|Wen, J - WAGENIGEN UNIVERSITY|
|Bindlish, R - SSAI|
|Su, Z - WAGENIGEN UNIVERSITY|
Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 15, 2005
Publication Date: December 1, 2005
Citation: Wen, J., Jackson, T., Bindlish, R., Hsu, A., Su, Z. 2005. Retrieval of soil moisture and vegetation water content using SSM/I data over a corn and soybean region. Journal of Hydrometeorology. 6:854-863. Interpretive Summary: The ability of a Special Sensor Microwave/Imager (SSM/I) based soil moisture algorithm under significant levels of vegetation was evaluated. SSM/I is operational but not optimally designed for soil moisture retrievals and as a result it has limitations in this application. Previous studies with the SSM/I have involved sparsely vegetated or semi-arid regions. In this investigation, a unique database from a large-scale field experiment (SMEX02) allowed us to develop and test a method for soil moisture retrieval that used the 37 GHz channels to estimate the vegetation water content and the 19 GHz for soil moisture. Validation using the ground observations of soil moisture showed an acceptable level of estimation error for the study site using the retrieval method. A statewide extension of the algorithm produced soil moisture patterns that were credible when compared to the daily antecedent precipitation. This study concludes that soil moisture retrieval is possible using the SSM/I observations, but the accuracy is limited under high-level vegetation and cloud conditions. Results of this investigation provide new insights on how we might operationally correct for vegetation effects using high frequency microwave observations.
Technical Abstract: The potential for soil moisture and vegetation water content retrieval using SSM/I brightness temperature over a corn and soybean fields region was analyzed and assessed using datasets from the Soil Moisture EXperiment 2002 (SMEX02). Soil moisture retrieval was performed using a dual polarization 19.4 GHz data algorithm that requires the specification of two vegetation parameters, single scattering albedo and vegetation water content. Single scattering albedo was estimated using published values. A method for estimating the vegetation water content from the microwave polarization index using SSM/I 37.0 GHz data was developed for the region using extensive datasets developed as part of SMEX02. Analyses indicated that the sensitivity of the brightness temperature to soil moisture decreased as vegetation water content increased. However, there was evidence that SSM/I brightness temperatures changed in response to soil moisture increases resulting from rainfall during later stages of crop growth. This was partly attributed to the lower soil and vegetation thermal temperatures that typically followed a rainfall. Comparisons between experimentally measured volumetric soil moisture and SSM/I retrieved soil moisture indicated that soil moisture retrieval was feasible using SSM/I data, but the accuracy highly depended upon the levels of vegetation and atmosphere atmospheric precipitable water; the standard error of estimate over the 3-week study period was 5.49%. The potential for using this approach on a larger scale was demonstrated by mapping the state of Iowa. Results of this investigation provide new insights on how we might operationally correct for vegetation effects using high frequency microwave observations.