Surface irrigation is one of the oldest irrigation methods, but due to its low efficiency, water losses in this method is considerable. Accordingly, the more accurate the irrigation design and the more correct the water management, the greater the efficiency of surface irrigation. In order to properly design surface irrigation, it is essential to determine the uniformity of the irrigation system. Since it is difficult, time-consuming, and in some cases impossible to measure the required parameters in the field to calculate the uniformity coefficient, this study presents a mathematical model for calculating the uniformity coefficient in border irrigation. In this model, the uniformity coefficient was developed as a function of length, width, inlet flow rate, slope, roughness coefficient, infiltration equation coefficients, and irrigation duration, and by combining them, dimensionless parameters were defined. To prepare this model, the simulation of the advance stage was performed using the numerical solution of the Saint-Venant equations (hydrodynamic model) in MATLAB software, then a nonlinear model was fitted to the dimensionless parameters. In order to verify the mathematical model, field data was used in the border irrigation eva‎luation model, which resulted in a uniformity coefficient error of 3.99% RMSE. Based on the presented mathematical model, border irrigation can be designed with maximum uniformity. In addition to the mathematical model, an artificial intelligence model was developed. The gradient boosting algorithm with an RMSE error of 1.29% was used in the border irrigation eva‎luation and an RMSE error of 1.95% in the border irrigation design. Finally, by presenting graphs of the effect of each of the different parameters on the uniformity coefficient, it was concluded that the uniformity coefficient shows the greatest sensitivity to length, inlet flow rate, and irrigation duration.

محمد شايان نژاد

مهدي قيصري , شمس اله ايوبي