Click here to load reader
Click here to load reader
Apr 06, 2018
Hydrol. Earth Syst. Sci., 13, 20552068, 2009www.hydrol-earth-syst-sci.net/13/2055/2009/ Author(s) 2009. This work is distributed underthe Creative Commons Attribution 3.0 License.
Hydrology andEarth System
Comparison of different base flow separation methods in a lowlandcatchment
A. L. Gonzales1, J. Nonner1, J. Heijkers2, and S. Uhlenbrook1,3
1UNESCO-IHE, Department of water engineering, P.O. Box 3015, 2601 DA Delft, The Netherlands2Hoogheemraadschap de Stichtse Rijnlanden HDSR, P.O. Box 550, 3990 GJ Houten, The Netherlands3Delft University of Technology, Water Resources Section, P.O. Box 5048, 2600 GA Delft, The Netherlands
Received: 11 March 2009 Published in Hydrol. Earth Syst. Sci. Discuss.: 27 April 2009Revised: 18 September 2009 Accepted: 14 October 2009 Published: 4 November 2009
Abstract. Assessment of water resources available in differ-ent storages and moving along different pathways in a catch-ment is important for its optimal use and protection, and alsofor the prediction of floods and low flows. Moreover, un-derstanding of the runoff generation processes is essentialfor assessing the impacts of climate and land use changeson the hydrological response of a catchment. Many methodsfor base flow separation exist, but hardly one focuses on thespecific behaviour of temperate lowland areas. This paperpresents the results of a base flow separation study carriedout in a lowland area in the Netherlands. In this study, fieldobservations of precipitation, groundwater and surface waterlevels and discharges, together with tracer analysis are usedto understand the runoff generation processes in the catch-ment. Several tracer and non-tracer based base flow separa-tion methods were applied to the discharge time series, andtheir results are compared.
The results show that groundwater levels react fast to pre-cipitation events in this lowland area with shallow ground-water tables. Moreover, a good correlation was found be-tween groundwater levels and discharges suggesting thatmost of the measured discharge also during floods comesfrom groundwater storage. It was estimated using tracer hy-drological approaches that approximately 90% of the totaldischarge is groundwater displaced by event water mainlyinfiltrating in the northern part of the catchment, and onlythe remaining 10% is surface runoff. The impact of remoterecharge causing displacement of near channel groundwa-ter during floods could also be motivated with hydraulic ap-proximations. The results show further that when base flowseparation is meant to identify groundwater contributions to
Correspondence to: A. L. Gonzales([email protected])
stream flow, process based methods (e.g. the rating curvemethod; Kliner and Knezek, 1974) are more reliable thanother simple non-tracer based methods. Also, the recursivefiltering method (proposed by Eckhardt, 2005) can be cal-ibrated well using the results of tracer investigation givinggood results. Consequently, non-tracer based base flow sep-aration methods that can be validated for some events mayprovide a powerful tool for groundwater assessment or modelcalibration/validation in lowland areas.
Understanding runoff generation processes, i.e. source areas,pathways and retention times, is important for the predic-tion of water quantities, including floods and low flows (baseflows), and water quality in a catchment (e.g. Bonell, 1998;Uhlenbrook, 2006; Eckhardt, 2008). However, these pro-cesses continue to be difficult to quantify and conceptualize(McDonnell and Tanaka, 2001; Uhlenbrook and Hoeg, 2003)and the direct measurement of each discharge component, ina continuous way and at a sufficient number of locations ispractically impossible (e.g. Tardy et al., 2004). The accurateanalysis of water flow pathways from rainfall to streams isalso needed for the optimal protection of surface and ground-water resources (e.g. Wenninger et al., 2004). Understandingof the runoff generation processes is also essential for assess-ing the impacts of changes (e.g. land use changes, climatechange) on the hydrological response of a catchment (e.g.Uhlenbrook et al., 2008).
In many catchments, base flow is an important compo-nent of stream flow and, therefore, base flow separationshave been widely studied and have a long history in the sci-ence of hydrology (Hall, 1968; Tallaksen, 1995). Base flow
Published by Copernicus Publications on behalf of the European Geosciences Union.
2056 A. L. Gonzales et al.: Comparison of different base flow separation methods in a lowland catchment
separation methods can be divided in two main groups: non-tracer-based and tracer-based separation methods. However,most of the studies focused on mountainous catchments, andlittle attention has been given to lowland areas and areas thathave been strongly modified by man. Thus, in this study wewill focus on such a lowland study area.
1.1 Non tracer based base flow separation
The first base flow separation methods focused on the anal-ysis of the recession or depletion curves (e.g. Linsley et al.,1975; Szilagyi and Parlange, 1998) and they are capable ofidentifying the point where direct runoff (presumably surfacerunoff) finishes but they do not try to reconstruct the tempo-ral variable base flow hydrograph during floods (Dingman,2002). Later, the first filtering base flow separation meth-ods were developed to standardize the graphical base flowseparation methods (see also methodological section below):fixed interval, sliding interval and local minimum methods(Pettyjhon and Henning, 1979; Sloto and Crouse, 1996). Ba-sically, these methods take the minimum values of the hy-drograph within a pre-defined interval by following differ-ent criteria and connect them. The discharge under the con-structed line is defined as base flow accordingly. More re-cent filtering methods assume that base flow, associated asit is with discharge from groundwater storage, produces thelong wave responses of the hydrograph. Hence, low pass fil-tering of the hydrograph can be used to separate base flow(Eckhardt, 2005, 2008). Other base flow separation methodsuse the unit hydrograph method (e.g. Su, 1995). Here, thebase flow is determined by fitting a unit hydrograph modelwith information from the recession limbs of the hydrographand extrapolating it backwards. Another group of methodsfor base flow separation are the envelope and rating curvemethods (Kliner and Knezek, 1974; Sellinger, 1996; Holkoet al., 2002), which assume that a close relation exists be-tween groundwater levels and stream flows during recessionperiods due to the hydraulic connection between the streamand aquifer. Therefore, observed groundwater levels are usedto calculate base flow contributions based on previously de-fined relationships between groundwater levels and streamflows.
1.2 Tracer-based base flow separation
Hydrograph separations using hydrochemical tracers and en-vironmental isotopes offer the possibility to gain a better un-derstanding of the runoff generation processes (e.g. Bonell,1998). For example, the use of natural tracers demonstratedthat the retention of water in small catchments can be verylong (e.g. Kirchner et al., 2000; McGuire and McDonnell,2006). However, how and where the water is stored for solong in these catchments, while the hydrodynamic reactionduring rain events can be very quick (cf. hydrological para-dox: Kirchner, 2003) is not completely understood.
This type of hydrograph separation is based on a mass bal-ance approach, which assumes that the composition or chem-ical signature of water coming from various sources is con-stant and unique (different from each other) and that conser-vation of mass applies to the water quantities and water qual-ity including conservative mixing of different water compo-nents (e.g. Weiler et al., 1999; Uhlenbrook and Hoeg, 2003).However, relatively large uncertainties may be present in thequantification of the runoff components due to a number offactors (e.g. Joerin et al., 2002). Some of these uncertain-ties are the product of tracer analyses and discharge mea-surements, intra-storm variability of isotopic concentration,elevation effect on the isotopic composition of rain, chemicalreactions during runoff formation and the mixing of compo-nents, and spatial heterogeneity of tracer concentrations (seeUhlenbrook and Hoeg, 2003, for further discussion). Eventhough some of these uncertainties can be reduced by theuse of transfer function methods (e.g. Weiler et al., 2003),the assumption of a constant and uniform signature for ev-ery component is often fulfilled within short intervals (e.g.within an event). Then, tracer based hydrograph separationscan provide valuable information about the groundwater con-tributions to stream flow. However, little experience with thistechnique exists for low land areas.
The objectives of this paper are (i) to compare different ap-proaches for base flow separation in a lowland area, (ii) todemonstrate how the application of different methods in con-junction with additional experimental investigation can leadto a better understanding of the runoff generation processes,and (iii) to discuss the applicability of different base flowseparation methods in lowland areas. The study was carriedout in a typical lowland area at Langbroekerwetering in theNetherlands.
2 Material and methods
2.1 Study area
The study area is located in the central part of The Nether-lands, in the province of Utrecht (Fig. 1). The Langbroek-erwetering area is limited by the rivers Neder Rijn in thesoutheast and the Kromme Rijn towards the southwest. Inthe North, the area is bounded by an ice pushed ridge calledthe Utrechtse Heuvelrug (topographic divide). The surfaceof the study area is 51.7 km2.