Assessing Aquacrop Model for Deficit Irrigation Management on Sorghum (Sorghum bicolor L. Moench)

Abstract

The study was conducted in 2023, during which the sorghum crop (Sorghum bicolor L. Moench) was cultivated in the biocontrol garden of the Faculty of Agricultural Engineering at the University of Damascus. The aim was to evaluate the AquaCrop mathematical model for managing deficit irrigation of the sorghum crop. Three irrigation treatments (I1, I2, and I3) were applied, corresponding to 100%, 80%, and 60% of the total net water requirement of the crop, respectively, with each treatment replicated three times. Treatment I1 was used to calibrate the model, and the field evaluation of the model's performance was conducted.

The results of the statistical analysis showed that the simulation of crop growth using the AquaCrop model was highly accurate for vegetation cover (CC), biomass (B), and grain yield (Y), despite the crop being subjected to varying levels of water stress. This was supported by high values of the Willmott  index (d), which reached 99%, with a variance (D) ranging from -8.46% to 17.39%, and a normalized root mean square error (NRMSE) ranging from 15.52% to 19.02% for vegetation cover, and from 4.63% to 20.93% for biomass and grain yield.

However, for evapotranspiration (ET), the prediction accuracy decreased as water stress increased. The NRMSE value for treatment I3 rose to 51.3%, and the d value dropped to 0.95.