Click for next page ( 17


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 16
Diffuse and Focused Recharge and Discharge Depending on the hydrogeologic setting, recharge may occur fairly uniforndy over the subsurface or may be focused by entering the subsurface through depressions such as sinkholes. Recharge can also be focused by unsaturated zone flow instabilities (e.g., gravity fingering); geologic features like fractures, faults, ciastic dikes and karstic sinkholes; and manmade infrastructure including recharge wells and waste disposal facilities. While one mechanism typically dominates, sometimes diffused and focused recharge are both important, especially at regional scales (e.g., Izbicld et al., 2002~. Similarly, discharge may be spread out over relatively large areas or may be focused into springs, including submerged springs discharging into rivers, wetlands, lakes, and the ocean. Naturally occurring discharge is also focused through transpiration by plants (e.g., Meyboom, 19661. Furthermore, discharge may be focused by pumping from wells or drain- age systems. The distinction between diffuse and focused discharge can be difficult to determine since what may be termed diffuse discharge often occurs preferentially along curvilinear features such as rivers, shore- lines, and slope breaks. Recharge and discharge rates and locations can also vary greatly over time. Shallow flow systems are particularly responsive to seasonal changes, especially where the infiltration capacity of the soil and the hydraulic conductivity of the underlying rock or sediment are high (Winter, 2001~. Focused recharge and discharge tends to be more episodic than diffuse flow. Recharge and discharge events in arid zones gener- ally tend to be more episodic than in humid settings where recharge and discharge fluctuate in a relatively predictable way in response to seasonal changes in precipitation and evapotranspiration. Recharge and dis- charge fluxes also change temporally in response to climate change. These Tong-term temporal trends are discussed in Chapters 3 and 4. Motivations for quantifying recharge and discharge fluxes arise from a variety of water resource and ecosystem management problems. The net balance between total recharge and discharge for a groundwater basin is the primary control on the amount of water stored in aquifers that are used for water supply. Thus, quantifying natural recharge and discharge fluxes is an essential step in evaluating the groundwater resource and provides a baseline assessment for managing/developing groundwater resources. Additionally, recharge is the primary mechanism by which contaminants enter the groundwater system. The spatial distribution of recharge fluxes and the travel time for movement of water from the land surface to the water table largely determine the vulnerability of aquifers to contamination including agricultural contaminants (e.g., BohIke, 2002; Fogg et al., 1999) and are used to define zones of capture for pumping wells in welThead protection programs (Reilly and Pollock 1993~. The distinction between focused and diffuse recharge mechanisms is particularly important to contaminant transport because these two end-member recharge regimes can lead to large differences in travel times, concentrations, and contaminant plume geometries. 16

OCR for page 16
Diffuse and Focused Recharge and Discharge 17 A variety of aquatic ecosystems, including wetlands, lakes, and streams, are sustained by groundwa- ter discharge. Distinct vegetation and aquatic communities are likely to be associated with focused and dif- fuse discharge (e.g., Rosenberry et al., 2000; Lodge et al., 1989~. Biogeochem~cal processes that modify water quality during discharge (e.g., denitrification) are also likely to be affected differently by these two discharge modes. An understanding of both the physical and biochemical processes occurring at the inter- face between groundwater and river systems within the hyporheic zone and between groundwater and lakes and wetlands within the hypolentic zone is key to assessing and managing aquatic ecosystems. Spatial and temporal variability in recharge and discharge fluxes has been studied in many different environments, from the prairie potholes (Winter and Rosenberry, 1995; van der Kamp and Hayashi, 1998) to coastal plains (Logan and Rudolph, 1997~. However, there are no uniformly applicable methods for quanti- fying recharge/discharge fluxes in space and time. For example, measurement difficulties arise because fo- cused recharge fluxes can vary widely over space and diffuse discharge fluxes may be of relatively low magnitudes. In most cases we estimate recharge and discharge indirectly, for example by measuring head or using baseflow estimates or tracers (Table 1-~. Such indirect estimates typically are affected by uncertainty in the estimates of other parameters used in the analysis (e.g., Winter, 1981~. Distribution of recharge/discharge fluxes may also be estimated by using mathematical models of the watershed in conjunction with field measurements to solve the inverse problem (e.g., Stoertz and Bradbury, 1989; Levine and Salvucci, 1999; Stone et al., 2001; Sanford, 2002; and Lin and Anderson, 2003~. However, fluxes estimated in this way are also plagued by the uncertainty associated with estimating all the other parameters used in the model as well as uncertainty in the heads used to calibrate the model. When comparisons of different techniques to estimate recharge and discharge have been made, results from various techniques are often different (e.g., Flint et al., 2002; Burnett et al., 20024. Only when discharge is discretely focused and accessible to measurement, for example, springs that emerge at the land surface, is it possible to make relatively accurate direct measurement of a groundwater flux. Direct measurements of submerged springs can be made using seepage meters (Lee, 1977; Taniguchi and Fukuo, 1993; Paulsen et al., 2001) but these measurements can be logistically difficult to make and are often affected by measure- ment errors. This chapter explores some of the research needs for understanding and characterizing the timing, spatial distribution, and rates of focused and diffuse recharge and discharge in major hydrogeologic settings. With an improved process-level understanding of controls on these fluxes, it may be possible to anticipate changes in the timing of recharge events and changes in the spatial patterns of recharge and discharge in response to climate variability, changes in land use, and water resource development. In particular, this chapter focuses on two questions discussed at the workshop that formed the basis of this report. The first is "How do the landscape setting and climate fundamentally control the nature of recharge and discharge in any given hydrogeolog~c setting?" The discussion of this question is woven throughout the chanter. The second question is "Is diffuse or focused fresh groundwater discharge a significant source of fresh water recycling to estuaries and the oceans?" This issue is discussed in the last section on coastal/estuarine sys- tems, in Box 2-3, and also in Chapter 3 in the context of climate. . . EXPERIMENTAL BENCHMARK SITES In effect there are presently a large number of generally uncontrolled recharge and/or discharge ex- periments taking place each year, as hundreds of researchers, practitioners, and managers collect and ana- lyze data, or build models of systems, in which groundwater recharge and discharge are quantified. In most of these studies, estimation of recharge/discharge fluxes is not the primary motivation for the study, and in many of them, inclusion of recharge/discharge is an afterthought, and few or no field measurements are

OCR for page 16
18 Groundwater Fluxes Across Interfaces made to quantify the fluxes or to verify guesses. There are few experimental sites at the scale proposed in this report (see discussion later in this section), and essentially no field campaigns, in which these processes were the focus of attention. Data and modeling results from other existing studies cannot be reliably synthe- sized, since recharge/discharge was not the focus and, in any event, the methods, scales, and uncertainties vary widely and may not even be documented. In summary, little is known about the spatial and temporal variability of these fluxes and little is known about the relative importance or magnitude of diffuse vs. fo- cused fluxes. We do know that the relative importance of diffuse vs. focused flux and the variability of fluxes depend on the hydrogeologic setting. One promising way to achieve consistency in addressing recharge and discharge issues would be to establish a network of experimental benchmark sites that sample a wide range of landscape and geohy- drologic types, and climatic regimes, building on the concept of hydrogeologic conceptual models that is fundamental to hydrogeologic analysis. An example of a framework for organizing hydrogeologic concep- tual models or generic hydrogeolog~c settings is the Fundamental Hydrologic Landscape Unit (FHEU) in- traduced by Winter (2001) (Figure 1-3~. The proposed experimental sites would be selected to differentiate the relative importance of disuse versus focused recharge and discharge processes at different scales and to test different methods of estimating recharge and discharge in disparate settings. Care would be taken to distinguish modem-day recharge rates from paTeorecharge rates (see Chapter 3~. Research at the experimen- tal sites would facilitate the development and testing of new concepts and new tools for measurement of recharge and discharge fluxes. Prior to and during startup of each experimental setting, generic modeling studies would be used to aid in selection of sites, design field instruments and instrumentation networks, and design experiments. Studies at each experimental site logically would lead to quality assured results that are transferable to sim~- lar settings elsewhere. Studies at the network of sites, using consistent science and technology, would per- m~t synthesis. All data and information generated at these sites should be maintained at a centralized facility and made available to the scientific community. Advances in technology including m~cro-instrumentation, access to online digital data bases, avail- ability of remotely sensed data sets, high resolution tomography, hydrolithology, soil and other landscape data, improved computational tools, data telemetry from remote locations, automated sampling capabilities, and isotope and other novel geochem~cal tools, make beneficial detailed studies of recharge and discharge possible. With these advanced technologies incorporated into a network of benchmark sites, important sci- ence questions can be successfi~ly addressed. What size would be ideal for a benchmark site? The relevant scientific issues and methods of measurement change with scale (Chapter 4~. Some experiments could be run at the 'point" scale of, say, a meter or could be fixed on the hilisiope scale of, perhaps, lO-lOOm, the scale used in many field studies of other hydrologic and geomorphic problems such as runoff generation and slope stability. However, many of the methods, questions, and issues demand a larger scale, perhaps that of a watershed, or even an entire river basin. Appropriate spatial scales also depend on the landscape, hydrogeology, and climate. ideally each benchmark site would cover a spectrum of spatial scales, possibly using a nested design where the bench- mark site itself is at the hilIsIope or watershed scale but smaller scale sites would be located within the benchmark site, while larger scale studies might also include more than one benchmark sites. What is the appropriate temporal scale for studies at benchmark sites? New observations should be designed with a sufficient sampling frequency to address the relevant scientific issues and to test the meas- urement methods. Benchmark sites should be located where historical data such as that from streamgages and Tong-term monitor wells can be used to extend the record, and preferably where the record can be ex- tended into the past even further using paTeohydrolog~cal and paTeoecological data. While fixed benchmark sites would dramatically improve the scientific community's ability to ad- dress recharge and discharge issues, and to develop and test measurement methods, the sites would not have

OCR for page 16
Diffuse and Focused Recharge and Discharge 19 the geographic coverage necessary to answer all questions, or to work cooperatively with large scale hydro- climatologic field experiments and campaigns, such as the Hydrologic Atmospheric Pilot Experiment (NAPEX) or the more recent Global Energy and Water Cycle Experiment (GEWEX). While soil moisture issues were an important part of these experiments, groundwater recharge and discharge were effectively ignored and an important opportunity to improve understanding of these processes and the interconnection with other water reservoirs (Figure lot) was missed. The hydrogeologic community should develop the ability to lead and/or participate in such large-scale campaigns in order to take advantage of these opportuni- ties in the future and thereby address the research questions that a fixed network of observatories cannot fully address. Further discussion of these opportunities (e.g., coupled aquifer-land surface-atmosphere mod- els) may be found throughout Chapter 3. A network of benchmark sites for groundwater recharge and discharge studies is an ambitious undertaking. A few hilIsIope scale sites might be readily achievable, but integrated watershed scale sites will require coordination by the hydrologic community. In particular, the community must leverage other opportunities such as existing federal experimental watershed programs, which have related long term data- bases, infrastructure support, and an interest in many of the same issues. A second opportunity lies with federally sponsored field sites like the system of Long Term Ecological Research Centers (www.Iternet.edu) or the Consortium of Universities for the Advancement of Hydrologic Science's (CUAHS~ proposed net- work of hydrologic observatories (www.cuahsi.org). The benchmark sites would fit nicely into the proposed hierarchical structure and central management of the CUAHST observatones. The following sections illustrate how benchmark sites can be used to conduct research on recharge and discharge issues and methods by focusing attention on four generic hydrogeologic settings: karst, glaci- ated Midwest, mountain and valley, and coastaVestuarine. These settings represent four different types of landscapes and hydrogeologic conditions, and are used to suggest specific focal areas for research at each type of site. These four terrains represent a wide range of hydrogeologic conditions including some of the most problematic settings with respect to recharge and discharge but they are by no means a comprehensive representation of the types of settings found in the U.S. or elsewhere. The purpose is to illustrate the nature of the issues and questions that need to be addressed relative to groundwater fluxes and to show how the spatial and temporal variability and occurrence of disuse vs. focused fluxes is dependent on the hydro- geologic setting. Important generic hydrogeolog~c settings that are not considered herein include alluvial valleys, permafrost regions, volcanic terrain (e.g., the Columbia River Plateau and the Hawaiian Islands), crystalline rock terrain (the Precambrian Shield), and unglaciated plains (i.e., the High Plains including the Ogallala Aquifer and the Sand Hilis). For a more comprehensive description of the hydrogeologic settings found in North America, the reader is referred to Back et al. (1988~. Similar expositions of the research needs relative to recharge and discharge could be written for each of these settings, and benchmark sites would ideally be located in many of them. Karst Karst is a general term for a wide range of landscape settings in which the underlying rocks have been modified by solutional processes. Rocks that are subject to dissolution include gypsum, rock salt, and carbonate rocks such as limestone and dolostone. Carbonate rocks are common in most karst settings, and the "classic" karst landscape contains such characteristic landforms as sinkholes, closed depressions, hum- mocky topography, caves, and springs. However, the continuum of karst landscapes ranges Tom minor, poorly developed karst features to major, integrated conduit networks. These environments are often ex- tremely vulnerable to groundwater contamination and contain unique ecological niches, but understanding

OCR for page 16
20 Groundwater Fluxes Across Interfaces and quantifying recharge and discharge in karst environments poses significant scientific challenges (Ap- pendix B). Groundwater movement through karst systems is predorn~nately along secondary features, such as fractures, conduits, or caves that are formed or enlarged by solution. Groundwater flow rates can be very rapid, and flow can be turbulent. Storage of groundwater in karst can be large or small depending on the nature of secondary void space (Figure 2-~. High-intensity precipitation events may move water through karst systems rapidly from recharge to discharge. Focused recharge to karst systems depends less on evapotranspiration and plant cover than in other hydrogeologic settings and occurs when the surface topog- raphy routes runoff to specific locations, such as sinkholes, exposed fractures, or closed basins. Thus, the structural, sedimentological, and geomorphic history of the rock, its mineralogical composition, and overly- ing soil thickness and composition all influence the extent to which focused recharge can occur. Many karst terrains contain springs, and groundwater discharge from conduit-flow karst springs can be very large and extremely variable. Focused recharge and discharge can be fairly readily identified in many karst systems from geo- morphic features such as sinkholes, stream networks, and vegetation patterns defining fracture traces (Sasowsky, 1999~. Operationally, however, the three-dimensional distribution of conduits and fracture net- works in karst systems can be very complicated, and discrete Towpaths are difficult to map. Often in karst terrains subsurface potentiometric divides do not coincide with surface topographic basins, making the computation of water and chemical balances very difficult. Discharges from springs, however, represent an integrated rn~x of water that has followed different flow paths. Potentially, a large amount of flow is cap- tured by direct measurement of spring flow and under steady-state conditions recharge can be estimated in- directly from discharge measurements. For example, Plummer and others (1998) estimated the percentage of river water recharging the Floridan aquifer using chloride and isotopes. Urbanization and agncultural practices have compromised karst systems throughout North America because of the ease with which water can be recharged in these settings and because of high yields to pump- ing wells. Dewatering has led to subsidence features (e.g., Wilson and Beck, 1992; Halliday, 1998) and ma- jor ecological stress and agncultural practices and urbanization has led to widespread contamination of karst aquifers by nutrients (USGS, 1999), solvents (Wolfe and Haugh, 2001), and other chemicals. Remote sensing of vegetation and soil water content, geomorphological analysis, and field observa- tion and mapping of focused recharge and discharge relative to fracture systems over the watershed would be a first step towards evaluating recharge and discharge in the karst setting. Numerous studies of karst hy- drology in many climatic settings show that diffuse recharge and discharge is a very small component of a karst water budget (White, 1988~. A benchmark watershed to study karst recharge and discharge at a scale of ~100 km2 might include the following field components: scales; years; 1. development of Tong-term water balances for the watershed at several different spatial 2. measurement of discharge and geochemical parameters at all major springs over a period of 3. measurement of travel times and residence times using environmental isotopes, geochem~cal parameters, temperature, and tracer tests; 4. installation of monitoring wells and meteorological stations; continuous measurement of water- level and geochemical changes in these wells; 5. detailed mapping of surface topography, soils, and individual karst features such as sinkholes, exposed fractures, and closed depressions; identification of vegetation to find possible correlations with dis- charge zones at seeps.

OCR for page 16
Disuse and Focused Recharge and Discharge Surface-water / . I . 1 ~ Water table UPLAND I I- l l ; - ~ ~ , . Carbonate LOWLAND ~ ===~ FIGURE 2-1 One type of karst Hydrogeologic Setting found In carbonate rock. The setting shown here is typical of well-developed karst with an integrated subsurface network but many other types of karst are found. SOURCE: Re- printed, wi~penmission, Tom Winter (2001~. (a) 2001 byAmencan Water Resources Association. Glaciated Midwest 21 Storage of groundwater in the glaciated Midwest can be large or small depending on whether the aquifer is confined or unconfined. High-intensity precipitation events commonly generate overland flow; in some instances the overland flow can converge on a topographic depression causing a type of focused re- charge. Low-intensity rains provide diffuse groundwater recharge (Figure 2-2) provided the water is not intercepted by vegetation or used to replenish soil moisture. Antecedent soil moisture conditions and soil texture help determine whether rainfall results in overland flow (and associated focused recharge) or diffuse recharge. Generally, late fall rains and spring snowbelt events provide the majority of the annual recharge. Focused recharge often occurs after extended periods of wetness such as snowmelt or a series of intense rains. The scale of focused recharge may be even smaller when considering the effects of macro- pores in the shallow soil column. Focused discharge areas are commonly characterized by springs or distinct hydrophytic (wetland) vegetation (e.g., Rosenberry et al. 2000~. The presence of heterogeneities (i.e., zones of high or low hydraulic conductivity) can result in groundwater flow systems with complex three- dimensional distributions. In some cases distinct local and regional flow systems can develop (Toth, 1962~. The geochemistry of recharge water can be maintained along its entire flow path until it mixes with surface water at discharge points in both glacial outwash (Walker and Krabbenhoft ~ 998) and glaciated bed- rock settings QIunt and Steuer, 2000; Box 2-1~. Vertical variation in geochemistry can be important where discharge is diffuse and redox conditions change (Box 2-2~. While care is needed to characterize the spatial nature of recharge and discharge, temporal variability is less problematic in glacial sediments that have high hydraulic conductivity and storage. In these settings, a large amount of flow is captured by direct measure- ment of baseflow in streams and under steady state conditions an areally averaged recharge rate can be esti- ~ ~ A

OCR for page 16
22 Grour~dwater Fluxes Across Interfaces HUMMOCI OCR for page 16
Diffuse and Focused Recharge and Discharge 23 BOX 2-1 Horizontal Variation in Geochemistry of Focused Discharge into a Regional Spring Complex One might assume that in areas of focused groundwater discharge relatively few samples would be needed to characterize water chemistry because focused discharge consists of an integrated, well mixed water volume channeled from a large collection area. However, this may not be true. Hunt and Steuer (2001) stud- ied a regional spring discharge complex near an urbanizing fringe in south-central Wisconsin using numerical modeling and geochemical investigation (Figure 2-3~. The spring had a discharge of about 3400 liters/minute (2 cubic feet per second), that drained an area of about 1000 hectares. Within the spring complex, large dif- ferences in spring water chemistry were noted even when vents were within 15 m of each other. Many con- stituents from carbon species to stable isotopes of water and strontium showed significant spatial variation within the spring complex. In the case of nitrate+nitrite, areas on one side of the spring were above Linking water standards and below them on the other. Particle tracking simulations demonstrated that the chemical variation was owing to distinct recharge areas that maintained their chemical signatures when discharged into the spring complex. 11 liters/sec| ~ , , 2a : : : : ~ : : : ~ : t'2b : : :, 1 /1 3 Y / 1 ~ ]Nc \` I: 5~ ~ . _ TC~ ~ : ~ :2 litess/sec' I~ i:: ~:~ I.' r6 ; ~ : 62 liters/sec : :~ 1r _ _ _ _ 6 m 62 liters/sec = flow measured 4/6/2000 * = flow too low to measure FIGURE 2-3 Map of spring complex with sampling locations. SOURCE: Reprinted, with permission, from Hunt et al. (200 1~. ~ 2001 by National Ground Water Association.

OCR for page 16
24 Groundwater Fluxes Across Inte7faces BOX 2-2 Vertical Variation in Geochemistry of Diffuse Discharge into a Wetland In areas of diffuse groundwater discharge, vertical variation in water chemistry may be greater than hori- zontal variation. Hunt (1996) and Hunt et al. (1997) investigated a groundwater dominated wetland system in southwestern Wisconsin, in order to examine the type of sampling network needed to characterize the spatial and temporal patterns in the geochemical processes operating within the wetland. Four sites in the wetland complex were instrumented with typical water table wells and piezometers to capture relatively large scale trends, an array of suction lysimeters and mini-piezometers with 15 cm spacing for intermediate scale trends, and close interval membrane equilibrators with approximately 1.5 cm spacing for small scale trends. In this wetland, the large-scale instrumentation is insufficient to capture chemical variability, which can be substan- tial on the centimeter scale (Figure 2~), and was generally larger than the horizontal variability. The vertical variability was also larger than the temporal variability over the growing season. Tc~ P (Tr0) 0 2 4 6 8 10 12 14 16 18 20 depth al I I (cm) ~ _ 15 _~ _ GO Hatural Wetland 45 ~ ~ Tctal fi I Total Filterable Fe 55 , 1 , 1, 1, 1, 1, 1, 1, 1, 0 10 20 30 40 ~iO 60 70 80 90 TO Fe In) FIGURE 2-4 Chemical profile of total filterable phosphorus and iron within the wetland obtained using a close interval membrane equilibrator. Large changes in profile concentration were likely due to effects of plant roots. SOURCE: Reprinted, with permission, from Hunt et al. (1997~. ~ 1997 by Kluwer Academic Publications.

OCR for page 16
Diffuse arid Focused Recharge and Discharge 25 2. stream hydrograph separation (using both physical and chemical methodologies) to discern base- flow, storTnflow, and new and old water components; 3. identification of Towpaths using environmental isotopes, temperature, and tracer tests; 4. installation of monitoring wells and meteorological stations throughout the study watershed; 5. remote sensing with ground-truth of vegetation and moist soils to find possible correlations with discharge zones; and 6. synthesis of these data with coupled groundwater and surface water models that encompass the entire hydrologic system. The models can be used to gain insight into processes and to identify data needs even when parame- ter uncertainty precludes accurate characterization and prediction. The models should be at the appropriate scale for the data collected. Thus, nested models would be used for simulating site-scare data and resulting parameters would be used directly or up-scared for a larger scale mode! that includes the regional data (such as streamfiow3 (Chapter 4~. Finally, it should be noted that comprehensive data sets with large spatial and temporal coverage would be useful to assess scaling and the effects of heterogeneity at various scales. Mountain and Valley Mountain-dom~nated terrains constitute 20 percent ofthe earth's surface (Forster and Smith, 1988a). Mountainous terrain is found in several parts of the U.S.: the Western mountain ranges, the Basin and Range province, the Pacific Coastal Range/Central Valley of California, and the Appalachian region. The mountainous portions (uplands) are often rugged and are composed of exposed rock or may have a weath- ered zone of a few tens of meters and be covered with vegetation, making direct measurements of ground- water recharge difficult (Forster and Smith, 198Sa). The water table associated with the valley sides (sIop- ing bedrock and alluvial fans) tends to be located in the bedrock or found near the base of thick alluvial de- posits (Figure 2-5) making direct measurements costly Worker and Smith, l98Sa, b). Valley lowlands are covered by alluvial sediment. Sources of recharge to mountain and valley settings include diffuse direct infiltration of precipitation, including melting snowpack, to the mountain mass, valley slopes and lowlands. Discharge in these systems occurs locally as focused spring flow and seeps, and as diffuse inflow to streams and as focused discharge as water is transpired or lost to the atmosphere by direct soil evaporation in the nparian zone. Discharge measurements in these settings generally rely on standard stream gauging tech- niques and estimates of evapotranspiration. In arid regions of the Western U.S., especially the Basin and Range province, closed basins sur- rounded by mountains may contain dried up lakes known as playas. ~ this setting focused recharge infi~- trates the alluvium around the edge of the basin from streams that enter the basin from the surrounding mountains. Groundwater converges toward the playa where phreatophytes discharge groundwater. In many mountain-valley groundwater system investigations and watershed models, water balance approaches are used to compute groundwater recharge or discharge rates by difference. Fluxes are then re- lated to precipitation by developing a factor used to convert precipitation data to groundwater recharge (e.g., Maxey and Eakin, 1949~. Discharge rates are measured when possible (e.g., changes in groundwater base- flow to streams) and estimated when the process is diffuse (e.g., evapotranspiration). For an example of the application ofthese methods see Lambert and Stolp (1999) for the Tooele Valley, Utah. A benchmark watershed to study recharge and discharge in this hydrogeologic setting might include the following field components.

OCR for page 16
26 Grour~dwater Fluxes Across Interfaces ran Water \ table ...... , ' ; ~.2 ~. 1 . ~ Irk of ~ ~ $ (B) MOUNTAIN VALLEY Seepage face l Land surTa~ /Water / table ~ .., vl,=~,,~,,~, ...., , ,, _. ~ I ':~.: . groundwater flow ' :': :: , ~ ., ., Hi; ~ | FIGURE2-5 Mountain valley and playahydrogeologic. SOURCE: Reprinted,withpe~mission,firomWinter(2001). 2001 by American Water Resources Association. I. Quantification of focused and diffuse recharge and discharge requires quantification of precipita- tion, snowpack storage, vegetation and runoff. Application of appropriate analytical techniques requires consideration of climatic data, distribution of soils, vegetation types and densities, and topographic setting. Mass and energy budget techniques need to be applied to define the fate of basin sources and sinks of water (Winter, 1981; Reiner et al., 2002~. 2. Micrometeorological data such as net radiation, temperature, humidity, rainfall, snowpack and wind data combined with soil moisture and soil heat flux information should be collected using remote re- cording meteorological stations strategically placed in the lowland, valley slope and upland portions of a study site. Land use and soil types, and vegetation types, densities and properties such as leaf area indexes, need to be spatially and temporally identified. Basin wide data collection over a number of annual cycles could be accomplished using a variety of remote sensing techniques, including airborne and satellite plat- forms supported by ground truth studies. 3. Quantifying the portion of basin recharge entering the bedrock mountain region would be based on the collection of hydrogeologic data sets at and immediately adjacent to the alluvial-bedrock contact within the primary recharge area. Nests of monitoring wells paralleling the mountain front and located along downgradient Bowlines outfitted with continuous water level recorders could generate time dependent three-dimensional images of groundwater potential. Aquifer testing would provide estimates of hydraulic properties. Geophysical surveys would determine the depth of valley fill and its skatigraphy. In addition, the chemistry of the precipitation, steam flow (where present), and mountain mass (spring discharge) and alluvial groundwater at and downgradient of the mountain mass-alluvial fan interface would be obtained.

OCR for page 16
Diffiuse and Focused Recharge and Discharge 27 Physical and geochem~cal data would be used to compute flux rates from the mountain bedrock into the al- luvial sediments. Analysis of isotopic signatures of snowpack might prove useful in detennining the role of melting snowpack in recharging groundwater. 4. Focused recharge occurs as water from streams originating in the mountains flows across the ad- jacent alluvial fans. Recharge could be estimated from stream flow losses measured by perfonning standard seepage runs of these streams in conjunction with electrical resistance sensors (Blasch et al., 2002) to indi- cate streamflow timing. A research basin would include a number of continuous recording stream gauging stations to provide continuous information on seepage rates. 5. In the discharge areas, stream groundwater exchange in valley systems with rivers could be quantified using seepage runs and gauging stations. Riparian vegetation mapping, using airborne and satel- lite platforms, tied with climatic data and modeling would be used to estimate diffuse discharge. Coastal/Estuarine Coastal areas (Figure 2-6) are the terminus of groundwater flow systems; in some cases reflecting discharge from a local-scale flow system bordering the coast, and in other cases reflecting the distal end of a confined or unconfined regional flow system. Environments of interest include open coastlines, embay- ments where discharge may be focused, and estuaries at the mouths of rivers. Freshwater seepage below the high tide line is called submarine groundwater discharge or SGD (e.g., Burnett et al., 2002; Taniguchi et al., 2002~. Two factors greatly complicate the analysis of g~oundwa- ter discharge in the immediate vicinity of the coastline: the position of the salt water interface influences the location of discharge sites, and tidal fluctuations cause reversals in the groundwater flow direction. Estuar- ies can pose an additional challenge if salt water moves in and out of the river channel during tidal cycles and with seasonal variations in river discharge. A large body of literature exists on the behavior of the salt-water interface, especially the response of the interface to pumping from coastal wells. Most research has been on the landward movement of salt water in a coastal aquifer. Surprisingly less attention has been paid to the complementary case of freshwater discharge into the near-shore marine environment. In part this may reflect the difficulty of measuring direct discharge into the coastal zone (Burnett et al., 2001~. The importance of SGD in the near-shore marine env~- ronm~ent is primarily related to the delivery of nutrients such as nitrogen or phosphorous to the marine eco- system. In areas with industrial development, SOD can be a significant pollutant vector in the coastal zone. The salt-water interface can be viewed as a hydraulic barrier to freshwater discharge. Where the freshwater flux is higher, it is able to displace the interface seaward and create a region within and beyond the intertidal zone where freshwater discharges through the seabed (Figure 2-7~. It is unlikely that this process wall be spatially uniform. Furthermore, tidally-driven and convective mixing processes within the upper few meters of the seabed cause this Eeshwater to co-mingle with the infiltrating saline water, so that observations of discharge at the seabed nearly always carry a significant seawater component (known as re-circulated sea water). In an unconfined aquifer, the magnitude of the freshwater discharge generally decreases exponen- tially with distance onshore, and is likely to be limited to a distance onshore that roughly corresponds to a length two to three times the thickness of the unconfined aquifer (McBnde and Pfannkuch, 1975~. In set- tings with lower seaward groundwater fluxes, the freshwater discharge is likely to be predominantly occur- ring on the seepage face that forms during a falling tide. In a coastal setting where a confined aquifer ex- tends offshore, it is possible for freshwater to migrate substantial distances beyond the coastline, perhaps several tens of kilometers or more. This freshwater may leave the system as either a diffuse upward leakage

OCR for page 16
28 Groundwater Fluxes ~4 cross Interfaces - ~~ Regional upland Water table COASTAL TERRAIN Terrace . ~.~ ..~,.- .. -, . ~- _~,__-,_~,_~Ocean . .~ e. -: . : a. ~ .~. . . . I.. ~> ~ _ ~ ~ , , ~ ~ ~ ~ ~ ~ hi, ~ \ . :. : - :~: :, . :: : * : ..~r:.: : .... \ ~ .d.~.~.~` at, . . : ~ . : . ~ ~ . . : ~ . . ~ . ~ . - ~ . . . ~~ Direction of regional ground-water flow ...- .;... Direction of local :...:-:;;.:.... ..:. :: :. .: ground-water flow . .. . :~ .: :.:. .: ..;. .:~:.:.: .;: Is .::.: ~ ..:;,. . : .. .: :. ~ :: .. An:. Fresh/Saline FIGURE 2-6 Coastal hydrogeologic setting. SOURCE: Reprinted, with permission, from Winter (2001~. ~ 2001 by American Water Resources Association. ocean 7 DRIVING FORCES: c = con vectl'on h _ hydraulic head t = tidal pumping W = wave sef-oP c ~ Cafe t ~ I. _ ~ ~ ~ ~ ~ ~ ''brx~7 wamr ~ t ~` at; ~'%`' ~ ~'~'~''~,~, ,:~`'~",~'~'`""bra~'sh watery, water table ;7 fresh water ~ unconfined aquifer ~ ., confine ag~ffer fresh water ~ FIGURE 2-7 = Schematic depiction of processes associated with submarine groundwater discharge. Arrows indicate fluid movement. SOURCE: Modified from Taniguchi et al. (2002~.

OCR for page 16
Diffuse and Focused Recharge and Discharge 29 BOX2-3 Measuring Submarine Groundwater Discharge A team of researchers sponsored by the Scientific Committee on Oceanic Research, the Land-Ocean Interaction in the Coastal Zone Project, and the Intergovernmental Oceanographic Commission, has been carrying out experiments to compare different methods of estimating submarine groundwater discharge (SOD) in the near-shore marine environment. To date, measurements have been made at three sites; the Florida State University Marine Lab on Apalachee Bay in the northeastern Gulf of Mexico, in Cockburn Sound south of Perth, Australia, and in Long Island Sound, New York. The experiment in Florida (Figure 2-8) illustrates the approach of this interdisciplinary team of oceanographers and hydrogeologists (Burnett et al., 2002~. Three principal techniques have been compared to estimate the magnitude of submarine groundwater discharge: (1) direct measurement with manual and automated seepage meters, (2) isotopic techniques that use natural geochemical tracers to estimate the component of SGD present in the water column above the seabed, and (3) hydrogeologic modeling. Manual seepage meters to measure flow directly, and both heat-pulse and ultrasonic automated recording meters were placed on the seabed. The main tracers examined were radon and radium isotopes. A series of piezometers on the beach and in the offshore zone provided the key data for constructing a hydrogeologic model of the site. The three types of seepage meters yielded similar estimates in side-by-side deployments. Automated meters provide the advantage of yielding a continuous record of the seepage that can be correlated to the tidal cycle (Figure 2-99. There was also reasonable consistency between the estimates of SGD derived by integrating the point estimates of SGD from the seepage meters across a zone extending 200 m offshore, with the estimates derived from geochemical tracers. Predictions of SGD using a density-dependent groundwater flow model yielded values for the offshore discharge that were approximately one order of magnitude smaller than these estimates. Work is continuing to understand this discrepancy. One possibility is that the hydrogeologic model did not account for transient processes that occur at the seabed, such as tidal pumping, the effects of which should be recorded by the other techniques. W- ~~ it* ~ ~ ~ ~ p,~2 : ~ ~ ~ ABS _ ~AE)1 _~ I ~ Bare A- ~ lUlear, Ticle dirge X1 X.4 XS ~ , X~ 1 Be ~ Cal B5i \ B4 133 ~ BC A: F2C^; BY BO3 Cal ~ Y3 1 C3 ~ Lv4 ~ Y4A f1 Vi'~1 Y2 ~ rv`ETl~R~S Dl D3 4 -Y5 ~ Ye FIGURE 2-8. Schematic view of the near-shore experimental site on Apalachee Bay, northeastern Gulf of Mexico. Closed circles are seepage meters; circles with crosses are nested piezometers. SOURCE: Reprinted, with permission, from Burnett et al. (2002~. (a) 2002 by American Geophysical Union.

OCR for page 16
30 Groundwater Fluxes ~4 cross Interfaces 75 ~ ha: 5C' a' ~ 25 Q .O m 75 a' . _ I >a 50 loll _ ~ O (3 25 2 = .~ Q 14-Aug 15-Aug 16-Aug 17-Aug 18-Aug }' ~ 1 ~- .< .. .' ~ 'if i to 1.{ ~ i ~ s of j~ t\5 l,) I O 1 4-AUg ~ 5-AU g: ~ 6-AU8 1 7-AUg ~ f3~AUg (b) ~ :! , :: f Date 2000 r 150 100 . _ :~ .o ~ _ G - O ide Auto Meter Manual Meter ~~ Radon Mode! FIGURE 2-9. Records of (a) automated and manual seepage measurements at station Y4 (left-hand scale) and water depth (right-hand scale); and (b) seepage flux estimates based on modeling radon measurements. The gray dashed lines in (b) are 25% uncertainties of the estates. SOURCE: Reprinted, win permission, from Burnett et al. (2002~. ~ 2002 by American Geophysical Union. across the confining bed, or as freshwater springs on the seabed where fractures or solution features focus discharge. SGD occurs in a dynamic hydrologic environment. Driving forces originate both on land (e.g., fluid pressures beneath the sea bed that exceed the seafloor hydrostat, or a response to a local precipitation event in an unconfined system), and in the ocean (wave action, diurnal and spring / neap tides). Measure- ment techniques for SGD have been developed by both oceanographers and hydrogeologists. Oceanogra- phers have emphasized the interpretation of various natural geochem~cal tracers as an indicator of the pro- portion of freshwater in the marine water column (Burnett et al., 2002~. Examples of these tracers include methane, radon gas and radium isotopes. Tracer techniques provide a spatially averaged estimate of SGD. Hydro-geologists have used direct measurement techniques such as seepage meters placed on the seabed (Taniguchi, 2002), or calibrated discharge using land-based groun`dwater flow models (Burnett et al., 2002~.

OCR for page 16
Diffuse and Focused Recharge and Discharge 31 They have also inferred the magnitude of SGD using borehole temperature data (Taniguchi et al., 1998), pore pressure measurements (Davis et al., 1991; Schultheiss, 1990) or other techniques in conjunction with electrical conductivity (Harvey et al., 1997; Vanek and Lee, 1991) or temperature mapping (Henry et al., 1992). Where comparisons have been made among these various approaches or techniques, there has not always been uniform agreement among the estimates (see Box 2-3). Airborne thermal mapping has been successful in locating areas with enhanced SGD. Direct measurement of SGD is more problematic if the seafloor is rocky and bare of sediment, unless it occurs as focused discharge at a karst spring. At times or at sites with high surf, it is difficult if not impossible to seat any measurement devices on the seabed. There is only limited understanding of the character of spatial and temporal variations of SGD, which creates difficulties in deciding upon sampling plans for direct measurement of SGD. It is possible that relatively small differences in the silt content of the seabed sediments may create substantial focusing of discharge below the low tide line. Other, still ill defined processes may also be involved. At Waquoit Bay on Cape Cod Massachusetts, measurements of seepage fluxes using a large array of seepage meters yielded a pattern of discharge that could not be explained with simple hydrologic models (Michael et al., 20031. There is a clear need for the accumulation of field observations in a number of different coastal settings, with differing geology, and differing magnitudes of the seaward movement of freshwater. To a degree, benchmark sites to study discharge in this hydrogeologic setting already exist (Box 2- 3~. Several coastal LTER sites also exist - Florida Coastal Everglades, Georgia Coastal Ecosystems, Vir- ginia Coast Reserves, and Plum Island Ecosystem. Research on small-scale flow systems has already been done at several of these sites.