Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 24, 2005
Publication Date: May 30, 2006
Citation: Cosh, M.H., Jackson, T.J., Starks, P.J., Heathman, G. 2006. Temporal stability of surface soil moisture in the Little Washita River Watershed and its applications in satellite soil moisture product validation. Journal of Hydrology. 323(1-4):168-177.
Interpretive Summary: In order to validate satellite soil moisture products, it is necessary to monitor in situ soil moisture across a large variety of land surfaces at large scales. One method of simplifying this problem is to identify points within an area of interest, which accurately represent the overall area in regards to average surface soil moisture. Finding one of these representative sites can be accomplished with temporal stability analysis. This system of statistical tests can help to identify areas or sites within an area, such as a watershed, which has the same average soil moisture as the area as a whole. Examining a 21 month study period for the Little Washita River Micronet in Chickasha, OK, temporal stable or representative sites are identified and the characteristics of that stability are examined. The soil moisture pattern across the watershed was shown to be stable or persistent over time regardless of season and invariable throughout shorter time periods (a 24-hour period).
The concept of temporal stability can be used to identify persistent soil moisture patterns and estimate the large scale average from select representative sensor locations. Accurate and efficient estimation of large-scale surface soil moisture is a primary component of soil moisture satellite validation programs. However, monitoring the soil surface at large grid scales is difficult. As part of the Aqua satellite Advanced Microwave Scanning Radiometer (AMSR) Validation Program, a soil moisture sensor network was installed in the Little Washita River watershed in Oklahoma, USA in 2002. Along with data from the Soil Moisture Experiment 2003 (SMEX03), this network will provide a valuable dataset for satellite soil moisture product validation. Analysis shows that most of the network sensors are temporally stable and four sites are identified as representative with negligible bias and small standard deviation to the watershed mean. As part of this analysis, the protocols established for large-scale soil moisture sampling campaigns such as in the Soil Moisture Experiments (SMEX) are validated. This analysis showed that basing grid scale estimates on six sampling points is reasonable and accurate. Temporal stability is shown to be a valuable tool for soil moisture network analysis and can provide an efficient means to large-scale satellite validation.