Visual landsat analysis in land cover inventory in a topical country (Venezuela)

Abstract

The purpose of this study was to evaluate three sources of geographical information, False Color Composite, Band 5 and Band 7 images of Satellite LANDSAT-2, applied in the detection and delimitation of land cover areas in a case study in the Republic of Venezuela. Previous bibliographic references stated the existence of deciduous, semi-deciduous and evergreen woody vegetation communities in different stages of openess (forest, woodlands, shrublands and thickets) and grasslands, in addition to some kinds of tropical agricultural practices (grazing and cropping areas). The identification of these land cover types was accomplished by visual discrimination of colors in the FCC image and of tones in the Bands 5 and 7. FCC image gave the most complete identification of vegetation communities, agricultural practices and burned areas. Band 7 provided the most accurate delimination of burned areas, water bodies and evergreen forest. Band 5 permited the clearest identification of stages of openess in the vegetation.