As one of natural resource systems, land system has the closest man-land relationship, which is changed dynamically by interleaving interactions between man and land. As one of the most strongly influenced environments by human activity, mining area environment differ from all others, especially in land use change, which not only has uniqueness in characteristic, process and mechanism of dynamic change, but also interleaves interactions between man-land relationships.Beginning with the developments and applications of Geographical Simulation Systems (GSS), theories of combining descriptive and quantitative (represented by Geo-Information science in modern geography) methods were introduced, then basic concepts of Cellular Automata (CA), which is regarded as the new generation of Geographical Simulation Systems (GSS), and highlighted the similar aspects between Cellular Automata (CA) and Geographical Grid Systems to outstand its applicability in geography.Based on the theories, fundamental driving forces depend on national climax driving forces, and effect according to direct forces, which are put forward by Liu Jiyuan on the basis of studies on spatial - temporal dynamic changes of land use and driving forces analyses of China in the 1990s', data of equal-time interval was picked up from existing data, and reclassified according to actual situation.Changing scale, speed, extent quantified by mathematical statistical methods as well as characteristics of spatial-temporal pattern and spatial ring structure of land use were analyzed by geo-information science technology based on Geo-information Photographical Methodology from 1987 to 2007 in Nantun coal mining area separately to summarize process, structure, dynamic and mechanism of land use in study area.Using geo-statistical/statistical model and determined rules by spatial analysis, land use change was simulated by Cellular Automata (CA) in Nantun coal mining area. The result of that in 2007 was verified with existing data, and then analyzed the trend of change from 2007 to 2017 according to the prediction of 2017.Combing quantitative description from static mathematical statistical methods and spatial-temporal dynamic change from geo-information science technology, process of land use change was analyzed to summarize general trend of spatial-temporal expansion and spatial structure change. Last but no means the least, predicting was made to analyze the trend of future.During the study period, infrastructural land increased rapidly, especially in stage of expanded reproduction from 1997 to 2007, in which production and construction land rose in all rings and compelled to move to edge becoming new residential area, whose curve converted from single-peak to double-peak; Trend of extremely unstable in land use change was caused by mining collapse aggravating, which has enlarged 3.35 times during that 20 years, and tressed all others far away from center; Gratifying achievements in ecological reconstruction has success initially, manifested morphologically as: trend of decreasing in mining collapse, stabilizing in land use of all rings and rationalizing and equalizing in structure of land use.Based on the description of Geographical Simulation Systems (GSS) and Cellular Automata (CA), model in this paper was described detailedly. Situation of land use in 2007, 2017 was simulated by Markov and Cellular Automata (CA) simultaneously, and the result in 2007 was used to validate the simulations of the two. The validation showed that the accuracy of Markov was 84.9292% while that of Cellular Automata (CA) was 81.5357%, and the Kappa coefficient was 0.6609. Both of these two methods simulated land use change preferably, however, the accuracy of the latter was better than that of the former.This paper provided early-warning for production and life in study area and basis of spatial determination. Moreover, the prediction supplied the programming of economic development and ecological reconstruction. |