Change Of Support For Recoverable Resource Estimation

In the estimation of recoverable resources, geostatistical techniques such as Multiple Indicator Kriging require a change of support correction to derive theoretical recoveries based on an assumed level of mining selectivity. The change of support models most widely used in practice entail model based generalized corrections either on a global or local basis. Examples include the affine and indirect lognormal methods. Change of support corrections derived empirically through short-scale conditional simulation realizations often show notable differences compared with theoretical corrections determined using model variogram parameters. The existing changes of support models are reviewed and an alternative method is proposed.