Sensitivity analysis and auto-calibration of ORYZA2000 using simulation-optimization framework

There has been an increasing interest to employ crop growth simulation models for taking decision on irrigation water management. The effectiveness of such decisions mainly lies on the efficiency of the model in simulating the crop growth and the yield, which are influenced by the value of the parameters of the model. Therefore, calibration of such models is necessary before it can be employed for any application. This study proposes an auto-calibration procedure for ORYZA2000, a rice crop growth simulation model, for its application in South India. The data employed for calibration is taken from a field experiment conducted for 2 years in an experimental farm in South India. The ORYZA2000 model was integrated within Genetic Algorithm optimizer, which calls the simulator during each generation to evaluate the objective function. The auto-calibrated model was tested for its performance using a validation data set from the same experimental data. The results showed that the calibrated ORYZA2000 model is capable of simulating the full irrigation and water stress condition of rice crop effectively, and can be used to develop deficit irrigation management schedules.