- © 2000 Geological Society of South Africa
Sixteen articles were carried in the Northern Miner in Canada over nine months in 1998—99, supporting or refuting the methodology of geostatistics in one aspect or another. The appearance of such articles has slowed, but is certain to begin again, because the discussion has never evolved into a high enough level of dialogue. The crux of the problem appears to be the various good and bad experiences concerning reliability.
The erratic results of geostatistics as we understand them, with their supporters and detractors, strongly suggest to some of us that industry standard split drill core samples are too small to be reliable. There can be no doubt that adequate sample weight should be established and used, as detailed by Tilsley in the PDAC Workshop on Due Diligence in 1995. Moreover, an adequate sample must be customized within each deposit. The inadequate sample is a function of core size selection, sample length selection, core splitting, pulverization, splitting again, weight of aliquot, etc. These sample size reductions can yield.the ‘conditional bias’ (Krige, 1999) and the ‘nugget effect’ in general parlance.
Consider, were we …