【摘要】 土地覆盖研究是全球变化的一个重要内容。本文针对中国土地覆盖实际情况,从遥感制图角度和陆地生态系统观点出发,建立了一种基于陆地生态系统特点的土地覆盖遥感分类体系。充分利用目视解译的精确性和自动分类的快速性,在高精度目视解译数据提供的基本覆盖类型及其位置、边界的控制下,以250m分辨率的2005年MOD IS植被指数产品为遥感分类信息源,获取主要类型的详细分类信息,并结合人机交互判读分析与相关资料等的综合运用,实现2005年全国土地覆盖遥感制图。对比实地调查结果的精度分析表明:土地覆盖制图的类型分类准确,基本上反映了各地区的土地覆盖特点和分布趋势,图斑格局与影像地理特征基本吻合,全国土地覆盖制图总体准确率为91%,各类型准确率也达89%以上。更多还原

【Abstract】 Land cover change is an important part of global change.Aiming to present situation of land cover in China,according to terrestrial ecosystem characteristic and remote sensing mapping,in this paper we put forward a remote sensed land cover classification system.In order to combine the accuracy of interpretation and speediness of self-classification,we integrated remote sensing classification,alternating interpretation and assistant information analysis,under the control of polygon’s position and borderline which provided by higher accurate land use data,with the data source of MODIS vegetation index products in 250m resolution,to realize national land cover remote sensing mapping in 2005.Using field survey data,the accuracy assessment is also implemented.Results show that: each class of land cover map has an exact precision,and the map can represent the characteristic and distribution trend of land cover in each region on the whole;the polygon patterns matched the image geography characteristics quite well.The overall precision of national land cover map is 91%,and the precision of each class is over 89%.更多还原