Reducing the Geological Model Uncertainty in Miduk Copper Deposit by Post-processing

Abstract

Geostatistical modeling, in recent years, has been widely and increasingly used in analyzing underground structures. Using the combination of uncertainty reduction techniques and modeling processes could increase the identification accuracy of underground structure models. In this study, Indicator Kriging (IK) method is employed to model four geological units comprising leached, oxidized, supergene and hypogene zones. The aforementioned units are all located within Miduk Copper Deposit. To evaluate the uncertainty index based on the modelled geological structure, Bernoulli variance is employed. In order to reduce the uncertainty amount of the executed geological model, post-processing procedure is performed using resampling of the interpolated model. This procedure is reiterated as many times as possible until a geological model with an acceptable uncertainty amount is acquired. At the final step, blast holes data is used in order for the validation of the results. The yielded results indicate that the amount of uncertainty dropped from 5.87% to 2.98% after implementing the post-processing procedure.