Key points are not available for this paper at this time.
Abstract Measurements of the cover of vegetation were made at a network of 58 permanent 30 ha sites within a semi-arid shrub rangeland from 1981 to 1984. These vegetation data were used to calibrate Landsat data to produce images of 'COVER', a composite variable comprising all the percentage cover of living (shrubs, grass, etc.) and non-living (litter) plant material. The complementary variable was 'BARE' soil (100 – COVER). The regression equations included only Multispectral Scanner band 5 (MSS5) and were used to generate images of COVER for the years 1981, 1982 and 1984 for each of the five major rangeland types within the 25000 km2 test area. These COVER images were incorporated into a Geographical Information System (GIS) and used to assess and monitor the behaviour of individual rangeland types and pastoral properties over the 1981–1984 period which included a severe drought in 1982–3. The Landsatderived data for the region mirrored that measured on the ground. Drought conditions produced a universal decline in COVER levels but the size of the decline was specific to rangeland type being determined by both the composition and condition of the pasture. Similarly, individual properties showed differing responses to the drought. The capacity of detect and display differences in COVER between pre-drought (1981) and post-drought (1984) conditions illustrated the different ecological responses of rangeland types.
Graetz et al. (Fri,) studied this question.