Evaluation of CERES-Rice Model in Simulation of Rice Growth under Constraint Irrigation and Nitrogen Fertilizer Conditions

摘要

CERES-Ricemodel was used to evaluatethe effect ofnitrogenfertiliser and irrigation onthe growth of riceinIran. To simulate the responsefunctionin a split plotexperiment, a randomised completeblock design withthreereplications was conducted. Threeirrigationtreatments(continuous, around 5 and 8 days) were considered asthemainfactors and the amount ofnitrogeninfourlevels(0,90,120 and150 kg N/ha) as a sub-factor.Based ontheevaluationresults, itcan beconcluded that the Ceres-Ricemodel with the Normalized Root MeanSquared Error(RMSE n)8 and 6% withexplanation factor, (R2) 0.82 - 0. 95 tosimulate thegrainseed riceand also RMSE n 10 and 9 %, R2 0.87 - 0.85 tosimulate the biological yield, had appropriateaccuracyofsimulations. Also RMSE n 24 and 16 %, R2 0.76 - 0.87 tosimulate the Max LAI, had a moderateaccuracy of simulations. These outputEvaluation resultsshowed a reasonable estimate ofthemodel asthe efficiencyof amodel forthe grain yield, biological yield and LAI were0.89, 0.75, 0.38respectively. Thus, these modelscould beused by researchersas useful tools to support the resultsunder the managementof irrigation, improved qualityofthe consumed water and nitrogenfertiliserin rice.