A Generalized Model for Cellular Urban Dynamics

Cities have long been recognized as complex and dynamic entities. However, the mathematical description of cities has been traditionally based on aggregate observations and on their static properties. Although some new approaches derived from the science of complex systems, particularly Cellular Automata (CA), have been explored in recent years, many urban researchers think that current forms of these models are cursory, fragmentary, and unrealistic in terms of theoretical and practical solidity. This paper generalizes an urban model to deal with questions of dynamic urban evolutionary modeling (DUEM). The DUEM situates its conceptual and technical foundations on Couclelis's CA spaces, Dendrinos' urban selection, and contemporary techniques of GIS. The DUEM provides a generic modeling technique that is capable of capturing the processes of evolutionary reproduction at the finest level and handling interactions between automaton self-reproduction and environmental and socioeconomic influences through the feedback among neighborhood, field, and region. Here we first review developments of dynamic systems theories and CA in particular. We then discuss emerging issues of CA applications in urban fields, and lay out conceptual foundations for this research. Next, we propose a generic urban CA model—DUEM for describing urban complexities and dynamics. We finally show how this generic model can be used to simulate growth dynamics for multi-sector land uses in a suburban area of Buffalo, New York.