INFERRING ROOT ZONE SOIL WATER CONTENT BY ASSIMILATING REMOTELY SENSED DATA INTO A SOIL WATER MODEL

Authors: Starks, Patrick; Jackson, Thomas

Submitted to: Interagency Conference on Research in the Watersheds
Publication Type: Proceedings
Publication Acceptance Date: 10/1/2003
Publication Date: 10/27/2003
Citation: STARKS, P.J., JACKSON, T.J. INFERRING ROOT ZONE SOIL WATER CONTENT BY ASSIMILATING REMOTELY SENSED DATA INTO A SOIL WATER MODEL. PROCEEDINGS OF THE FIRST INTERAGENCY CONFERENCE ON RESEARCH IN THE WATERSHEDS. 2003. p. 522-527.

Interpretive Summary: Increased demand for available water supplies necessitates that tools and techniques be developed to quantify soil water reserves over large land areas as an aid in management of water resources and watersheds. Microwave remote sensing can provide measurements of volumetric water content of the soil surface up to about 10cm deep. The objective of this study was to examine the feasibility of inferring content of the root zone (down to 60cm) by assimilating remotely sensed estimates of surface (0-5cm layer) soil water content into a simple soil water budget model. Two modeling scenarios were evaluated at four tallgrass prairie sites located in central and south-central Oklahoma. In scenario 1, the model was run without being provided with remotely sensed updates of surface soil water content. In scenario 2, the model was provided updates of surface soil water content derived from remotely sensed data. Scenario 1 model simulations underestimated measurements at all sites. Simulations from scenario 2 agreed well with measured data at two study sites, but underestimated measurements at the remaining sites. The underestimation was due largely to inaccurate remotely sensed soil surface water content values. These preliminary results suggest that it is feasible to infer root zone soil water content in tallgrass prairies by assimilating remotely sensed estimates of surface soil water into soil water models, provided that the remotely sensed data correctly estimates surface conditions.

Technical Abstract: Increased demand for available water supplies necessitates that tools and techniques be developed to quantify soil water reserves over large land areas as an aid in management of water resources and watersheds. Microwave remote sensing can provide measurements of volumetric water content of the soil surface up to about 10cm deep. The objective of this study was to examine the feasibility of inferring the volumetric water content of the root zone by combining remotely sensed estimates of surface soil water content and modeling techniques. A simple soil water budget model was modified to estimate root zone soil water content from remotely sensed estimates of surface soil water content. Two modeling scenarios were evaluated at four tallgrass prairie sites located in central and south-central Oklahoma: 1) model simulations without assimilation of remotely sensed estimates of soil water content, and 2) model simulations with assimilation of soil surface water content estimated from remote sensing. In scenario 1, the model simulations underestimated measurements at all sites. Simulations from scenario 2 agreed well with measured data at two study sites, but underestimated measurements at the remaining sites. The underestimation was due largely to inaccurate remotely sensed soil surface water content values. These preliminary results suggest that it is feasible to infer root zone soil water content in tallgrass prairies by assimilating remotely sensed estimates of surface soil water into soil water models, provided that the remotely sensed data correctly estimates surface conditions.