Abstract: Daily weather records and observations from paddy rice field experiments at two sites in Jianghan plain (Wuhan and Jingzhou) were used to calibrate and validate the recently released version 4.5 of CERESRice model.The experiments included singleseason rice and doubleseason early rice and late rice.The model was firstly calibrated using the GLUE parameter estimation calculator and trial and error method as a complement, then validated through comparing the observed dynamic leaf area index (LAI) and biomass, development stage (flowering and maturity), total above ground biomass and grain yield with that of simulated. The results showed that the model performed well in development stage simulation, the errors between simulated and measured flowering and maturity date was no more than 3 days, with early rice the most accurate and late rice the least accurate, owing to the differences in photoperiod sensitivity. The simulated and observed of dynamic LAI and biomass during the growing stage showed a consensus trend, with the Index of Agreement(D) of 0.98 and 1.0 respectively, although slight difference existed in specific data. The predicted and observed grain yields and total aboveground biomass were also very close on the whole, with the relative Root Mean Square Error (RMSE) of 8.72% and 5.91% respectively, but a relative large RMSE of 11.6% for late rice for the chilling injury during the flowering stage. The results suggested that the CERES Rice model was able to simulate rice production with good performance in the Jianghan plain under a normal weather condition. This research supports the model application in the Jianghan plain for operational use, such as crop phenology prediction, as well as meteorological effect assessment and yield forecasting provided that effect of extreme weather condition on yield formation was taken into consideration.

Key words: Paddy rice, CERESRice model, Adaptability evaluation, Jianghan plain