Water is a critical resource, and groundwater is an important component of that resource. Water differs from other ground-based resources in that it is renewable: other extractive resources are expected to be depleted eventually, whereas it is possible to develop groundwater so that it will last indefinitely—a very attractive possibility.
In this paper we explore under what circumstances a groundwater system can be developed for perpetual use. The principles are well established; however, there are practical problems of implementation that involve the incompatibility of the time horizon of human decision making (years or decades) with the dynamic response of groundwater systems over a much longer period (often hundreds of years). In many instances the incompatibility of the two time frames creates groundwater developments that mine the resource—that is, permanently remove groundwater from the system.
The basic principles of groundwater physics and chemistry are well understood. Hydraulic head in an aquifer can be defined as the elevation that water stands in a piezometer inserted into the groundwater system at a point; groundwater flows from high to lower hydraulic head. Often an analogy can guide one’s intuition: the flow of groundwater is analogous to the flow of heat in a solid body, groundwater hydraulic head is analogous to temperature in heat flow, hydraulic conductivity is analogous to thermal conductivity, heat capacity is analogous to aquifer storativity. With these definitions of analogous quantities, the mathematics that describe groundwater flow are directly analogous to the mathematics that describe heat flow in solids.
The first principle of quantitative groundwater hydrology is that groundwater systems in their natural state are at equilibrium: the long-term rate of recharge is balanced by the long-term rate of discharge. Lohman (1972, p. 8) provides a concise explanation of the response of a groundwater system when a well is pumped and defines the key concept of capture:
Water withdrawn artificially from an aquifer is derived from a decrease in storage, a reduction in the previous discharge from the aquifer, an increase in the recharge, or a combination of these changes (Theis 1940). The decrease in discharge plus the increase in recharge is termed capture. Capture may occur in the form of decreases in ground-water discharge into streams, lakes, and the ocean, or decreases in that component of evapotranspiration derived from the saturated zone. After a new artificial withdrawal from the aquifer has begun, the head in the aquifer will continue to decline until the new withdrawal is balanced by capture.
This description, introduced by Theis (1940), contains the essence of quantitative groundwater hydrology and is elegant in its simplicity.
Pumping initially lowers the hydraulic head at the well, creating a gradient in hydraulic head that causes flow to the well. Soon after starting to pump the hydraulic head takes on an inverted cone shape, with the well at the center of the cone. This is what hydrologists call the cone of depression. Later the area of reduced head caused by pumping, still commonly referred to as the cone of depression, spreads out over expansive areas of the aquifer.
Capture is concerned with the changes in recharge and/or discharge created by pumping—not the initial values of recharge and/or discharge. Pumping removes water from storage in the aquifer when the hydraulic head in the groundwater system declines. At some point the head declines in the discharge area and the reduced discharge is captured by pumping.
Capture can occur in different ways. Pumping causes water table declines in areas of phreatophytes (deep-rooted plants that get their water from below the water table), which may cause the plants to die, reducing transpiration of groundwater. If pumping lowers heads in the vicinity of springs, the spring flow declines. If it lowers heads near streams that receive base flow from groundwater, the stream flow declines (Alley and Leake 2004). Pumping can also increase recharge by drawing from surface water bodies.
The groundwater system goes through a period of transition in which the capture increases over time. Ideally, the system eventually reaches a new equilibrium when the capture equals the pumping rate—no more water is drawn from storage and water levels throughout the system are stable. The transition process was described by the Nevada State Engineer (1971, p. 13):
Transitional storage reserve is the quantity of water in storage in a particular ground water reservoir that is extracted during the transition period between [undisturbed] equilibrium conditions and new equilibrium conditions under perennial-yield concept of ground water development.
In the arid environment of Nevada, the transitional storage reserve of such a reservoir means the amount of stored water which is available for withdrawal by pumping during the non-equilibrium period of development (i.e., the period of lowering of water levels).
The quote shows recognition that the system had to undergo a period of transition before reaching a new equilibrium, but in 1971 there was a lack of information about the duration of that period. It is now known that the transition period can be very long, especially if the system under development is dominated by a water table aquifer with large storativity.
Development of Groundwater Resources
Sustainable development was defined in 1987 by the World Commission on Environment and Development (known as the Brundtland Commission) as “development that meets the needs of the present without compromising the ability of future generations to meet their own needs” (chapter 2, I-1). Alley and Leake (2004, p. 13) elaborated: “groundwater sustainability commonly is defined in a broad context, and somewhat ambiguously, as the development and use of groundwater resources in a manner that can be maintained for an indefinite time without causing unacceptable environmental, economic, or social consequences.”
In the current groundwater lexicon, however, “sustainable” often has a much more restrictive definition. Pumping groundwater is considered sustainable if the system can reach a new equilibrium in which no more groundwater is removed from storage and water levels stabilize throughout the system.
We propose that the long-term yield of a groundwater system is the largest quantity of capture that can be achieved without undesired impacts. In some situations this quantity is equal to the total undisturbed discharge from the system, and since in the undisturbed state the system is in balance, the discharge is equal to the recharge. This leads to the often heard statement that the safe yield cannot exceed the recharge. But the idea that pumping should not exceed undisturbed recharge is not ironclad. Pumping near lakes or streams can induce additional recharge so that the capture, in some instances, is greater than the undisturbed recharge. While recharge may indicate an upper limit to long-term development, in many situations other factors such as effects on surface water, water quality, and subsidence can limit groundwater development (Alley 2007).
Continuum of Development Approaches
Groundwater developments exist in a continuum (Pierce et al. 2013): at one extreme are those that can be maintained indefinitely, and at the other are those that clearly mine the resource. These two extremes are relatively easy to identify.
The development of the High Plains Aquifer, which spans Nebraska, Wyoming, Colorado, Kansas, New Mexico, Oklahoma, and Texas, illustrates these two extremes (Figure 1). The southern part of the aquifer (in Kansas, Oklahoma, and Texas) is a well-known area of groundwater mining. Groundwater pumping for irrigation in the area began in the 1930s and really got going after World War II to support the region’s dominant crop, cotton. It was recognized from the early days that this was a groundwater mining operation and that once the area’s groundwater is depleted, farmers will have to return to dryland farming. In contrast, in Nebraska, where the aquifer underlies almost the entire state, the quantity of groundwater in storage remains so large that although there are many wells, there has not been a big percentage change in water levels.
Less clear are intermediate developments. For these systems, the difficulties of analyzing the development of groundwater so that pumping can be maintained indefinitely and capture much of the entire potential resource are manifold:
- The conceptual model of the system is often uncertain.
- The geology of the system is invariably complex.
- The hydrologic data for even a well-investigated groundwater system leave uncertainties; for example, recharge is often estimated with an uncertainty of +/−30 percent.
- The yield of a system can be defined only for a specified development scheme.
Given the uncertainties, there is room for differing analyses and interpretations by qualified experts, and arguments about whether an intermediate system is sustainable often result in litigation.
Understanding the Importance of Time
So far our discussion has been almost entirely an algebraic, budget exercise in which inflow equals outflow, but groundwater systems are dynamic and time plays an all-important role in them.
When development occurs, the groundwater system, as described above, transitions to a new equilibrium state (assuming the development is not so big that a new equilibrium is impossible). Establishing the new equilibrium takes time. For many groundwater systems, the time to reach the new equilibrium is quite long—often hundreds of years—especially for water table aquifers. This long time poses a problem.
When Theis (1940) and Lohman (1972) elaborated these ideas they had no good way to estimate the time required for complete capture—the tools did not exist. As the profession has gained experience with modeling, there is a new appreciation of the length of time required for a developed groundwater system to reach a new equilibrium (Bredehoeft and Durbin 2009; Walton 2010).
For example, modeling of systems under consideration for development in Nevada indicates that it may take more than several hundred years for these systems to reach a new equilibrium (Bredehoeft 2011). This means that as water is drawn from storage in the system, groundwater levels continue to decline. These systems in Nevada are not unique; it is becoming increasingly clear that many groundwater systems take a very long time to reach a new equilibrium state.
Responsible water managers seek to create groundwater systems capable of long-term yield. However, as hydrogeologists inform these managers, a stable system (i.e., one that does not experience continuing significant water level declines) can be established only over a very long time, so the groundwater system must be managed during the period of transition, when water levels decline as large quantities are pumped from storage. If the pumping continues, this water will not be replaced—it is mined.
In this context, the fact that the pumping meets some definition of long-term sustainability has very little relevance for the water manager who is concerned with the system over the next several decades. Often the only inkling that the system may be behaving according to some criteria of long-term equilibrium is a groundwater model analysis. Hydrogeologists place great faith in these analyses.
Many of these principles and associated challenges are illustrated in two examples—a groundwater development in Paradise Valley, Nevada, and pumping on the Colorado Plateau of Arizona—that fall in the middle of our development continuum.
Paradise Valley, Nevada
Paradise Valley is a typical basin and range valley that opens at the Humboldt River Valley and extends north from Winnemucca for approximately 50 miles. The valley floor is about 10 miles wide and is underlain by permeable alluvial deposits in which high-yield irrigation wells can be drilled.
Irrigated farming of potatoes in Paradise Valley began in the late 1960s and by 1980 more than 30,000 acre-feet (ac-ft) of groundwater was pumped annually to irrigate the crops. Inevitably, there has been pressure to increase the development. The US Geological Survey (USGS) undertook an investigation to assess the impact of doubling the size of the development and produced a two-layer groundwater model to forecast future development under various scenarios (Prudic and Herman 1996). Scenario five called for pumping 72,000 ac-ft/yr from the valley, much of it in the vicinity of the present development.
Figure 2 is a plot of the trajectories in input and output indicated by the model analysis. There are several features to observe in the figure:
- Pumping is held steady at 72,000 ac-ft/yr throughout the 300-year simulation.
- Initially much of the water pumped comes from storage; at 100 years the rate of water removal from storage is approximately 10,000 ac-ft/yr (14 percent of pumping), at 200 years the rate is approximately 5,000 ac-ft/yr (7 percent), and at 300 years it is approximately 2,500 ac-ft/yr (3 percent).
- The lowering of the water table diminishes evapotranspiration, much of it by phreatophytes.
- The cone of depression lowers the water table beneath the stream in the valley and captures all of the streamflow in Paradise Valley.
- The cone of depression created by pumping reaches the Humboldt River where it induces flow from the river, which increases over time.
The USGS analysis suggests that after 300 years the system will be close to equilibrium, as only 3 percent of the pumping will still come out of storage (Prudic and Herman 1996). Figure 3, an isometric projection of the model-calculated drawdown of the Paradise Valley water table after 300 years of pumping, shows water table declines over much of the valley and a large cone of depression in the southern part where the decline is more than 150 feet. All of the water in the aquifer associated with the decline is removed from storage and will not be replaced as long as the pumping continues. The State of Nevada referred to this water as transition storage but it is water that is mined from the system. The creation of the cone of depression is an integral part of system response and is necessary for the system to reach a new steady state.
The hypothetical development in Paradise Valley could reach a new equilibrium that could be maintained indefinitely. However, that equilibrium would induce flow from the Humboldt River. This is a problem because Humboldt River water is already fully appropriated. Therefore, if this development proceeded the diversion of Humboldt River water would pose a problem for the Nevada State Engineer and affected downstream users.
Groundwater is the principal source of water for much of northeastern Arizona. A groundwater model was developed to assess the impact of increased pumping proposed in an area south of the community of Leupp (Figure 4). Of particular concern was the potential impact on two perennial streams in the area, Clear Creek and Chevelon Creek. Although the scenario shown in Figure 4 stops the pumping after 50 years, streamflow capture of water from both creeks continues to rise for another 20 or more years. Thus if one were monitoring these two streams with the expectation that streamflow depletions would decrease once pumping ceased, it would actually be at least 20 years before such a decline. This is a significant challenge to the monitoring of hydrologic system development for making groundwater management decisions.
Since the 1950s groundwater depletion has spread from isolated pockets to large areas in countries throughout the world. This phenomenon is well recognized, even if poorly documented in many areas. Much less recognized is how the long response time of groundwater systems can complicate the already difficult task of managing this shared resource. In fact, some assessments of groundwater depletion simply ignore capture (e.g., Wada et al. 2010).
To manage groundwater resources sustainably, it is critical to consider the time scale of the consequences. But society is poorly adapted to balance environmental issues and economic development over intergenerational timescales. Likewise, there are significant limitations in current predictive capabilities.
There is no silver bullet to address these challenges, but the following measures can help mitigate long-term problems: (1) longer groundwater policy horizons, (2) integrated monitoring and modeling, and (3) adaptive management with explicit recognition of its limitations. These measures should supplement other socioeconomic groundwater governance principles (not addressed here) such as enhanced local community involvement.
Groundwater policy horizons, when they exist, have typically been 5 to 20 years. Gradually, these time frames are being increased; for example, the Texas Water Development Board requires groundwater management areas to set goals for 50-year horizons. Gleeson and colleagues (2012) suggest setting groundwater sustainability goals for many aquifers on a multigenerational time horizon (50 to 100 years), while also acknowledging longer-term impacts. But, as we’ve illustrated, these are still much shorter than the response time of many aquifers. Perhaps the greatest value of setting longer groundwater policy horizons is in fostering greater awareness of the long-term effects of pumping.
Monitoring and Modeling
Monitoring and computer modeling are complementary activities but too often are treated separately, ignoring important linkages and feedbacks. An idealized framework for integration of monitoring and modeling includes a long-term network that is systematically monitored over time.
Models and their periodic updates can integrate new information, address questions as they arise, and advance understanding of how the aquifer system responds to human development. This is an iterative and reciprocal process: long-term monitoring provides input to modeling, and the latter provides insights into the adequacy of and gaps in monitoring data. Unfortunately, monitoring networks are rarely evaluated at the conclusion of a modeling study.
Figure 5 shows that every simulation model is built on a conceptual model. Because data typically fit more than one conceptual model (Bredehoeft 2003), the appropriateness of the model is tested as a groundwater model is built and field observations are compared to the model simulations. Reevaluation of the conceptual model is an important part of updating simulation models.
In addition to long-term monitoring networks, periodic studies should be integrated into each stage of model development (Figure 5). For example, information about the age of the water (time since recharge) and water sources obtained from environmental tracers can be compared to groundwater ages and flow paths inferred from modeling. And geologic and geophysical studies may yield new insights into the hydrogeologic framework.
The third mitigating measure, adaptive management or staged decision making, is an approach to making choices about long-term management under uncertainty. At predetermined decision points or triggered by certain indicators, management approaches are evaluated by all interested and affected parties, and choices are made about whether to proceed along the current path or reconsider next steps. For example, a minimum water level might be set below which pumping would be deemed to have a significant deleterious effect on spring flow; pumping would be curtailed if this water level were breached.
But the effectiveness of adaptive management for addressing groundwater depletion problems remains untested. There is concern that it may become a rationale for early inaction. Conditions will likely continue to deteriorate after any trigger point is reached, and it is difficult to cease pumping once it begins. For example, if a water-level indicator of spring flow is triggered, proponents of continued pumping may argue that unusually dry conditions are the cause—that it is the climate, not the pumping, that is the problem. Spring flow also illustrates the difficulty of predicting local phenomena of interest with a reasonable degree of reliability.
Groundwater development can be viewed as a continuum. At one extreme are developments that quite deliberately mine the resource, at the other are developments that can clearly be maintained indefinitely. Between these extremes are developments that seek to maximize use of the resource while allowing pumping to continue indefinitely. It is this third type of development that most challenges hydrogeologists and water managers. Once pumping of an aquifer begins, a new equilibrium state must be established in order for the pumping to be maintained indefinitely. Establishing the new equilibrium state is a dynamic process that often takes long periods of time. The time response of a groundwater system is often much slower than the planning horizons that society is used to; therein lies the dilemma faced by many attempting to manage groundwater.
The challenges in effectively managing groundwater continue to lead to groundwater developments that are oversubscribed, even with the best intentions of management to create a system that accommodates goals of sustainability. The pressure is invariably to pump more rather than less, and this is unlikely to change as population and resource needs continue to grow.
Alley WM. 2006. Tracking US groundwater: Reserves for the future? Environment 48(3):10–25.
Alley WM. 2007. Another water-budget myth: The significance of recoverable ground water in storage. Groundwater 45(3):251.
Alley WM, Leake SA. 2004. The journey from safe yield to sustainability. Groundwater 42(1):12–16.
Bredehoeft JD. 2003. From models to performance assessment: The conceptualization problem. Groundwater 41(5):571–577.
Bredehoeft JD. 2011. Report on the Hydrogeology of Proposed Southern Nevada Water Authority Groundwater Development. Prepared for Office of the Nevada State Engineer on behalf of Great Basin Water Network. Sausalito, CA: The Hydrodynamics Group.
Bredehoeft JD, Durbin T. 2009. Ground water development: The time to full capture problem. Groundwater 47(4):506–514.
Gleeson T, Alley WM, Allen DM, Sophocleous MA, Zhou Y, Taniguchi M, VanderSteen J. 2012. Towards sustainable groundwater use: Setting long-term goals, backcasting, and managing adaptively. Groundwater 50(1):19–26.
Leake SA, Hoffmann JP, Dickinson JE. 2005. Numerical ground-water change model of the C aquifer and effects of ground-water withdrawals on stream depletion in selected reaches of Clear Creek, Chevelon Creek, and the Little Colorado River, Northeastern Arizona. US Geological Survey Scientific Investigations Report 2005-5277. Reston, VA: USGS.
Lohman SW and others. 1972. Definitions of selected ground-water terms: Revisions and conceptual refinements. US Geological Survey Water-Supply Paper 1988. Washington: US Government Printing Office.
McGuire VL. 2013. Water-level and storage changes in the High Plains aquifer, predevelopment to 2011 and 2009–11. US Geological Survey Scientific Investigations Report 2012-5291. Reston, VA: USGS.
Nevada State Engineer. 1971. Water for Nevada, Report #3: Nevada’s Water Resources. Carson City: State Engineer’s Office.
Pierce SA, Sharpe JM Jr, Guillaume JHA, Mace RE, Eaton DJ. 2013. Aquifer-yield continuum as a guide and topology for science-based groundwater management. Hydrogeology Journal 21(2):331–340.
Prudic DE, Herman ME. 1996. Ground-water flow and simulated effects of development in Paradise Valley, a basin tributary to the Humboldt River in Humboldt County, Nevada: Regional aquifer system analysis—Great Basin Nevada-Utah. US Geological Survey Professional Paper 1409-F. Washington: US Government Printing Office.
Theis CV. 1940. The source of water derived from wells. Civil Engineering 10(5):277–280.
Wada Y, van Beek LPH, van Kempen CM, Reckman JWTM, Vasak S, Bierkens MFP. 2010. Global depletion of groundwater resources. Geophysical Research Letters 37:L20402, doi: 10.1029/2010GL044571.
Walton WC. 2010. Aquifer system response time and groundwater supply management. Groundwater 49(2):126–127.
World Commission on Environment and Development. 1987. Our Common Future. New York: Oxford University Press. Available at www.un-documents.net/ocf-02.htm#I.