Skip to main content
ARS Home » Research » Publications at this Location » Publication #106846
Title: ASSIMILATING LANDSAT IMAGERY IN A GRASSLAND GROWTH MODEL: A CASE STUDY IN ARIZONA

Author
item Moran, Mary
item NOUVELLON, Y - US WATER CONSERVATION LAB
item BRYANT, R - UNIV OF ARIZONA
item NI, W - UNIV OF ARIZONA

Submitted to: Pecora Conference Land Satellite Information in the Next Decade
Publication Type: Proceedings
Publication Acceptance Date: 10/4/1999
Publication Date: N/A
Citation: N/A

Interpretive Summary: Rangeland managers need accurate information about the health and extent of grasslands for proper management of renewable resources. Considering the variation in grassland conditions, it is preferable to have such information in a map-like format covering the entire region of interest. In this study, we combined a grassland growth model with images of surface conditions obtained from a satellite-based sensor to produce maps of grassland health over a ten-year period in southeast Arizona. The modeled results were within 5% of ground-based measurements of grass density at a typical rangeland site. The study also addressed several issues associated with the combined use of models and satellite-based sensors including long computation times for model simulations, preprocessing of satellite images to remove unwanted noise, and determination of air temperature and wind speed variability over the region. With this model validation and with these issues addressed, this approach has potential to produce daily, spatially-distributed information on grassland growth to improve resource management.

Technical Abstract: This study presents a modeling approach to combine a soil-vegetation- atmosphere transfer model with Landsat imagery to produce images of spatially-distributed living plant and root biomass. The model was run for a 10-year period with 25 Landsat Thermatic Mapper (TM) images over a USDA watershed ( 150 km 2) in Arizona for grassland-dominated regions. This manuscript addresses the special issues associated with applying the approach on a spatially-distributed basis, including minimizing computer run time, processing TM images for atmospheric and other unwanted noise, and obtaining spatially-distributed meteorological conditions for model input. The results from this work will provide a blueprint for similar studies with other models and images at other locations.