基于逻辑回归模型的环北京地区土地退化态势分析

Study on Land Degradation Trend by Applying Logistic Multivariate Regression Model in Northwest Region of Beijing

以环北京地区为研究区域,从土地利用类型变化和农用地质量下降两个方面分析了环北京地区土地退化状况.并利用逻辑回归模型分析了土地退化与气候、地貌、植被、土壤、人口、土地利用变化、社会经济因子之间关系,探讨土地退化的发展趋势。研究结果表明,风蚀土地是环北京地区土地退化的主要类型.主要分布在监测区内蒙古高原西北部、坝上高原北部及阴山北部地区.从监测区的西北到东南方向.土地退化数量逐渐减少;由牧草地退化而增加沙地是土地利用类型变化造成退化的重要类型.主要集中在浑善达克沙地监测区。基于建立的沙地增加、风蚀土地和水蚀土地逻辑回归模型.年均降雨量和土壤有机质对环北京地区沙地增加的影响最强,沙地增加最可能的区域将主要分布在浑善达克沙地的边缘地带和科尔沁沙地的东部区域;NDVI和土壤有机质对环北京地区风蚀土地影响最强.风蚀土地最可能发生的区域集中在浑善达克沙地南部:NDVI、坡度和土壤湿度对环北京地区水蚀土地影响最强,水蚀土地最可能发生的区域集中在监测区的东南部.并且所预测的最可能发生土地退化的区域与已存在的退化土地区域具有一定的相关性。通过以上分析.最后提出了环北京地区生态环境建设的建议。

Land degradation processes, which imply a reduction of the potential productivity of the land （e.g., soil degradation and accelerated erosion, reduction of the quantity and diversity of natural vegetation）, result from a long history of human pressure upon land resources as well as from interactions between varying climatic characteristics and ecologically unbalanced human intervention. The north-west region outside of Beijing, is one of the most important regions where many departments invest most and pay environmental problems in this region affect not paper analyzed characteristics of land degradati most attention. The land degradation and other only Beijing but also the surrounding area. This on actuality situation in the NW region of Beijing, based on TM （ETM） in 2002. The wind-eroded land was mainly distributed in north of Yin Shan Mountain. Due to degradation of grassland, the sandy land increased from 1991 -2002, mostly distributed in the monitoring zone of Hunshandake sandy land. The water-eroded land was mainly distributed in monitoring zone of the south of Yin Shan Mountain and south of monitoring zone of Horqin sandy land. The salination-land was mainly distributed in lake surrounded area and the drainage basin of Sanggan River. And To better understand the drive forces of land degradation processes in study area, a multivariate spatial model associated with land degradation is found by the explanatory variables of Logistic Multivariate Regression model （LMR）. The explanatory variables include wind speed, soil humidity, soil organic matter, NDVI, average precipitation, soil slope, et al. The value of the parameter estimated by model with their corresponding standard error, chi-square statistics, and significance probability are analyzed to find the driver of land degradation in studied area. And the probability of land degradation is predicted. Finally, suggestions for the eco-environment construction of the studied region have been put forward.