Regional groundwater discharge: phreatophyte mapping, groundwater modelling and impact analysis of land-use change

Abstract

The relationship between groundwater recharge and discharge is one of the most important aspects in the protection of ecologically valuable areas. Knowledge of groundwater systems is therefore a pre-requisite for up-to-date integrated land and water management. A methodology is presented for assessing the relative importance of different recharge–discharge systems, with respect to ecological status or development, including mapping of regional groundwater systems, and recharge and discharge areas. This methodology is applied to a land-use planning project in the Grote-Nete basin, Belgium. Discharge regions are delineated on the basis of their spatial discharge contiguity, position in the landscape and alkalinity of the plants habitat. The simulated discharge areas are verified by field mapping of phreatophytic vegetation. Particle tracking is used to delineate the recharge area associated with each discharge area, and to characterize each recharge–discharge groundwater system. Three groundwater flow and two vegetation parameters are used in a cluster analysis to obtain four different clusters of groundwater discharge systems. It is shown that the discharge clusters are significantly different in discharge intensity and alkalinity. The effects on the groundwater system due to anthropogenic impacts on the land-use are studied by simulation of the present, pre-development, and future situation. The results indicate the sensitivity and impact of the changes on the recharge and discharge areas, and groundwater discharge fluxes. The impact of the changes for the different areas for both the pre-development and the future situation appears to differ from large decrease to large increase in total groundwater discharge. Of additional ecological importance is the fact that some areas show an opposite behaviour regarding the changes in groundwater discharge area and fluxes. The delicate shifts in the groundwater systems, which cause the changes in the recharge and discharge, clearly show the need for hydrological modelling. The synergy of hydrological modelling and vegetation mapping proves advantageous and reveals some of the ecological differences in the catchment.

Introduction

One of the main objectives of modern land planning is the protection of ecologically valuable areas and land-use that supports integrated water management. Special attention should be given to the effect of land-use changes on the hydrological cycle and the protection of groundwater systems, especially discharge and recharge areas (Boeye and Verheyen, 1992, Bernáldez et al., 1993, Pucci and Pope, 1995).

Groundwater discharge, seep or spring wetlands have often developed in discharge areas (Mitsch and Gosselink, 2000). These are, from an ecological point of view, very valuable wetlands, since they mostly have an almost permanent shallow water table, and a constant lithotrophic water quality.

In order to be able to formulate a sound land-use planning strategy, analysis of the groundwater flow system connecting recharge and discharge areas is required. This information can be derived from different technologies like hydrological mapping, vegetation mapping, groundwater modelling and hydrochemical analysis, usually in combination with geographical information system (GIS) techniques and remote sensing.

Groundwater interacts with surface water in nearly all landscapes, ranging from small streams, lakes, and wetlands in headwater areas to major river valleys and sea coasts. It is generally assumed that groundwater recharge occurs in topographically higher areas, and groundwater discharge in topographically lower areas. This is true primarily for regional flow systems, but the superposition of local flow systems makes the interaction between surface and groundwater more complex (Winter, 1999).

Discharge areas occur in that part of the drainage basin where, the net saturated flow of groundwater is directed upward towards the water table. In these areas the groundwater level is at or near the surface. On the other hand, recharge areas occur where, the net saturated flow of groundwater is directed away from the water table (Freeze, 1969). In these areas the groundwater usually is situated at a deeper level below the soil surface.

Different landscape physiographic features can be used to obtain information on the groundwater systems and especially the groundwater discharge areas. Tóth (1966) used a hydrological mapping procedure to identify regions with upward and downward groundwater movement. He found a correlation between mapped physiographic features, as springs, seepages, vegetation types, salt precipitates etc., and the direction of natural groundwater movement. Tóth (1971) explained a variety of naturally occurring geologic, morphologic and physiographic phenomena by having a common generator: groundwater discharge at the ascending end of gravity flow systems. de Vries, 1977, de Vries, 1994, de Vries, 1995 showed by way of a theoretical model that for the sandy Pleistocene part of The Netherlands the stream network is genetically coupled with groundwater discharge systems of various extents. In addition to physiographic features of the landscape also different direct measurements of the groundwater can be used to obtain information on the groundwater system. Salama et al. (1993) explained the distribution of recharge and discharge areas from four observed trends in water level changes. Shedlock et al. (1993) showed the complex dependence of wetlands on multiple groundwater systems, based on sedimentological, geophysical and hydrochemical data from test holes and wells. Roulet (1990) mapped groundwater discharge in a small headwater wetland by flownets derived from measured piezometry. Hunt et al. (1996) determined groundwater inflow to a wetland from water levels, stable isotope mass balances, and temperature profiles.

It is long known that there is a clear relation between phreatophytes and groundwater discharge in arid and semi-arid regions (Tóth, 1971, Nichols, 1994). For more humid regions this relation is less obvious and has only been investigated recently. van Wirdum (1991) pioneered the relation between fen vegetation and hydrology. Klijn and Witte (1999) reviewed the new research field of ecohydrology and paid special attention to the influence of groundwater seepage on site factors in plant ecology. They concluded that plant species may be used as seepage indicators in rapid assessments and surveys, but that constant awareness of the limitations is required. Rosenberry et al. (2000) also came to this conclusion, using the Marsh marigold plant species as an indicator of focused groundwater discharge to a Minnesota lake. De Becker et al. (1999) mapped in detail phreatophytes in an entire floodplain. The spatial distribution of the plant species was statistically explained in relation to the groundwater regime, chemistry, soil texture, chemical composition and land management.

Although analytical and numerical modelling of groundwater discharge areas were developed several decades ago, only a limited number of studies deal with this approach. Tóth, 1962, Tóth, 1963 developed a theoretical understanding and analytical formulation of groundwater systems and distribution of recharge–discharge areas in small drainage basins. Freeze and Witherspoon, 1966, Freeze and Witherspoon, 1967, Freeze and Witherspoon, 1968 extended Tóth's work with analytical and numerical solutions for regional flow under different conditions. By simulation of theoretical examples they showed that groundwater discharge occurs under the influence of at least six distinguishable cases of water table configuration and geologic setting. In their two-dimensional hypothetical models, the discharge areas occupied between 7 and 40% of the total basin. Winter (1978) simulated three-dimensional groundwater flow near lakes and investigated the conditions of seepage.

Bronders and De Smedt, 1985, Bronders and De Smedt, 1986 developed and applied regional groundwater models specifically aimed at predicting groundwater discharge and the location of the saturated source areas. They found for the Demer and Dijle basins (Belgium) groundwater discharge area-fractions of, respectively, 28 and 30%. Batelaan et al., 1993, Batelaan et al., 1996 used the same approach to simulate regional flow to a concentrated groundwater fed wetland and to predict discharge with a high spatial resolution on a regional scale. Ophori and Tóth (1989) simulated the groundwater flow in a basin by formulation of stream functions. The water table in the basin was assumed to coincide with the topography and by analysing the stream lines the location and amount of groundwater discharge was calculated. The calculated groundwater discharge area covered 24% of the basin.

Stoertz and Bradbury (1989) used the budget calculation of MODFLOW with a specified, measured water table configuration, to calculate flows to (discharge) and away from (recharge) the water table. Hunt et al. (1996) followed this approach for simulation of groundwater inflow to a wetland system. Gilvear et al. (1993) quantified water balance terms and hydrochemistry of a small groundwater fed wetland. They showed that MODFLOW simulated upward head differences below the wetland, which confirmed the groundwater discharge to the wetland. Reeve et al. (2001) used MODFLOW in modelling the regional groundwater flow to peatlands, and the DRAIN package to remove surface runoff and to constrain the simulated phreatic water level to the land surface. A particle tracking methodology was applied by Buxton et al., 1991, Modica et al., 1997, Modica et al., 1998 to analyse the configuration of theoretical and realistic flow systems, groundwater residence times, and recharge areas.

Groundwater chemistry and isotopes are often very supplemental in characterising groundwater discharge areas (Pedroli, 1990). Wassen et al., 1989, Schot, 1990 explained vegetation gradients in a fen by the chemistry of the seeping groundwater. Gerla (1992) showed that the chemical characteristics of discharging groundwater in the Red River Valley (North Dakota) suggested mixing of spatially varying proportions of local recharge and more evolved, deeper groundwater. Batelaan et al. (1998) delineated different zones in a regional groundwater discharge wetland and linked these to specific groundwater paths through different aquifers. Kehew et al. (1998) used chemical and isotopic composition of shallow groundwater around a wetland to spatially delineate areas of groundwater discharge to the wetland and groundwater recharge from the wetland. These data alleviated the inconclusive hydraulic head data with respect to the determination of the recharge–discharge function of wetlands.

The objective of this study is to set up a methodology for mapping regional groundwater systems, recharge and discharge areas. The methodology combines both hydrological modelling as well as vegetation mapping, integrated in a GIS environment. Purpose of the methodology is to allow assessment of the relative importance of the quantitative flow characteristics of different recharge–discharge systems, which should contribute to the evaluation of the ecological status or development of an area. The methodology is tested for the land planning project in the Grote-Nete basin, Belgium. Discharge areas and associated recharge areas are identified using groundwater modelling and vegetation mapping. Differences and advantages of these methods are discussed. The second objective of the study is to assess the effects of anthropogenic impacts on the groundwater system, in particular the size and intensities of discharge areas.