Assimilation of Remotely Sensed Soil Moisture and Vegetation with a Crop Simulation Model

Abstract:

When a crop model is used to predict crop yields early in the growing season, two sources of uncertainties prevail those coming from climate and model uncertainties. Climate uncertainty is greatest early in the growing season and tends to decrease as weather data become available in the growing season. Model uncertainty due to errors in model structure, modeling assumptions and other ancillary data, generally remains constant through the growing season. Skillful climate forecasts can reduce climate uncertainty especially at the earlier stages of the growing season, while assimilating remote sensing (RS) data within the growing season can reduced model uncertainty. In this talk, we focus on the development, application and verification of a crop modeling-data assimilation framework capable of ingesting RS soil moisture and vegetation parameters, in this case, leaf area index for predicting aggregated crop yields. We discuss the lessons learned from our case studies in Iowa, with more homogenous rainfed agricultural system, and Georgia, more heterogeneous mixed rainfed/irrigated agricultural system. One of our goals is to show the utility of better soil moisture products, e.g. from SMAP, for improving the prediction of agricultural/hydrological variables with actionable lead-times.