Design of open drains by solving Richards equation and artificial neural networks

The perfect design of open drains requires understanding the effect of the different design parameters on groundwater table. The paper objective is to perform sensitivity analysis to show the effect of the drainage flow rate, the distance between the bed of the open drain and the bottom impervious layer, the drain bed width, the spacing between drains, and the soil type (design parameters) on the maximum height of groundwater table above the open drain bed. HYDRUS software, Dupuit equation, and artificial neural networks (ANNs) are used. Comparison is performed between HYDRUS and Dupuit equation for simulating water flow in unconfined aquifer between two water ways. The predicted groundwater table using HYDRUS is above the other one from Dupuit equation. ANNs were used to predict the maximum height of groundwater table from its design parameters. MATLAB code is written to simply use the previous ANNs to find the best drain design.