A new framework for exploration targeting: Integrating multifractal geochemical analysis, structural controls and fuzzy C-means unsupervised clustering

In mineral exploration, detecting weak geochemical anomalies in covered areas remains a significant challenge due to overlapping backgrounds and anisotropic mineralization controls, particularly in tectonically complex regions like the Takhte-Soleyman district, northwest Iran, part of the Urumieh-Dokhtar magmatic arc (UDMA). This study addresses these issues by introducing a novel framework that enhances exploration targeting for gold (Au) and lead‑zinc (PbZn) deposits. The methodology integrates multifractal geochemical anomaly separation, including multifractal inverse distance weighting (MIDW), spectrum-area (S-A) and local singularity mapping (LSM) models, alongside distance-distribution analysis (DDA) and fuzzy c-means (FCM) clustering, to account for structural and geochemical complexities. Factor analysis identifies key elemental associations—Au with As and Sb, and PbZn with Ag and Cd—while DDA delineates northwest (NW) and northeast (NE) trending faults as critical controls, with approximately 85 % of Au and 77 % of PbZn deposits located within 1 km of these faults. The framework refines prospectivity models by optimizing FCM clustering with the V_XB index, targeting high-potential zones. Results demonstrate FCM-based LSM's superior anomaly detection, achieving area under curve (AUC) values of 94.8 % for Au and 88.08 % for PbZn, with strong spatial correlations to known deposits, offering a practical tool for exploration in challenging environments. This approach advances mineral prospectivity mapping (MPM) by overcoming limitations of traditional methods, with potential applications to other metallogenic provinces.