Author(s): M. Tamer Ayvaz; Ulas Tezel; Elcin Kentel; Recep Kaya Goktas
Linked Author(s): Tamer Ayvaz
Keywords: River Flow Prediction; Artificial Neural Network; Ergene River
Abstract: The objective of this study is to develop an artificial neural network (ANN) based solution approach to predict the weekly flows of Ergene River which is the largest river in Thrace region of Turkey. In the developed approach, precipitation–flow data relationships have been investigated in order to establish the best model structure to predict streamflow at the selected basin. The developed relationships are then evaluated using a feed forward neural network where back propagation algorithm is used to determine the associated network weights. The performance of the developed ANN based solution approach is evaluated by using the weekly precipitation and flow data collected from different monitoring sites in Ergene River basin. The model results are also compared with HEC-HMS model outputs which is calibrated using the same precipitation and flow data. Results indicate that the proposed ANN based solution approach can be effectively used to predict the weekly flows of Ergene River.
Year: 2018