An EasyCA model with few steady variables and clone stamp strategy for simulation of urban growth in metropolitan areas

It is critical to predict urban growth for understanding of urbanization process. Traditional urban CA mainly focused on discovery of conversion rules from a series of biophysical and social-economic factors. Many varying factors are generally treated as constant, and the simulation result is none of great reference value for decision-making. In order to eliminate these limitations, an Easy Cellular Automata (EasyCA) was proposed with integration of seed searching and clone stamp growth strategy, it aims to provide feasible scenario analysis for decision-making. Only four steady variables namely distance to current urban patches, difficulty coefficient of land conversion, density of urban land and topographic conditions were applied to estimate urban suitability. Policy intervention was further considered for adjustment of urban suitability, and patch growth templates were collected as clone stamps for updating the status of cells during iteration. This EasyCA was then tested with example area of Guangzhou City located in the Pearl River Delta of China. Accuracy indices of averaged fuzzy kappa (FuzzyK) and figure of merit (FOM) were calculated with the values of 0.8419 and 0.2874, respectively, and both of them were higher than those from cell growth-based CA, indicating that EasyCA is efficient in utility and reliability. Comparison analysis verified that few steady variables can obtain reliable result of urban suitability, and it is greatly necessary to integrate policy intervention into estimation of urban growth probability. And the simulation pattern in 2035 is maximally close to actual status compared with the planning layout, it shows that EasyCA is potentially applicable and effective for planning practices in metropolitan areas.