基于MCE-CA-Markov的三峡库区土地利用演变模拟及预测  

Simulation and prediction of land use change in Three Gorges Reservoir Area based on MCE-CA-Markov

作　　者：张晓娟 周启刚 王兆林[2] 王福海[3] 

机构地区：[1]重庆工商大学环境与资源学院,重庆400067 [2]重庆工商大学旅游与国土资源学院,重庆400067 [3]重庆工商大学融智学院,重庆400055

出　　处：《农业工程学报》2017年第19期268-277,共10页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家自然科学基金项目(41101503);重庆市教委科技项目(KJ1601901);重庆工商大学融智学院科研项目(20158001);重庆工商大学研究生创新型科研项目(yjscxx2017-066-63)

摘　　要：为深入研究三峡库区蓄水以来土地利用演变规律及驱动机制,该研究以库区2000、2007及2014年的Landsat TM遥感影像作为数据源,在运用ARCGIS10.1对库区土地利用现状进行分析的基础上;基于多标准评价(multi-criteria evaluation,MCE)的元胞自动机和马尔可夫链(CA-Markov)模拟库区自组织的土地利用演变;并利用多时相的遥感分类图像检验MCE-CA-Markov的模拟精度,经检验,其Kappa系数达到0.971 2;利用检验过的土地利用演变限制条件及因子组合来对库区2028年土地利用进行预测。研究结果表明:运用MCE-CA-Markov模拟土地利用演变规律是可行的;在2000—2014年间,库区土地利用变化图谱面积由大到小排序为稳定变化型>后期变化型>前期变化型>反复变化型>持续变化型;模拟得到库区2014年土地利用格局与解译的土地利用格局一致,相对精度达到84.48%;预测到2028年时,库区耕地、林地及草地继续减少,水域、建设用地及未利用地持续增加,预测结果表明三峡库区土地利用变化十分活跃,生态环境恢复和重建的压力很大。因此,迫切需要加强三峡库区耕地、林地和草地的保护,控制和阻止盲目开垦耕地的行为,促进库区生态环境的保护和改善。研究结果以期为三峡库区土地利用规划、城市建设以及生态环境修复提供理论支撑。In order to achieve better coordination between land use and ecological environment in the Three Gorges Reservoir Area, this paper explored the pattern and mechanism of the evolution of land use since the reservoir was built and impounded. Based on Landsat TM image data of the Three Gorges Reservoir Area in 2000, 2007 and 2014, the first step of the research was to obtain basic information of present land use through ArcGIS 10.1, then generate land use change map to analyze land use changes in 2000-2014, and calculate land use transfer matrix to analyze the flow characteristics of land use in the Three Gorges Reservoir Area in 2000-2007 and 2007-2014, respectively. Second, multi-criteria evaluation （MCE） - cellular automata Markov chain （CA-Markov） model was used to simulate self-organizing evolution of land use in the reservoir area. In particular, the MCE model was used to make maps of land use suitability and the CA-Markov model was used to calculate the land use transfer matrix. The accuracy of MCE-CA-Markov simulation was tested against multi-temporal remote sensing images and the Kappa coefficient was found to be 0.971 2. Third, land use structure in 2028 was forecasted with the tested constraints and the combinations of factors of land use evolution. Overall, findings included： 1） It was feasible to simulate land use evolution with MCE-CA-Markov. 2） During 2000-2014, types of land use change areas, sorted from large to small, were stable change 〉 late change 〉 early change 〉 repeated changes 〉 continuous change. Especially, areas of cropland and forestland were mainly changed to water area and construction land, while the changes of the areas of grassland and unused land were not very obvious and scattered in space due to their smaller areas. 3） The land use pattern in 2014 from the simulation was consistent with that from interpreted remote sensing images and the relative accuracy reached 84.48% after verification. 4） Our model predicted that, up to 2028, cropland and forestl

关 键 词：土地利用 演变 模拟 MCE-CA-Markov 三峡库区 

分 类 号：F301.2[经济管理—产业经济]