New distance measures: The route toward truly non-Gaussian geostatistics

Abstract

The projection or minimum error norm algorithm does not require that the distance measure be a variogram. In non-Gaussian cases, the traditional variogram distance measure leading to minimization of an error variance offers no definite advantage. Other distance measures, more outlierresistant than the variogram, are proposed which fulfill the condition of the projection theorem. The resulting minimum error norms provide the same data configurations ranking as traditionally obtained from kriging variances. A case study based on actual digital terrain data is presented.