论文部分内容阅读
以科尔沁沙地西缘内蒙古翁牛特旗为对象,利用3个时相(1989、1999和2007年,分别表示三北防护林体系工程建设近10年、20年和30年)Landsat ETM/TM数据,采用决策树分类方法对该区土地利用/覆盖动态变化进行监测。结果表明:基于决策树分类方法进行遥感影像分类的总体精度能达到89.4%以上,Kappa系数达到0.84以上;近20年,草地和耕地是主要土地利用类型(两项合计占64%以上),林业用地比重为4.8%~5.3%,面积较少;各土地类型面积变化大小依次为耕地>未利用土地>林地>居民及建设用地>水域>草地。
Taking the Wengniute Banner of Inner Mongolia, the western margin of Horqin Sandy Land, as an example, Landsat ETM / TM data of three phases (1989,1999 and 2007, respectively, representing the construction of the Three North Shelterbelt Project in recent 10, 20 and 30 years) , The decision tree classification method is used to monitor the dynamic changes of land use / cover in this area. The results show that the overall accuracy of remote sensing image classification based on decision tree classification method can reach 89.4% and the Kappa coefficient reaches above 0.84. In the past 20 years, grassland and cultivated land are the main types of land use (the two combined accounted for more than 64%), forestry The area proportion of land use is 4.8% -5.3%, and the area is small. The area change of each land type is arable land, unused land, forest land, resident and construction land, water area, grassland.