利用K-SVD字典学习算法压制海洋大地电磁噪声

海洋大地电磁数据处理的结果会影响数据反演和解释的准确性.在浅水中, 由海浪引起的强电磁干扰和大地电磁场叠加在一起, 使得噪声所在频段内的视电阻率和相位曲线出现畸变.针对这种强干扰, 本文将K-SVD字典学习算法应用于海洋大地电磁信号处理中, 通过稀疏表示海浪感应磁噪声实现大地电磁磁场信号的重构, 并结合视电阻率信息进行相位校正.仿真数据和南黄海实测数据的处理结果表明, 该方法能够有效压制海浪感应磁噪声, 恢复后的视电阻率和相位曲线更加连续、光滑.

A noise suppression method of marine magnetotelluric data using K-SVD dictionary learning

The result of marine magnetotelluric (MT) data processing affects the accuracy of data inversion and interpretation. In shallow water, strong electromagnetic interference caused by ocean waves stacks with the MT fields, which makes the apparent resistivity and phase distortion in the frequency band where noise occurs. To reduce the influence of this noise, K-SVD dictionary learning algorithm is applied to marine MT signal processing and the MT signal is reconstructed. Phase correction is applied in data processing combined with the apparent resistivity information. The processing results of both the simulated data and the field data show that the proposed method can effectively suppress the wave-induced magnetic noise. The apparent resistivity and phase curves are more continuous and smooth after denoising.