Runoff Prediction Using Artificial Neural Networks

Abstract

The concept of rainfall-runoff variation is a non-linear event and highly tedious and continuously changing process. It involves different parameters which include i.e. mainly rainfall, soil, morphology and vegetation. The modelling for runoff prediction requires many engineering applications. The biggest obstacles in prediction of runoff are the estimation of extreme values. Unless models are not able to get the dynamics of rainfall-runoff process accurately, accurate prediction of these extremes is not possible. Various approaches have been adopted to represent rainfall-runoff process. The best among those approaches is using Artificial Neural Network techniques for the development of long -term and short-term forecasting models. Comparision is held between various approaches for runoff forecasting using ANNs.