Speculations on an information theoretic approach to spatial representation

Abstract

One of the themes which emerges from the theories and techniques of spatial representation presented so far in this book involves the important relationship between the variation in the phenomenon characterising the system of interest, and the way that phenomenon is measured. In essence, the problem of designing an appropriate spatial system by partitioning a space into zones is the problem of constructing an appropriate measure which detects the phenomenon in some meaningful way. It is therefore clear that a zoning system can never be studied in isolation from the spatial phenomena it seeks to measure, for changes to the system lead to changes in the variation of the phenomena, which in turn have important repercussions on the zoning system itself. In a sense, it is this interrelationship between zoning system and phenomenon which makes the problem most challenging, for any theory of optimal zoning will certainly have to address this issue by reference to some external criteria. Indeed, for any kind of theoretical progress to be made in this area, some fairly strong assumptions about the form of the phenomenon or zoning system will have to be made so that each can be studied in relation to the other.