Assessing a semi-empirical model performance to predict the wetting patterns in subsurface drip irrigation

Context

Because of the complexity of soil water distribution, accurate prediction of wetting pattern is not easily accessible and this has led to a inefficiency in some proposed models in the literature. These models do not consider the hydraulic characteristics of the irrigation system and are developed solely on the basis of the water volume or infiltration rate, soil hydrodynamic properties, and other conditions of the percolation environment.

Objective

Due to the importance of the estimation of wetting front beneath Subsurface Drip Irrigation (SDI) laterals, a semi-empirical model using easily accessible data of the SDI system is proposed to predict the wetting patterns in both distribution and redistribution phases. The aforementioned model has been developed based on a novel approach that couples the governing equations of the lateral hydraulics with empirical equations derived by dimensional analysis.

Methods

To develop the model and evaluate its performance, three 16-mm drip line pipes with 0.2, 0.4, and 0.5 m emitter spacing, and 2–5 l/h discharge were placed at 0.20 m depth in a soil box filled with clay loam soil. Water was applied for 3 hours at 50, 100, and 150 kPa operating pressures, and the wetting patterns’ geometries were measured in each lateral after 1, 2, 3, and 24 h. The performance of the model was then assessed and compared with that of three other models.

Results and conclusion

The result demonstrated that the proposed model provides the most accurate estimations of the wetting depths and widths. RMSE and MAE statistical indexes of the wetting depth were 0.001–0.002 m and 0.004–0.009 m, respectively, whereas those associated with the wetting width were 0.001–0.003 m and 0.005–0.016 m, respectively. These values resulted in the lowest error when compared with the corresponding obtained from other well-known models. Consequently, the model allows acceptable predictions of the wetting patterns using accessible hydraulic parameters of the SDI system.

Significance

Uniquely, the results of the lateral hydraulic analysis were applied to determine the wetting front dimensions in this study. Also, the results demonstrated that the model was successful in wetting pattern prediction beneath the lateral as a line source. While other models usually are used to estimate the wetting front around a point source.