The Influence of the Calibration Interval on Simulating Non-Stationary Urban Growth Dynamic Using CA-Markov Model

The temporal non-stationarity of land use and cover change (LUCC) processes is one the main sources uncertainty that may influence calibration validation spatial path-dependent LUCC models. In relation to that, this research aims investigate on urban growth modeling accuracy based an empirical approach uses past LUCC. Accordingly, development in Rennes Metropolitan (France) was simulated using fifteen intervals which are set from six training dates. study used Idrisi’s Cellular Automata-Markov model (CA-Markov) inductive pattern-based software package. demand for simulation year estimated Markov Chain method. Model carried out by assessing quantity change, allocation, patterns accuracy. disagreement analyzed taking into consideration rate over prediction intervals, ability reproduce amount future, time duration interval. results show interval significantly influenced allocation change. addition CA-Markov depended highly basis image rather than observed transition during period. Therefore, provides useful insights role dates non-stationary