Integration of various geophysical data with geological and geochemical data to determine additional drilling for copper exploration

In this paper, three data-driven methods (i.e., Bayesian, k-nearest neighbour (k-nn) and neural network classifiers) are used to generate a prospectivity map for porphyry-Cu deposits. Different data layers of geological, geophysical and geochemical themes are integrated to evaluate the Now Chun porphyry-Cu deposit, located in the Kerman province of Iran, and to prepare a prospectivity map for mineral exploration. Both the Bayesian and k-nn methods showed correct classification rates (CCR) of 52.38% for 21 boreholes divided into five classes. Three types of the neural networks including multi-layer perceptron, radial based function (RBF) and probabilistic neural network are applied to evaluate the result. Among neural networks used, the RBF neural network generated the highest CCR equal to 80.95%. Multi-classification of the prospect for detailed study could increase the resolution of the prospectivity map and decrease the drilling risk.

► We apply three different data-driven methods for mineral prospectivity mapping. ► Different data layers, including geological, geophysical and geochemical data, are used to prepare a prospectivity map. ► Porphyry copper deposit prospectivity map shows acceptable correct classification rate. ► We increase resolution of prospectivity map into multi-class.