Integration of artificial neural networks with conceptual models in rainfall-runoff modeling

A hybrid form of rainfall-runoff models that integrates artificial neural networks (ANNs) with conceptual models is proposed in this study. Based on this integrated approach, the spatial variation of rainfall, the heterogeneity of watershed characteristics and their impacts on runoff can be investigated by the development of a semi-distributed form of conceptual rainfall-runoff models. As a result, in each subcatchment, the runoff generation and water budget among different runoff components including surface runoff and groundwater can be simulated with consideration of the spatially distributed model parameters and rainfall inputs. In the runoff routing, instead of a linear superposition of the routed runoff from all subcatchments in the formation of total runoff output at the entire watershed outlet as traditionally performed in a semi-distributed form of conceptual models, artificial neural networks as effective tools in nonlinear mapping are employed to explore nonlinear transformations of the runoff generated from the individual subcatchments into the total runoff at the entire watershed outlet. The feasibility of this new approach has been demonstrated in the study based on the selection of three different types of conceptual rainfall-runoff models. The verification results from the three conceptual models indicate that the approach of integrating artificial neural networks with conceptual models presented in this paper shows promise in rainfall-runoff modeling.