Modeling deep mineralization using surface exploration data of undrilled regions in Kal-e-Kafi prospect area, Isfahan Province

Document Type : Original Article

Authors

1 Isfahan University of Technology

2 PhD. Student, Mining Engineering dep. Urmia University, Urmia. Iran

Abstract

Prediction of prospective subsurface mineralization areas and determining operational priorities for detailed studies are key goals in mineral exploration engineering. Kal-e-Kafi metallogenic system is located in the north-east of Isfahan province and in the eastern border of Anark district, Naein city. In this region, hydrothermal ore-forming processes have led to the emergence of polymetallic gold mineralization in vein types. Although parts of the study area have been drilled, other areas still need detailed studies to design deep explorations. Therefore, this research has been carried out with the aim of using exploration data of the earth's surface to model the pattern of subsurface mineralization in areas without drilling. The proposed solution to solve this is to use new data processing technologies to train machine learning networks in order to recognize subsurface mineralization patterns. In this approach, by producing training data in drilling points and using machine learning models, linear or non-linear statistical relationships between depth data (gold grades recorded in drilling) and surface data (predictive exploration information) are extracted and classification functions are applied to exploration data in undrilled areas in order to obtain patterns of gold mineralization in the deep areas of the study area. Mineralization potential maps from different learning models are combined with each other and produce the final exploration model to define new exploration targets. Spatial distribution modeling of mineralization probabilities and uncertainties related to it has been able to bring efficient practical insight for the management and optimal design of future discoveries within the Kal-e-Kafi area.

Keywords