Mapping of potential Cu and Au mineralization using EBF method

摘要

Theory of data-driven evidential belief function (EBF) is based on generalized Bayesian probabilities. It computes the reliability of evidences according to the presence or absence of known indices. This method has a reliable performance even in the case of limitation of data layers diversity. The study aims to detect the most probable regions for copper and gold mineralization in Takhte-h-Soleiman, Iran using EBF method. The available data layers including lithological map, faults, geochemical anomalies and known indices maps were processed to generate a potential map for the study area. The Bbelief^, Bdisbelief^, Buncertainty^ and Bplausibility^ functions were therefore calculated based on data-driven EBF for each pattern of evidences in two different classification types. The data layers were then combined with the use of the EBF AND-OR operators, and the Bbelief^ map was proposed as a target recognition map for following detail explorations. As a validation criterion, the ratio of indices number to the total area of the most favourable regions in each map was computed. According to the highest ratio for the belief map resulted by integrated AND-OR operators on equal interval classification (0.62), it was selected as the target map for primary explorations.