EOS Validation Program

Institution: USDA ARS Hydrology Laboratory and Remote Sensing Lab
                    Bldg. 007, BARC West, Rm 104
                    Beltsville, MD 20705
Phone: (301) 504-8511
FAX: (301) 504-8511
E-mail: tjackson@hydrolab.arsusda.gov

Venkat Lakshmi, University of South Carolina
David D. Bosch, USDA ARS Southeast Watershed Research Lab
David C. Goodrich, USDA ARS Southwest Watershed Research Center
Mark Seyfried, USDA ARS Northwest Watershed Research Center
Patrick J. Starks, USDA ARS Grazinglands Research Lab

February 2003
April 2004

The EOS Aqua AMSR-E soil moisture product will be the first attempt at routinely mapping surface soil moisture. Soil moisture products from the AMSR-E have to be validated because the retrieval algorithms utilize formulations, parameters and ancillary data that have not been thoroughly developed and verified. Validating these soil moisture products at the scale of the AMSR-E footprint will be difficult. The current validation plan adequately addresses some aspects through episodic field campaigns. Longer-term observations in diverse environments are needed to understand variations in the soil moisture-brightness temperature relationships that arise from seasonal variations of vegetation and validate the retrieval for a range of conditions. Addressing these issue with existing networks is not adequate since the spatial density of measurements these networks provide cannot be used to estimate the mean and variance of the AMSR footprint retrieval, which is required for validation.

We propose to extend and strengthen the current validation plan by adding well calibrated, real time, and publicly available observations of the average soil moisture for U.S. watersheds in different climate and physiographic regions that will be used to quantify the accuracy of the AMSR-E products on a continuous long term basis of this product. Four ARS watersheds will provide data in this project. These are located in Arizona, Georgia, Idaho, and Oklahoma. The project will provide well calibrated real time surface soil moisture from the SCAN stations in each watershed, provide a watershed average soil moisture using supplemental sampling within thirty days, and contribute to the development of a sound theoretical basis for scaling from a single station (SCAN) to the integrated watershed/footprint average.

Results of the project will contribute to quantifying the accuracy of the soil moisture product over the entire year. A robust validation will lead to increased reliability, acceptance and use of the soil moisture data in land surface hydrology and climate studies.