Unconsolidated Aquifers

Erosion and deposition are two earth processes that shape much of the surface and near surface of the Earth.  Much of the Earth’s near-surface material consists of geologically recent deposits that are unconsolidated (loosely arranged or consisting of particles are not cemented together).  Coarsely grained near-surface deposits are among the most important aquifers because they are an active part of the water cycle and because they generally yield water more readily than their consolidated equivalents.   Most unconsolidated aquifers are composed of sand and gravel that was carried by water either in streams flowing from runoff or from melting ice, especially over the last three million years during continental glaciation.

 Unconsolidated aquifers can be grouped as follows:

1. Sand and gravel aquifers of glacial origin that include:
   1. Outwash plains deposited near the end of continental glaciation
   2. Coarse-grained deposits buried beneath till, including buried stream networks
2. Unconfined aquifers along rivers and streams, some of which are associated with glacial meltwater runoff
3. Basin-fill and blanket alluvial aquifers formed by erosion of mountain slopes

Discussion Questions:

After going through the following text and graphics, answer the following questions about aquifers and their connections to geology:

Question 1: Briefly summarize the processes that create unconsolidated and semi-consolidated sand and gravel aquifers. Describe the implications in terms of hydrologic properties of these types of aquifers.

Question 2: Choose 5 features from the following list of aquifer and aquitard geologic settings, and briefly explain the processes that produce the following aquifer and aquitard features, paying special attention to the energy with which the sediments are deposited, the environment in which they are deposited, the manner of deposition, sequence of deposition, and/or sorting processes relevant to the landforms produced. Discuss implications in permeability, storage properties, variability, trends, and anisotropy. 

• valley/basin fills
• alluvial fan
• floodplain sediments,
• lake bottom sediments
• lacustrine deposits 
• channel/river deposits
• braided river deposits
• braided bars
• point bars
• meandering river deposits
• floodplain deposits
• overbank deposits
• upland deposits fluvial deposits fining upwards
• glacial sediments
• glacial outwash
• proglacial outwash
• ice contact outwash
• outwash plain
• esker 
• kettles
• kames
• drumlins
• end moraine or terminal moraine
• lateral moraine
• ground moraine

Sand and Gravel Aquifers of Glacial Origin

Figure 1 -- Sand and gravel aquifers of glacial origin are north of the line of continental glaciation.  U.S. Geological Survey (https://water.usgs.gov/ogw/aquiferbasics/sandgravel.html) [accessed 4/03/2020]

Glacial Outwash

Outwash plains are extensive areas of glaciofluvial sediments deposited in the proglacial zone (Figure 2). They occur at the margins of ice sheets or piedmont glaciers, distinct from the valley trains or valley bottom deposits that occur in front of valley glaciers in confined settings.

Sediment is deposited primarily by braided meltwater streams that migrate across the surface of the outwash plain.  During floods, previously deposited sediments may be reworked and redeposited.  Although they are commonly highly productive aquifers, outwash plains have poorly developed soils and tend not to be used either for agriculture or for urban development (Figure 3).

Figure 2 -- Schematic of melting continental glacier showing an outwash plain (from Indian Academy of Science)

Because the ice sheets would commonly advance and retreat, in some locations, outwash sediments could be overrun by ice but not be scoured.  Thus, it is common in till plains to find outwash or other glaciofluvial deposits buried under till from advancing ice.  This results in confined permeable sand and gravel units that form localized aquifers referred to as inter-till or intra-till aquifers. 

Outwash plains are associated with areas where the continental ice sheets stagnated (also called ice still stands or ice marginal positions).  These stagnation places and their resulting moraines are most often related to the bedrock surface beneath the moraine.  Additional ice buildup was needed to allow the ice to move over or around an obstacle such as bedrock that was not easily eroded.  Thus, knowledge of the bedrock topography and rock type is needed to comprehend the glacial ice movement and resulting deposition.

Figure 3 -- View of an outwash plain in central Wisconsin (Photo from Wisconsin Department of Natural Resources).

Outwash plains also occur in valley glacial settings but demand for water in these settings precludes them from being commonly used aquifers (Figure 4).

Figure 4 -- Valley outwash downstream of Red Glacier, Alaska.  (Photo from U.S. National Park Service).

Buried Glacial River Valleys

          Ancient river valleys that were buried by advancing continental glacial ice are a truly “hidden” groundwater resource.  They were formed when previously developed riverbeds were overrun by continental glaciers thus burying the old river valley beneath till which acts as a confining layer for permeable sand and gravel units in the buried valley.  One of the best understood buried valleys is the Teays River in Illinois, Indiana, Ohio, Kentucky, West Virginia, and Virginia in the United States (Figure 5).  This old river system was probably formed as an ice marginal river during an earlier continental glacial advance.

Figure 5 – Buried valley aquifer in the central United States (from nmanchesterhistory.org) [accessed 4/03/2020]

Unconfined aquifers along rivers and streams

          Many of the world’s largest rivers have floodplains that are much larger than the modern stream that flows within their valley walls.  These are named “underfit streams.”  The depth of sediment in the valley is also much deeper than the river’s bottom.  In Northern Europe and North America, underfit streams are common because they once carried much larger volumes of water from melting glacial ice due to melting glaciers as they retreated due to changing climatic conditions.  The Missouri and Ohio Rivers are classic examples of rivers that formed at the most southern margin of the last glacial ice sheet.  They scoured wide valleys bounded by dramatic limestone cliffs in places.  As the volume of melt water reduced about 10,000 years before present, the flow velocity also reduced and large sediment loads filled the valleys (Figure 6).  During the filling process, coarser-grained sediment built up first followed by increasingly finer-grained sediment.  This is called a fining upward sequence.  Wells drilled into the coarser material can be very productive.  Municipalities along the rivers often tap these productive units for public water supply because less water treatment is needed than withdrawing the water directly from the river.  In addition, agricultural production in the river valleys also utilize these productive wells for crop irrigation during dry periods.

Figure 6 -- Typical cross section of unconsolidated material in the Lower Missouri River Valley (Grannemann, N.G., 1974)

Basin-fill and Blanket-alluvial Aquifers

          Erosion of mountains can produce large volumes of sediment over geologic time.  This ongoing process takes place constantly but is amplified when thick ice sheets fill mountain valleys simultaneously with continental glaciation.  These mountain glaciers scraped large volumes of sediment that was eventually carried downstream by glacial meltwater and rain.  The sediment filled adjacent basins or covered nearby plains with blankets of sediment.  These filled valleys and blanketed plains became important aquifers outside of the areas covered by continental glaciation (Figure 7). 

Figure 7 – Location of basin-fill and blanket-alluvial aquifers in the United States (https://water.usgs.gov/ogw/aquiferbasics/uncon.html) [accessed 4/03/2020]

Basin-fill aquifers

          Basins have underlying structural features in bedrock such as the basin-bounding faults are illustrated in Figure 8.  The resulting basin creates a vessel into which sediment can collect.  Many such basins can be thousands of feet deep.  Groundwater within the resulting unconsolidated aquifers can be thousands of years old at depth (6,000 years old in the Rio Grande Basin in New Mexico) but of modern age at the mountain fronts where modern-day recharge is prevalent even if sporadic.  Deep wells within the basin mix old and recent water when pumped.  In dry environments, the unsaturated zone in the middle of the basin can be hundreds of feet thick.  The basins have coarser-grained deposits near the mountain front and finer-grained deposits in the center of the basin away from the sediment source.  Most basins have flowing rivers near the basin center where recent alluvial deposits occur.

Figure 8 – Schematic showing basin-fill aquifer setting (USGS Circular 1358, 2014).

Blanket Alluvial Aquifers

The High Plain Aquifer in the western United States is one of the most important blanket-alluvial aquifers in the world.  It represents a classic example of blanket-alluvial deposition from erosion of the Rocky Mountains west of the aquifer.  The aquifer is almost naturally divided into three areas – Northern, Central, and Southern portions.  About 30 percent of the irrigated acreage in the United States is in the High Plains.  Nearly 200,000 irrigation wells have been drilled into the Aquifer and the resulting aquifer depletion is illustrated in Figure 9.

Figure 9 – The High Plains aquifer as an example of blanket alluvial unconsolidated aquifer (https://www.usgs.gov/news/usgs-high-plains-aquifer-groundwater-levels-continue-decline) [accessed 4/03/2020]

Estimating Hydraulic Properties of Unconsolidated Deposits

          Darcy’s law can be represented as:

          Q = -KiA

       Where: Q is the rate of flow, K is effective hydraulic conductivity, i is the gradient, and A is the cross sectional area perpendicular to flow.  The minus sign is used because water flows in the direction of decreasing head.  K has units of velocity, i is dimensionless, and A has units of area so that Q has units of volume per unit time.  We can usually calculate A and we often know at least a few head measurements to determine i.  However, if we do not know either K or Q, it is clear that there are many different combinations of K and Q that will match the gradient.  Q can sometimes be determined by measuring pumping rates or by doing low-flow measurements of streams, thus reducing the range of K values to help evaluate the accuracy of a groundwater flow model.  When this is not possible, estimating K is sometimes the best option.  Unfortunately, K values can range greatly (even orders of magnitude), even for similar looking granular material.  In 1892, while working on sand filtration for water purification, Allen Hazen established a correlation between intrinsic hydraulic conductivity and the square of mean grain size diameter:   k = Cd² where k is only a function of the granular medium, d is mean grain size, and C is a constant of proportionality.  Hazen later modified this to be the 10^th percentile grain size.  While this relationship has great value, it still has a proportionality constant that is not easy to determine and, to some extent, it replaces one proportionality constant with another.  The constant, C, incorporates the arrangement, shape, and assortment of the grains.  That is, arrangement and assortment may adjust the size of void space, and the shape may influence the ease with which water may move through void spaces because of roughness.  In addition, any clay-sized particles will lower the medium’s ability to transmit water. 

       Beginning in the 1900’s, as wells were being used more and more for supply, engineers developed well tests that could provide accurate estimates of localized hydraulic conductivity as a way of determining pumping rates for well fields.  These approaches consist of long-term pumping tests (also called aquifer tests) and short-term slug tests.  More recently, downhole tests, including geophysical tests, have been developed to determine hydraulic properties of aquifers without averaging the results of the whole open borehole.  As the need to trace and model contaminant transport has increased, these techniques have improved. 

The following paragraph illustrates the complexity of estimating hydraulic conductivity in a sandy aquifer in Cape Cod, Massachusetts.  More information about using statistical methods to evaluate flow in complex aquifers is provided in the lessons on Darcy’s Law.

    Most of the sediments in the upper 100 feet of the Cape Cod unconfined aquifer are coarse-grained sand and gravel glacial outwash deposited during the retreat of continental glacial ice between 21,000 and 19,000 years ago.  Kettle lakes are common.  The sediments generally become finer with depth and include thin layers of fine grained lake sediments.  Many studies of the area have been conducted over the last 50 years and considerable data are available about the hydrogeologic properties of the aquifer.  One study published in 1996 compared estimates of hydraulic conductivity in eastern Cape Cod using grain-size analysis with results from slug tests (Desimone and others, 1996).  Grain-size analysis of 40 cores in a fine to coarse-grained unit ranged from 130 to 194 feet per day while the slug test for a well in this unit was 14 feet per day.  It is possible that placement of the open interval for the slug test was not representative of the unit.  Grain-size analysis of 5 samples of fine to medium sand ranged from 20 to 26 feet per day while analysis of two slug tests in this unit were 59 and 110 feet per day.  While these two methods do not correlate very well, they are both helpful even though they highlight the difficulty of obtaining consistent data on hydraulic conductivity. 

        Below is a table of data from a 305-foot well in western Cape Cod.  Data analyzed from cores from this well shows considerable variation in hydraulic conductivity (Table 1).  Highest values are in the first 107 feet with generally lower values below that depth due to deposition in a calmer water environment.  Look at the information in this table and evaluate:

1. Could you develop a model to simulate groundwater flow in two dimensions?If so, what would be your starting K value?
2. If you were to develop a quasi-three-dimensional groundwater flow model, how many layers would be appropriate and what horizontal K values would you use for each layer?What vertical K values would be appropriate between layers?

Table 1 – Depth, description, and hydraulic conductivity values from grain-size analysis for core samples from a well in southwestern Cape Cod.  BLS = below land surface. (Hull and others, 2019)

References

Desimone, L.A., Barlow, P.M., and Howes. B.L., 1996, A Nitrogen-Rich Septage-Effluent Plume in a Glacial Aquifer Cape Cod, Massachusetts, February 1990 Through December 1992: U.S. Geological Survey Water-Supply Paper 2456, 89 p.

Hazen, A, 1892. Physical properties of sands and gravels with reference to their use infiltration, Massachusetts State Board of Health, Boston, MA.

Hull, R.B., Johnson, C.D., Stone, B.D., LeBlanc, D.R., McCobb, T.D., Phillips, S.N., Pappas, K.L., and Lane, J.W., 2019, Lithostratigraphic, Geophysical, and Hydrogeologic Observation from a Boring Drilled to Bedrock in Glacial Sediments Near Nantucket Sound in East Falmouth, Massachusetts: U.S. Geological Survey Scientific Investigations Report 2019-5042, 27p.