Assimilation of Satellite-Based Soil Moisture into the USDA Global Crop Production Decision Support System
John
D.
Bolten, USDA-ARS Hydrology and Remote Sensing Lab, john.bolten@ars.usda.gov
(Presenting)
Wade
T.
Crow, USDA-ARS Hydrology and Remote Sensing Lab, wade.crow@ars.usda.gov
Xiwu
Zhan, USDA-ARS Hydrology and Remote Sensing Lab, xiwu.zhan@noaa.gov
Tom
J.
Jackson, USDA-ARS Hydrology and Remote Sensing Lab, tom.jackson@ars.usda.gov
Curt
A.
Reynolds, USDA-FAS Production Estimates and Crop Assessment Division, curt.reynolds@fas.usda.gov
Global estimates of soil moisture are a primary component of crop yield fluctuations provided by the US Department of Agriculture (USDA) Production Estimation and Crop Assessment Division (PECAD). The system utilized by PECAD estimates soil moisture from the Palmer two-layer water balance model based on precipitation and temperature data from the World Meteorological Organization (WMO) and US Air Force Weather Agency (AFWA). The accuracy of this system is highly dependent upon the data sources used, particularly the accuracy, consistency, and spatial and temporal coverage of the land and climatic data input into the models. However, many regions of the globe lack observations at the temporal and spatial resolutions required by PECAD. This study aims to improve upon the existing system by incorporating NASA&rsquos soil moisture remote sensing product provided by the EOS Advanced Microwave Scanning Radiometer (AMSR-E) to the U.S. Department of Agriculture Crop Assessment and Data Retrieval (CADRE) decision support system. A quasi-global-scale operational data assimilation system utilizing ensemble Kalman filtering (EnKF) has been designed and implemented to provide CADRE with a daily product of integrated AMSR-E soil moisture observations. In addition to the above effort using the Palmer model for AMSR-E soil moisture data assimilation, we have also implemented the same ensemble Kalman filter (EnKF) in NASA's Land Information System (LIS) for potential improvement and comparison. With direct meteorological forcing data input from NCEP's operational Global Data Assimilation System (GDAS) and the implemented EnKF, we have used the Noah land surface model (LSM) in LIS to assimilate AMSR-E soil moisture retrievals. The analysis results of soil moisture from the Noah LSM assimilation system are compared and shown to agree with the field measurements better than both AMSR-E retrievals and the open loop Noah LSM simulations. A methodology of system design and an evaluation of the system performance over the Conterminous United States (CONUS) will be presented.
NASA Carbon Cycle & Ecosystems Active Awards Represented by this Poster: