Spatial adaptive filters are techniques for mathematically modelling changes in causal relationships among geographical variables through time. They modify original causal relationships by using negative feedback, applied to subsequent of data observations. This project will further develop the modelling technique and its applications to temporal data. The case study used will be changes in the locations of physicians among the 48 continguous states between 1963 and 1985. Spatial adaptive filtering techniques offer the promise of greatly improving our ability to analyze complex geographically distributed systems through time. Robust techniques of this kind would improve the effectiveness of geographic information systems.
