Application of the CERES-Wheat Model to Winter Wheat Yield Forecast in Beijing

Abstract

Wheat yield is important for the economy and food requirements of the state, so there is a need for reliable estimates of wheat production under varied environments. On the basis of early-warning information of successive yield predictions, producers and marketing boards could, for example: adjust use of fertilizers, revise marketing plans, and decide on options for following seasons. The dynamic simulation model CERES-Wheat (crop estimation through resource and environment synthesis-wheat) was applied to simulate yields from 2005 to 2007 at Xiaotangshan, which located in Changping of northern Beijing. Experiment datum required of CERES-Wheat model such as soil data, weather data and management data were all collected and checked. Simulation results of 2005 were used to calibrate Model and which of the other two years were gained for validation. Model calibration was made through comparing the field-observed and model-simulated results of five stages: ① dates of anthesis and maturity; ② values of LAI (leaf area index); ③ biomass yields of anthesis and maturity; ④ dry matter of leaf, stem, and grain; ⑤ final wheat yield. The difference of simulated and actual dates of anthesis is less than 6 days, which of the autumn is less than 5 days. The values of LAI and yield also agree well with measured data. This study revealed that the calibration of CERES-Wheat can be used for the prediction of wheat growth and yield in Beijing.