Friday, August 10, 2007

PS 72-119: Predicting evapotranspiration from sparse and dense vegetation communities in a semiarid environment using NDVI derived at different spatial resolutions

Malika Baghzouz and Dale A. Devitt. University of Nevada, Las Vegas

The proposed research is a part of a long term ongoing study in the Great Basin (Northern Nevada) where normalized difference vegetation index (NDVI) derived from Landsat 5TM imagery is being used to estimate basin wide Evapotranspiration (ET). Utilization of remote sensing technology to estimate, scale or structure long term environmental processes such as ET should be associated with different spatial, temporal and spectral resolutions. Repeatable and continuous ground measurements are needed in order to gain further insights into the biophysical processes driving the NDVI signature. The proposed approach is based on the integration of various remote sensing sensors with different spectral resolutions consisting of Landsat 5TM imagery for large area land cover, Red and NIR light sensors (Skye 1800) mounted above plant canopies and bare soil surfaces for continuous data collection and a portable ground-based spectroradiometer (350-2500 nm). The study is being conducted in two different sites in the basin chosen for having sparse and dense vegetation covers. The sparse vegetation cover site consists of a combination of Greasewood (Sarcobatus vermiculatus), Rabbit Brush (Chrysothamnus viscidiflorus), Big Sage (Artemisia tridentata) and Shadscale (Atriplex confertifolia). While in the dense cover site, Greasewood represents the dominant species. Various plant and soil measurements are being taken every two weeks to estimate a number of physiological and biological parameters. All remote sensing measurements obtained from various resolution sensors will be analyzed and investigated at different temporal and spatial scales. The aim of this study is to provide a stable remote sensing estimation model of ET to be tested in future research and be used to generate ET maps.