Simulation of wetted area in slope lands in drip irrigation system

Soil wetted pattern plays an important role in the design of drip irrigation systems. Wetting pattern depends on various factors such as soil texture, soil structure, emitter discharge, hydraulic conductivity, applied water volume and land slope. In this research, two physical rectangular cubic models were constructed to measure the soil moisture advance front. The smaller model was used for experiments with lower discharge and the larger model was used for experiments with higher discharge. These experiments were carried out for four different slopes (0, 10, 20 and 30%), three soil types with different textures (light, medium, heavy) and three emitter discharges (2, 4, and 6 lit/hour). The purpose of this study was to simulate the wetting area in a drip irrigation system on slope land. The artificial neural network (ANN) model was used to simulate the wetted area. The results of the comparison between the measured and simulated values showed that the ANN model has an acceptable performance in estimating the moisture front area. Coefficient of determination (R 2) in training phase ranged from 0.994 to 0.997, root mean square error (RMSE) ranged between 18.84 to 42.70 cm2 and mean absolute error (MAE) ranged between 12.28 to 31.96 cm2. These values for the test phase were 0.993 to 0.997, 26.87 to 47.02 cm2 and 20.14 to 31.45 cm2, respectively.