基于CLUE_S的精河流域绿洲LUCC变化模拟

发布时间：2019-01-25 10:43

【摘要】：LUCC是人类活动与自然环境相互作用的直接反应，是全球环境变化的重要组成部分，也是研究全球环境变化的核心内容。特别是地处干旱半干旱地区的绿洲，具有高度的封闭性，分布具有高度的分散性，生态承载力具有高度的脆弱敏感性。以地处全国最大的半固定沙漠西缘，同时也是新疆最大的咸水湖流域中心的精河流域绿洲为研究靶区，运用3S技术对研究区近40年来LUCC规律做了详细分析，并引入CLUE_S模型对该区未来LUCC做出预测。对生态环境脆弱区的土地利用可持续发展具有重要的实践意义。研究结论如下：过去的40年间，精河流域绿洲土地利用格局发生了显著的变化。耕地和建设用地面积持续增大，耕地由105.15km~2增至398.45km~2，建设用地从10.07km~2增至33.43km~2；水域和中植被覆盖区面积持续减少，水域由622.6km~2降至463.26km~2，中覆盖度植被由347.95km~2降至56.57km~2；高覆盖度植被区和未利用土地面积基本保持稳定。土地利用面积比例最大的三类是未利用土地、水域和低覆盖度植被，三种地类比例，从1972年的24.21∶34.01∶20.73，转换到2010年的18.01∶33.64∶26.39。低覆盖度植被区面积波动增加，水域面积持续减少，未利用土地面积总体减少大，但是减少幅度非常小。 LUCC格局变化幅度总体前期大于后期。1972-1990年期间与1990-2010年期间相比较，就具体地类而言：前期建设用地变化幅度最大104.22%，后期耕地变化幅度157.465%远高于其他地类，植被覆盖区前期变化幅度也大于后期，最为明显的是中覆盖度植被区，变化幅度为64.02%。近40年来土地利用变化幅度最大的地类是耕地和建设用地，变化幅度为278.94%和222.11%。人文因素是导致变化的主要影响子，其中人口增长是主要原因。运用CLUE_S模型对土地利用格局模拟的核心是各个空间和非空间驱动力与各地类之间的回归分析，，分析的结果是模型参数文件设置的必要条件。选取对土地利用格局影响显著的道路、渠系、居民点、距离湖面距离、铁路等因素对精河流域绿洲2010-2025年土地利用格局模拟，ROC检验结果较理想：耕地0.881、水体0.757、建设用地0.895、低覆盖度植被0.766、中覆盖度植被0.833、高覆盖度植被0.744、未利用土地0.756，拟合度均大于0.7,可以做为LUCC模拟的研究模型。以1990年为基期，输入驱动因子模拟2010年的土地利用格局，将结果与2010年的土地利用现状做比较，模拟正确的栅格数量正确率为85.3%，Kappa指数为0.832，说明CLUE_S模型可以较好的模拟精河流域绿洲土地利用格局。对研究区2025年LUCC情景设计2个假设需求方案，分别将相应的限制区文件和需求方案输入到模型中。结果显示参考多方面的因素，设计合理的需求方案并给予一定的限制条件约束下的模拟结果更理想，因此，方案1的模拟结果更具参考价值。
[Abstract]:LUCC is a direct reaction of the interaction between human activities and natural environment, is an important part of global environmental change, and is also the core content of studying global environmental change. The oasis, which is located in arid and semi-arid area, has a high degree of closure, a high degree of dispersion and a high sensitivity of ecological carrying capacity. Taking the oasis of the Jinghe Basin, which is located in the western edge of the largest semi-fixed desert in China and also the center of the largest saline lake basin in Xinjiang, as the research target, the LUCC law of the research area in the past 40 years has been analyzed in detail by using 3s technology. CLUE_S model is introduced to predict the future LUCC in this area. It has important practical significance for the sustainable development of land use in the fragile ecological environment. The results are as follows: in the past 40 years, the land use pattern of the Jinghe Oasis has changed significantly. The area of cultivated land and construction land continued to increase, from 105.15km~2 to 398.45km2, and from 10.07km~2 to 33.43kmm2; The area of water area and middle vegetation cover area decreased continuously, the water area decreased from 622.6km~2 to 463.26 km2, the middle coverage vegetation decreased from 347.95km~2 to 56.57 km2, and the area of high coverage vegetation and unused land remained stable. The largest proportion of land use area is unused land, water area and low coverage vegetation. The proportion of three land types has changed from 24.21: 34.01: 20.73 in 1972 to 18.01: 33.64: 26.39 in 2010. The area fluctuation of low coverage vegetation area increases, the area of water area decreases continuously, and the area of unused land decreases greatly, but the range of decrease is very small. Compared with the period of 1990-2010, the range of change of LUCC pattern in the early stage is larger than that in the later period. In terms of specific land types, the range of change of construction land in the early stage is 104.22, and the difference between 1972-1990 and 1990-2010 is significant. The change range of cultivated land in late stage was 157.465% higher than that of other land types, and the range of change in the early stage of vegetation cover area was larger than that in later period, the most obvious change was in middle coverage vegetation area (64.02%). In the last 40 years, the largest changes in land use are cultivated land and construction land, with a range of 278.94% and 222.1111%. Human factors are the main factors leading to change, among which population growth is the main reason. The core of land use pattern simulation using CLUE_S model is regression analysis between spatial and non-spatial driving forces and local classes. The result of the analysis is a necessary condition for setting up model parameter files. Selecting the road, canal system, residential area, distance from lake to lake, railway and so on, which have significant influence on land use pattern, are selected to simulate the land use pattern of Jinhe Oasis from 2010 to 2025. The results of ROC test are satisfactory: cultivated land 0.881, water body 0.757, etc. The construction land was 0.895, the low cover was 0.766, the middle cover was 0.833, the high cover was 0.744, the unused land was 0.756, and the fitting degree was more than 0.7, which could be used as the model of LUCC simulation. Taking 1990 as the base period, the input driving factor was used to simulate the land use pattern in 2010, and the results were compared with the present land use situation in 2010. The correct rate of simulating the correct number of grids was 85.3% and the Kappa index was 0.832. The results show that CLUE_S model can simulate the land use pattern of Qinghe River Basin. Two hypothetical requirement schemes were designed for the LUCC scenario in 2025, and the corresponding restricted area files and requirement schemes were input into the model respectively. The results show that it is better to design a reasonable demand scheme and give some constraints to the simulation results, so the simulation results of scheme 1 are more valuable.
【学位授予单位】：新疆师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F301;P208

【参考文献】