The use of auxiliary variable sulphide factor to improve the estimation results of copper by collocated cokriging and kriging with external drift

In projects of reserves estimation, we would like to reduce the estimation variance and related uncertainty. This reduction usually requires extensive and costly drilling. Multivariate geostatistical methods could be noted as an inexpensive and budget saving method to reduce the estimation variance. Often, there exists much secondary data that must be considered in a geostatistical reservoir modeling. All secondary data being used as a secondary variable must be highly correlated with the primary one. This work introduces the collocated cokriging and kriging with an external drift to incorporate sulphide factor as secondary information to estimate the copper content in a porphyry system. The study area was one of the desirable areas of copper, located in Kerman Province. Sulphide factor as highly correlated and fully covered data was selected to improve the estimation of copper in this region. In collocated cokriging, the number of the secondary information used for estimation is reduced to the estimation location. However, the datum colocated with primary data implementations of collocated cokriging is often limited to a single secondary variable. Improved models would be constructed if multiple variables were accounted for simultaneously. Contrary to cokriging, kriging with an external drift does not make explicit the structural link between the target variable and the auxiliary variable, for the latter is considered to be deterministic. After determining the mean of the estimation variance, experimental results showed that both methodologies incorporating secondary information led to better results than ordinary kriging that did not incorporate any sulphide factor data. The ordinary kriging method uses only cupper assay information‚ while in collocated cokriging and kriging with an external drift, both cupper assay and sulphide factor information are used as an auxiliary variable. The validation sample was used to compare the performance of the methods. Collocated cokriging and kriging methods with an external drift based on mean absolute error (MAE), root mean square error (RMSE) and the correlation coefficient of real and estimated values (R) illustrate better results.