Submitted to: Agricultural and Forest Meteorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 6, 2000
Publication Date: N/A
Interpretive Summary: Information on regional soil moisture conditions is important for mapping rainfall events, monitoring differential drying patterns, and assessing water availability for plant growth. Though the demand for such information is high, the means for mapping soil moisture are few. Conventional measurement techniques are for only a small area, and require on-site operators and tedious processing. On the other hand, there is some evidence that information from orbiting satellite-based sensors could provide a regional assessment of surface soil moisture content. These sensors detect the reflection from a radar beam directed at the earth's surface, and this signal is related, in part, to variations in soil moisture. The signal is also related to surface topography and the amount of green vegetation. In our study, we devised a means to minimize the effects of varying topography and vegetation, and thus, enhance the link between the radar signal and surface soil moisture. With this technique, it may be possible to use the orbiting radar sensor to map soil moisture over large areas with reasonable accuracy. This will lead to a better understanding of weather conditions and improvements in management of scarce resources.
Efforts to utilize satellite-based Synthetic Aperture Radar (SAR) to determine soil moisture conditions of rangeland regions have been confounded by variations in topographic features, surface roughness, and vegetation density. We designed an experiment to investigate the sensitivity of C-band SAR backscatter to soil moisture in a semi-arid rangeland and to test a data-fusion approach based on both optical (Landsat TM) and radar (ERS-2 SAR) to improve regional estimates of soil moisture. The data-fusion approach used the difference between dry- and wet-season SAR backscatter to normalize roughness effects and used surface reflectance in optical wavelengths to account for differences in vegetation density. We focused the study on three flat, uniformly-vegetated sites of known surface roughness and monitored variations in soil moisture, vegetation density, and SAR signal over time. We found that this optical/SAR approach improved the correlations between SAR backscatter and soil moisture from r2=0.27 to r2=0.93. We confirmed that the SAR signal was insensitive to both sparse green vegetation cover and dense, standing, brown vegetation cover. This study also raised two concerns: the overall sensitivity of SAR backscatter to soil moisture was relatively low, and the approach required a high level of accuracy in the estimate of green leaf area level that may not be obtainable with optical remote sensing algorithms. In any case, the positive results from this study should encourage the use of optical/SAR remote sensing for monitoring range conditions and improving management of scarce resources.