基于CLUE-S模型的南昌市土地利用变化模拟研究

发布时间：2018-12-10 08:21

【摘要】：土地利用／土地覆被变化(LUCC)作为当今研究全球环境变化的核心和热点之一,受到区域自然资源约束和经济社会发展驱动,具有自然和人文双重属性,反映了人类与自然界相互作用、相互影响的复杂过程。LUCC模型可定性、定量地反映了区域土地利用／覆被变化成因和变化过程,CLUE-S模型是LUCC模型中成熟且应用广泛的一种,是研究区域土地利用／覆被变化的重要分析工具。本文选取南昌市为研究区域,以2005年和2011年土地利用现状图为数据源,运用GIS和数理统计技术,并设置了四种情景模式模拟2020年南昌市土地利用格局。主要结论有：(1)基于GM(1,1)改进的Markov模型可以对南昌市的未来土地利用需求量进行预测,以Markov模型的状态特性,避免了经济、自然等因素的干扰,使得南昌市土地需求的预测结果更为准确可靠。(2) CLUE-S模型可以较好地模拟南昌市土地利用／覆被变化特征。以2005年土地利用数据为基期数据,选取了13个自然和社会经济因素作为驱动因素,应用Logistic回归模型确定南昌市土地利用变化模拟的最佳尺度规模：100*100m,以及各种影响因素的概率分布,模拟了2011年的土地利用类型空间格局,并用2011年的土地利用现状图加以验证,验证结果表明：正确模拟的比例达到85.37%,由此计算出的Kappa指数为0.83,说明CLUE-S模型能够有效模拟南昌市土地利用／土地覆被变化。(3)对南昌市历史土地利用情况及利用趋势分析,设置了自然演替情景、耕地保护情景、生态安全保障情景、规划指标控制情景四种情景方案,用于模拟2020年南昌市的土地利用格局。模拟结果表明在不同情景方案下的土地利用格局存在显著的空间差异。南昌市的发展均表现为“沿江发展,沿国道发展,向昌南及昌北扩张”的趋势。差异在于,各发展模式中,地类面积及转化方向的不同。
[Abstract]:Land use / land cover change (LUCC), as one of the core and hotspot of global environmental change, is constrained by regional natural resources and driven by economic and social development. LUCC model can qualitatively and quantitatively reflect the cause and process of regional land use / cover change. CLUE-S model is a mature and widely used LUCC model. It is an important analytical tool to study regional land use / cover change. In this paper, Nanchang City is selected as the research area, land use status maps in 2005 and 2011 are taken as data sources, GIS and mathematical statistics techniques are used, and four scenarios are set up to simulate the land use pattern in Nanchang City in 2020. The main conclusions are as follows: (1) based on the improved Markov model of GM (1t1), the future land use demand of Nanchang city can be predicted, and the state characteristics of Markov model can be used to avoid the interference of economic and natural factors. It makes the prediction results of land demand in Nanchang more accurate and reliable. (2) CLUE-S model can well simulate the characteristics of land use / cover change in Nanchang city. Based on the data of land use in 2005, 13 natural and social economic factors were selected as driving factors, and the optimal scale of land use change simulation in Nanchang was determined by using Logistic regression model: 100m. As well as the probability distribution of various factors, the spatial pattern of land use types in 2011 is simulated and verified by the current land use map of 2011. The results show that the proportion of correct simulation reaches 85.37%. The calculated Kappa index is 0.83, which indicates that the CLUE-S model can effectively simulate the land use / land cover change in Nanchang. (3) based on the analysis of the historical land use and the trend of land use in Nanchang, a natural succession scenario is set up. Cultivated land protection scenarios, ecological security scenarios and planning index control scenarios were used to simulate the land use pattern in Nanchang in 2020. The simulation results show that there are significant spatial differences in land use patterns under different scenarios. The development of Nanchang city shows the trend of "developing along the river, developing along the national highway and expanding to the south and north of Changchang". The difference lies in the difference of ground area and transformation direction in each development mode.
【学位授予单位】：江西农业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F301.2;F224

【相似文献】