Regional groundwater flow in an area mapped as continuous

permafrost, NE Alaska (USA)

Douglas L. Kane & Kenji Yoshikawa &

James P. McNamara

Abstract Fundamental knowledge of groundwater sys- Introduction

tems in areas of permafrost is often lacking. The
likelihood of finding good quality groundwater resour- Permafrost is extensive in the northern hemisphere. Williams
ces of acceptable quantities generally decreases as the (1970) described frozen ground as an impermeable layer
areal coverage of permafrost increases. In areas of which (1) restricts recharge, discharge and the transmission
continuous permafrost, the probability of finding areas of groundwater, (2) acts as a confining layer and (3) limits
of groundwater recharge and discharge are minimal. the volume of aquifers where liquid water can be stored.
Still, in northeastern Alaska (USA), the presence of Generally, the hydraulic conductivity of ice-rich frozen soils
numerous springs and associated downstream aufeis is so low in areas of continuous permafrost (Burt and
formations clearly indicates that there has to be a Williams 1976; Horiguchi and Miller 1980; Kane and Stein
groundwater system with the required complementary 1983a, b) that near-surface water is hydraulically discon-
areas of groundwater recharge and transmission. Re- nected from deeper subpermafrost groundwater systems.
charge zones and transmission pathways in this area of Despite the low hydraulic conductivity of permafrost, it
extensive permafrost, however, are essentially un- is evident from the presence of springs and icings in
known. This study shows that the recharge occurs on northeastern Alaska (Fig. 1), that groundwater flow is
the south side of the Brooks Range in northeastern active in this area mapped as continuous permafrost. The
Alaska, where extensive limestone outcrops are found. origin and pathways of groundwater to these springs are
The transmission zone is beneath the permafrost, with unknown. Springs can arise from local recharge sources
discharge occurring through the springs via taliks traveling through short, suprapermafrost pathways, or
through the permafrost (where faults are present) and non-local recharge areas traveling through complex sub-
also likely at the northern edge of the permafrost along permafrost pathways. Unfortunately, few groundwater
the Beaufort Sea coast. observations exist in this and other continuous permafrost
regions.
Keywords Hydrochemistry . Groundwater The goal of this study is to understand the source of
flow . Permafrost . Springs . Alaska (USA) spring water in northeastern Alaska as a first step towards
gaining more comprehensive knowledge of groundwater
flow systems in continuous permafrost systems. In this
initial reconnaissance, data are used that were collected as
part of other investigations. Information is drawn from
Received: 31 May 2012 / Accepted: 13 November 2012 surface observations (springs) and measurements (flow
Published online: 6 December 2012 rate, spring water age, water quality, temperature, etc.),
* Springer-Verlag Berlin Heidelberg 2012 geochemical and isotopic data, and the results of the few
groundwater studies that have been carried out in
Published in the theme issue “Hydrogeology of Cold Regions” continuous permafrost environments. Based on these
sources of information, it is argued that subpermafrost
D. L. Kane : K. Yoshikawa
groundwater originates from the other side (south side) of
Institute of Northern Engineering, University of Alaska Fairbanks, the Brooks Range (Fig. 1), as illustrated in Fig. 2, and that
Fairbanks, AK, USA transmission of groundwater crosses the topographic
divide. The likelihood of permafrost is much greater on
J. P. McNamara
Department of Geosciences, Boise State University, the north side of the Brooks Range than the south side
1910 University Drive, Boise, ID, USA (Brown et al. 1998); plus, limestone outcrops are more
extensive on the south side (State of Alaska 2008).
D. L. Kane ())
Water and Environmental Research Center, Following a summary of the present knowledge on
Room 525 Duckering, Fairbanks, AK 99775, USA groundwater in permafrost regions, observations and
e-mail: [email protected] interpretations that support this premise are presented.

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

42

Fig. 1 Map showing the location of springs on the North Slope in eastern Alaska that were used extensively in this article. Springs
numbered from west to east

Previous knowledge larger quantities of water are needed. Williams and van
Everdingen (1973) produced a review paper on ground-
An understanding of groundwater dynamics in northern water investigations in permafrost regions of North
Alaska and other continuous permafrost regions has been America for the Second International Conference on
slow to develop. Part of the lack of progress is due to Permafrost. They presented material on the following
limited information on permafrost distribution, and a lack topics; permafrost and movement of groundwater, avail-
of demand for water-resource information in this sparsely ability of groundwater, icings, pingos and artesian
populated region. In the small villages with minimal water pressures, influence of permafrost on water quality,
needs, surface water (streams and lakes) generally is the methodology, studies of basin hydrology, investigations
source of choice. Interest in groundwater development is for pipeline projects, hydrologic model studies and
generally associated with resource development where research needs. Since their review, the number of

Fig. 2 Simple conceptual representation of the groundwater flow system in an area mapped as continuous permafrost in NE Alaska,
assuming that groundwater recharges through permafrost free areas of limestone on the south side of the Brooks Range and discharges
through taliks (probably coinciding with faults), extending through the permafrost north of the Brooks Range. Some discharge is apt to
discharge along the northern boundary of the permafrost and the coast

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

 43
hydrologic studies in regions with continuous and discon- Along such flowpaths, permafrost impacts the direction of
tinuous permafrost has expanded, mostly because of flow and can also act as a confining layer leading to artesian
concern about climate change and its impact on perma- conditions (Linell 1973). Although ice significantly reduces
frost. While the primary interest of this study is in soil hydraulic conductivity of seasonally frozen soils,
recharge, transmission and discharge of subpermafrost snowmelt can infiltrate partially frozen soils at rates sufficient
groundwater in continuous permafrost regions, first there to generate groundwater recharge (Kane and Stein 1983a, b).
is a review of groundwater in discontinuous regions where This is particularly true for well-drained soils. Deep lakes and
more information is available. The role of suprapermafrost other water bodies can create taliks that serve as discharge
groundwater is not addressed here as it does not constitute zones for subpermafrost groundwater. For example, Kane
a large enough subsurface storage reservoir for ground- and Slaughter (1973) demonstrated that permafrost was
water development. absent beneath a small lake, 4.5 m maximum depth, located
in a valley bottom with permafrost extending down 40–70 m
depth adjacent to the lake. They documented the discharge of
Subpermafrost groundwater associated subpermafrost groundwater by measuring the hydraulic head
with discontinuous permafrost at various depths beneath the lake bottom with piezometers.
In areas of discontinuous permafrost, south-facing slopes are The hydraulic head was above the lake water-surface
typically permafrost free, while north facing slopes and valley elevation for all piezometers. Some nearby shallower lakes
bottoms are generally underlain by permafrost (US Depart- were found to have no hydraulic connection to the deeper
ment of Agriculture, Soil Conservation Service 1963). East- subpermafrost groundwater.
and west-facing slopes represent a transition from no
permafrost to permafrost with variable amounts of permafrost.
Whether permafrost exists or not is more complex as drainage, Subpermafrost groundwater associated
soils and vegetation may influence specific site conditions. with continuous permafrost
Where permafrost is near the thawing temperature, generally Continuous permafrost is generally found north of the
near the southern limit of permafrost, it is less likely to be continental divide in the Brooks Range of northern
present on east- and west-facing slopes, while it is more likely Alaska; an area referred to as the North Slope. Prior to
as one progresses northward and the mean annual air the early 1970s (when oil was discovered in Prudhoe
temperature decreases. More on the distribution of permafrost Bay), very little was known about water resources on the
can be found in Brown et al. (1998) for the circumpolar Arctic North Slope, either surface or subsurface. The USGS
and Ferrians (1998) specifically for Alaska. initiated a stream-gauging program in the vicinity of
Groundwater systems typically consist of recharge on Prudhoe Bay area shortly after oil was discovered.
permanently unfrozen south-facing slopes (and those areas Childers et al. (1977) also initiated a hydrological
on east- and west-facing slopes without permafrost) with reconnaissance program in the eastern North Slope in
transmission from the point of recharge to the valley 1975 in response to potential construction of oil and gas
bottom. This water discharges at the surface where either pipelines. Two of the hydrologic characteristics (although
taliks (unfrozen zones) exist through the permafrost in the there were others) that they targeted were icings (aufeis)
vicinity of lakes and streams or permafrost is simply and springs. In their survey, they sampled 18 springs with
absent. Also, discharge springs can often be found at the discharges ranging from 0.03 to 2.5 m3/s (30 to 2,500 l/
base of slopes where groundwater first encounters perma- sec) and water temperatures from 0 to 33 °C. Comple-
frost in the valley bottom. mentary to springs were aufeis formations found down-
Areas of permafrost are unlikely candidates for subper- stream from these springs during the winter and spring
mafrost groundwater recharge due to ice in the soil pores seasons (with some formations persisting over the summer
significantly reducing the hydraulic conductivity. The to the next winter some years).
hydraulic conductivity of frozen soils can be several orders These springs are indicative of subpermafrost groundwa-
of magnitude lower than comparable unfrozen soils; the ter discharge. If these springs were emanating from fractures
amount of reduction in the hydraulic conductivity of frozen in the bedrock, where was this water being recharged and
soils depends upon the amount of ice in the pores or the what was the pathway from the area of recharge to the area of
thickness of the unfrozen film on the surface of the soil discharge? The entire North Slope is mapped as continuous
particles with generally high ice contents (thin unfrozen permafrost (Brown et al. 1998; Ferrians 1998), but for the
films), resulting in low hydraulic conductivities (Burt and origin of the water for these springs to be subpermafrost
Williams 1976; Horiguchi and Miller 1980; Kane and Stein groundwater there has to be taliks penetrating through the
1983a, b). Anderson and Morgenstern (1973) showed that entire depth of permafrost. Also, there must be areas free of
the amount of unfrozen water in a frozen soil was related to permafrost for recharge to occur.
the soil-particle surface area—the greater the surface area the An earlier study in the central Alaskan Arctic (area
greater the unfrozen water content (assuming of course that west of Dalton Highway to the western boundary of the
the moisture content equals or exceeds the predicted Chandler River basin), a region not noticeably rich in
unfrozen water content). observable springs like those found in the eastern Alaskan
Recharge in unfrozen zones can flow under permafrost to Arctic, was published on regional groundwater flow based
discharge zones that penetrate up through the permafrost. on the measured thermal regime in exploratory oil and gas

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

44
wells (Deming et al. 1992). It should be noted that, in Aufeis
areas of recharge, advective heat transport by the Aufeis, also referred to as naleds (Russian) and icings, is a
groundwater is counter to the geothermal gradient, while hydrologic winter process of high latitude streams and
downstream they are acting in concert, as illustrated in rivers whereby water, from in and/or beneath the stream,
Fig. 2. They describe a thermal pattern coherent with under pressures greater than hydrostatic pressure (Kane
forced convection due to a topographically driven ground- 1981), flows out on top of the ice where it subsequently
water flow system at the regional scale. Further, they freezes and adds to the thickness of the ice in the channel/
found that both heat flow and subsurface temperatures floodplain. This process is repeated several times during
varied consistently from relatively low values in the the winter and, at the end of the cold season, the channel
foothills to much higher values on the coastal plain. and adjacent floodplain can be completely filled with ice.
Because there are few readily discernible signs of The thickness of the aufeis is much greater than the depth
subpermafrost groundwater discharge from the continental of the water in the channel prior to freeze-up. For
divide to the Arctic Ocean, it is hypothesized for their substantial aufeis formations to develop, there needs to
study area that recharge occurs in the Brooks Range and be a source of subpermafrost groundwater or water from
discharge occurs into the Arctic Ocean with flow occur- near surface taliks; the greater the source volume, the
ring as subpermafrost groundwater flow underneath the greater the volume of the downstream aufeis formation.
permafrost. This is essentially the same model for regional
groundwater flow that is used here for the northeastern
part of the state (Fig. 1); however, the conclusions Setting
presented here are not based on heat flow, but on the
visible evidence of numerous springs indicating subper- The North Slope of Alaska (Fig. 1) is an area of
mafrost discharge and the water quality of these springs. 230,000 km2 consisting of three physiographic units from
During recent field excursions, late winter evidence of south to north: mountains, foothills and coastal plain. The
relatively small aufeis formations and associated springs drainages are predominantly north-draining, flowing out
in the central Alaskan Arctic, especially in the valleys at of the mountains, through the foothills and over the
the transition from the mountains to the foothills, has been coastal plain before discharging into the Beaufort Sea. The
seen. It should be noted that the mountains in the central Brooks Range (an extension of the Rocky Mountains) has
part of the Alaskan Arctic are much farther from the basically an east–west-trending continental divide with an
Arctic Ocean than in the northeastern part of the North approximate maximum elevation of 2,500 m above sea
Slope (Fig. 1). level; the mountains are closer to the Arctic Ocean
In permafrost, heat conduction is the main heat-flow coastline in the northeastern part of the state than in the
mechanism. Where permafrost is lacking or taliks may northwestern part (Fig. 1). The elevation range for the
exist through the permafrost and groundwater flow is foothills is approximately 150–1,200 m with the coastal
present, advective heat transport may occur in concert plain being less than 150 m and the mountains greater
with conduction (Kane et al. 2001; Rowland et al. than 1,200 m.
2011). Clearly the thaw bulb beneath a lake in an area Permafrost is assumed to be continuous in this area
of continuous permafrost could be a subpermafrost with taliks extending through it in the vicinity of
groundwater discharge pathway. However, Brewer subpermafrost springs. The permafrost thickness is poorly
(1958) found that the thaw bulb beneath an Alaskan documented inland away from the coast (because of few
Arctic lake on the coastal plain did not come close to exploratory wells) but is estimated to be approximately
penetrating the permafrost; the permafrost was too 250 m thick. Along the coast where there have been
thick and the temperature of the permafrost too low (< extensive drilling programs, permafrost depths as great as
10 °C). Arcone et al. (1998) found that the thaw bulb 600 m have been observed. Lachenbruch and Marshall
beneath Arctic streams appears to be a thin layer (1986) reported that the long-term mean annual tempera-
extending across the channel; this conclusion has been ture at the top of the permafrost ranges from −4 °C in the
supported by many other observations (Bradford et al. Brooks Range to −12 °C along the coast. The depth of the
2005; Brosten et al. 2009a, b). Therefore, it is not active layer on the North Slope averages about 50 cm, but
likely that surface-water bodies and associated thaw generally decreases from south to north and is greater in
bulbs in the Alaskan Arctic represent pathways for well-drained sites. Surficial active layer soils generally are
subpermafrost groundwater discharge; however, for composed of an organic layer (approximately 25 cm thick,
larger, deeper lakes (like Teshekpuk Lake in the although quite variable) over a mineral layer which is
northwestern sector of the North Slope, maximum generally near saturation. Thaw lakes are much more
depth 6 m and surface area of 430 km2), it is a prevalent on the coastal plain than elsewhere; there are
possibility. It is clear from the presence of numerous lakes of glacial origin, especially in the foothills closer to
springs that an Arctic groundwater system is present in the Brooks Range.
this part of Alaska. Hall and Roswell (1981) and Vegetation is almost continuous over the coastal plain
others suggest that faults may represent the most likely and foothills with tussock tundra over the foothills and
pathway for groundwater discharge in this permafrost sedge tundra on the coastal plain. Shrubs (1–2 m in
environment. height) are common in the foothills and southern extent of

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

 45
the coastal plain and expanding (Tape et al. 2006). A few rock exposures give insight to the geology farther north
patches of trees can be found, especially in riparian areas under the coastal plain. Around 400 Ma ago (ma), the pre-
in the foothills. Alpine communities are found in the Mississippian sequence formed a wide range of rock types
mountains along with bare rock outcrops. including argillites (compact rock of siltstone, claystone or
Numerous studies have shown that climate warming is shale), graywackes (coarse grained sandstone), quartzite
ongoing on the North Slope of Alaska (Stafford et al. (metamorphic rock of mainly quartz), carbonate (mainly
2000; Chapman and Walsh 1993; Serreze et al. 2000) as limestone and dolomite) and some local deposits. Repeat-
well as other high latitude environments. It is also ed uplift and faulting due to compressional stresses around
established that most of this warming is seasonal, 145 ma produced the Brooks Range; continued compres-
occurring in winter. There is some statistically weak sion pushed older rocks over younger rocks producing
evidence (Stafford et al. 2000) that precipitation is unconformities. The thin, structurally less sound beds
decreasing, particularly in summer. Even though the were aggressively deformed and overturned. Thicker,
hydrologic cycle is likely intensifying (Rawlins et al. more competent beds are stacked up like leaning books
2010) through positive trends in hydrologic fluxes, the (Bureau of Land Management and Alaska National
impact of climate change on the processes studied here are History Association 1993). For the area of study, this
probably not measureable, as it took thousands of years resulted in a series of four parallel east–west-trending
for the permafrost to form. mountain systems from the continental divide northward.
One of the obvious indicators of groundwater dis- It is in the valleys of these progressively lower, down-
charge in areas of permafrost is the formation of aufeis in stream drainages that many of the springs reported here
a channel and adjacent floodplain, usually somewhere are found (Fig. 1); it is assumed that the springs originate
downstream of a spring (Kane 1981; Hall and Roswell along faults in these up-thrusted unconformities.
1981; Yoshikawa et al. 2007; Carey 1973; Harden et al. Williams (1970) confirms the existence of various
1977). There is a lot of speculation concerning why aufeis types of faults in limestone (Mississippian and Pennsyl-
forms where it does and why the volume of the formation vanian age) associated with the springs in northeastern
varies from year-to-year although the spring flow is Alaska. Also, Beikman and Lathram (1976) and others
relatively steady; unfortunately there have been very few have mapped the distribution of faults across northern
detailed studies on various aufeis processes. It is clear Alaska. They show that there are many more faults in
from the study of Kane (1981) that aufeis growth occurs northeastern Alaska than in the northwestern sector; this
during episodic warm periods during the winter and not also matches the distribution of major icing formations.
cold spells. During these warm periods, pressures build up Finally, Hall and Roswell (1981) have produced a map
under the ice, deforming the ice and inducing cracks that that shows the colocation of springs and faults in
let water flow out onto the surface and freeze. As the ice northeastern Alaska.
thickens during the winter, the pressure needed to produce
another overflow must be greater than the last overflow
event. Observations
Figure 1 shows the location of many of the springs on
the northeastern North Slope of Alaska. Many of these Presented below are physical and chemical data collected
springs have been visited in the past, first in detail by at various springs, both by others and by University of
Childers et al. (1977). Visibly, a majority of the springs Alaska Fairbanks, Water and Environmental Research
are located at the base (Fig. 1) of many of the outer Center (UAF/WERC) personnel on numerous visits to the
mountain ranges (like the Shublik and Sadlerochit area. Most of the visits to the springs were made in late
mountains). Childers et al. (1977) examined the flow winter when the springs are more easily spotted because
rates, temperature, and chemistry of the more predominant the relatively warm discharge melts the surrounding snow.
springs. There are many springs in this region and not all Many of the springs emanate off the side hillslopes or
are presented here. Some are very close to others, often from the base of hillslopes and then travel a short distance
the exact location of springs is confused with neighboring to a river or stream. In some cases, all of the flow is
springs (particularly for studies carried out before the collected in a single channel; however, in some cases the
advent of GPS). flow is quite dispersed making it difficult to get accurate
measurements of discharge rates. It is likely that there are
springs discharging from the bed of streams or rivers that
Geology are undetected and are rapidly diluted by base flow from
upstream. Flood Creek (spring No. 6, Table 1) would
Although the presence of permafrost in this environment appear to fall into this category; the channel remains ice-
has a significant impact on the groundwater flow system, free for a long distance during winter with only the one
the regional geology also is quite important. The scrutiny obvious spring adjacent to the stream. The distribution of
of the geology of this region has been driven, not by water springs is shown in Fig. 1. Except for a few springs on the
resources, but oil and gas development. The Brooks coastal plain, most of the springs are located on the
Range constitutes the southern boundary of the study northern fringe of the Brooks Range, including the outer
area, to the north are lesser mountains and foothills whose ranges of the Shublik and Sadlerochit mountains.

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

46
Table 1 Physical properties of springs on the eastern North Slope of Alaska where meaningful measurements have been made either by
UAF/WERC or the USGS (Childers et al. 1977). Springs ordered from west to east
Spring No. and name Lat (deg/min) Long (deg/min) Elev (m asl) Discharge Specific Conductivity pH Temp
Q (l/s) (μS/cm) (°C)
1 May Creek 68 41.570 −150 29.018 423 263 6.5 0.4
2 Kuparuk River 68 59.150 −149 42.400 408 251 6.4 0.1
3 Lupine 68 53.512 −148 4.920 497 42.5 298 8.1 1.0
4 Saviukviayak West 68 53.389 −148 4.832 507 2,520 259 8.0 4.8
5 Saviukviayak Trib 68 56.333 −147 58.750 678 1,529 239 7.9 3.5
6 Flood Creek 68 58.813 −147 52.862 425 2,350 237 8.2 7.7
7 Ivishak Hillside 69 1.4120 −147 43.284 406 5,494 238 8.2 7.4
8 Echooka River West 69 15.583 −147 22.833 391 623 257 7.9 7.0
9 Echooka Hillside 69 16.041 −147 20.706 404 250 8.1 4.7
10 Echooka Valley 69 16.051 −147 20.454 403 249 8.1 4.7
11 Shublik 68 28.333 −146 11.833 414 680 275 8.0 5.5
12 Red Hill 69 37.617 −146 1.633 357 24 1,000 7.0 33.0
13 Katakturuk Trib 69 41.700 −145 6.550 422 122 245 8.2 1.0
14 Sadlerochit 69 39.383 −144 23.617 308 991 410 7.9 13.0
15 Hulahula River Icing 69 45.650 −144 9.250 166 207 240 8.0 1.0
16 Okerokovik River 69 43.100 −143 14.417 160 736 300 7.3 1.0
17 Aichilik River 69 31.100 −143 2.000 318 42.4 338 8.0 3.6
18 Ekaluakat River 69 35.450 −142 18.000 203 144 350 7.9 6.4
19 Kongakut River 69 32.600 −141 49.633 141 368 276 7.9 1.0
20 Kongakut River Above Delta 69 42.833 −141 47.500 30 2,503 210 6.7
21 Kongakut River Delta 69 43.600 −141 46.117 17 1,048 215 7.9 1.0
22 Clarence River 69 30.733 −141 11.617 151 133 250 6.8 0.0

Eastward, the springs and mountains are located closer to known. For example, Red Hill spring with a discharge of
the coast. This results in a general decreasing trend in the 24.1 l/s (Childers et al. 1977) has been measured at
general elevation of springs. temperatures ranging from 29 to 33 °C and Okpilak spring
(69°17.227′–144°1.063′) with even lower flow has been
measured at 38 °C. However, Fig. 3 shows that there are
Discharge from springs several springs near 0 °C with rather low flows that appear
The measured temperature of springs (Table 1, Fig. 3) to be able to maintain the required taliks. A plot of
ranges from 0 to 33 °C. There are a large number of discharge versus elevation (Fig. 4) shows no discernible
springs that have temperatures at or just above the pattern, but there is a tendency for spring elevation to
freezing point of water. Most of these springs are located decrease from west to east.
either on the west or east ends of the study area. When Collectively for all the springs citied in this study, the
sampling in the winter with snow on the ground, it is discharge is 20,000 l/s, which would require a groundwater
possible that the measurements are not made at the point recharge area of 3,150 km2. This calculation is based on
where the spring first emerges from the ground, but some annual precipitation in the Brooks Range of around 400 mm
distance downstream. In late spring during snowmelt or in (Kane et al. 2012) and a runoff coefficient of 0.5 for the north
the summer, it is possible that spring flow is also mixed side of the Brooks Range (Kane et al. 2000), which would
with summer precipitation/runoff both diluting and alter-
ing the quality and temperature of the spring water.
During ablation, mixing will reduce the temperature if it
is not already at 0 °C; during the summer, precipitation
could alter the spring water temperature either up or down.
While Fig. 3 shows the ranges of temperature and
discharge of springs, there is no relationship between the
two variables.
The flow rate (Table 1) from springs varies from high
(5,494 l/s, Ivishak Hillside spring No. 7) to low (24 l/s,
Red Hill spring No. 12). What is not known is how much
heat is required to keep the taliks open for the springs to
exist; the amount of advected heat will depend upon the
groundwater water temperature and the flow rate. A quick
rumination of this yields the conclusion that the amount of
advected energy needs to offset the heat conducted from Fig. 3 A plot of water temperature versus spring discharge;
the talik by the colder permafrost and this will depend Ivishak Hillside has the highest flow rate at 5,500 l/s (T=7.4 °C) and
upon the geometry of the talik which is presently not Red Hill has the highest temperature at 33 °C (Q=24 l/s)

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

 47
magnesium, sodium, potassium, bicarbonate, sulfate and
chloride; calcium and bicarbonate are the two most
significant contributors to specific conductivity of these
springs, while magnesium and sulfate also influence the
measured values (Tables 1 and 2).

Spring-water chemistry
The data presented here is a mix of data collected by the
USGS and the University of Alaska Fairbanks, Water and
Environmental Research Center (UAF/WERC). It is
assumed that the USGS followed their protocol for testing
back in the 1970s. In the measurements UAF/WERC
personnel made, the procedures outlined in APKA-
Fig. 4 Plot of elevation versus discharge for springs in northeast-
ern Alaska where detailed data have been collected AWWA-WEF (1998) were followed. Where both parties
collected the same data for a common site, the magnitude
of the measurements was quite similar. This indicated that
leave a maximum of 200 mm of potential recharge. This the data could be collectively used together and the
200-mm estimate of annual recharge is obviously conserva- properties had not significantly changed over the 25-year
tive as it does not include evapotranspiration. period between sampling campaigns.
The State of Alaska (2008) geological map shows that the The pH of the spring water ranged from a low of 6.4 to
areal coverage of surficial limestone is much greater (at least a high of 8.2 (Table 1). The low values (6.4–6.8) were
one-half of the eastern one-third of the Brooks Range) than found at both the east and west ends of the study area. The
the recharge area derived in the preceding paragraph. pH in the middle of the study area was above 7.0, with 16
Although potential recharge areas are widespread through of 22 values in the narrow range of 7.9–8.2. The pH
the whole of the Brooks Range, the distribution of surficial values found here are typical of groundwater in general.
limestone is greater, both on the south side of the Brooks Calcium levels (Table 2; Fig. 6) vary from 36 to 78 mg/
Range and east of the Dalton Highway. l for the springs with one outlier at 7.4 mg/l (Kuparuk,
spring No. 2). Calcium levels are generally low for
groundwater in igneous and metamorphic bedrock and
Specific conductivity high for carbonate sedimentary formations. With the
Specific conductivity was measured at all of the springs extensive limestone bedrock in the Brooks Range (State
(Table 1; Fig. 5) visited by personnel either from the of Alaska 2008), especially in the eastern sector, high
University of Alaska Fairbanks in the last decade or the levels of calcium should be expected if that is the pathway
USGS in the early 1970s. There was little difference between of the groundwater. As mentioned previously, the
similar measurements taken by the two organizations at the Kuparuk River spring (No. 2) has the low outlier value
same springs. The values ranged from a low of 210 μS/cm to of calcium; this spring has many features that separate it
a high of 1,000 μS/cm. Most of the measured values were in from the quality of most of the other springs. For example,
the high 200s. The high value of specific conductivity was at it has a low pH, temperature near 0 °C, and low
a hot spring (Red Hill, spring No. 12, Table 1) with a magnesium and potassium levels.
measured temperature of 33 °C. The ionic constituents that Magnesium levels (Table 2; Fig. 6) range from 0.7 to
increase specific conductivity of water are calcium, 21 mg/l. The springs in the middle of the study area have
the highest values with Red Hill (No. 12) at 21 mg/l and
Sadlerochit (No. 14) at 18 mg/l. About one-half of the
springs are in the 7–9 mg/l range. Sodium levels are
generally low (0.3–2.8 mg/l) except for the same two
springs, Red Hill (120 mg/l) and Sadlerochit (7.8 mg/l).
These two springs are also warmer than the other springs
that were analyzed in detail; the higher sodium and
temperature values are likely a result of deeper ground-
water flow. Potassium, which is not very soluble, ranges
from 0.1 to 5.8 mg/l, with Red Hill the highest.
Sadlerochit spring has the highest chloride level at
4.0 mg/l, while most of the remaining springs are below
1.0 mg/l.
Groundwater with high bicarbonate levels are generally
Fig. 5 Relationship between specific conductivity and spring an indication of limestone or dolomite bedrock. Bicar-
temperature showing an increase in specific conductivity with bonate levels (Table 2; Figs. 6 and 7) of the springs ranged
temperature from 62 to 322 mg/l with an average of 146 mg/l. The hot

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

48
Table 2 Chemical constituents for springs sampled on the eastern North Slope of Alaska, data from UAF/WERC and USGS (Childers et
al. 1977)
Spring No. and name Ca (mg/l) Mg (mg/l) Na (mg/l) K (mg/l) Sulfate (mg/l) Cl (mg/l) Bicarbonate (mg/l)
1 May Creek 50.8 7.95 1.9 0.3 2.15 0.24
2 Kuparuk River 7.4 1.12 0.83 0.1 23.1 0.19
3 Lupine 51 7.7 0.4 0.1 20.8 0.22 177
4 Saviukviayak West 40 9.2 0.7 0.1 8.5 0.7 155
5 Saviukviayak Trib 39 7.3 0.8 0.5 12 0.5 137
6 Flood Creek 52 8.8 0.3 0.5 13.1 0.18 136
7 Ivishak Hillside 47 8.72 0.3 0.1 14 0.16 128
8 Echooka River West 36 9.8 1.3 0.2 24 1.3 131
9 Echooka Hillside 40.1 9.48 0.47 0.1 33.6 0.15
10 Echooka Valley 39.8 9.17 0.47 0.1 31.6 0.13
11 Shublik 38 11 1.5 0.3 37 1.3 127
12 Red Hill 55 21 120 5.8 150 1.1 322
13 Katakturuk Trib 52 0.5 0.5 1.8 0.8 130
14 Sadlerochit 78 18 7.8 1 71 4.0 156
15 Hulahula River Icing 51 0.7 0.2 27 1.3 116
16 Okerokovik River 1.1 0.3 22 1.3 163
17 Aichilik River 65 2.8 2.1 39 2.0 148
18 Ekaluakat River 69 4.5 0.6 25 3.6 165
19 Kongakut River 50 1 0.5 25 0.9 134
20 Kongakut River Above Delta 1.1 17 1.0 62
21 Kongakut River Delta 46 1.1 1.1 17 0.5 122
22 Clarence River 53 0.7 0.5 12 0.7 124

spring (Red Hill, No. 12) had the highest and the compared with the other springs. Alkalinity was measured
Kongakut River above the delta (No. 20) had the lowest. at several springs; most alkalinity (capacity to neutralize
The lowest value of calcium (Table 2) was at the acids) is due to carbonate and bicarbonate. Figure 8 shows
Kuparuk River aufeis (spring No. 2) at 7.4 mg/l, which the strong relationship between specific conductivity and
compares to the area average for all springs of around alkalinity.
50 mg/l. Sulfate was the major contributor to the specific
conductivity at this site. Sadlerochit (spring No. 14) has
the highest concentration of calcium at 78 mg/l (Table 2) Radiocarbon dating
and the second highest specific conductivity of 410 μS/ A few samples of spring water were collected for
cm. Except for the two hot springs, Red Hill and Okpilak, radiocarbon dating. There are many possible sources of
Sadlerochit is the warmest spring at 13 °C with a flow of error when performing 14C dating of groundwater; the
approximately 1,000 l/s. Red Hill, with the highest three major sources of error are: (1) the 14C level of
specific conductivity (Table 2), has average Ca levels atmospheric CO2 is not constant, (2) recharge entering the
(55 mg/l) and high levels of magnesium (21 mg/l), sodium groundwater system may mix with other water of different
(120 mg/l), potassium (5.8 mg/l) and sulfate (150 mg/l) ages, and (3) carbon from other sources (some much older

Fig. 6 A plot of spring-water chemistry for data collected by the USGS (Childers et al. 1977) and UAF/WERC. Some data is absent in
cases where there were no measurements; also note the log scale on the y-axis

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

 49

Fig. 9 Plot of the percent of modern carbon (pMC) versus δ13C

Fig. 7 Water temperature versus the bicarbonate level of springs. mixing modeled age in northeastern Alaska where detailed data
All of the bicarbonate levels are fairly high and are probably due to have been collected (Table 3). Most of the spring water was less
limestone rocks than 2,000 years old except two hot springs (Okpilak and Red Hill
hot springs), which are the farthest from the hypothesized recharge
area
such as limestone and other carbonate rocks) may dissolve
into the groundwater. There are models (Pearson 1965;
Pearson and Hanshaw 1970) that require additional data percent modern carbon, the younger the water. The age of
and assumptions to make adjustments to the groundwater discharging water at the two hot springs—Red Hill (No.
age. The net result of all these errors is that the 12) and Okpilak—appears to be in the 3,000–7,000 year
groundwater is younger than determined by a straightfor- range. From a physical standpoint, it would seem logical
ward analysis of 14C. In this case, it was only intended to that the water for the two hot springs would travel longer,
get a rough idea of the age to determine if there was farther and deeper, thus warming due to geothermal heat.
sufficient time for the groundwater to flow beneath the
permafrost and then surface somewhere downstream
through taliks. After completing this exercise for a few Spring-water isotopic composition
of the springs, it was found that the age of the discharging The last contribution to this study was the measurement of
spring water was around 1,500–1,000 years BP (Fig. 9). isotope ratios (δ2H (or δD) and δ18O) of some of the springs
Some of the springs—such as Kuparuk (No. 2) and on the North Slope and snowpack and some rainfall on a
Hulahula (No. 15)—that are distant from the mountains south–north transect along the Dalton Highway both north
appear to be younger (Table 3); the two hot springs—Red and south of the Brooks Range (Fig. 10). Although the
Hill (No. 12) and Okpilak—seem to be the oldest isotope concentration of individual precipitation events is
measured (Table 3); this seems reasonable as they are variable, the isotopic signatures of the precipitation on the
the most distant from the identified recharge area. It is near south side of the Brooks Range and the few springs
hypothesized that the Kuparuk spring functions differently observed north of the Brooks Range were similar. They
than most of the other springs as the pH, Ca, Mg and K indicated that the recharge for the springs most likely came
are much lower, while the percent modern carbon (pMC) from the south side of the Brooks Range. Generally, the
values (Fig. 9) are much higher (Table 3). The higher the isotopic signature of precipitation north of the Brooks Range
is lighter than the heavier spring water (Fig. 10).

Discussion
Interest in potential groundwater flow on the North Slope
of Alaska for this study resulted from various field
observations during a number of related hydrologic
research projects. UAF/WERC hydrologic investigations
concentrated on active layer and surficial hydrological
processes (Kane et al. 1989, 2000; McNamara et al.
1998). There are, however, two aufeis deposits in the
Kuparuk watershed that were of interest; why were they
there and what was the origin of the water that allowed
them to form? There also existed several USGS open file
Fig. 8 Plot of alkalinity versus specific conductivity of springs in
northeastern Alaska. The data plotted here are a mix of some of the reports about aufeis deposits in more remote areas (away
springs in Table 1 and new springs (like Okpilak) that had not been from the road system) of the North Slope. It was not until
sampled before the last decade that UAF/WERC personnel were able to

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

50
Table 3 Results of radiocarbon dating of springs on the eastern North Slope of Alaska
Name and spring No.a δ13C (‰) Libby ageb (years) pMC (%) δ13C model age (years)
Kuparuk River (No. 2) −19.0 850±30 89.96±0.35 700
Ivishak Hillside (No. 7) −3.1 9,740±40 29.72±0.15 852
Flood Creek (No. 6) −3.6 7,280±40 40.35±0.19 789
Echooka Valley (No. 10) −6.1 4,080 ±30 60.17±0.27 867
Saviukviayak River (No. 5) −3.9 9,820±40 29.43±0.16 1,187
Lupine (No. 3) −7.5 5,350±40 51.36±0.23 1,449
Well at Pump station −6.6 11,740±40 23.17±0.13 2,739
Hulahula River (No. 15) −7.4 2,360±40 74.5±0.37 629
Sadlerochit River (No. 14) −4.0 11,440±40 24.1±0.12 1,430
Kavik River −10.8 2,330±40 74.8±0.37 951
Okpilak Hot Spring −7.5 15,610±50 14.3±0.09 3,002
Red Hill Hot Spring 1 −9.2 18,350±80 10.2±0.10 6,270
Red Hill Hot Spring 2 −9.8 18,950±80 9.5±0.09 6,948
Shublik River (No. 11) −5.0 7,230±40 40.6±0.20 1,205
Slope Mountain spring −11.3 3,200±40 67.1±0.33 1,373
a
Note that some of these springs are not numbered
b
Method of carbon dating developed by Willard Libby, University of Chicago

get helicopter support for their research projects and could This 591 mm would be the equivalent groundwater recharge
visit some of these more remote sites. Individually, the required above the spring in the watershed to produce the
springs may not seem very significant hydrologically, with spring discharge. Best estimate of annual precipitation
the largest spring discharging about 5,500 l/s (5.5 m3/s). measured over a 4-year period for this area by this study is
Collectively the 22 springs presented in Table 1 discharge about 400 mm (Kane et al. 2012). In an adjacent creek
over 20,000 l/s (20 m3/s). The significance of these (spring No. 6, Flood Creek, drainage area 207 km2), the
springs at the watershed scale can be better appreciated depth of hypothetical groundwater recharge over the basin
by the size of the aufeis formations that develops above the spring would be 233 mm in comparison again with
downstream at winter’s end (Sloan et al. 1976; Yoshikawa average annual precipitation around 400 mm. In addition to
et al. 2007). Aufeis is an important water-storage spring discharge, there is both surface runoff and evapo-
component in the winter months and influences the local transpiration from the watershed above the spring. It should
geomorphology and ecology. Sloan et al. (1976) reported be recalled that these watersheds also have extensive
that the larger aufeis formations generally developed at the permafrost that will further prevent recharge. These two
same locations each year, but the depth and areal coverage cases demonstrate that a significant amount of the ground-
varied from year-to-year. water recharge for these two springs has to originate outside
Clearly, some of these large springs discharge more water the watersheds from where the discharge occurs. The
annually than seems possible from a water-balance view- headwaters of these two watersheds border on the continen-
point at the watershed scale. In the Saviukviayak River tal divide, so the likely source (some may come from
tributary spring (spring No. 5, Table 1), the discharge was adjacent watersheds) for groundwater recharge is on the
measured at 1,529 l/s; assuming the discharge is constant south side of the continental divide in the Brooks Range.
over a period of 1 year (measurements show this to be One of the most extensive surficial rock formations in
generally true), this would be equivalent to a depth of the Brooks Range, especially on the south side of the
591 mm over the 83 km2 drainage area above the spring. divide, is limestone, conglomerate, shale, and dolomite

Fig. 10 Plot of deuterium (δD or hydrogen (δ2H)) versus oxygen (δ18O) isotopes of spring runoff in northern and central Alaska. Spring
water samples (blue circles) are similar to values for the central or southern parts of the Brooks Range. Water samples north of the
continental divide in the Brooks Range (green squares) are lighter than spring water in the same area

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

 51
(State of Alaska 2008) of the Pennsylvanian and Missis- groundwater recharge is generally much more spatially
sippian periods. As one progresses west in the Brooks distributed over larger areas.
Range, this predominantly limestone bedrock starts to thin Data on the water balance of two watersheds (close to
out around the Dalton Highway. West of the highway to the continental divide) with springs on the north side of
the Alaska coastline (Chukchi Sea) there is limited the Brooks Range have been presented here. The
limestone like bedrock outcrops; this is also an area of conclusion is that there is not sufficient recharge area
minimal environmental (meteorology, hydrology, etc.) within the watersheds to sustain the flow of the springs.
study. Springs are known to exist in the area because of Recharge could come from watersheds on each side of the
the aufeis formations that are observed at winter’s end. catchment accommodating the spring, but is more likely
However, these formations are not as extensive as those that water is coming from the other side of the continental
observed east of the Dalton Highway. This is the same divide. Also, the south-facing slopes of the Brooks Range
area where Deming et al. (1992) carried out their are more prone to having areas free of permafrost than
evaluation of regional groundwater flow based on the north-facing slopes because of the warmer climate; the
thermal data garnered from oil/gas exploration holes. They continental divide in the Brooks Range represents the
concluded that there was subpermafrost groundwater flow, transition from discontinuous to continuous permafrost.
although they did not expound on areas of recharge or The water chemistry of the springs indicates that, with
discharge. Instead, they assumed that the recharge was one or two exceptions, the groundwater has had extensive
somewhere in the Brooks Range and the discharge was at contact with limestone aquifers. Also, the chemical
the northern terminus of the permafrost along the coast. differences between springs show that different ground-
The spring chemistry (specific conductance, bicarbon- water processes are ongoing. The small amount of 14C age
ate and alkalinity) demonstrates that the pathway of dating that was carried out indicates that there is sufficient
groundwater flow is likely through bedrock dominated travel time for groundwater to travel through limestone
by limestone. The geologic map (State of Alaska 2008) aquifers from higher elevations on the south side of the
shows that there is widespread limestone in the area, Brooks Range to the points of spring discharge on the
particularly on the south side of the continental divide. north side. The two warmest springs are located the
While the areas of groundwater discharge (springs) occur farthest away from the recharge area; this fits the concept
in taliks of limited surface area, groundwater recharge of geothermal heating of deeper groundwater flow.
zones must be much more expansive and permafrost free. Finally, stable isotope ratios (δ2H and δ18O) hint that
Obviously, there is a much greater chance of permafrost- there is a match between precipitation (mostly snow) on
free surficial rock formations on the warmer south side of the south side and springs on the north side of the Brooks
the Brooks Range than on the north side. Range.
The travel time of 1,000–1,500 years seems reasonable. There is some doubt about the source of groundwater
If a typical distance (45–85 km) from the point of recharge for springs located north and out from the Brooks Range,
to the point of discharge is divided by the travel time like the Kuparuk River spring. One common characteristic
(1,000–1,500 years), a Darcy velocity range of 30–85 m/ of these springs is that their temperature is just above
year is obtained. Also, if the hydraulic gradient is 0 °C. Many of these springs also have other character-
approximated (change in elevation divided by travel istics that differ such as pH (generally lower) and Ca,
distance) for this mountainous region and divided into Mg, K and sulfate (generally lower). If the water for
the Darcy velocity, a hydraulic conductivity can be these springs is not subpermafrost groundwater, then
estimated. In this case, the range is approximately 1× the only likely source is water stored in taliks within
10−6–1×10−9 m/s. These values are in the range of values glacial moraines along the drainages, a concept that
reported for limestone and dolomite in the literature has not been discussed in the literature yet.
(Freeze and Cherry 1979). So, the estimated travel time
of 1,000–1,500 years seems plausible for this setting Acknowledgements The results reported here were not the out-
come of one specific funded project, but instead come from
measurements collected from several projects carried out in the
Conclusions area and the USGS contribution of Childers et al. (1977) in the
1970s. This work was funded in part by National Science
Foundation grants OPP-9814984, OPP-0335941, ARC-0229705.
The springs in the eastern North Slope of Alaska are clear We also acknowledge the US Fish and Wildlife Service for
evidence that a groundwater flow system exists in an permitting us to enter the wildlife refuge for sampling of springs.
environment mapped as continuous permafrost. Attempts
are made here to demonstrate that the recharge for the
groundwater system feeding the springs occurs at the
higher elevations on the south side of the Brooks Range in References
areas where limestone bedrock at the surface is extensive
and the water flows northward under the permafrost to Anderson DM, Morgenstern NR (1973) Physics, chemistry and
mechanics of frozen ground: a review. North American
discharge through taliks in the permafrost or ultimately contribution, 2nd International Conference on Permafrost,
into the Arctic Ocean (Fig. 2). While groundwater Yakutsk, U.S.S.R. National Academy of Sciences,
discharge can occur at concentrated areas like springs, Washington, DC, pp 257–288

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0

52
APHA-AWWA-WEF (1998) Standard methods for the examination Kane DL, Hinzman LD, Benson CS, Everett KR (1989) Hydrology
of water and wastewater, 20th edn. American Public Health of Imnavait creek, an arctic watershed. Holarct Ecol 12:262–
Association, Washington, DC 269
Arcone SA, Chacho EF, Delaney AJ (1998) Seasonal structure of Kane DL, Hinzman LD, McNamara JP, Zhang Z, Benson CS (2000)
taliks beneath arctic streams determined with ground-penetrat- An overview of a nested watershed study in Arctic Alaska.
ing radar. Proceedings, Permafrost, Seventh International Nord Hydrol 4(5):245–266
Conference, Yellowknife, Canada, June 1998, No. 55, Kane DL, Hinkel KM, Goering DJ, Hinzman LD, Outcalt SI (2001)
Collection Nordicana, Quebec, pp 19–24 Non-conductive heat transport associated with frozen soils.
Beikman HM, Lathram EH (1976) Preliminary geologic map of Glob Planet Chang 29(3/4):275–292
northern Alaska. MF 789, US Geological Survey, Reston, VA Kane DL, Youcha EK, Stuefer S, Toniolo H, Schnabel W,
Bradford J, McNamara JP, Bowden W, Gooseff M (2005) Gieck R, Myerchin-Tape G, Homan J, Lamb E, Tape K
Measuring seasonal thaw depth beneath arctic streams using (2012) Meteorological and hydrological data and analysis
ground penetrating radar. Hydrol Process 19:2689–2699 report for Foothills/Umiat Corridor and Bullen Projects:
Brewer MC (1958) Some results of geothermal investigations of 2006–2011. Report INE/WERC 12.01, University of Alaska
permafrost in northern Alaska. Trans Am Geophys Union 39 Fairbanks, Water and Environmental Research Center,
(1):19–26 Fairbanks, AK, 260 pp
Brosten TR, Bradford JH, McNamara JP, Zarnetske J, Bowden WB, Lachenbruch AH, Marshall BV (1986) Changing climate: geother-
Johnston ME (2009a) Estimating 3D variation in active-layer mal evidence from permafrost in the Alaskan Arctic. Science
thickness beneath arctic streams using ground-penetrating radar. 234:689–696
J Hydrol 373:479–486 Linell KA (1973) Risk of uncontrolled flow from wells through
Brosten TR, Bradford JH, McNamara JP, Gooseff MN, Zarnetske permafrost. North American contribution, 2nd International
JP, Bowden BW, Johnston ME (2009b) Multi-offset GPR Conference on Permafrost, Yakutsk, U.S.S.R. National
methods for hyporheic zone investigations. Near Surf Geophys Academy of Sciences, Washington, DC, pp 463–468
247–257 McNamara JP, Kane DL, Hinzman LD (1998) An analysis of stream
Brown J, Ferrians OJ Jr, Heginbottom JA, Melnikov ES (1998) flow hydrology in the Kuparuk River Basin, Arctic Alaska: a
Revised February 2001. Circum-Arctic map of permafrost and nested watershed approach. J Hydrol 206:39–57
ground-ice conditions. National Snow and Ice Data Center/ Pearson FJ (1965) Use of C13/C12 ratios to correct radiocarbon
World Data Center for Glaciology, Boulder, CO, Digital media ages of material initially diluted by limestone. In: Proceedings
Bureau of Land Management and Alaska Natural History of the 6th International Conference on Radiocarbon and Tritium
Association (1993) Riches from the earth, a geologic tour along Dating, Pulman, WA, June 1965, 357 pp
the Dalton Highway, Alaska. Anchorage, AK, 128 pp Pearson FJ, Hanshaw BB (1970) Sources of dissolved carbonate
Burt TP, Williams PJ (1976) Hydraulic conductivity in frozen soils. species in groundwater and their effects on carbon 14 dating. In:
Earth Surf Proc 1:349–360 Isotope hydrology 1970, IAEA Symposium 129, Vienna, March
Carey K (1973) Icings Developed from Surface and Groundwater. 1970, pp 271–286
CRREL Monograph, vol III-D3, US Army Cold Regions Rawlins MA et al (2010) Analysis of the arctic system for
Research and Engineering Laboratory, Hanover, NH, 67 pp freshwater cycle intensification: observations and expectations.
Chapman WL, Walsh JE (1993) Recent variations in sea ice and air J Clim 23:5715–5737
temperature in high latitudes. Bull Am Meteorol Soc 74(1):33–47 Rowland JC, Travis BJ, Wilson CJ (2011) The role of advective
Childers JM, Sloan CE, Meckel JP, Nauman JW (1977) Hydrologic heat transport in Talik development beneath lakes and ponds in
Reconnaissance of the Eastern North Slope, Alaska, 1975. US discontinuous permafrost. Geophys Res Lett 38:L17504.
Geol Surv Open-File Rep 77–492, 65 pp doi:10.1029/2011GL048497
Deming D, Sass JH, Lachenbruch AH, De Rito RF (1992) Heat Serreze MC, Walsh JE, Chapin FS III, Osterkamp T, Dyurgerov M,
flow and subsurface temperature as evidence for basin-scale Romanovsky V, Oechel WC, Morrison J, Zhang T, Barry RG
ground-water flow, North Slope of Alaska. Geol Soc Am Bull (2000) Observational evidence of recent changes in the northern
104:528–542 high-latitude environment. Clim Chang 46:159–207
Ferrians O (1998) Permafrost map of Alaska, USA. National Snow Sloan CE, Zenone C, Mayo L (1976) Icings along the Trans-Alaska
and Ice Data Center/World Data Center for Glaciology, Boulder, Pipeline route. US Geol Surv Prof Pap 979, 31 pp
CO, Digital media Stafford JM, Wendler G, Curtis J (2000) Temperature and
Freeze RA, Cherry JA (1979) Groundwater. Prentice Hall, precipitation of Alaska: 50year trend analysis. Theor Appl
Englewood Cliffs, NJ Climatol 67:33–44
Hall DK, Roswell C (1981) The origin of water feeding icings on State of Alaska (2008) Regional geology of the North Slope of
the Eastern North Slope of Alaska. Polar Rec 20(128):433–438 Alaska. Department of Natural Resources, Division of Oil and
Harden D, Barnes P, Reimnitz E (1977) Distribution and character Gas, State of Alaska, Juneau, AK
of Naleds in north-eastern Alaska. Arctic 30(1):28–40 Tape K, Sturm M, Racine C (2006) The evidence for shrub
Horiguchi K, Miller RD (1980) Experimental studies with frozen expansion in Northern Alaska and the Pan-Arctic. Glob Chang
soils in an “Ice Sandwich” permeameter. Cold Reg Sci Technol Biol 12:686–702
3:177–183 United States Department of Agriculture, Soil Conservation Service
Kane DL (1981) Physical mechanics of aufeis growth. Can J Civ (1963) Soil survey, Fairbanks, Alaska. USDA, Soil
Eng 6(2):186–195 Conservation Service, Washington, DC
Kane DL, Slaughter CW (1973) Recharge of a Central Alaska Lake Williams JR (1970) Ground Water in the permafrost regions of
by subpermafrost groundwater. North American contribution, Alaska. US Geol Surv Prof Pap 696, 83 pp
2nd International Conference on Permafrost, Yakutsk, U.S.S.R. Williams JR, van Everdingen RO (1973) Groundwater investiga-
National Academy of Sciences, Washington, DC, pp 458–462 tions in permafrost regions of North America: a review. North
Kane DL, Stein J (1983a) Field evidence of groundwater recharge American contribution, 2nd International Conference on
in interior Alaska. Proceedings, 4th International Permafrost Permafrost, Yakutsk, U.S.S.R. National Academy of Sciences,
Conference, Fairbanks, AK. National Academy of Sciences. Washington, DC, pp 435–446
National Academy Press, Washington, DC, pp 572–577 Yoshikawa K, Hinzman LD, Kane DL (2007) Spring and aufeis
Kane DL, Stein J (1983b) Water movement into seasonally frozen (icing) hydrology in the Brooks Range, Alaska. J Geophys Res
soils. Water Resour Res 19(6):1547–1557 112:G04S43. doi:10.1029/2006JG000294

Hydrogeology Journal (2013) 21: 41–52 DOI 10.1007/s10040-012-0937-0