Standardized pairwise relative variogram as a robust estimator of spatial structure / Стандартизированная парная относительная вариограмма как надежный инструмент оценки пространственной структуры

A requirement in geostatistical modeling is to find an appropriate and stable variogram. The traditional variogram is sensitive to outliers and sparse data. A number of robust alternatives have been proposed. Although the correlogram, indicator variogram and normal scores variograms sometimes work, the pairwise relative variogram is a remarkably robust estimator of the true underlying spatial structure. The variogram is adjusted by the squared mean of each data pair. A challenge in the interpretation of the pairwise relative variogram is that the sill of the pairwise relative variogram has not been well documented. The goal of this paper is to understand the sill of the pairwise relative variogram in order to standardize it for improved interpretation. <...>