Understanding the amount of instantaneous peak flow in watersheds is one of the most important factors that plays important role in planning and designing of projects related to water and river engineering. The purpos...Understanding the amount of instantaneous peak flow in watersheds is one of the most important factors that plays important role in planning and designing of projects related to water and river engineering. The purpose of this study is to compare the efficiency of artificial neural network and empirical methods for estimating instantaneous peak flow in Kharestan Watershed located northwest of Fars Province, Iran. For this purpose, 25 years of daily peak and instantaneous peak flow of Jamal Beig Hydrometric Station was considered. Then the estimation was done based on empirical methods including Fuller, Sangal and Fill-Steiner and artificial neural network and were compared based on RMSE and R2 . Results showed that estimation of artificial neural network is more accurate than empirical methods with RMSE = 13.710 and R2=0.942 which indicated the lower errors of artificial neural network method compared with empirical methods.展开更多
文摘Understanding the amount of instantaneous peak flow in watersheds is one of the most important factors that plays important role in planning and designing of projects related to water and river engineering. The purpose of this study is to compare the efficiency of artificial neural network and empirical methods for estimating instantaneous peak flow in Kharestan Watershed located northwest of Fars Province, Iran. For this purpose, 25 years of daily peak and instantaneous peak flow of Jamal Beig Hydrometric Station was considered. Then the estimation was done based on empirical methods including Fuller, Sangal and Fill-Steiner and artificial neural network and were compared based on RMSE and R2 . Results showed that estimation of artificial neural network is more accurate than empirical methods with RMSE = 13.710 and R2=0.942 which indicated the lower errors of artificial neural network method compared with empirical methods.
