Get reliable rangeland science

Satellite imaging systems do not produce images that are in perfect planimetric accord with the Earth's surface, nor do the images coincide with the deliberately distorted geometrics of various map projections. For many purposes there is a requirement that satellite images be georeferenced or transformed according to the constraints of some convenient map projection. Other requirements may call for one satellite image to be transformed to the geometry of another. In this paper consideration is given to the geodetic correction and registration of polar-orbiting satellite data with particular reference to the monitoring of vegetation over large areas. Attention is given to the variation of the field of view with change in sensor view angle associated with the Advanced Very High Resolution Radiometer (Avhrr) on the Tiros-n series of meteorological satellites. Avhrr data, with a nominal resolution of 1.1 km on the ground, have been compared with multi-spectral scanner (Mss) data from the Landsat series of satellites which have a resolution of approximately 80 m on the ground. As the deterministic correction of image distortions is limited by nonsystematic deviations of the satellite from its known position, methodologies have been developed for emperically determining the appropriate transformation with optimal accuracy. The results presented suggest that polar-orbiting meteorological satellite data enable the monitoring of vegetation over vast areas. Such monitoring, not conceivable previously, could contribute to a timely and appropriate response in the case of calamitous drought, given the experience of Ethiopia and the Sahel in recent years.