基于CLUE-S模型的潜江市土地利用时空格局模拟
发布时间：2019-06-13 14:02
【摘要】：近年来,学术界对土地利用／土地覆被研究的关注程度越来越高。在众多土地利用变化模型中,基于过往土地利用研究经验、未来年份土地利用需求数量的CLUE-S模型适用于空间尺度较小的土地利用变化模拟。 本文以潜江市为研究对象,以潜江市2005年、2010年土地利用现状图为基础,在综合了影响潜江市土地利用变化自然、社会驱动因子的基础上,结合灰色预测GM(1,1)模型土地需求预测结果以及CLUE-S模型的相关参数设置,利用CLUE-S模型对潜江市2010年、2020年两期土地利用空间分布状况进行模拟预测。研究有利于揭示潜江市土地利用变化规律,为潜江市优化区域土地资源配置提供参考,同时,今后在江汉平原地区开展类似研究具有一定的参考价值和借鉴意义。本研究的主要成果如下： (1)本研究通过对不同模拟尺度栅格数据的Logistic回归分析结果进行Roc对比检验,以得到的Roc值大小为判断准则来选择适合模拟尺度。通过对比不同尺度的Roc值可以发现,耕地、交通水利用地、城乡建设用地、水域、其他农用地这些潜江市主要用地类型的Roc值在尺度200*200下达到最大值,故CLUE-S模型适用于栅格尺度200*200的县级尺度潜江市土地利用空间格局模拟。Logistic回归分析结果还表明,在不同时间段内,影响潜江市不同地类空间格局变化的主要驱动因子不完全相同。 (2)利用CLUE-S模型,模拟预测了潜江市2010年、2020年两期土地利用空间格局。模拟结果显示,潜江市2010年土地利用空间模拟正确土地面积为175977公顷,即正确模拟栅格个数为43994个,占栅格总数的88.26%；Kappa指数的计算结果为0.7617；模拟结果表明,CLUE-S模型仿真模拟结果可信度强,适用于模拟县级尺度的潜江市土地利用空间格局变化。潜江市2020年土地利用空间模拟正确土地面积为190729公顷,即正确模拟栅格个数为47682个,占栅格总数的95.66%,Kappa指数的计算结果为0.9146。研究显示,CLUE-S模型对于研究区2020年土地利用空间格局的模拟与其2020年土地利用总体规划高度一致。 (3)2010年到2020年,各地类的变化存在较大差异。耕地、林地、其他农用地面积呈现减少趋势,城乡建设用地增长较快,水域面积略有增加,交通水利用地保持不变。各地类的变化幅度从大到小依次为城乡建设用地林地耕地其他农用地水域交通水利用地。其中,耕地因基数面积较大,故变化幅度较林地略小
[Abstract]:In recent years, academic circles pay more and more attention to land use / land cover research. Among many land use change models, based on the past land use research experience, the CLUE-S model of land use demand quantity in the future year is suitable for the simulation of land use change with small spatial scale. Taking Qianjiang City as the research object, based on the present land use map of Qianjiang City in 2005 and 2010, on the basis of synthesizing the natural and social driving factors of land use change in Qianjiang City, combined with the prediction results of land demand of grey prediction GM (1, 1) model and the setting of relevant parameters of CLUE-S model, the spatial distribution of land use in Qianjiang City in 2010 and 2020 is simulated and predicted by using CLUE-S model. The study is helpful to reveal the law of land use change in Qianjiang City and provide a reference for Qianjiang City to optimize the allocation of regional land resources. at the same time, similar research in Jianghan Plain will have certain reference value and significance in the future. The main results of this study are as follows: (1) the results of Logistic regression analysis of raster data with different simulation scales are compared and tested by Roc, and the suitable simulation scale is selected by using the obtained Rocs value as the criterion. By comparing the Roc values of cultivated land, traffic water conservancy land, urban and rural construction land, water area and other agricultural land, it can be found that the CLUE-S model is suitable for simulating the spatial pattern of land use in Qianjiang city with raster scale 200. Logistic regression analysis also shows that in different time periods, the CLUE-S model is suitable for simulating the spatial pattern of land use in Qianjiang city with raster scale 2000.Logistic regression analysis also shows that the CLUE-S model can be used to simulate the spatial pattern of land use in Qianjiang city with raster scale 2000.Logistic regression analysis also shows that the CLUE-S model is suitable for land use spatial pattern simulation at county level. The main driving factors affecting the change of spatial pattern of different lands in Qianjiang City are not exactly the same. (2) using CLUE-S model, the spatial pattern of land use in Qianjiang City in 2010 and 2020 is simulated and predicted. The simulation results show that the correct land area of land use spatial simulation in Qianjiang City in 2010 is 175977 hectares, that is, the number of correctly simulated grids is 43994, accounting for 88.26% of the total grid, and the calculated results are 0.7617. the simulation results show that the simulation results of CLUE-S model are credible and suitable for simulating the spatial pattern change of land use in Qianjiang City on a county scale. The correct land area of land use spatial simulation in Qianjiang City in 2020 is 190729 hectares, that is, the number of correctly simulated grid is 47682, accounting for 95.66% of the total grid, and the calculated result of Kappa index is 0.9146. The results show that the simulation of land use spatial pattern in 2020 by CLUE-S model is highly consistent with its overall land use planning in 2020. (3) from 2010 to 2020, there are great differences in the changes of different categories. The area of cultivated land, forest land and other agricultural land showed a decreasing trend, the urban and rural construction land increased rapidly, the water area increased slightly, and the traffic water use land remained unchanged. The variation range from large to small is urban and rural construction land, woodland, cultivated land, other agricultural land, water conservancy land. Among them, due to the large base area of cultivated land, the change range of cultivated land is slightly smaller than that of forest land.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F301.2;F224
[Abstract]:In recent years, academic circles pay more and more attention to land use / land cover research. Among many land use change models, based on the past land use research experience, the CLUE-S model of land use demand quantity in the future year is suitable for the simulation of land use change with small spatial scale. Taking Qianjiang City as the research object, based on the present land use map of Qianjiang City in 2005 and 2010, on the basis of synthesizing the natural and social driving factors of land use change in Qianjiang City, combined with the prediction results of land demand of grey prediction GM (1, 1) model and the setting of relevant parameters of CLUE-S model, the spatial distribution of land use in Qianjiang City in 2010 and 2020 is simulated and predicted by using CLUE-S model. The study is helpful to reveal the law of land use change in Qianjiang City and provide a reference for Qianjiang City to optimize the allocation of regional land resources. at the same time, similar research in Jianghan Plain will have certain reference value and significance in the future. The main results of this study are as follows: (1) the results of Logistic regression analysis of raster data with different simulation scales are compared and tested by Roc, and the suitable simulation scale is selected by using the obtained Rocs value as the criterion. By comparing the Roc values of cultivated land, traffic water conservancy land, urban and rural construction land, water area and other agricultural land, it can be found that the CLUE-S model is suitable for simulating the spatial pattern of land use in Qianjiang city with raster scale 200. Logistic regression analysis also shows that in different time periods, the CLUE-S model is suitable for simulating the spatial pattern of land use in Qianjiang city with raster scale 2000.Logistic regression analysis also shows that the CLUE-S model can be used to simulate the spatial pattern of land use in Qianjiang city with raster scale 2000.Logistic regression analysis also shows that the CLUE-S model is suitable for land use spatial pattern simulation at county level. The main driving factors affecting the change of spatial pattern of different lands in Qianjiang City are not exactly the same. (2) using CLUE-S model, the spatial pattern of land use in Qianjiang City in 2010 and 2020 is simulated and predicted. The simulation results show that the correct land area of land use spatial simulation in Qianjiang City in 2010 is 175977 hectares, that is, the number of correctly simulated grids is 43994, accounting for 88.26% of the total grid, and the calculated results are 0.7617. the simulation results show that the simulation results of CLUE-S model are credible and suitable for simulating the spatial pattern change of land use in Qianjiang City on a county scale. The correct land area of land use spatial simulation in Qianjiang City in 2020 is 190729 hectares, that is, the number of correctly simulated grid is 47682, accounting for 95.66% of the total grid, and the calculated result of Kappa index is 0.9146. The results show that the simulation of land use spatial pattern in 2020 by CLUE-S model is highly consistent with its overall land use planning in 2020. (3) from 2010 to 2020, there are great differences in the changes of different categories. The area of cultivated land, forest land and other agricultural land showed a decreasing trend, the urban and rural construction land increased rapidly, the water area increased slightly, and the traffic water use land remained unchanged. The variation range from large to small is urban and rural construction land, woodland, cultivated land, other agricultural land, water conservancy land. Among them, due to the large base area of cultivated land, the change range of cultivated land is slightly smaller than that of forest land.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F301.2;F224
【参考文献】
相关期刊论文 前10条
1 张永民,赵士洞,张克斌;科尔沁沙地及其周围地区土地利用变化的时空动态模拟[J];北京林业大学学报;2003年03期
2 周锐;苏海龙;王新军;李月辉;胡远满;张凤娥;许博涵;;CLUE-S模型对村镇土地利用变化的模拟与精度评价[J];长江流域资源与环境;2012年02期
3 王丽艳;张学儒;张华;王卫;;CLUE-S模型原理与结构及其应用进展[J];地理与地理信息科学;2010年03期
4 王静,郭旭东;我国县级尺度土地可持续利用的科学调控[J];地理科学进展;2002年03期
5 蔡玉梅,刘彦随,宇振荣,Peter H.Verburg;土地利用变化空间模拟的进展——CLUE-S模型及其应用[J];地理科学进展;2004年04期
6 蔡玉梅,董祚继,邓红蒂,杨枫;FAO土地利用规划研究进展评述[J];地理科学进展;2005年01期
7 吴健生;冯U，
本文编号：2498575
本文链接：https://www.wllwen.com/jingjilunwen/jingjiguanlilunwen/2498575.html
