Biophysical attributes estimation from satellite images in arid regions

To efficiently manage limited rangeland resources, land managers and rangeland users require accurate and timely geospatial data products concerning the health, productivity and biomass of the rangeland area. Traditional approaches to deriving these products are limited in and and semi-arid regions due to several reasons. First, in and and semi-arid regions, vegetation is often very sparse and therefore signals from vegetation are often much smaller than those from soil backgrounds. Second, vegetation in and and semiarid regions is often in senescent form and therefore is unable to be related to traditional spectral vegetation indices that were specifically designed to be sensitive to green materials. In this study, we developed new techniques that circumvent these limitations by exploring the use of the shortwave infrared spectral bands of the Landsat images. These spectral bands are very sensitive to water content of pixel elements. The inclusion of SWIR spectral bands in spectral vegetation indices results in sensitive indices to both green and senescent vegetation. The products derived from this study include fractional green grass cover, fractional senescent grass cover, biomass or forage, and canopy height in and grass dominated rangelands in the Southwest from Landsat7 ETM+ imagery. Because the algorithms used are simple, they can be used operatically to produce these geospatial products to assist rangeland managers in making optimal management decisions.