Evaluation of the CSM-CERES-Wheat Model Using Variety Trial Data.

Selection of planting date and irrigation management are critical components of successful wheat production. In the southeastern US, yield potential can be reduced with very early or late planting, and due to water deficit. The goal of this study was to evaluate the CSM–CERES–Wheat model for growth, development and yield using variety trial data and to determine the best planting date and irrigation practice for winter wheat in the southeastern US. The observed phenological and yield data collected from 1997 to 20007 for three locations were used. For model calibration the 2002 and 2005 data were used, while the remaining years were used for model evaluation. The model was then used to determine the impact of inter–annual weather variability on potential yield, and irrigation water productivity for seven planting dates, ranging from late October to late November, under both rainfed and irrigation conditions using historical weather data from 1977–2007. Model calibration and evaluation showed good agreement between simulated and observed for both phenology and yield with index of agreement (d) ranging from 0.85 to 0.98 and the root mean square error (RMSE) ranging from 5.3 to 8.5 days for phenology and the d and RMSE values ranging from 0.89 to 0.94 and 649 to 1205 kg ha^-1 for yield. In general, the long–term simulations indicated that yield and total biomass were positively affected by irrigation and by an earlier planting date. Yield showed a notable increase for earlier planting dates. Simulated yields clearly reflected the response among cultivars and locations due to weather variability.   These findings indicate that the optimum planting date and irrigation management strategy for winter wheat production can be determined to maximize yield using the CSM-CERES-Wheat model.