Studies of the interaction of groundwater and surface water in hummocky terrane are largely concerned with the hydrology of lakes and wetlands in glacial terrane and in sand-dune terrane. Although fewer studies have been made of the interaction of lakes and wetlands with groundwater than of the interaction of streams and groundwater, studies of the former have increased markedly since the mid-1970s.
Generalized numerical models of steady-state ground-water flow with respect to lakes enclosed by a groundwater divide were discussed by Winter [1976] for vertical sections and by Winter [1978] for three-dimensional settings. Townley et al. [1992] developed a generalized model for determining groundwater capture zones associated with shallow water bodies that have groundwater inflow on one side and seepage loss to groundwater on the other side. In these studies, the geologic framework was simplified to consider only major geologic features. Guyonnet [1991] used a finite-element vertical section model to study the effect of small-scale sedimentary structures on groundwater flow into lakes. He showed that thin continuous or discontinuous low-permeability layers can significantly affect flow paths and that advective movement of contaminants into lakes can be highly variable in space and time.
In a paleoecological application of groundwater flow modeling near lakes, Almendinger [1990] used the analytic element model developed by Strack [1989] to determine lake and groundwater interaction during the extreme dry interval in the upper Midwest about 5000 years ago. Lake levels during that time were determined by analysis of the physical properties of lake sediments [ Digerfeldt et al., 1992].
Generalized models of transient groundwater flow with respect to lakes were discussed by Winter [1983] for vertical sections and for variably saturated conditions. Results of the study indicated that groundwater recharge is highly variable in space and is initially focused directly adjacent to surface-water bodies where the unsaturated zone is thin. Field studies in a number of different hydrogeologic and climatic settings have confirmed the presence of transient water-table mounds resulting from near-shore focused recharge. For example, Cherkauer and Zager [1989] mapped a transient water-table mound on the downgradient side of Lower Nashotah Lake, Wisconsin. As the mound increased in height, it caused an expanding zone of inseepage in an area where outseepage is normally present. Transient water-table mounds resulting from focused recharge were also documented by Anderson and Munter [1981] near Snake Lake, Wisconsin, by Phillips and Shedlock [1993] near small lakes in Delaware, by Shedlock et al. [1993] near interdunal wetlands on the south shore of Lake Michigan in Indiana, by Lee and Swancar [1994] near Lake Lucerne, Florida, by Winter and Rosenberry [written communication] near prairie wetlands in North Dakota, and by Mills and Zwarich [1986] near prairie wetlands in Saskatchewan. Two studies of nutrient input to lakes have also indicated the dynamic movement of groundwater in the nearshore region of lakes. Shaw et al. [1990] and Vanek [1991] both indicated that most phophorus input to the lakes was derived from the nearshore riparian zone.
Conversely, having a shallow water table near the edges of surface water results in a process that has the opposite effect on seepage patterns compared to focused recharge. Transpiration directly from groundwater near the edges of surface water results in cones of depression that enhance seepage out of the surface-water body. The effect of this process on water levels in prairie wetlands was documented by Meyboom [1967] several decades ago, and has been documented in a number of studies in recent years, such as those by Doss [1993], Phillips and Shedlock [1993], and Winter and Rosenberry [written communication]. The effect of these transient flow processes on the geochemical processes associated with lakes has been discussed by Phillips and Shedlock [1993] for coastal areas and by Arndt and Richardson [1993] for prairie wetlands.
Instruments commonly used in field studies of the interaction of lakes and wetlands and groundwater include seepage meters [ Lee, 1977] and minipiezometers [ Winter et al., 1988]. Although studies that discuss the results of using these devices are common, several studies in recent years are notable. For example, Cherkauer and Nader [1989] used seepage meters to measure the distribution of groundwater seepage into lake Michigan and found that geologic heterogeneities have a significant control on seepage patterns. Although seepage meters are commonly used, the validity of the seepage values obtained continues to be questioned and studies to evaluate the meters and how the data are interpreted continue. Shaw and Prepas [1990] evaluated seepage meters and the field design of their placement. Belanger and Montgomery [1992] also evaluated seepage meters recently in what was perhaps the most extensive evaluation of the devices. The experiments were conducted using multiple seepage meters and minipiezometers in a large tank where hydraulic gradients could be controlled.
As is true of streams, many studies of the interaction of lakes and wetlands with groundwater are prompted by concerns over water chemistry; therefore, the use of chemical methods to understand the interaction is common. LaBaugh [1988] indicated a close relation of hydrogeologic setting to lake water type within a climatic gradient in the north-central United States. Stauffer [1985] indicated the usefulness of solute tracers to estimate groundwater flow into lakes. Krabbenhoft et al. [1990] used a 3-dimensional solute transport model to evaluate the interaction of groundwater with a lake in northern Wisconsin. Kenoyer and Anderson [1989] and Anderson and Bowser [1986] discussed the role of groundwater in minimizing the effects of acid precipitaion on lakes. Siegel [1983] indicated that traditional hydrogeological modeling, field, and chemical transport approaches to studying lakes in hummocky terrane could also be used effectively in the study of extensive peatlands. His study indicated that local flow systems are common within the peatlands, and that the distribution of biological communities reflect the chemical characteristics of these small local flow systems.
Although the usefulness of isotopes for understanding hydrologic processes of lakes was discussed by Dincer [1968], Krabbenhoft et al. [1994] recently indicated the usefulness of environmental isotopes for understanding the interaction of lakes and groundwater. Alpers and Whittemore [1990] also used environmental isotopes to gain insight into the role of groundwater in the hydrology of two lakes in northern Chile, and Herczeg et al. [1992] did the same for Lake Tyrrel in Australia.
Crowe and Schwartz [1981] used a total hydrologic system approach to understanding the interaction of lakes and groundwater by developing a watershed model for predicting water and solute exchange with lakes. Crowe [1993] recently used this approach to evaluate the effect of climate change on the hydrology and chemistry of a prairie lake in Canada.