Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/30/2007
Publication Date: 2/15/2008
Citation: Vivoni, E.R., Gebremichael, M., Watts, C.J., Bindlish, R., Jackson, T.J. 2008. Comparison of ground-based and remotely-sensed surface soil moisture estimates over complex terrain during SMEX04. Remote Sensing of Environment. 112:314-325. Interpretive Summary: An approach for understanding soil moisture patterns during the North American monsoon season was developed through the use of field sampling and passive microwave remote sensing in a study conducted in Sonora, Mexico. Relatively few ground or remote sensing observations of soil moisture are available in semiarid areas characterized by complex terrain and monsoonal climates. As a result, our current understanding of the spatiotemporal variability of soil moisture in the mountainous monsoon rainfall regions is quite limited. Furthermore, soil moisture can be especially variable in the region due to the seasonal effects of the monsoon and terrain controls on rainfall, vegetation and soil conditions. The observational data analysis and interpretations presented in this study identified the spatiotemporal distribution of soil moisture in mountain landscapes during the North American monsoon. Lessons learned from this study will be used for aggregating soil moisture estimates to remote sensing footprints and designing sampling networks to take advantage of the basin hypsometric relation. These results contribute to the accuracy and reliability of the soil moisture retrievals, which should lead to increased acceptance of the soil moisture products in applications involving hydrology and agriculture.
Technical Abstract: Comparisons to ground-based surface soil moisture estimates are necessary to evaluate the capability of remote sensors to determine soil moisture and its spatiotemporal variability. Soil moisture can be especially variable in regions of complex terrain which exhibit large variations in vegetation, soil properties and hydrologic conditions. The objective of this study was to evaluate the spatiotemporal variability of soil moisture in a mountainous basin in northwestern Mexico. Soil moisture estimates from ground sampling over a topographic transect and high resolution retrievals from the Polarimetric Scanning Radiometer were compared during a two week period in August 2004 as part of the Soil Moisture Experiment 2004. Results indicated that the soil moisture estimates exhibit similar variability with mean water content. Statistical analysis, however, revealed clear differences in soil moisture in the basin, in particular for wet periods and high elevations. Despite these differences, the temporal persistence of soil moisture from the estimates agrees well and indicates locations that capture the basin-averaged conditions. Furthermore, the spatiotemporal soil moisture characteristics from the two products are linked to terrain attributes. As a result, a hypsometric technique is shown to improve comparisons between basin-averaged values derived from ground data and remote sensing, as compared to arithmetic averaging. To our knowledge, this study is the first attempt to evaluate PSR/CX retrievals over a region of high terrain and vegetation variability using statistical, time-stability and terrain analysis techniques.