Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Assimilation of Thermal Remote Sensing-Based Soil Moisture Proxy into a Root-Zone Water Balance Model

Authors
item Crow, Wade
item Kustas, William

Submitted to: American Geophysical Union
Publication Type: Abstract Only
Publication Acceptance Date: May 1, 2006
Publication Date: June 1, 2006
Citation: Crow, W.T., Kustas, W.P. 2006. Assimilation of thermal remote sensing-based soil moisture proxy into a root-zone water balance model [abstract]. EOS Transactions, American Geophysical Union. 87(36) Joint Meeting Supplement, Abstract, H31A-01.

Technical Abstract: Two types of Soil Vegetation Atmosphere Transfer (SVAT) modeling approaches are commonly applied to monitoring root-zone soil water availability. Water and Energy Balance (WEB) SVAT modeling are based forcing a prognostic water balance model with precipitation observations. In constrast, thermal Remote Sensing (RS) observations of canopy radiometric temperatures can be integrated into purely diagnostic SVAT models to predict the onset of vegetation water stress due to low root-zone soil water availability. Unlike WEB-SVAT models, RS-SVAT models do not require observed precipitation. Using four growings seasons (2001 to 2004) of profile soil moisture, micro-meteorology, and surface radiometric temperature observations at the USDA's OPE3 site, root-zone soil moisture predictions made by both WEB- and RS-SVAT modeling approaches are intercompared with each other and availible root-zone soil moisture observations. Results indicate that root-zone soil moisture estimates derived from a WEB-SVAT model have slightly more skill in detecting soil moisture anomalies at the site than comporable predictions from a competing RS-SVAT modeling approach. However, the relative advantage of the WEB-SVAT model disappears when it is forced with lower-quality rainfall information typical of continental and global-scale rainfall data sets. Most critically, root-zone soil moisture errors associated with both modeling approaches are sufficiently independent such that the merger of both information from both proxies - using either simple linear averaging or an Ensemble Kalman filter - creates a merge soil moisture estimate that is more accurate than either of its parent components.

Last Modified: 10/25/2014
Footer Content Back to Top of Page