ARTIFICIAL NEURAL-NETWORK MODELING OF THE RAINFALL-RUNOFF PROCESS

An artificial neural network (ANN) is a flexible mathematical structur e which is capable of identifying complex nonlinear relationships betw een input and output data sets. ANN models have been found useful and efficient, particularly in problems for which the characteristics of t he processes are difficult to describe using physical equations. This study presents a new procedure (entitled linear least squares simplex, or LLSSIM) for identifying the structure and parameters of three-laye r feed forward ANN models and demonstrates the potential of such model s for simulating the nonlinear hydrologic behavior of watersheds. The nonlinear ANN model approach is shown to provide a better representati on of the rainfall-runoff relationship of the medium-size Leaf River b asin near Collins, Mississippi, than the linear ARMAX (autoregressive moving average with exogenous inputs) time series approach or the conc eptual SAC-SMA (Sacramento soil moisture accounting) model. Because th e ANN approach presented here does not provide models that have Physic ally realistic components and parameters, it is by no means a substitu te for conceptual watershed modeling. However, the ANN approach does p rovide a viable and effective alternative to the ARMAX time series app roach for developing input-output simulation and forecasting models in situations that do not require modeling of the internal structure of the watershed.