Application of artificial neural networks to the design of subsurface drainage systems in Libyan agricultural projects

The study data draws on the drainage design for Hammam agricultural project (HAP) and Eshkeda (EAP), located in south of Libya, north Sahara Desert. results this are applicable to other arid areas. This aims improve prediction saturated hydraulic conductivity (Ksat) enhance efficacy system data-poor Artificial Neural Networks (ANNs) were developed estimate Ksat compared with empirical regression-type Pedotransfer Function (PTF) equations. Subsequently, ANNs PTFs estimated values used EnDrain software subsurface systems which evaluated against designs using measured values. Results showed that more accurately predicted than PTFs. Drainage based predictions (1) result a deeper water-level (2) higher density, increasing costs. gave drain spacing water table depth equivalent those data. indicate can be existing under-utilised sets applied successfully As is time-consuming measure, basing ANN generated from alternative datasets will reduce overall cost making them accessible farmers, planners, decision-makers least countries.