APPLICATION OF A NEURAL-NETWORK TECHNIQUE TO RAINFALL-RUNOFF MODELING

This paper deals with the application of a neural network technique in the context of rainfall-runoff modelling. The chosen form of neural n etwork is tested using different types of input information, namely, r ainfall, historical seasonal and nearest neighbour information. Using the data of six catchments, the technique is applied for four differen t input scenarios in each of which some or all of these input types ar e used. The performance of the technique is compared with those of mod els that utilize similar input information, namely, the simple linear model(SLM), the seasonally based linear perturbation model (LPM) and t he nearest neighbour linear perturbation model (NNLPM). The results su ggest that the neural network shows considerable promise in the contex t of rainfall-runoff modelling but, like all such models, has variable results. (C) 1997 Elsevier Science B.V.