Location: Location not imported yet.Title: Potential for Improved Crop Yield Prediction Through Assimilation of Satellite-Derived Soil Moisture Data) Author
Submitted to: Meeting Abstract
Publication Type: Abstract only
Publication Acceptance Date: 6/1/2010
Publication Date: 9/27/2010
Citation: Mladenova, I., Doraiswamy, P.C., Teng, B., Crow, W.T. 2010. Potential for improved crop yield prediction through assimilation of satellite-derived soil moisture data [abstract]. Remote Sensing and Hydrology Symposium. 2010 CDROM. Interpretive Summary:
Technical Abstract: Crop yield estimates have a strong impact on dealing with food shortages and on market demand and supply; these estimates are critical for decision-making processes by the U.S. Government, policy makers, stakeholders, etc. Most of the decision making is based on forecasts provided by the U.S. Department of Agriculture (USDA). Therefore, there have been continual efforts by the USDA towards improving the accuracy of the operational crop yield production models. Given the crop yield sensitivity to root-zone soil moisture and considering that the latter is a critical component of crop production models, it is reasonable to expect that enhancing the model soil moisture information will improve its yield forecasting capabilities. In this study, we evaluate the potential to improve the accuracy of the USDA Environmental Policy Integrated Climate (EPIC) model forecasts, through assimilation of surface layer soil moisture data derived from the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) using an Ensemble Kalman Filter (EnKF). Initial assessments will be carried out over the entire state of Iowa, USA, where corn and soybeans are the two major agricultural crops. EPIC crop growth is simulated as a function of climate, soil/ground conditions, and management practices. Model simulations will be performed for a period of six years (2003 to 2008). The EnKF will be utilized to update the model soil moisture and soil temperature at all soil layers. In general, model soil moisture is strongly dependent on the quality of the precipitation input data. In this study, the model will be run using two different precipitation sources: (1) gauge data obtained from the Midwestern Regional Climate Center and (2) coarse resolution precipitation estimates derived from the Tropical Rainfall Measuring Mission. This approach will allow us to examine whether assimilation of the AMSR-E-derived soil moisture into EPIC can improve crop yield forecasting.