Application of Artificial Neural Networks for estimating index floods

This article presents an application of Artificial Neural Networks (ANNs) andmultiple regression models for estimating mean annual maximum discharge (index flood)at ungauged sites. Both approaches were tested for 145 small basins in Slovakia in areasranging from 20 to 300 km2. Using the objective clustering method, the catchments weredivided into ten homogeneous pooling groups; for each pooling group, mutually independentpredictors (catchment characteristics) were selected for both models. The neuralnetwork was applied as a simple multilayer perceptron with one hidden layer and witha back propagation learning algorithm. Hyperbolic tangents were used as an activationfunction in the hidden layer. Estimating index floods by the multiple regression modelswere based on deriving relationships between the index floods and catchment predictors.The efficiencies of both approaches were tested by the Nash-Sutcliffe and a correlationcoefficients. The results showed the comparative applicability of both models with slightlybetter results for the index floods achieved using the ANNs methodology.