Forecast of Dynamic Change of Land Use Based on Cellular Automata

Land use change is a very complex process of evolution. On the basis of the principle of cellular automata, this article presents a kind of method that we can first mine state transition rule from historical map data, and then conduct forecast by virtue of Monte-Carlo method, achieving spatial dynamic forecast from map to map. We interpret TM remote sensing image in Ji'nan City in 2004 and 2006 to get present land use map for empirical research, and forecast land use map in 2012 and 2016, respectively. Studies show that this method of using spatial data to mine state transition rule, has advantages of simpleness, accuracy, strong real-time characteristic etc. in the simulation of dynamic change of land use, the results of which are roughly in line with the actual results, therefore, it can provide reference for land use planning