Capture Versus Capture Zones: Clarifying Terminology Related to Sources of Water to Wells

Abstract

The term capture, related to the source of water derived from wells, has been used in two distinct yet related contexts by the hydrologic community. The first is a water‐budget context, in which capture refers to decreases in the rates of groundwater outflow and (or) increases in the rates of recharge along head‐dependent boundaries of an aquifer in response to pumping. The second is a transport context, in which capture zone refers to the specific flowpaths that define the three‐dimensional, volumetric portion of a groundwater flow field that discharges to a well. A closely related issue that has become associated with the source of water to wells is streamflow depletion, which refers to the reduction in streamflow caused by pumping, and is a type of capture. Rates of capture and streamflow depletion are calculated by use of water‐budget analyses, most often with groundwater‐flow models. Transport models, particularly particle‐tracking methods, are used to determine capture zones to wells. In general, however, transport methods are not useful for quantifying actual or potential streamflow depletion or other types of capture along aquifer boundaries. To clarify the sometimes subtle differences among these terms, we describe the processes and relations among capture, capture zones, and streamflow depletion, and provide proposed terminology to distinguish among them.

Introduction

Many groundwater studies are concerned with the determination of the source of groundwater pumped by wells. Such determinations focus mainly on two distinct yet related questions, the first being “What are the impacts of groundwater pumping on aquifer storage and the inflows and outflows of water along boundaries of the groundwater system?” and the second “What portion of the groundwater system contributes water discharged by pumping wells?” Reilly and Pollock (1995, 1996) defined the first question as the “water‐budget context” of the source of water to wells and the second as the “transport context.” The water‐budget context is concerned with how rates of release of water from aquifer storage and flow rates into and out of the aquifer along certain types of boundaries are changed in response to a volumetric rate of withdrawal at a well; the water‐budget context says nothing about which water particles flow to and are discharged by a well or about their particular flowpaths. In fact, water‐budget analyses of capture can be carried out without knowledge of directions and rates of groundwater flow (Leake 2011). Conversely, the transport context is concerned with the specific flowpaths of water from groundwater source locations to a discharging well; in most cases, transport analyses are less focused on the effects of pumping on system‐wide water budgets (Reilly and Pollock 1995, 1996). Although the distinction between the two contexts was noted more than 20 years ago, in our experiences with stakeholders and at training classes we have found that there continues to be some confusion among hydrologists, water‐resource managers, and the public regarding which process is being addressed. Furthermore, we know of no publications other than Reilly and Pollock (1995, 1996) that compare and contrast differences between these concepts. This confusion stems in part from the development over time of a similar terminology used in conjunction with the two processes—“capture” for the water‐budget context and “capture zones” for the transport context. Although the concept of “capture” predates that of “capture zones,” in our experience hydrologists tend to have a better understanding of the latter than the former.

An important water‐resource and environmental issue that is closely related to the source of water to wells is that of the effects of groundwater pumping on nearby streams and springs and their associated ecosystems. Groundwater pumping reduces streamflow, a process that is commonly referred to as streamflow depletion by wells. In addition to reducing the amount of flow in streams, streamflow depletion can change the mix of the contributions of groundwater discharge (baseflow) and surface‐water runoff to a stream, with associated changes in water quality and stream temperature and consequent impacts on aquatic and riparian ecosystems. Because many source‐of‐water investigations are primarily concerned with the effects of groundwater pumping on streamflow depletion, the processes and terms associated with the source of water to wells have become intertwined with those of streamflow depletion.

The purpose of this article is to contrast and attempt to clarify the processes, terminology, and relations among capture, capture zones, and streamflow depletion by wells. The next two sections provide background on the water‐budget and transport contexts, including discussions of the modeling approaches typically used for each type of analysis and examples that illustrate some of the important concepts associated with each process. These background discussions are followed by a few comments on existing terminology and a suggested terminology to describe the various processes.

Water‐Budget Context: System‐Wide Changes in Aquifer Storage and Boundary Flows (Capture)

The system‐wide effects of groundwater pumping by wells have been described in the literature for more than 100 years, beginning at least as far back as the study of Lee (1915) in the Owens Valley of California. Lee (1915) and later C. V. Theis in his 1940 paper “The Source of Water Derived from Wells—Essential Factors Controlling the Response of an Aquifer to Development” (Theis 1940) were clearly concerned with the water‐budget context of the source of water to wells. Theis, for example, noted in his summary (p. 280) that “All water discharged by wells is balanced by a loss of water somewhere.” Theis described the physical factors of the aquifer and pumping well that affect this balance, including conditions along the boundaries of an aquifer, and emphasized the importance of time to the source of water to a discharging well. As described by Theis, initially all of the water discharged by a well is derived from a decrease in aquifer storage. As pumping continues, the cone of depression that forms in the water surface around the well will expand outward until it reaches one or more boundaries of the aquifer from which recharge to the aquifer can be increased or discharge from the aquifer can be decreased. At equilibrium, a new steady‐state condition will be reached in which there are no further declines in aquifer storage and discharge at the well is balanced completely by increases in aquifer recharge, decreases in aquifer discharge, or a combination of the two. Although Theis did not use the term “capture,” it was subsequently adopted to describe these pumping‐induced decreases in discharge and increases in recharge and has now been used for decades. Lohman et al. (1972) defined the term as follows:

“Water withdrawn artificially from an aquifer is derived from a decrease in storage in the aquifer, a reduction in the previous discharge from the aquifer, an increase in the recharge, or a combination of these changes… The decrease in discharge plus the increase in recharge is termed capture.”

Capture occurs in many forms. The primary sources of captured discharge are groundwater that would have flowed to streams, springs, or the oceans in the absence of pumping, or have been evapotranspired from the water table in low‐lying areas such as riparian zones and wetlands. Examples of captured recharge are induced infiltration of streamflow and infiltration of “rejected recharge.”11 That is, water that recharges an aquifer where the water table was previously at or near land surface for at least part of the time but has been lowered by pumping, thereby providing pore space that was previously fully saturated and through which recharge can occur.

Capture can occur only at boundaries at which pumping‐induced groundwater‐head changes cause changes in inflow to or outflow from the aquifer. Those boundaries commonly are surface water in hydraulic connection with the groundwater system, areas of groundwater evapotranspiration, and springs. In groundwater analyses, these boundaries are referred to as head‐dependent boundaries. Capture cannot occur along no‐flow boundaries or where water enters the aquifer through an unsaturated zone. An exception to the latter case is when a water table at land surface is lowered to allow infiltration of previously rejected recharge.

Streamflow depletion is a unique type of capture because it can consist of both captured groundwater discharge to a stream and captured recharge in the form of induced infiltration of streamflow into the aquifer. Both captured groundwater discharge and induced infiltration of streamflow lead to streamflow reductions; streamflow depletion is therefore equal to the sum of the two pumping‐induced processes. Capture that is a reduction in the rate of groundwater inflow to a stream reduces the baseflow of the stream. Capture in the form of induced infiltration likely includes a component of runoff entering the aquifer. Although there is no difference in the change in streamflow caused by either a reduced rate of inflow of groundwater to the stream or an increased rate of induced infiltration from the stream, a distinction may be important for analyses of water chemistry or temperature.

Many factors affect the timing and rates of capture in response to pumping from a particular well, including the geologic structure, dimensions, and hydraulic properties of the groundwater system, and the locations and hydrologic conditions along the boundaries of the system, including proximity of streams (Leake 2011; Barlow and Leake 2012). An important and often misunderstood concept related to streamflow depletion is the effect of the ambient (i.e., prepumping) rates and directions of groundwater flow on total streamflow depletion and on the individual components of streamflow depletion. Specifically, although the rates and directions of ambient groundwater flow do not affect the timing or rates of total streamflow depletion caused by pumping, they do affect the individual components of captured groundwater discharge and induced infiltration of streamflow. A consequence of this is that although the rate of total streamflow depletion is independent of the recharge rate to an aquifer, the individual components of streamflow depletion are not. For example, although higher rates of recharge to a particular aquifer will not lower the total rate of streamflow depletion caused by pumping at a well, higher rates of recharge will result in lower rates of induced infiltration of streamflow relative to the rates of captured groundwater discharge by the well (see figure 28 in Barlow and Leake 2012, p. 37). Of course in many systems, particularly those in humid settings, groundwater recharge sustains streamflow in the form of baseflow provided by groundwater discharge; consequently, the recharge rate to an aquifer ultimately affects the amount of streamflow that is available for capture. Moreover, in a gaining stream network, more baseflow is available for capture as the drainage area of the system increases downstream toward the outflow point of the basin.

There are important links between the quantity and timing of the reductions in baseflow and total streamflow due to groundwater pumping and the related field of environmental flows, which is concerned with the hydrological regime required to sustain freshwater and estuarine ecosystems (Acreman 2016). The environmental‐flows literature is extensive and the topic beyond the scope of this paper; Poff et al. (2010), Acreman (2016), and Mott Lacroix et al. (2017) describe the scientific frameworks and specific methods used for environmental‐flow assessments and the critical role of groundwater in sustaining freshwater ecosystems.

Water‐budget components that are affected by groundwater pumping most frequently are determined by use of analytical or numerical groundwater‐flow models. An example of an analytical model commonly used to determine time‐dependent streamflow capture is the “Glover” or “Jenkins” solution (Glover and Balmer 1954; Jenkins 1968), which quantifies time‐variant rates and volumes of streamflow depletion in a single bounding stream in response to pumping from a single well withdrawing water at a constant rate. The solution calculates total streamflow depletion and cannot be used to differentiate between captured groundwater discharge and induced infiltration or to determine the effects of the withdrawal on multiple nearby streams.

Because of the inability of analytical models to account for conditions typically found in real‐world settings, most water‐budget analyses are done with numerical groundwater‐flow models such as MODFLOW (Harbaugh 2005). Numerical models can calculate pumping‐induced changes to the inflow and outflow rates to aquifer storage and to all simulated boundary features simultaneously (streams, springs, drains, evapotranspiration, and so forth), either directly by use of superposition (or “change”) models or, more typically, in a three‐step process by use of models that have been calibrated to specific field conditions (Leake et al. 2010; Leake 2011). In the first step, a simulation is made to establish baseline values of all water‐budget components in the absence of pumping from the well (or wells) of interest. Such baseline conditions might represent predevelopment conditions throughout a groundwater system prior to any groundwater pumping, or conditions prior to initiation of pumping at a specific well or group of wells. In the second step, the simulation is rerun with no other changes than the added withdrawal or withdrawals at the well or wells of interest. In the final step, changes in the flow rates to and from aquifer storage and the simulated boundary features are determined by subtraction of the pumping‐affected flow values from the baseline values. Using, for example, MODFLOW as the groundwater‐flow simulator, rates of capture and aquifer‐storage change can be determined for the entire model domain from the budget table provided at the end of each time step or stress period. Alternatively, capture rates can be determined for specific boundary features such as a particular stream, spring, or agricultural drain. For the case of streamflow depletion, detailed water‐budget analyses can be done for individual model cells (or for groups of cells) to quantify the individual components of streamflow depletion or to differentiate streamflow depletion from other sources of capture (Konikow and Leake 2014; Ahlfeld et al. 2016; Feinstein et al. 2016; and Nadler et al. 2017).

Many groundwater models are not fully integrated with watershed models capable of simulating overland runoff and interflow, both of which can contribute to total streamflow. Moreover, many models treat the stream boundary as a specified head or do not track the amount of flow within simulated streams and stream reaches. As a consequence of the limitations of these modeling approaches, total streamflow available for depletion either does not include overland runoff and interflow or can exceed the amount that would be available if simulated streams were allowed to go dry. These limitations are particularly relevant to first‐order (headwater) streams that have relatively little flow and are most vulnerable to pumping (Feinstein et al. 2016; Fienen et al. 2016). In some cases, fully integrated groundwater/surface‐water models such as ParFlow (Kollet and Maxwell 2006) and GSFLOW (Markstrom et al. 2008) that simulate overland runoff and interflow may be required to fully understand the sources of water to pumped wells and the effects of pumping on groundwater‐dependent ecosystems. For example, integrated models can be used to simulate capture of rejected recharge.

Numerical simulation of the effects of pumping on aquifer‐storage changes and capture has found many uses in groundwater‐system analyses and management, including generation of response functions and capture maps. Although response functions have been defined and used in different ways and are referred to by different names, all response functions have the common characteristic that they represent a change in the state of a system variable (such as total capture, streamflow, or aquifer storage) that results from a change in pumping rate at a single well or group of wells, independently of other pumping or recharge stresses that may be occurring simultaneously within an aquifer. Fienen et al. (2017) introduced the term depletion potential to refer specifically to the ratio of model‐calculated potential reduction in streamflow resulting from pumping at a unit rate from an existing or proposed well or group of wells.

Over the past decade, capture maps have received interest as a water‐management and educational tool by hydrogeologists, resource managers, stakeholder groups, and the public. Capture maps display the spatial distribution of expected capture fractions (i.e., the ratio of capture to simulated pumping rate) from selected boundary features and times of interest from theoretical pumping wells distributed over large regions of a groundwater system (Leake et al. 2010). They can be used to illustrate the effects of pumping location on capture within a large set of possible pumping locations within an aquifer. Two general types of capture maps can be created: “Global capture maps” characterize total capture from all boundaries within an aquifer; alternatively, “local capture maps” characterize capture from a particular feature within a basin such as streamflow depletion from a particular stream, stream tributary, or stream segment. Feinstein et al. (2016) and Fienen et al. (2016) have extended this concept of local capture by defining a local area near a well over which the pumping effect on surface water is most concentrated; in their work, local sources of capture can comprise multiple streams and other water bodies within a defined area around each well, as opposed to a single surface‐water feature.

An example of a global capture map is shown on Figure 1 for the lower basin‐fill aquifer within the upper San Pedro River Basin of southeastern Arizona and northern Sonora, Mexico. Capture within the simulated area consists primarily of changes in streamflow in the San Pedro and Babocomari Rivers but also includes minor components of reductions in groundwater evapotranspiration and groundwater discharge to springs. The map shows total capture from these boundary features as a fraction of the pumping rate at each location after 50 years of pumping. The map was constructed by running a groundwater‐flow model developed for the basin hundreds of times; in each run, pumping was simulated at a single well and the effects of that pumping on total capture were determined by use of the three‐step process described previously. As expected, capture fractions are generally largest for wells pumping near the rivers (orange to red areas in Figure 1).

Figure 1

Example of global capture map for the upper San Pedro River Basin, Arizona, and Sonora, Mexico. In this example, capture consists of the total change in streamflow, riparian evapotranspiration, and springflow throughout the basin that would result for pumping at a constant rate for 50 years from a well located in the lower basin‐fill aquifer within the mapped area. The color at any location represents the fraction of the pumping rate by a well at that location that can be accounted for as changes in outflow from and (or) inflow to the aquifer for model boundaries representing streams, riparian vegetation, and springs (reproduced with permission from Leake et al. 2008).

Transport Context: Flowpaths, Capture Zones, and Contributing Areas

Since the mid‐1980s, concurrently with the 1986 Amendments to the Safe Drinking Water Act and enactment of the U.S. Environmental Protection Agency's Wellhead Protection Program (U.S. Environmental Protection Agency 1987), the term source of water derived from wells often has been used in the transport context, referring specifically to the determination of the flowpaths of water from their points of entry into a groundwater system to a discharging well. An extensive terminology has developed to describe both the areal extent and volumetric portion of a groundwater system that contributes discharge to a particular well, including “contributing recharge area,” “wellhead protection area,” “zone of contribution,” and “capture zone.” The term that has become most widely accepted in the literature is “capture zone,” which is defined as the three‐dimensional, volumetric portion of a groundwater flow field that discharges water to a well (Anderson et al. 2015, p. 360; Figure 2). Capture zones often are delineated with respect to a specific period of travel time to a well (Zheng and Bennett 2002)—such as the 5‐year capture zone—particularly for wellhead protection and regulation (Rayne et al. 2014; Anderson et al. 2015). The two‐dimensional areal extent of that portion of a capture zone that intersects the water table or surface‐water features is most often referred to as the contributing area (or area contributing recharge) to the well (Figure 2); a related term is wellhead‐protection area, which is frequently used when there is a regulatory aspect to the area. The two‐ and three‐dimensional portions of the aquifer that comprise a well's contributing area and capture zone, respectively, are schematically contrasted in Figure 2 with the capture of streamflow along the aquifer's boundary in response to pumping at the well.

Figure 2

Capture zone and contributing area for a single discharging well in a simplified hypothetical groundwater‐flow system. The capture of groundwater discharge to the bounding stream in response to pumping is illustrated by the dashed sections of the groundwater‐flow arrows at the stream. (Figure modified from Paschke et al. 2007.)

Capture zones and contributing areas are determined by use of transport models, particularly particle‐tracking (advective‐transport) models such as MODPATH (Pollock 1989, 2016) that track the advective component of water particles through a simulated aquifer. Other transport‐modeling techniques, such as volumetric‐tracking methods (Potter et al. 2008; Black and Foley 2013; Foley and Black 2013) and solute‐transport models that simulate both advection and dispersion (e.g., Sousa et al. 2013; Okkonen and Neupauer 2016) also have been used. In particle tracking, groundwater heads calculated by a flow model are used in conjunction with aquifer properties to determine the velocity field throughout the model domain; particles are then tracked within the model domain according to their average linear velocities. Particles can be tracked in either the forward or reverse flow direction. In forward tracking, particles are tracked in the direction of groundwater flow from their point of entry into the groundwater system (i.e., their source location) to a pumping well; in reverse tracking, particles are tracked in the direction opposite to flow from a pumping well to their source location. Detailed descriptions of the theory and application of particle‐tracking techniques are provided by Zheng and Bennett (2002) and Anderson et al. (2015). Although particle tracking is frequently used to determine capture zones and contributing areas, as a general rule it is not a useful method for analysis of capture (in the water‐budget context) or of capture maps, such as that for the upper San Pedro River Basin shown on Figure 1.

Capture zones frequently intercept streams, lakes, and other surface‐water bodies. For conditions in which a well located near a stream or river pumps at a rate that is large enough to capture induced infiltration of streamflow, the quality of the induced surface water will affect the quality of water in the underlying aquifer, and possibly that of the pumped well itself. Newsom and Wilson (1988) developed an analytical model to determine the fractions of captured groundwater discharge and induced infiltration discharged by a well; they also presented a terminology to describe the stream and aquifer capture zones for the case in which a well captures part of its discharge by induced infiltration from a nearby stream. As illustrated on Figure 3, water discharged by a well can consist of both induced infiltration within the “stream capture zone” and captured groundwater discharge (which they refer to as ambient groundwater flow) within the “aquifer capture zone.” The third zone shown on Figure 3 is the “zone of induced throughflow,” which consists of streamflow that has been drawn into the aquifer by induced infiltration but does not actually reach the well; this water flows back to the stream (or possibly another discharge location from the groundwater system) downgradient from the well. The factors that affect the relative amounts of induced infiltration and ambient groundwater discharge captured by a particular well are the distance of the pumping well from the stream, the direction and rate of ambient groundwater flow toward the stream, and the pumping rate of the well.

Figure 3

Flow net around a well pumping from a homogeneous, two‐dimensional groundwater‐flow system bounded by a fully penetrating stream. The well captures part of its discharge from ambient groundwater flow within the “aquifer capture zone” and the remainder from induced infiltration of streamflow within the “stream capture zone.” The “zone of induced throughflow” consists of induced infiltration that flows back to the stream downgradient from the well. (Flow net based on the analytical model developed by Newsom and Wilson 1988 for an initially gaining stream and ambient groundwater‐flow direction that is oriented at an angle 60° counter‐clockwise from the horizontal axis. The stream is losing between points B and C and gaining between points C and D. Figure modified from Newsom and Wilson 1988.)

Particle‐ and volumetric‐tracking methods can be used as alternatives to the analytical model developed by Newsom and Wilson (1988) to identify and in some cases quantify the contributions of captured groundwater discharge, induced infiltration of streamflow, and induced throughflow to a well. With particle tracking, particles are tracked from source locations at the water table and bounding streams to discharging wells. For models that do not include weak sinks (see Anderson et al. 2015, pp. 349 to 351), volumetric‐flow rates can be assigned to individual particles based on either the recharge rate specified to model cells for particles that originate at the water table or the streamflow‐leakage rate from stream cells for particles that originate in losing stream reaches. The contributions of captured groundwater discharge and induced infiltration to each pumping well can then be determined by summing the individual flow rates of particles captured by the well from each source location. Potter et al. (2008), Foley and Black (2013), and Black and Foley (2013) describe the use of MODALL and FlowSource for volumetric‐tracking approaches to identify and quantify the source of water derived from wells; Fienen et al. (2017) describe the use of FlowSource to identify contributing areas to wells and the Little Plover River in central Wisconsin. In these volumetric techniques, the individual model‐cell water budgets are used to track fluxes throughout the model domain for either steady‐state or transient conditions. With an identified source or sink location, all fluxes to or from the feature can be tracked throughout the model domain efficiently without the need for particle tracking. The result is a map of the model domain expressing the fraction of water in each model cell that emanated from an identified source or left the model through an identified sink.

There is an important implication to the presence of a zone of induced throughflow that points to differences between the water‐balance and transport contexts in the accounting of the sources of water derived from a well. Specifically, as illustrated on Figure 3, it cannot be assumed that the rate of induced infiltration from a stream as calculated from the changes in flow rates (water balance) along the stream‐aquifer boundary will be equivalent to the rate of induced infiltration discharged by a well; the amount of induced infiltration that is actually captured by a well must be determined by one of the transport‐modeling approaches listed previously. Moreover, because the groundwater and surface water may be distinct geochemically, the relative proportions of captured groundwater discharge and induced infiltration that are captured by a well ultimately will affect the quality of water pumped by the well.

Another example of how a transport analysis can help to understand and visualize the individual components of streamflow depletion is that described by Fienen et al. (2017), who found it useful to expand upon the concept of captured groundwater discharge to differentiate between groundwater that is directly intercepted and discharged by a pumping well from groundwater that is diverted away from a stream but not actually discharged by the pumping well. Figure 4A, for example, shows two hypothetical wells located near a tributary stream. Well A is located very close to the tributary (and within the prepumping contributing area to the tributary). The ratio of the steady‐state rate of streamflow depletion in the tributary to the pumping rate at the well (i.e., the depletion potential of the well to the tributary) is 1.0; that is, the rate of streamflow depletion in the tributary will be 100% of the pumping rate at the well. In contrast, well B is located some distance downgradient to the tributary. Well B's impact on the tributary is less than that of well A; the depletion potential of the well to the tributary is 0.25 (i.e., the rate of streamflow depletion in the tributary will be 25% of the pumping rate of the well and the remaining 75% occurs along the main stem and other tributaries to the river). Figure 4B illustrates contributing areas to the two wells and to the tributary for hypothetical pumping rates of 10 L/s at each well. The contributing area to well A is completely within the prepumping contributing area to the tributary—all of the groundwater that is discharged by the well consists of intercepted groundwater discharge that would have discharged to the tributary in the absence of pumping. In contrast, the contributing area to well B is completely outside the prepumping contributing area to the tributary, even though pumping at the well has changed the flowfield near the tributary and diverted groundwater that previously discharged to the tributary to another sink (in this case, the main stem and other tributaries to the river). Fienen et al. (2017) refer to the zone in which water is diverted away from the tributary as the “zone of diverted discharge,” and provide a real‐world example of the phenomenon for the Little Plover River in central Wisconsin. Figure 4 illustrates that contributing areas cannot be used alone to determine whether a particular well will deplete streamflow in a particular stream, reach, or tributary; calculations of depletion potential are needed to fully understand a well's impact on streamflow.

Figure 4

A. Contributing area to a tributary stream for prepumping conditions. B. Contributing areas to the tributary and to wells A and B for pumping at rates of 10 L per second at each well. Pumping at well B has created a “zone of diverted discharge,” which is groundwater that discharged to the tributary for nonpumping conditions but that is diverted away from the tributary because of changes in the flowfield caused by pumping at the wells. None of the groundwater within the prepumping contributing area to the tributary is discharged by well B. (Figure based on concepts presented in Fienen et al. 2017.)

Comments on Existing Terminology

The preceding sections have attempted to describe the historical development and current uses of the primary terms associated with the water‐budget and transport contexts of the sources of water derived from wells, and to clarify the relations among terms used to describe processes that occur in response to pumping at the well itself from those that occur within and along the boundaries of an aquifer. The discussions noted that a similar terminology has developed around the two contexts and has led to some confusion related to the particular processes that are being analyzed; specifically, capture for the water‐budget context and capture zone for the transport context.

The term capture refers to changes in the flow rates into and out of head‐dependent boundaries of an aquifer in response to pumping. In contrast, capture zone refers to the three‐dimensional, volumetric portion of a groundwater‐flow field that discharges water to a well. The authors maintain that capture is a well‐established term that clearly describes the physical processes that occur along the boundaries of an aquifer in response to pumping. Moreover, capture is a broad term that describes the many forms in which changes in boundary flows can occur, and is therefore useful for differentiating between the general process of capture and the specific forms in which capture is manifested in a particular basin. Davids and Mehl (2014), for example, differentiate between a “sustainable capture threshold” that quantifies the total amount of water that can be captured from all sources and sinks within a system from “sustainable capture fractions” that quantify the acceptable limits of the effects of pumping on each individual capture component (e.g., streamflow depletion or groundwater evapotranspiration) within a system.

A process closely related to the sources of water to wells is streamflow depletion. In contrast to the general term capture, streamflow depletion refers specifically to changes in flow rates into and out of an aquifer at stream and river boundaries. Streamflow depletion can consist of both reductions in the outflow rates of groundwater discharge from an aquifer and increases in the inflow rates of streamflow to an aquifer. Traditionally, reductions in groundwater outflow rates have been referred to as captured groundwater discharge and increases in the inflow rates as induced infiltration of streamflow. However, it can be helpful to differentiate between the reductions in groundwater outflow rates caused by groundwater that is actually intercepted and discharged by a pumping well (i.e., groundwater within the capture zone of the well) from reductions caused by groundwater that is diverted away from the stream, stream segment, or tributary but not actually withdrawn at the well.

Suggested Terminology

With these issues and historical development in mind, we propose the following definitions for the terms most often used in the context of the source of water to wells and streamflow depletion. In our opinion, the terms provide a comprehensive vocabulary that is consistent with current usage.

Capture: pumping‐induced increases in aquifer recharge, decreases in aquifer discharge, or a combination of the two at head‐dependent boundaries of an aquifer.

Capture analysis: a water‐budget analysis of pumping‐induced increases in aquifer recharge, decreases in aquifer discharge, or a combination of the two at head‐dependent boundaries of an aquifer.

Streamflow depletion: a reduction in streamflow caused by groundwater pumping. The total amount of streamflow depletion equals the sum of the reductions in the outflow rate of groundwater from an aquifer (captured groundwater discharge) and increases in the inflow rate of streamflow to an aquifer (induced infiltration of streamflow). Reductions in groundwater discharge can consist of both groundwater that is intercepted and discharged directly by pumping wells and groundwater that is diverted away from a stream, stream segment, or tributary to another sink. Although streamflow depletion refers specifically to streams, its use can include depletion of flow in other surface‐water features including rivers, springs, and lakes.

Depletion potential: the ratio of model‐calculated potential reduction in streamflow resulting from pumping at a unit rate from an existing or proposed well or group of wells.

Capture zone: the three‐dimensional volumetric portion of a groundwater‐flow field that discharges water to a well.

Capture‐zone analysis: a transport analysis, often by use of particle‐tracking methods, of the three‐dimensional portion of a groundwater‐flow field that discharges water to a well.

Contributing area (or area contributing recharge): the two‐dimensional areal extent of that portion of a capture zone that intersects the water table and surface‐water features where water entering the groundwater‐flow system is discharged by a well.

Zone of induced throughflow: induced infiltration of streamflow that has been drawn into the aquifer but does not actually reach the well; this water flows back to the stream (or other discharge location from the groundwater system) downgradient from the well.

Zone of diverted discharge: that part of the capture zone to a boundary sink (such as a stream or stream tributary) in which water that would have discharged to the sink in the absence of pumping is diverted to another boundary sink.

Conclusions

An extensive and somewhat confusing terminology has developed around discussions of the source of water to wells. Some of this confusion has resulted from the different uses of the term capture. In the water‐budget context, capture refers to changes in the inflow and outflow rates to and from an aquifer in response to pumping. In the transport context, a capture zone refers to the three‐dimensional, volumetric portion of a groundwater‐flow field that discharges water to a well. A closely related water‐resource and environmental issue that is often associated with the source of water to wells is streamflow depletion, which is the reduction in streamflow caused by pumping. Streamflow depletion is a type of capture that consists of the sum of the reductions in the outflow rates of groundwater from an aquifer and increases in the inflow rates of streamflow to an aquifer in response to pumping. In any discussion of the sources of water to wells, it is important that the processes being described are clearly articulated from the perspective of either the well or impacted boundary feature.

Rates of capture and streamflow depletion are calculated by use of water‐budget analyses of the changes in the inflow and outflow rates of water along head‐dependent boundaries of an aquifer that occur in response to pumping, most often by use of numerical groundwater‐flow models. Transport models, particularly particle‐tracking methods, are used to determine capture zones and contributing areas to wells; they also may be used to identify and in some cases quantify the amounts of captured groundwater discharge and induced infiltration that are actually discharged by a well. In general, however, transport methods are not useful for quantifying actual or potential streamflow depletion or other types of capture along aquifer boundaries. Analyses of capture in the water‐budget context and capture zones in the transport context are both useful, but for different purposes. If the question at hand involves quantifying groundwater‐pumping induced changes in flow in connected surface‐water features, capture analyses are used. Capture‐zone analyses using transport models cannot easily quantify changes in flow to surface features in all cases. On the other hand, capture‐zone analyses are useful for analyses involving possible movement of dissolved chemical constituents to particular wells, as well as developing an understanding of induced throughflow and diverted discharge.

Acknowledgments

We thank our USGS colleagues William Cunningham and Daniel Feinstein, Abraham Springer, and two anonymous reviewers for their insightful comments on earlier drafts of this paper. We also thank John Wilson, New Mexico Institute of Mining and Technology, for permission to use Figure 3.

1 That is, water that recharges an aquifer where the water table was previously at or near land surface for at least part of the time but has been lowered by pumping, thereby providing pore space that was previously fully saturated and through which recharge can occur.