Abstract

This article presents a computer program to directly simulate continuous regionalized variables such as mineral grades over a block support. Simulation is performed in the scope of the discrete Gaussian model and does not rely on a block discretization. The realizations can be made conditional to point-support data by use of simple or ordinary kriging, depending on whether or not the average value of the data is considered known. The proposed program can account for an information effect (misclassifications between ore and waste) by co-simulating the true block-support grades together with the grades that will be predicted at the production stage to discriminate between ore and waste. © 2009 Elsevier Ltd. All rights reserved.