Terrain: Slope Influence on QuikSCAT Backscatter

Abstract

Soil moisture (SM) is an important variable in determining streamflow, agricultural productivity, weather, and climate. An effective way to map SM over large areas on a regular basis is by using active microwave observations. This paper examines the influence of topography on radar backscatter measurements for a range of vegetation conditions and on the development of a normalization technique for the correction of topography-induced variability. Radar backscatter observations derived from the QuikSCAT sensor were analyzed to investigate the effect of sloping terrain over the North American Monsoon Experiment region that is characterized by heterogeneous surface conditions and complex topographic terrain. A digital elevation model, along with local incidence angle and slope, was used to investigate the backscatter dependence on topography variation for eight main vegetation classes. Pearson product-moment correlation (R) analysis showed strong backscatter dependence on the local incidence angle caused by changes in slope. The overall average reduction in variances after correction for August 2004 depended on vegetation type and ranged between -16% to -42% and -18% to -37% for horizontal and vertical polarizations, respectively. The corrected sigma-0 was also evaluated using in situ SM observations obtained during the Soil Moisture Experiment 2004 field campaign. The computed percent change in R between sigma-0 and SM demonstrated significant improvement after correction when using vertically observed sigma-0. The standard errors of estimate for these two vegetation classes were lowered by about 12% and 5%, respectively, after applying the proposed topographic normalization technique to the QuikSCAT observations.