. Geological surveys targeting the bedrock surface are indispensable in the development process of offshore wind farms, as they directly affect the site selection and design of the wind farms. Single-channel seismic surveying is a towed survey method that offers advantages in exploration efficiency and accuracy. However, strong interference from multiple reflections is a unique challenge in offshore seismic exploration. To address this issue, we propose a sparse-promoting multi-channel predictive deconvolution method. This method begins by establishing a framework for multichannel predictive deconvolution, and then leverages the high redundancy and sparsity benefits of the Shearlet transform to enhance the sparsity of seismic signals. Through the processing of field data, we have verified that our method can effectively suppress multiple reflections, improve signal-to-noise ratio, and enhance the accuracy and efficiency of seismic exploration, providing more accurate and reliable technical support for the preliminary environmental investigation of offshore wind power projects.