The accuracy evaluation of artificial neural networks in estimating instantaneous peak flow, Case study: Fars province watersheds

Estimating instantaneous peak flow in watershed is one of the most important problems that cause hydrologists and experts to work seriously on it. One the new methods in river engineering and suspended sediment estimation is application of artificial neural networks which uses the same algorithm of human brain to find the internal relation between data based on the training process. The objective of this study is to compare the efficiency of artificial neural network method and experimental methods for estimating instantaneous peak flow in Fars province watershed. For this purpose، 24 years of daily peak and instantaneous peak flow of seven hydrometric stations were considered and tested for outlier data. Then the estimation was done based on experimental methods including Fuller، Sangal and Fill-Steiner and artificial neural network method and were compared based on RMSE، MAE and R2. Results showed that estimation of artificial neural networks is more accurate than experimental methods in all stations which indicated the lower errors of artificial neural network method compared with experimental methods.