Modeling of the daily rainfall-runoff relationship with artificial neural network

An approach for modeling daily flows during flood events using Artificial Neural Network (ANN) is presented. The rainfall-runoff process is modeled by coupling a simple linear (black box) model with the ANN. The study uses data from two large size catchments in India and five other catchments used earlier by the World Meteorological Organization (WMO) for inter-comparison of the operational hydrological models. The study demonstrates that the approach adopted herein for modeling produces reasonably satisfactory results for data of catchments from different geographical locations, which thus proves its versatility. Most importantly, the substitution of the previous days runoff (being used as one of the input to the ANN by most of the previous researchers), by a term that represents the runoff estimated from a linear model and coupling the simple linear model with the ANN may prove to be very much useful in modeling the rainfall-runoff relationship in the non-updating mode. (C) 2003 Elsevier B.V. All rights re served.