Capability of Updated SWAP Model with Satellite Images and AquaCrop Model in Simulating the Hashemi Rice Yield in Guilan Province

Nowadays, the use of simulation models is of great importance for optimal management of the production process and yield on farm scale. Due to the fact that combination of remote sensing with a model increases the accuracy of simulation, this study was conducted to evaluate the updated SWAP model combined with Sentinel 2 satellite images and AquaCrop model to simulate Hashemi rice yield in one of the research farms at the National Rice Research Institute of Iran. The measured parameters that were compared with two models included the quantity of dry matter of plant components during growth period. Evaluation of simulated and measured values indicated that the root mean square error of paddy and rice yield simulated with SWAP model updated with Sentinel 2 satellite images in two stages of reproductive and ripening were 6.28 and 142.40 kg/ha, respectively. Also, the RMSE values obtained for AquaCrop model were 9.55 and 338.46 kg/ha, respectively. Further evaluation of obtained statistics revealed that both SWAP updated with satellite images and AquaCrop models, can predict the yield and biomass of Hashemi cultivar with reasonably high accuracy. However, the AquaCrop model with a slight difference (R2=0.96), was less accurate in simulating the quantity of paddy than the SWAP model updated with satellite images. Furtheremore, when SWAP’s inputs were updated by data obtained from satellite images, at different stages, provided a higher accuracy to simulate the yield of paddy. Thus, it can be fairly used to simulate the rice yield.