ABSTRACT
This chapter suggests that geostatistics, and in particular variogram analysis and kriging, are suitable tools for Exploratory Spatial Data Analysis and could provide some of the analytical needs of geographic information systems (GIS). Interpolation is an important feature of GIS because geographically referenced data are often fragmentary. M. A. Oliver and R. Webster suggested that GIS could be improved by incorporating kriging as the principal method of interpolation. The chapter describes two case studies; binomial kriging of a rare disease and disjunctive kriging for environmental management. Disjunctive kriging should provide environmental scientists with a useful decision-making tool, especially where there are sensitive issues that could result in litigation or damage to health. In Gaussian disjunctive kriging, the most common form for analysing environmental properties, the data are transformed to a standard normal distribution using a linear combination of Hermite polynomials.
