Using of Hydrograph Analysis Methods for Base Flow Estimation (Case Study: Silakhor-Rahimabad Watershed)

Introduction

A significant portion of the flow of perennial rivers originates from groundwater. The changes that occur in a watershed due to natural and human factors are indicative of physical changes and artificial mismanagement of water resources. These situations change the contribution of groundwater to streamflow. Therefore, understanding baseflow allows for the identification of the potential and dynamics of the groundwater system. In principle, the separation of base flow and quick flow is difficult to distinguish from the measured discharge data in a river, because the measured discharge in a river is a combination of the two flow components. Separation of riverbed flow is essential for water resources management and can significantly contribute to the calculation of water availability in the dry season (relatively short discharge period). In addition, comparing different watersheds in terms of flow recession characteristics can provide valuable information about storage and recharge properties in the watershed. The main objective of this research is to estimate base flow using several hydrograph analysis techniques, as there has been neither organized research on groundwater resources at the watershed level nor studies on different methods for estimating the base flow contribution in these streams.

Materials and Methods

The current research includes estimating base flow from daily streamflow data using the Flow Duration Curve (FDC) technique, the Web-based Hydrograph Analysis Tool (WHAT), and the Baseflow Index (BFI+) program. Using the FDC technique, the long-term annual average fraction of flow from base flow is estimated after obtaining the values of Q90 and Q50. The Web-based Hydrograph Analysis Tool includes three algorithms: the local minimum algorithm, the one-parameter algorithm, and the two-parameter algorithm. The web-based WHAT system provides an efficient tool for hydrologic model calibration and validation. Baseflow information from the WHAT system can also play an important role in sustainable groundwater and surface water management, including irrigation and industrial uses, and estimating pollutant loads from both baseflow and direct runoff. The Baseflow Index program also uses the following algorithms: Fix Interval, Sliding Interval, Local Minima, Lynne-Hollick, Chapman, One Parameter Algorithm, Two Parameter Algorithm, Exp. Weighted Moving Average, Eckhardt, BFLOW, IHACRES, and Fure & gupta. For this purpose, daily streamflow and precipitation data were obtained from the Regional Water Organization of Lorestan Province. The time series of data for streamflow and precipitation was selected from 2009 to 2019.

Results and Discussion

The results of the data analysis indicate that most of the automatic filter techniques used with assumed parameters have overestimated the baseflow above the average compared to the FDC. In addition, the FDC analysis showed that the contribution of groundwater storage to streamflow was below average. The WHAT automatic digital filter tool has been widely used for long-term baseflow separation using a two-parameter digital filter (α and BFImax). In this study, the BFImax value was set to 0.80 for alluvial streams and the filter parameter (α) was set to 0.995 for the Rahim Abad stream. The BFI values for the Rahim Abad River are consistent and are estimated to be around 45% for the entire watershed. The results obtained from BFI+ showed that the calculated baseflow values for the one-parameter and two-parameter algorithms, except for RDF-IHACRES, BF-BFLOW, BF-Chapman, and BF-Furey, were higher than the mean flow. In addition, the long-term baseflow to flow ratio or BFI is equal to the ratio of Q90/Q50. This ratio indicates the discharge of groundwater or other delayed sources to the streamflow. Hydraulic structures built upstream of hydrological gauging stations can affect flow conditions. These BFI values are related to the geology and hydrogeology of the watershed. The Q90/Q50 ratio has an annually varying decreasing and increasing trend for flow, indicating that the contribution of groundwater to streamflow varies from year to year with decreasing and increasing changes. Finally, by employing all methods, the range of groundwater contribution to Rahim Abad streamflow was obtained between 2 and 84 percent, and its average value was also determined to be 63 percent.

Conclusion

Considering the average of all BFI values obtained from all methods with values less than the mean, an overall average of 45% was obtained, which provided a better estimate for the entire watershed. In this regard, the modified RDF methods (one-parameter and two-parameter algorithms), IHACRES, BF-BFLOW, BF-Chapman, and BF-Furey were selected as the top algorithms for the entire watershed. Further studies are necessary for future groundwater resource problems in the watershed. The interaction of groundwater and surface water and pollution problems, water quality management of rivers, estimation of groundwater potential using other techniques, and estimation of the contribution of groundwater under climate change are among those that can be mentioned. Introduction of suitable methods of separation of daily flow in hydrological modeling, regional analysis of minimum flows and determination of base flow share can be used. It is hoped that the output of this study will help the planning, development and management of water resources in the Silakhor watershed.