Landuse Inventory Using Landsat Thematic Mapper Imagery to Study Environmental Pollution

摘要

The possibility to obtain up-to-date information on agricultural landuse from Landsat TM imagery by means of classification was investigated in a pilot study on a 50.00 ha area. The pilot study was a preparation for a national exercise. The accuracy of the results for the most important classes varied between 80% and over 95%. The result was due to the excellent quality and timing of the image, the ecological homogeneity of the area and the spectral differentiation between the classes. The study was meant to supply information on agricultural landuse that could not be obtained in other ways. Landuse information was particularly required on the cultivation of maize and other crops, grasslands, coniferous and deciduous forest and heather and grass-invaded heather. The information was to be used to investigate and to predict the effects of manure on soils, surface water and ground water and the effects of national policies to solve the manure problems. The manure problems are caused by the rearing of large concentrations of animals in the Eastern and Southern part of the Netherlands.