Simulation of peri-urban growth dynamics using weights of evidence approach

The study aims to simulate the peri-urban growth dynamics in a growing region of India using Weights of Evidence (WOE) based cellular automata model. The growth process was expressed as a function of four causative variables corresponding to which seven data layers were generated in a Geographic Information Systems environment. The model was calibrated for the period 2000-2005 using Kappa indices and fuzzy set theory based two way comparison method. The Kappa value was 0.7, while the value of K-location and K-histo were 0.81 and 0.93, respectively. The fuzzy similarity values increased for small to large neighbourhood sizes which showed that the model was able to simulate the contiguous and dense growth. However, for dispersed and isolated growth the model showed less accuracy. The model was validated for period 2005-2010 and revealed a Kappa value of 0.88, while value of K-location and K-histo were 0.91 and 0.96, respectively.