Enhanced prediction of deep-seated Cu-Mo porphyry mineralization: A comprehensive interpretation based on 2D inversion of audio-magnetotelluric data

• We employed an innovative data processing methodology that incorporated machine learning technologies to remove noise from AMT data and improve subsurface imaging. • Utilized Audio-Magnetotelluric (AMT) technology in conjunction with 2D OCCAM inversion to improve the accuracy of predicting deep-seated Cu-Mo porphyry mineralization. • Medium-resistivity structures with a resistivity of approximately 2500 Ω.m have been detected at depths greater than 500 m associated with significant Cu-Mo mineralization potential in the Baohuashan region of Jiangsu Province, China. • Confirmed the existence of Cu-Mo mineralization by drilling holes, thereby verifying the reliability of the AMT approach in predicting ore bodies disseminated located at great depths and overcoming past constraints in exploration. • The findings enhance our comprehension of deep-seated porphyry systems and improve exploration targeting in the region. Copper and molybdenum are indispensable raw materials for modern industry and green energy. The rising demand and prices necessitate the exploration of new mineral resources. Porphyry copper-molybdenum (Cu-Mo) deposits are crucial suppliers of these metals, yet their deeper exploration (>500 m) remains underexplored in the study area. Baohuashan is known for its intricate geological composition, rich in Cu-Mo porphyry. Previous exploration methods struggled to map extensive underground resources accurately. To solve this problem, this study utilizes Audio-Magnetotelluric (AMT) technology to investigate deep Cu-Mo porphyry deposits in Baohuashan, Jiangsu Province, China, aiming to address the challenges of accurately characterizing mineralization in this complex geological setting. We employed the AMT method, conducting two long survey lines with 50 AMT stations to explore and characterize the second-space of the space of the study area. We used a machine learning tools call watex to process the data, filtered the initial 53 frequencies used to collected and obtained 27 valid frequencies (10.16 Hz-1032 Hz). The adaptive-moving-average spatial filtering has been performed to rectify the static shift. Furthermore, we completed a transverse electric (TE) and transverse magnetic (TM) modes joint inversion using OCCAM2D inversion with a grid of 79 x 65 cells for both lines following a regularization factor selection criterion our parameters used. The two-dimensional models obtained by the robust data processing and inversion enhanced the detection and visualization of resistivity anomalies indicative of potential Cu-Mo deposits. Our results reveal distinct medium-resistivity (∼2500 Ω.m) structures at depths exceeding 500 m, likely associated with Cu-Mo deposits and regional faults (F2 and F15) oriented NE-SW and NW-SE, characterized by low resistivity. These findings suggest significant mineralization potential previously undetected by conventional methods. The presence of mineralization has been confirmed by three drill holes proposed and executed.