|
|
What is Meant By “Surface and Ground Water Interaction”?
Aquifers exist beneath much of the land on which we live and work.
Ground water occurs within the pores between soil and rock particles and
in cracks and fractures in rocks. The aquifers are often partially
fed by seepage from streams and lakes. In other locations, these
same aquifers may discharge through seeps and springs to feed the streams,
rivers, and lakes. Outstanding examples of both situations can be
found in the Snake River Basin.
The Big Lost River is an example of a river feeding an aquifer.
The river flows out of a mountain valley on the northwest margin of the
Snake River Plain and entirely disappears through seepage into the permeable
lava of the Plain. The underlying Snake River Plain aquifer flows
to the southwest, ultimately discharging in the form of springs along the
wall of the Snake River canyon.
The Snake River provides an excellent example of a river fed by ground water. As the Snake River flows across southern Idaho much of the flow is diverted for irrigation. At Shoshone Falls, about 30 miles downstream of Milner Dam, the river may nearly dry up due to irrigation diversions. In the next 40 miles downstream, the river is again “reborn” in the impressive Thousand Springs area, where springs collectively discharge more than 5,000 cubic feet per second. Niagara Springs is an example of the many scenic springs in the Thousand Springs area. These river “gains” provide the majority of the downstream flow during summer.
Discharges from springs are often relatively constant, but may fluctuate
with the season and from year to year, depending upon natural weather patterns
and man-induced effects of pumping and irrigation.
Discharge of Blue Lakes Spring along the Snake River near Twin Falls
shows both seasonal and long-term variation. Much of the short and
long-term variation in the flow of Blue Lakes Spring is due to distribution
and application of water from the Snake River for irrigation and ground
water pumping.
In some situations,
river seepage (losses to the aquifer) may be affected by ground water pumping
and natural variations in aquifer water level. When the aquifer water
level is near land surface, seepage from the river is partially controlled
by the height of the aquifer water level, (see losing stream illustration).
Activities or events that result in a lowering of the water table, such
as ground water pumping, induce more seepage from the river.
Conversely, events that cause the aquifer water level to rise (recharge
events) will result in a decrease in river seepage. If aquifer water
levels rise above the level of the river, what was previously a losing
river reach will become a reach that is gaining water from the aquifer.
Another hydrologic
condition exists that is very important in understanding surface and ground
water interaction. A surface water body is “perched” above an aquifer
when aquifer water levels are well below the bed of the river, stream,
or lake (see perched stream illustration). Under these conditions,
water will seep from the surface water body to the ground water, but the
surface water body will not be affected by aquifer water levels and consequently
does not change in response to ground water pumping. Nearby ground
water pumping will cause a lowering of the water table, but will not affect
surface water supplies.
In summary, any of three conditions may exist that determine if, or
how, ground water use may affect surface water resources. These conditions
are: 1) an interconnected river (or lake) and aquifer, where the river
is losing water to the aquifer, 2) an interconnected river or lake in which
the river or lake is gaining water from the ground water, and 3) a perched
river which is losing water to the aquifer. In the first condition
river losses will increase in response to ground water pumping. In
the second condition, river gains will decrease in response to ground water
pumping. In either case, ground water pumping will result in a depletion
or capture of surface water. In the third case, ground water pumping
has no impact on surface water resources. All these conditions may
exist in the same river or lake at different locations or times of year.
What Controls
the Degree of Surface and Ground Water Interaction?
An analogy is often made between an overflowing horse trough and an aquifer. The horse trough has a continual source of water flowing in at a fixed rate. Obviously, the trough must also have an overflow that is flowing at the same rate. If a small pump is introduced into the trough and begins pumping continuously, then the overflow will soon be depleted by an amount equal to the rate of pumping. In many ground water systems, surface water supplies are ultimately depleted by an amount of water equal to volumes pumped and consumptively used. The effects of pumping on surface water sources are normally greatly attenuated relative to the horse trough analogy. The effects of pumping on surface water supplies may be distributed over years, or even decades, depending on the size and properties of the aquifer. Johnson and others (1993) demonstrates how the stream depletion effects for 30 years of continuous pumping from the Snake River Plain aquifer persist for decades after pumping ceases (see graph left).
Difficulties arise in determining the timing, location, and magnitude of the impacts. The degree to which ground water pumping depletes surface water supplies is dependent on several features of the particular basin. Considerations include: 1) the degree to which the river and aquifer are interconnected, 2) the distance between the river and the pumping source, 3) the rate of pumping, and 4) the physical characteristics of the aquifer. These factors are discussed in the following paragraphs.
The degree of river and aquifer interconnection is of great importance
in controlling the amount of surface water depletion resulting from ground
water pumping. If a river is perched above an aquifer, ground water
pumping has no effect on river flow. If the river is not perched,
but sediments have accumulated in the riverbed, or the river only slightly
penetrates into the aquifer, then the hydraulic communication between the
river and aquifer may be limited. Examples are shown in the
following illustrations: partially penetrating
river with silt deposition and fully penetrating
rivers. Spring discharge will nearly always be impacted by nearby
ground water pumping from the same aquifer.
The distance between a surface water body and a pumping location strongly
affects the timing and degree that pumping will impact stream depletion.
Pumping near an interconnected surface water body will have a nearly immediate
impact on the surface water source. The impact may be nearly equal
to the rate of ground water pumping. At greater distances, the effects
of pumping will be distributed over longer time periods and may be shared
with other hydraulically connected surface water bodies.
The rate of stream depletion associated with pumping from a given location is normally proportional to the rate of ground water pumping. If the rate of pumping from a given well is doubled, then the rate of stream depletion resulting from pumping that well also doubles. Stream depletion will be proportional to pumping rate unless aquifer water levels change so dramatically that springs are dried up, streams become perched, or aquifer properties change.
Ground water that is pumped, but not consumptively used (for example,
industrial pumping that is discharged to seepage ponds), may return to
the aquifer from which it was extracted and have little or no impact on
surface or ground water supplies outside the immediate vicinity.
Similarly, ground water pumped in excess of the amount required for crops
to grow may return to the aquifer and have little or no quantitative impact
on the surrounding resource. It is the amount of water that is permanently
extracted from the aquifer and consumptively used that is of significance.
Aquifer physical characteristics also affect the timing and magnitude
of stream depletion from pumping. Aquifer layering, water transmission,
and storage properties may have a strong influence on the direction and
rate of propagation of pumping effects. Wells completed in deeper
layers may have a more disbursed and delayed impact on surface water bodies
than wells completed in upper layers of an aquifer. Highly transmissive
aquifers with limited water storage capacity will transmit effects more
rapidly than aquifers of lower permeability or higher storage capacity.
A common misconception is that impacts of ground water pumping may be
projected along estimated flow paths through an aquifer. If this
were true, then only down-gradient streams and springs would be affected
by up-gradient pumping. In fact, the effects of ground water pumping
propagate radially in all directions (assuming aquifer properties are uniform),
regardless of the direction of ground water flow. This means that
pumping effects are felt upstream, laterally across the gradient, and downstream,
making conjunctive water rights allocation extremely difficult. In
the case of the Snake River Plain aqufier, this means that even down-gradient
pumpers have some impact on the upper river reaches.
Snake River Reach maps, which demonstrate the
effects of long-term pumping on various reaches of the Snake River, are
presented here for example.
How Can Pumping
Impacts be Measured or Estimated?
In a few cases, the impacts of ground water use on streams and springs may be measured. When this is not possible we must rely on theoretical methods to estimate the impacts of pumping on surface water resources.
Measurement
of the stream and spring depletion from ground water pumping requires controlled
testing in the field. The well of interest is turned on for a period
of relatively continuous pumping. The effects on a spring or stream
are then determined by measuring changes in the flow of the spring or stream.
This obviously requires that interference from other wells or recharge
activities be minimized during the test period.
The impacts of ground water pumping on surface water bodies can be measured in relatively few situations. Measurable impacts require situations where there is little interference from other wells, close proximity of the pumping location and the surface water, and a pumping rate that is a measurable proportion of stream or spring discharge. If nearby wells exist that have a capacity similar to that of the well of interest, and these wells are operating, it becomes difficult or impossible to isolate the effects of a single well on a nearby spring or stream. If the well of interest is a great distance from the spring or stream (perhaps a few miles or more), the effects may be so greatly attenuated that testing becomes infeasible. If the pumping rate is small relative to the discharge of the stream or spring, then changes in the spring or stream flow may be immeasurably small. Any of these conditions make the measurement of depletion due to pumping impracticable and require that theoretical methods be employed.
Ground water flow theory can be applied to estimate spring or stream depletion in several ways. Some of the most simple techniques were developed by Jenkins (1968) and Glover (1968). These methods use an analytical equation to calculate graphs that generically relate pumping to depletion. Depletion rates are a function of aquifer transmissivity and storativity and the distance between the stream and well. The depletion is assumed to be proportional to the pumping rate. Application of these methods involves numerous simplifying assumptions including: 1) the aquifer is infinite except where cut by the stream, 2) the stream fully penetrates the aquifer, 3) the aquifer characteristics are uniform, 4) the stream is approximately straight, and 5) aquifer thickness is not significantly changed by pumping. When these assumptions are too restrictive, more sophisticated methods must be employed.
Numerical modeling of stream depletion avoids many of the assumptions required in the more simple analytical techniques. Numerical modeling allows us to incorporate all of our understanding (which may be incomplete or flawed) of the real system into the process of calculating depletion. The sophistication of the model should be commensurate with the level of our understanding of the real system. Simple numerical models supported by limited data may still give more accurate estimates than the described analytical methods. Numerical modeling, however, does require more effort than application of the analytical techniques. In cases where entire aquifers are considered, and multiple reaches of streams are interconnected with the aquifer, the modeling process may become very complicated.
In some cases, the results of numerical models
may be generically represented by response functions or coefficients.
The response functions are determined from numerical models and quantify
the relationship between pumping and stream depletion for specific pumping
locations and stream reaches. Response functions are being determined
to relate depletion of surface flows of the Snake River due to ground water
pumping of the Snake River Plain aquifer. For more information on
response functions click
here.
How do Surface
Water Bodies Respond to Aquifer Recharge?
The previous discussion has focused on the impacts of ground water
pumping on stream and spring discharge. There is an equal and opposite
effect for conditions of aquifer recharge on spring and river flow.
As ground water pumping serves to deplete stream flow, aquifer recharge
enhances stream flow. All of the previous discussion related to pumping
also relates to recharge.
For more on recharge, see "Recharge of the
Snake River Plain Aquifer: Transitioning from Incidental to Managed."
