基于遥感的净耕地系数实验研究——以玛纳斯河流域为例

净耕地系数测算是土地利用现状调查中一项极其重要的工作，其目的是提高耕地面积调查的精度。通过对遥感影像数据进行目视解译，扣除各种细小地物，得出了基于遥感的净耕地系数，通过实地调查验证，得出了如下结论：利用遥感影像获取的净耕地系数与实地抽样调查获取的净耕地系数的相似度大都在95％以上，即由遥感影像获取的净耕地系数与真实值之间极其接近，数据的可靠性好；玛纳斯河流域耕地的总体垦殖水平较高，其中莫索湾灌区净耕地系数最高，为0．9216。新疆生产建设兵团农八师136团灌区系数最低，仅为0．6715；研究区内耕地的利用率与当地的地貌类型、母质与土壤类型、田块面积大小、破碎程度与平整程度、垦殖水平及人口密度等自然、人为条件密切相关。图3，表3，参13。

A net Arable Land Coefficient Study of Manas River Basin based on Remote Sensing

The measurement of net arable land coefficient is of great importonce in investigating the status of land-use, aiming at improving ac-curacy of the area survey. The net arable land coefficient was measured by visual interpretation of the remote seining images, eliminating kinds of tiny ground objects. Through the field investigation, some conclusions were reached as follows: the similarity of net amble land coefficient amounts to 95 % between the analysis of the remote sensing images and field sample survey, which means the remote sensing data are very near to true values. The overall reclamation levels highly in Manas River Basin. The Mosuowan has the highest net arable land coefficient as 0.921 6, and Xinjiang 136 Construction Corps has the lowest only as 0.671 5.The utilization rate of arable land within the research region has close relationship to natural or artificial conditions such as landform types, soil types, field size, fragmentation degree, reclamation degree, popula-tion density and so on.