Prediction of stream flow by utilizing artificial neural network in flood plain (Case study: Sepidroud watershed)

For knowing the hydrological behavior and water management of Sepidroud River (North of Iran-Gilan) the present study focused on stream flow forecasting with artificial neural network. Ten years (2000-2009) historical inflow data, observed from the Sepidroud River, were selected ; then 10 years inflow of the Sepidroud River have been forecasted by neural network. Finally, the results obtained from forecasted data compared with observed data. The results showed that neural network could predict stream flow with high precision and the maximum error between predicted and observed data was 3% approximately.