Discovering and evaluating urban signatures for simulating compact development using cellular automata

This paper provides anew method for retrieving, evaluating and modifying urban signatures for simulating compact development using urban cellular automata (CA). Urban CA usually adopt fixed transition rules for simulating urban dynamics in large complex regions. However, these regions can be segmented into sub-regions so that separate transition rules can be retrieved for generating better simulation results. Moreover, urban signatures or genes can be extracted from GIS data to assist the understanding of urban evolution for each sub-region. Good genes from a sub-region can be cloned to other sub-regions for producing better urban forms. A heuristic swapping technique is developed to modify existing genes so that compact development patterns can be generated for planning purposes. The proposed method has been applied to the simulation of compact development in the Pearl River Delta. The analysis indicates that this model can help improve the compactness of urban development in this fast growing region. (C) 2008 Elsevier B.V. All rights reserved.