URBAN GROWTH MODELING USING GENETIC ALGORITHMS AND CELLULAR AUTOMATA; A CASE STUDY OF ISFAHAN METROPOLITAN AREA, IRAN

This study integrates cellular automata (CA) and genetic algorithms (GAs) to model urban growth in the Isfahan Metropolitan Area in Iran. The simulation of urban growth through cellular automata models brings improved understanding of the complex dynamic process of land use change, which can not be achieved through conventional models. The cellular automata (CA) as a powerful spatial dynamic modeling tool are designed as a function of parameters whose calibration plays a crucial role for obtaining a suitable set of parameters in order for precise and reliable modeling. Genetic Algorithms are useful tools for decreasing the search space for finding the optimal solution of transition rules in cellular automata and reducing the simulation uncertainties and improving its locational accuracy in urban modeling. The considered objective function in this algorithm is percent correct match (PCM) obtained from error matrix between the simulated and the reference map. Historical land use/cover data of Isfahan Metropolitan Area were extracted from the 1990 and 2001 Landsat ETM+ images at 30m spatial resolution. Three different Moore neighborhood sizes have been considered for cellular automata model and simulation of urban growth for the year 2001 is performed. The simulation outcomes, evaluated with kappa statistic of 74.15% demonstrate that the integration of GA and CA could be suitable for dynamic urban growth modeling.