EXTREME STREAMFLOW FORECASTING USING ARTIFICIAL NEURAL NETWORKS
Forecasting of extreme stream flow is necessary for water resource planning and management at catchment scale. Artificial neural networks(ANN) have been widely used as models for a variety of nonlinear hydrologic processes including that of forecasting runoff over a watershed. In this study, ANN a data driven technique is used for forecasting the extreme streamflow. ANN architecture is optimized by selection of transfer function, training algorithm, hidden neurons, and initial weights. For ANN weights finalization LM algorithm is used. The performance of ANN model is validated using two different performance indices. It was found that the ANN model consistently gives superior predictions without any explicit consideration of different components of the hydrologic cycle during calibration and validation. Based on the results, ANN modeling appears to be a promising technique for forecasting the extreme streamflow in semiarid Saurashtra regions of Gujarat.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.