COMPUTING THE WATER-BALANCE OF A SMALL AGRICULTURAL CATCHMENT IN SOUTHERN ENGLAND BY CONSIDERATION OF DIFFERENT LAND-USE TYPES .1. LAND CLASSIFICATION USING REMOTELY-SENSED IMAGERY

被引:3
|
作者
ROBERTS, G [1 ]
FRANCE, M [1 ]
ROBINSON, M [1 ]
机构
[1] NERC,COMP SERV,WALLINGFORD,OXON,ENGLAND
基金
英国自然环境研究理事会;
关键词
D O I
10.1016/0378-3774(92)90089-F
中图分类号
S3 [农学(农艺学)];
学科分类号
0901 ;
摘要
Percentage land cover - grassland, arable and forest - in a small, 18.6 sq. km, agricultural catchment in southern England has been determined by a supervised classification of a Landsat remotely-sensed image. Training areas for the classification were obtained from the agricultural returns of three farms in the area. The land cover classifications for the individual farms generally agreed to within 10% of the agricultural returns in terms of grassland/arable land differentiation. However, the discrimination between permanent and temporary grassland, and between different types of cereal crops was poor. When the classification was extended to the entire catchment, the percentages of grassland, arable land and forested areas given from the Landsat imagery were almost identical to those given from parish statistics, though this degree of agreement may have been somewhat fortuitous.
引用
收藏
页码:145 / 154
页数:10
相关论文
共 3 条