Magnetic Properties as Proxies for Geochemical Prediction in Mining Tailings: A Semi-Supervised Spatial Approach

Abstract

Mine tailings require careful monitoring and management, but traditional geochemical characterization methods are costly and time-consuming. This study demonstrates that magnetic properties can serve as effective proxies for predicting copper concentrations in mine tailings through an innovative spatial modeling approach. Analysis of magnetic and geochemical measurements from a Chilean copper mine tailing showed that magnetic properties combined with spatial modeling techniques could predict copper concentrations with high accuracy (R2 = 0.873 ± 0.085). The spatial distribution of magnetic properties revealed coherent patterns that effectively predicted geochemical characteristics. This approach substantially reduces characterization costs compared to traditional methods while maintaining accuracy. Our findings establish magnetic properties as valuable screening tools for tailings characterization, offering mining operations a cost-effective approach to environmental monitoring and management.