Comparison of different base flow separation methods in a semiarid watershed (case study: Khorramabad watershed, Iran)

Base flow separation to identify groundwater contribution in river flow is important, and is one of the basic studies in hydrological science. Base flow separation approaches are important for determination of null index to calculate rainfall excess and rainfall losses in a watershed, unit hydrograph, and also estimating groundwater recharge. In this study, 15 base flow separation methods were used in "Chamanjir" outlet of Khorramabad watershed in Iran, and the results were compared with Recession Curve method as the base method. The results showed that most of these methods were significantly different at 95% confidence level. The results of comparing different methods' efficiency also revealed that sliding interval method, considered 91.6% of the river discharge as base flow, which was the highest value among all methods. Furthermore, this method with the lowest mean square error and the lowest absolute error was selected as the best method for Khorramabad watershed. Comparison of different methods' results showed that there are differences between the base flow separations in different ways. Eckhardet algorithm or two parameters algorithm with alpha = 0.925 due to considering the hydrological characteristics of the stream and watershed, and for its lower error, can be used as the best method of base flow separation in next studies in the study area. Moreover, based on the results, the highest fraction of runoff is related to the base flow in this watershed. The results of this study can be appropriate to determine the fraction of the groundwater discharged to the river flow, and also can be used for the drought studies.