In mathematical statistics, a sampling variogram is a graph that shows where a significant degree of causality (in this context, spatial dependence in sample spaces or sampling units) dissipates into randomness. A sampling variogram is obtained by plotting statisticallly significant variance terms of a temporally or in situ ordered set of measured values against the variance of the set and the lower limits of its asymmetric 95% and 99% confidence ranges. Corrected sampling variograms derive from uncorrected ones when extraneous measurement variances are subtracted before spatial dependence is verified. Bre-X's bogus gold grades for crushed, salted and in situ ordered core samples of Borehole BSSE198 in Busang's South-East zone give the following uncorrected sampling variogram. --> The corrected sampling variogram for Bre-X's bonanza borehole, its primary data set and test statistics are posted on this page. A temporally ordered set of on-stream data for mill feed to a mineral processing plant, its test statistics and sampling variogram are given in Appendix D of Sampling in Mineral Processing. A variogram or a semi-variogram, unlike the above sampling variogram, does not show where spatial dependence in sample spaces dissipates into randomness because kriging variances of sets of kriged estimates are invalid measures for variability, precision and risk.
|