AbstractThe focal point of this paper is the concept:the environmental unit is Earth itself. To analyze and predict global environmental phenomena, the dimensions of our domain of concern must match those of the planet's dynamic fluid envelope: atmospheric and oceanic. Traditional techniques of statistical diagnostic analysis and prediction must be cast in the natural coordinates of earth‐oriented systems and computer technology tasked to its present limits to handle global scale data bases.A notation is presented in which various approaches to estimating variables in large domains can be seen as adaptations to a common estimation objective. Their differences arise from differences in the scope and extent of information available to describe the system under study. The “universal estimator notation” serves to identify areas for significant further development. Examples are given which illustrate the potential impacts on spatial estimation algorithms of using available data to tune them to location and current