Progressive Kriging - A Method Of Testing The Geostatistical Hypotheses ? Introduction

Any geostatistical study involves (implicit) hypotheses about 1. The distribution function 2. The drift of the data 3. The (underlying) variogram Once these are known, most other results follow from geostatistical theory. However, to date, there have been few theoretical studies on testing these implicit hypotheses. Most of the published studies in this area have dealt with the fluctuation of the variogram or the sampling distribution of the variogram (Journel and Huijbregts, 1978; Davis and Borgman, 1979). Questions regarding these hypotheses regarding a given deposit arise often. Coal geologists often infer channels and washouts on the basis of thicknesses data. Zoning is recognized in a variety of metal deposits. A test as to whether these inferences are valid on just chance occurrences would have widespread application. .It is standard geostatistical practice to "eyeball" fit variograms. In the early exploration phase when just a few pairs are available, the variogram often displays wide fluctuations and there may be considerable doubt about which variogram model to use.