Nanhong agate, esteemed for its vivid color and natural shine, is experiencing a scarcity in supply despite its high demand. The primary deposits of agate, typically found near the surface, have not been extensively explored due to the predominance of traditional manual excavation methods. This research examined the Nanhong agate deposits in the Zhaojue–Meigu region of Liangshan, China, employing the integration of geological and geophysical surveys. Field geological surveys allowed us to outline the general areas where agate is found. Following this, using magnetic surveys, vertical electrical sounding, and controlled-source audio magnetotellurics, agate deposits were located within the conglomerate layer of the second member of the Feixianguan Formation from the Lower Triassic period at depths of less than 100 m. Our results identify mineralized layers, Xuanwei Formation mudstone, and the underlying bedrock, thus supporting the creation of a mineral prediction map. This research provides essential insights and guidance for agate exploration and the development of associated mineral resources.

Controlled-source electromagnetic method (CSEM), as a significant geophysical exploration technique, plays a crucial role in imaging subsurface structures. To enhance the accuracy and efficiency of CSEM simulations, this paper introduces a 3D unstructured vector finite element numerical simulation method based on the third-type boundary condition. Vector finite elements are particularly suitable for handling discontinuities in the electric field normal. They automatically satisfy tangential field continuity and zero-divergence requirements, providing a solid foundation for forward modeling in the CSEM. The adoption of the third-type boundary condition aims to reduce computational scale while ensuring higher simulation accuracy. Using this method, we conducted detailed numerical simulations on various models, including layered models, single-anomaly models, composite-anomaly models, and layered-anomaly models. The experimental results demonstrate that the algorithm accurately reproduces the electromagnetic responses of various geological models. It also exhibits superior computational accuracy under low-frequency conditions, outperforming traditional simulation methods. In summary, the 3D unstructured vector finite element numerical simulation method proposed in this study offers an efficient and reliable solution for CSEM, which is of great significance for advancing CSEM technology, especially in inversion techniques and data interpretation. Future work will focus on further optimizing algorithm performance and exploring its application potential in complex geological environments.