AI-GEOSTATS: Hall of Fame or Shame?

[JW]

Dear Dr van den Boogaart,   I do not blame you for responding the way you did because I like to ask questions that leave plenty of room for argument and debate. Most of all, I would like your response to the following proposition:   Each functionally dependent value has its own variance – true or false?   In my opinion, the trueness of this proposition would have far-reaching consequences for geostatistics as a stand-alone variant of mathematical statistics.   I was facetious when I talked about your belated analysis of Bre-X. My point is that Fisher’s F-test could have proved that a salting scam was in progress at Bre-X’s Busang property on the basis of original and duplicate bogus gold assays for the first three to five salted boreholes. Aggressive step-out drilling converted massive volumes of Busang’s barren rock into phantom ore. Journel’s doctrine of assuming spatial dependence without proof also contains his cryptic “(a decision rather)” note. So who decided that spatial dependence may be assumed? When? Why?   I certainly would not want you to apply the KWBLUP method to Bre-X data if I were to become complicit in a fraud! My son and I were more surprised than angry in the early 1990s that reviewers objected when we applied Fisher’s F-test to verify spatial dependence between gold grades for an ordered set of rounds in a drift. The more so because we were not even aware we had compiled a geostatistical paper.   Journel and Huijbregts’ 1978 Mining Geostatistics points out that the Cauchy algorithm is only of interest if the function K(u) is known a priori whereas Goovaerts’ 1997 Geostatistics for Natural Resource Evaluation does not refer to Cauchy anymore.   A complete chronology of events behind my complaints about IAMG’s brass and JMG’s brains has yet to be published.  The concepts of functional and spatial dependence and independence are of particular interest in my work. Please look at Fig 203 on page 286 of  David’s 1977 Geostatistical Ore Reserve Estimation and peruse the author’s test for geostatistical perspicacity. Do you see anything wrong in this figure or on this page? In case of doubt, read my retro-review of the first textbook on geostatistics   Kind regards, Jan W Merks   PS: In my opinion, it’s bound to become a hall of shame!