Integration of cellular automata and geographic information systems for modeling spatial dynamics

The primary goal of this research was to develop a prototype system that fully integrates CA simulators and GIS in order to improve the spatial dynamic modeling capability of GIS and provide an effective spatial dynamic modeling environment. An integrated CA-GIS system in which CA simulators (Cellular and CAM-6) are tightly coupled with GIS (IDRISI) and serve as the analytical engines for GIS was designed. In order to implement this tight coupling strategy, linkage modules (data model translators and an operation handler) and a common user interface were developed in this research.;Using this CA-GIS integrated system, this research was able to implement and develop both existing GIS operations/modeling functions and new GIS operations. Many existing GIS operations such as local, focal, and incremental operations have been successfully implemented within the integrated system. Some GIS operations implemented using the CA simulators showed better performance than the original GIS ones. More importantly, new types of spatial operations such as various shape of spatial filters local time-series, focal time-series filter, gas diffusion operators, and the diffusion-limited aggregation (DLA) operator that are not available in contemporary GIS were developed and implemented using the integrated CA-GIS system. Most significantly the demonstration of three spatial dynamic modeling examples (the Hagerstrand's innovation diffusion model, a forest fire model, and an urban growth simulation model) has shown that complex spatial dynamic models can be successfully implemented within this CA-GIS environment, providing increased ability of dynamic modeling in GIS.