Denoising land-based controlled-source electromagnetic data based on a same-site noise reference channel

Abstract

The applications of land-based controlled-source electromagnetic (CSEM) exploration are severely limited by strong noise interferences, particularly in mining areas. In this study, we introduce a novel denoising method for CSEM data using a same-site noise reference channel (NRC). While recording data through the normal survey channel (NSC), an additional set of the NRC was added at the same site. The NRC had a different surveying azimuth compared to the NSC and contained minimal or no useful signals. However, the noise characteristics in both the NRC and NSC were considerably similar due to their simultaneous acquisition at the same site. By establishing a set of overdetermined equations for the NSC based on quantified spectrogram characteristics of the NRC, the noise can be effectively eliminated from the NSC using least-squares inversion, resulting in enhanced signal-to-noise ratio data. The effectiveness of the proposed CSEM data-denoising method was validated through its application on real data, and the proposed method is applicable to other types of artificial source data.