With the advances in software speed and capability, many of us now are running multiple scenarios on entire models rather than just small areas or a few blocks. In SRK’s experience, both the geological and geostatistical academic theory and our rules of thumb often prove unuseable or incorrect in the real world. The only way to validate and finetune our models is to complete the entire model. If it lacks some property we were expecting, or contains some property we were not expecting, then we need to find out why and/or run different scenarios to see what changes. As we run several scenarios on models more often, we have come to understand that every deposit is different and requires its own parameters to obtain valid and useful results. We’ve been calling this process empirical geostatistics.
For example, the graph below shows how the grade estimate and the kriging slope of regression estimate change, for a group of ten blocks, when different maximum sample numbers are used in the search neighbourhood. Grade and estimation quality are not independent, as the estimation quality changes so does the estimated grade.
SRK is planning a series of articles and papers on the theme of empirical geostatistics focusing on what happens in reality with changes in different sets of parameters rather than what happens in theory.
