Advances In The Practical Implementation Of Indicator Geostatistics

Indicator Kriging (IK) is being commonly applied to mineral resource assessment problems. More recently, indicator simulation has also been used for sensitivity analysis and difficult change of support problems. The essential theory underlying indicator geostatistics has not changed since its introduction 10 years ago; however, the last decade has produced many important developments in its implementation. The objective of this paper is to consider some of the more important implementation details. Examples from the Porgera gold deposit in Papua New Guinea are used to illustrate the steps in an indicator geostatistical study and the importance of various implementation decisions. Specifically, in many precious metals deposits the continuity of the grades is typically quite anisotropic. This anisotropy varies in both magnitude and direction throughout the deposit. Building a model of this anisotropy can considerably improve the local kriged or simulated points. One approach to build this model is presented. In addition, there is typically more than one mineralization style and geological rock type; the spatial character of the mineralization can be quite different within each geological rock type. The rational treatment of anisotropy and geological subdivisions is common to all mineral resource assessment procedures; the importance of this qualitative geological information is not diminished with more sophisticated techniques such as indicator kriging. This paper presents an overdue comprehensive look at the practical aspects of indicator methodology. Many of the techniques and procedures documented in this paper have been developed during the course of detailed mine feasibility studies.