Uncertainty awareness in urban sprawl simulations: Lessons from a small US metropolitan region

Land use change modeling and simulation is a popular tool in land use planning and policy formulation. However, the outputs of land use change simulation are not always accompanied with information on uncertainty. The goal of this paper is to demonstrate the inherent uncertainty in sprawl simulation, which is attributable to error in the input parameters and to limitations in our understanding of land use systems. To reach this goal, the paper determines sprawl simulation accuracy and uncertainty for a small US metropolitan region as produced by the CLUE-S modeling framework. The model simulates sprawl location in the region accurately, but the certainty of sprawl location projections decreases with time. This uncertainty in the simulation suggests that modelers should report uncertainty with their output over all time horizons so that, on the one hand, land use planners and decision makers do not place too much confidence in any single sprawl simulation (which could lead to unwarranted and expensive urban growth management policies) and, on the other hand, do not place too little confidence in sprawl models (which could have severe socioeconomic and environmental consequences). Thus, reporting uncertainty with simulation output provides planners and decision makers with a platform for more informed land use policy.