Implementation of enhanced areal interpolation using MapInfo

The Enhanced Areal Interpolation method introduced by Flowerdew, Green, and Kehris, which is essentially an application of the EM Algorithm, postulates a more sophisticated statistical model and makes use of ancillary information available on the target regions.

Abstract

In the analysis of spatial data there may be a need to transfer data collected originally on one set of areal units (source regions) to a different set of areal units (target regions). The MapInfo Desktop Mapping Package currently enables the user to carry out this process by simple areal weighting. The Enhanced Areal Interpolation method introduced by Flowerdew, Green, and Kehris, which is essentially an application of the EM Algorithm, postulates a more sophisticated statistical model and makes use of ancillary information available on the target regions. It is an iterative method and uses the simple areal weighting estimates as its starting point. In this paper we shall describe the implementation of this method for count data in MapInfo, thus making it readily accessible to the general user.