The purpose of this study is to determine if Synthetic Aperture Radar (SAR) can be used independently, or in conjunction with Landsat Enhanced Thematic Mapper Plus (ETM+) to improve the classification accuracy of structural attributes of rangeland vegetation, particularly percent shrub cover and top shrub canopy height. Such information, if mapped accurately, can be used in models to better characterize fuel conditions and fire regimes, as well as to evaluate fire hazard status, called for by the U.S. National Fire Plan. The input datasets utilized in this investigation included eighteen bands of Landsat ETM+ path 38 / row 32 (three image dates, six bands each), backscattering and interferometic data derived from tandem ERS-1/2 SAR image pairs (C-band), and extensive field point data. The results showed the use of SAR data provided no significant improvement over the ETM+ data for estimating percent cover or shrub canopy height. The lack of improvement in classification accuracy is possibly due to the influence of topography on the radar backscattering signal. Additional results demonstrated improved model accuracies when a 3x3-averaging filter was applied to the eighteen bands of ETM+ imagery.
