Simulating sprawl

Suburban sprawl, a relatively recent phenomenon, is among the most important urban policy issues facing contemporary cities. To date, a well-accepted rationale has not been settled on for explaining and managing the causes of sprawl. Our contention is that consideration of geography is essential-that geographical explanations offer much potential in informing the debate about sprawl. Similarly, spatial simulation could support sprawl-related research, offering what-if experimentation environments for exploring issues relating to the phenomenon. Sprawling cities may be considered as complex adaptive systems, and this warrants use of methodology that can accommodate the space-time dynamics of many interacting entities. Automata tools are well-suited to representation of such systems, but could be better formulated to capture the uniquely geographical traits of phenomena such as sprawl. By means of illustrating this point, the development of a model for simulating the geographic dynamics of suburban sprawl is discussed. The model is formulated using geographic automata and is used to develop three sprawl simulations. The implications of those applications are discussed in the context of exploring geographic explanations of sprawl formation and the potential for managing sprawl by geographic means.