The Jiaodong region ranks as the world’s third-largest gold metallogenic province, where Late Mesozoic gold mineralization exhibits close genetic connections with cratonic destruction and multi-stage plate tectonic interactions. This study systematically deciphers the deep-seated architecture and metallogenic controls through integrated analysis of gravity, aeromagnetic, and magnetotelluric datasets. The key findings demonstrate the following: (1) Bouguer gravity anomalies reveal a “two uplifts flanking a central depression” tectonic framework, reflecting superimposed effects from Yangtze Plate subduction and Pacific Plate rollback; (2) zoned aeromagnetic anomalies suggest that the Sanshandao–Jiaojia–Zhaoyuan–Pingdu Metallogenic Belt extends seaward with significant exploration potential; (3) magnetotelluric inversion identifies three lithosphere penetrating conductive zones, confirming the Jiaojia and Zhaoyuan–Pingdu faults as crust mantle fluid conduits, while the Taocun–Jimo fault marks the North China–Sulu Block boundary; and (4) metallogenic materials derive from hybrid sources of deep Yangtze Plate subduction and mantle upwelling, with gold enrichment controlled by intersections of NE-trending faults and EW-oriented basement folds. Integrated geophysical signatures indicate that the northwestern Jiaodong offshore area (north of Sanshandao) holds supergiant gold deposit potential. This research provides critical constraints for the craton destruction type gold mineralization model.

Audio magnetotelluric (AMT) sounding is an essential tool for mineral exploration and subsurface electrical structure imaging. However, persistent cultural noise can severely degrade AMT data quality, particularly in mining areas with strong anthropogenic interference. In this study, we introduce an improved K-Singular Value Decomposition (K-SVD) dictionary learning method for time-domain AMT data denoising, incorporating three key innovations: (1) automatic identification of noise-contaminated segments using local kurtosis; (2) adaptive dictionary initialization combining principal component analysis with Gaussian perturbation, which accelerates convergence and avoids local minima; and (3) online atom screening and updating to maintain the noise specificity of the learned atoms. The proposed method learns noise morphology directly from raw AMT data without requiring reference stations or external training datasets. Synthetic experiments demonstrate that the method more effectively suppresses square-wave and charge-discharge noise compared with wavelet thresholding methods and conventional K-SVD, while better preserving signal morphology. Field applications further show that it significantly enhances the quality of observed data and apparent resistivity-phase curves compared with the widely used robust estimation method. The proposed method provides an efficient, reference-free solution for AMT data denoising in mining areas with strong cultural noise.