Rainfall-Runoff Modeling Using Artificial Neural Network

Abstract

The Artificial Neural Network (ANN) is a method of computation inspired bystudies of the brain and nervous systems in biological organisms. A neuralnetwork method is considered as a robust tools for modelling many of complexnon-linear hydrologic processes. It is a flexible mathematical structure which iscapable of modelling the rainfall-runoff relationship due to its ability togeneralize patterns in imprecise or ‘noisy’ and ambiguous input and output datasets. This paper describes the application of multilayer perceptron (MLP) andradial basis function (RBF) to predict daily runoff as a function of daily rainfallfor the Sungai Lui, Sungai Klang, Sungai Bekok, Sungai Slim and Sungai Ketilcatchments area. The performance of ANN is evaluated based on the efficiencyand the error. It has been found that the ANN has a potential for successfulapplication to the problem of runoff prediction.