The monitoring of global food supplies performed by the U. S. Department of Agriculture (USDA) Production Estimates and Crop Assessment Division (PECAD) is essential for early warning of food shortages, and providing greater economic security within the agriculture sector. Monthly crop yield and forecasting is calculated by PECAD through a combination of climatic and land surface data integrated into land surface models within the Crop Assessment Data Retrieval and Evaluation (CADRE) Decision Support System (DSS). The accuracy of this system is highly dependent on the data sources used; particularly the accuracy, consistency, and spatial and temporal coverage of the land and climatic data input into the models.

Soil moisture is a fundamental data source used in the crop growth stage and crop stress models. Currently, the PECAD DSS utilizes a modification of the Palmer two-layer soil moisture model to estimate surface soil moisture. Inputs into this model include soil parameter values of soil water holding capacity, daily precipitation and temperature estimates provided by weather data from the Air Force Weather Agency (AFWA) and precipitation observations from the world Meteorological Organization (WMO). These sources provide secondary estimates of soil moisture and may be improved by the addition of direct observations of soil moisture.

This study aims at improving the soil moisture estimates used by the PECAD by integrating soil moisture observations from the NASA EOS Advanced Microwave Scanning Radiometer (AMSR-E) into the USDA DSS. Launched in 2002, the AMSR-E instrument is capable of providing a full global coverage soil moisture product over lightly vegetated areas every 2-3 days. The improved spatial and temporal resolution of AMSR-E upon the current AFWA and WMO data will be beneficial particularly in areas where the AFWA and WMO data are sparse. Therefore, the integration of the AMSR-E soil moisture product into the PECAD FAS DSS is envisaged to provide a better characterization of surface wetness conditions at the regional scale and enable more accurate monitoring of boundary condition changes in key agricultural areas.

The application of the Ensemble Kalman Filter (EnKF) will be examined to assimilate a three-year dataset of AMSR-E observations in crop production regions. Before assimilation of the AMSR-E derived soil moisture can be accomplished, the modeled (PECAD) and observed (AMSR-E) data must be scaled to a common climatology to reduce the time-invariant errors. This analysis will demonstrate the bias removal and assimilation methodology used in the study. Results are forthcoming.