Content uploaded by Raymond S Bradley
Author content
All content in this area was uploaded by Raymond S Bradley on Jan 01, 2014
Content may be subject to copyright.
MASS BALANCE AND AREA CHANGES OF FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
Geografiska Annaler · 86 A (2004) · 1
43
MASS BALANCE AND AREA CHANGES OF
FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
BY
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Climate System Research Center, Department of Geosciences, University of Massachusetts,
Amherst, Massachusetts, USA
Braun, C., Hardy, D.R. and Bradley, R.S., 2004: Mass balance
and area changes of four High Arctic plateau ice caps, 1959–2002.
Geogr. Ann., 86 A (1): 43–52.
ABSTRACT. Small, stagnating ice caps at high lati-
tudes are particularly sensitive to climatic fluctua-
tions, especially with regard to changes in ablation
season temperature. We conducted mass balance
measurements and GPS area surveys on four small
High Arctic plateau ice caps from 1999–2002. We
compared these measurements with topographic
maps and aerial photography from 1959, and with
previously published data. Net mass balance (
b
n
) of
Murray Ice Cap was –0.49 (1999), –0.29 (2000), –0.47
(2001), and –0.29 (2002), all in meters of water
equivalent (m w.eq.). The mass balance of nearby
Simmons Ice Cap was also negative in 2000 (
b
n
=
–0.40 m w.eq.) and in 2001 (
b
n
= –0.52 m w.eq.). All
four ice caps experienced substantial marginal re-
cession and area reductions of between 30 and 47%
since 1959. Overall, these ice caps lost considerable
mass since at least 1959, except for a period be-
tween the mid-1960s and mid-1970s characterized
regionally by reduced summer melt, positive mass
balance, and ice cap advance. The regional equilib-
rium line altitude (ELA) is located, on average,
above the summits of the ice caps, indicating that
they are remnants of past climatic conditions and
out of equilibrium with present climate. The ice caps
reached a Holocene maximum and were several
times larger during the Little Ice Age (LIA) and their
current recession reflects an adjustment to post-LIA
climatic conditions. At current downwasting rates
the ice masses on the Hazen Plateau will completely
disappear by, or soon after, the mid-21st century.
Introduction
Small, stagnant ice caps at high latitudes without
appreciable iceflow are particularly sensitive to cli-
matic fluctuations, especially with regard to varia-
tions in ablation-season temperature (Paterson
1969; Hattersley-Smith and Serson 1973; Rosqvist
and Østrem 1989; Grudd 1990). In a general sense,
the position of the ice margin and the areal extent
of a stagnant ice cap are strongly related to its an-
nual mass balance (Paterson 1969). Here we report
the results of recent mass balance and GPS area
measurements on four small, stagnant plateau ice
caps located on the Hazen Plateau of Ellesmere Is-
land, Nunavut, Canada (Fig. 1). Most of the plateau
is currently unglacierized and the ice caps persist
today at about the same elevation as adjacent ice-
free areas, indicating that the plateau surface is
close to the regional equilibrium-line altitude
(ELA) or glaciation level (Miller et al. 1975).
Therefore, relatively small changes in climate
could lead to profound changes in the extent of
snow and ice cover on the Hazen Plateau. Aerial
Fig. 1. Ellesmere Island, Nunavut, Canada. For reference, the dis-
tance from Murray/Simmons Ice Cap to the St. Patrick Bay ice
caps is c. 110 km, to Alert c. 160 km.
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Geografiska Annaler · 86 A (2004) · 1
44
photographs and topographic maps from 1959 and
two earlier studies of Hazen Plateau ice caps (Hat-
tersley-Smith and Serson 1973; Bradley and Ser-
reze 1987a) provide a temporal context for the cur-
rent data. We are specifically interested in assessing
how snow and ice conditions on the ice caps and the
surrounding plateau have changed since they were
last visited some 20 years ago.
Background – previous studies
The Hazen Plateau forms a large upland region,
gently rising from c. 300 m above sea level (a.s.l.)
near Lake Hazen to over 1000 m a.s.l. along the
northeast coast of Ellesmere Island (Fig. 1). This
part of Ellesmere Island is characterized by some of
the lowest accumulation rates (<0.15 m; Koerner
1979) and highest glaciation levels or ELAs (c.
800–1000 m a.s.l.; Miller et al. 1975) in the Cana-
dian High Arctic. The plateau is largely ungla-
cierized today, except for two pairs of thin, stagnant
ice bodies along its northeastern margin (Figs 1, 2,
3) which we unofficially term the Hazen Plateau ice
caps. Murray and Simmons Ice Caps together
range in elevation between c. 960 and 1100 m a.s.l.
and are surrounded by ice-free plateau areas up to
c. 1030 m a.s.l. (Fig. 2). The St. Patrick Bay ice caps
(Fig. 3; unofficial name) are located c. 110 km to
the northeast at lower elevation (c. 750–900 m
a.s.l.), possibly related to local moisture sources
(Hattersley-Smith and Serson 1973). Our studies
(1999–2001) focused on Murray Ice Cap (e.g.
Braun et al. 2001) and also included mass balance
measurements on Simmons Ice Cap in 2000 and
2001. We visited the St. Patrick Bay ice caps in a re-
connaissance survey on 15 July 2001. We re-meas-
Fig. 2. Murray and Simmons Ice Caps, Ellesmere Island, Nunavut, Canada. The summits are at 81°21'N, 69°15'W and 81°21'N, 68°50'W
respectively. Ice extent on Canada National Topographic System (NTS) map sheet 120 C/6 (gray shading) is based on aerial photographs
from 6 July 1959. The lichen trim line (dashed line) was mapped by field observations using GPS.
MASS BALANCE AND AREA CHANGES OF FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
Geografiska Annaler · 86 A (2004) · 1
45
ured the main ablation stake transect on Murray Ice
Cap on 28 July 2002. Aerial photographs from 6
July 1959 show all four ice caps fully in the ablation
zone and the Hazen Plateau entirely free of season-
al snow.
Previous studies: Murray and Simmons Ice Caps
Prior to this study, no specific glaciologic studies
had been conducted on Murray Ice Cap. An abla-
tion stake network was established on nearby Sim-
mons Ice Cap (Fig. 2) in early June 1976 (Bradley
and England 1977) when winter snow accumula-
tion across the ice cap ranged between 0.1 and 0.18
meters of water equivalent (m w.eq.). The authors
inferred that Simmons (and Murray) Ice Cap had
probably gained mass over the 1975/76 balance
year (Table 1) and experienced overall positive
mass balance and lateral ice margin advance for
some time before 1976. Only six of the original 18
ablation stakes were located during a return visit on
11 July 1983 (Bradley and Serreze 1987a). They
assumed that the other 12 stakes had melted out and
estimated a minimum net mass loss of 0.49 m w.eq.
between 1976 and 1983 (Table 1). Field observa-
tions also indicated a recession of the 1983 Sim-
mons Ice Cap margin relative to its 1959 position
(Table 2) (Bradley and Serreze 1987a).
Previous studies: St. Patrick Bay ice caps
G. Hattersley-Smith and others visited the St.
Patrick Bay ice caps (Fig. 3) in July/August 1972
(Hattersley-Smith and Serson 1973). They estimat-
ed net accumulation on the larger (NE) ice cap for
the 1971/72 balance year of c. 0.14 m w.eq. (Table
1). The seasonal snowpack overlaid icy firn and su-
perimposed ice (c. 0.39 m w.eq.), which in turn
rested on a distinct older ablation surface. This
stratigraphy was interpreted as evidence that the ice
cap experienced net ablation for an extended period
until at least 1959 and more likely until the unusu-
Fig. 3. St. Patrick Bay ice caps, Ellesmere Island, Nunavut, Canada. The summit of the larger (NE) St. Patrick Bay ice cap (STPBIC-
NE) is at c. 81°57'N, 64°10'W; the summit of the smaller (SW) St. Patrick Bay ice cap (STPBIC-SW) is at c. 81°54'N, 64°25'W. Ice extent
on Canada National Topographic System (NTS) map sheet 120 C/16 (gray shading) is based on aerial photographs from 6 July 1959.
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Geografiska Annaler · 86 A (2004) · 1
46
ally warm summer of 1962. In contrast, c. 1963 to
1972 represented a phase of net accumulation on
this ice cap, possibly interrupted by some years
with net ablation. They reported that the ice cap in
1972 ‘appears to be in a healthy state and is spread-
ing laterally as well as thickening’ (Table 2). This
positive regime however did not persist, as net an-
nual mass balance was again negative for the 1974/
75 and 1975/76 balance years (Table 1). The orig-
inal 1972 stake network was re-surveyed in 1982
by a research group from the University of Massa-
chusetts (Bradley and Serreze 1987a) as part of a 2-
year topoclimatic study of the St. Patrick Bay ice
caps and surrounding Hazen Plateau (Bradley and
Serreze 1987b; Serreze and Bradley 1987). Net
mass balance between 1972 and 1982 was –1.3 m
w.eq. (Table 1); this mass loss led to a reduction in
area of both ice caps (Table 2). Mass balance con-
ditions and summer climate differed markedly be-
tween 1982 and 1983, with 1983 being notably
colder and having more summer snowfall, resulting
in a positive annual mass balance for the 1982/83
balance year on both ice caps (Table 1).
Methods
Ice-cap mass balance
We measured ice-cap mass balance using conven-
tional glaciological techniques as described by
Østrem and Brugman (1991). We established a
network of 11 ablation stakes on Murray Ice Cap
in 1999 and expanded the network in 2000 to 29
stakes (Fig. 2). We established a network of 15
stakes on Simmons Ice Cap in 2000 (Fig. 2). Win-
ter snow accumulation was measured each year in
late May (1999–2001) and summer ablation was
measured in late July/early August (1999–2002)
and confirmed the following May (2000/2001 on-
ly). Individual stake measurements for each ice
cap were grouped into 20 m elevation bands to
determine a linear ablation gradient for each year
(cf. Rosqvist and Østrem 1989). This ablation
gradient was used to integrate the net specific ab-
lation measurements over the entire ice cap sur-
face, based on a 10 m digital elevation model con-
structed from a 1:50000 topographic map (Fig. 2)
(cf. Jansson 1999). We consider ±0.2 m as a con-
servative uncertainty estimate for the annual net
mass balance values following Cogley and Ad-
ams (1998). We calculated minimum mass bal-
ance estimates for Simmons Ice Cap (1984–
1998) and the St. Patrick Bay ice caps (1984–
2000) using the mean remaining depth of stake
insertion into the ice in 1983 (M. Serreze, pers.
comm.) and assuming a relative ice density of 0.9
(Table 1).
Table 1. Net mass balances (m w.eq.) of the Hazen Plateau ice caps. Where a value represents
a multiyear period, the average annual value is shown in parentheses. * denotes a minimum es-
timate. Qualitative field observations are indicated by italics.
Balance year Murray Simmons
or period Ice Cap Ice Cap STPBIC-NE STPBIC-SW
1963–1971 0.39 (0.04)
1
1972 0.14
1
1973 0.14
2
1974 0.14
2
1975 –0.08
2
1976 positive
3
positive
3
–0.07
2
1972–1982 –1.3 (–0.14)
4
1976–1983 *–0.49 (–0.08)
4
1982 –0.14
4
1983 positive
4
0.14
4
0.17
4
1984–1998 *–0.49 (–0.03)
1984–2000 *–1.01 (–0.06) *–1.26 (–0.07)
1999 –0.49 negative
2000 –0.29 –0.40
2001 –0.47 –0.52 negative negative
2002 –0.29 negative
Sources:
1
Hattersley-Smith and Serson (1973),
2
Ommanney (1969, 1977),
3
Bradley and Eng-
land (1977),
4
Bradley and Serreze (1987a). Data for STPBIC-NE are available at
ftp://ftp.trentu.ca/pub/gghydro (Cogley and Adams 1998).
MASS BALANCE AND AREA CHANGES OF FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
Geografiska Annaler · 86 A (2004) · 1
47
Ice-cap area
We digitized the 1959 ice margins of the four
Hazen Plateau ice caps directly from available
1:50000 topographic maps (Figs 2, 3), scanned at
600 dpi and registered to UTM zone 19N (20N for
the St. Patrick Bay ice caps). The topographic maps
used are based on aerial photography from 6 July
1959. We visually confirmed the accuracy of the
ice-cap outlines depicted on the topographic maps
by detailed comparison with the original aerial
photographs (see below).
We surveyed the 1999–2001 ice-cap margins and
lichen trim lines on foot (or snowmobile) using a
portable GPS receiver, logging discreet positions
every 3–10 s (10–15 m). The points along each ice-
cap ‘trace’ were imported into a geographical infor-
mation system (GIS) software package and con-
nected as polygons for area calculations. The 1999
and 2000 GPS positions collected for Murray Ice
Cap were differentially corrected using data from
the nearest available GPS base station (Thule AFB,
Greenland, 76°20'N, 68°48'W). This ‘low-tech’
technique eliminates the need to operate a dedicated
GPS base station on-site, a significant advantage in
remote environments. The main disadvantages are
(1) greater uncertainties associated with the differ-
entially corrected GPS positions compared to more
sophisticated techniques, and (2) the dependence on
consistent base station data availability. The latter
problem was illustrated in 2001, when we were not
able to correct the four collected ice-cap traces be-
cause of partially missing base station data.
Ice-cap area – uncertainties
We assessed the uncertainties associated with our
ice-cap area measurements (Table 2) by first quan-
tifying each individual contributing error source
(Table 3) and then calculating the resultant uncer-
tainty for the position of the ice margin (Table 4).
It is important to note that some of the absolute val-
ues assigned to individual uncertainties listed in Ta-
ble 3 are themselves estimates. Furthermore, pos-
sible human errors and subjectivity associated with
the creation of the topographic maps from aerial
photographs cannot be rigorously quantified.
Uncertainties for the 1959 ice-cap area measure-
ments were a combination of (1) registration errors
of the scanned topographic maps relative to their
respective coordinate system, and (2) generaliza-
tion of the ice-cap margins during the digitization
process. Inherent in this type of study are errors and
uncertainties associated with the delineation of the
ice-cap margin, whether it is on the original aerial
photograph, the topographic map, or directly in the
field. A certain amount of subjective generalization
and human error is inevitable in this process and we
Table 2. Ice cap area (km
2
) of the Hazen Plateau ice caps and uncertainty estimates 1959–2002.
Qualitative field observations are indicated by italics.
Murray Simmons
Year Ice Cap Ice Cap STPBIC-NE STPBIC-SW
Area
1959 4.37
1
7.45
1
7.48
1
2.93
1
1972 advance
2
1976 advance
3
advance
3
1978 6.69 (89%)
4
2.74 (93%)
4
1983 recession
5
1999 3.28 (75%)
2000 3.15 (72%)
2001 3.05 (70%) 3.94 (53%) 4.61 (62%) 1.72 (59%)
2002 recession recession
Uncertainty estimate
1959 ±1.3% ±1.3% ±1.1% ±1.7%
1978 N/A N/A
1999 ±1.7%
2000 ±1.8%
2001 ±2.7% ±3.2% ±2.2% ±3.8%
Sources:
1
Topographic map,
2
Hattersley-Smith and Serson (1973),
3
Bradley and England
(1977),
4
Bradley and Serreze (1987a).
The updated values for 2003 are available at www.geo.umass.edu/climate/hazen/field23.html
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Geografiska Annaler · 86 A (2004) · 1
48
estimated this uncertainty at ±2 pixel or c. 5 m (Ta-
ble 3), based on careful comparisons of the original
aerial photographs, the topographic maps, and the
actual ice margin in the field. Wind-drifted snow
accumulations along the northeast margin of STP-
BIC-SW and along the terminus of STPBIC-NE,
both on the 1959 aerial photographs and in 2001,
made it difficult to determine the precise positions
of the ice margins at these locations. For consist-
ency, we mapped ‘maximum area’ solutions in both
cases in 2001. The uncertainty for each differen-
tially corrected GPS position collected in 1999 and
2000 was c. 5 m (Table 3), which represents the
maximum 99% confidence interval for the correct-
ed GPS positions (generated by the differential-
correction software). The horizontal error associat-
ed with the 2001 uncorrected GPS positions was
estimated to be c. 9.4 m (99% confidence limit of
22739 positions collected over 5 days at a fixed
point). It is interesting to note in this context that
the difference in Murray Ice Cap area between the
differentially corrected and the uncorrected 2000
trace was less than 100 m
2
(<0.1‰). We deter-
mined the resulting uncertainty for ice-cap area by
applying an area-buffer around the digitized ice-
cap margins using a GIS software package, calcu-
lated as the quadratic sum of the individual contrib-
uting uncertainties (Table 4). The final values for
the ice-cap area uncertainty estimates (Table 2) are
a function of the applied area-buffer, but are also af-
fected by ice-cap area and the length/irregularity of
each ice-cap margin. They clearly represent worst-
case estimates, as the area-buffer assumes that all
points defining the ice margin are systematically
displaced to induce maximum area change. In re-
ality, we can expect a certain amount of error can-
cellation in terms of total ice-cap area.
Results
Ice-cap mass balance
Murray and Simmons Ice Caps experienced highly
negative annual mass balances (–0.29 to –0.49 m
w.eq.) for at least the past four years (Table 1). Win-
ter snow accumulation on both ice caps was rela-
tively constant each year (0.06–0.1 m w.eq. 1999–
2001), and variations in annual net mass balance
were mainly a function of summer conditions.
Summer climatic conditions in 2000 and 2002 were
generally colder and snowier than in 1999 or 2001,
leading to less negative annual mass balance on the
ice caps (Table 1). We were not able to recover the
six ablation stakes remaining on Simmons Ice Cap
in 1983 (from the original 1976 network), but one
was found melted 10–20 cm horizontally into the
glacier surface. We were also unable to locate any
of the ablation stakes from the 1972 and 1982 net-
works on the St. Patrick Bay ice caps during our
visit on 15 July 2001 and assume that they had
melted out as well. These observations suggest an
overall negative mass balance for these three ice
caps since at least 1984 (Table 1). The Hazen Pla-
teau ice caps presently do not retain any accumu-
lation of snow, firn, or superimposed ice, even on
their highest or most-sheltered parts. The entire
surface of Murray and Simmons Ice Caps at the end
Table 3. Types and magnitudes of individual uncertainties associated with ice-cap area measurements.
1
Pixel size: 2.1 m.
Source of uncertainty Type of uncertainty Magnitude of uncertainty
Coordinate system registration Registration error ±2 pixel
1
(generated by GIS software)
Ice margin delineation Generalization/subjectivity ±2 pixel
1
(map), ±5 m (field) (both estimated)
Differential GPS Individual position ±5 m (1999), ±5.2 m (2000) (generated by DGPS software)
Uncorrected GPS Individual position ±9.4 m (2001) (estimated)
Table 4. Individual contributions of the uncertainties and resultant ice margin buffer.
Year Ice cap Registration Generalization DGPS GPS Buffer
1959 All ±4.2 m ±4.2 m ±6 m
1999 MIC ±5 m ±5 m ±7.1 m
2000 MIC ±5 m ±5.2 m ±7.2 m
2001 All ±5 m ±9.4 m ±10.6 m
MIC, Murray Ice Cap
MASS BALANCE AND AREA CHANGES OF FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
Geografiska Annaler · 86 A (2004) · 1
49
of each summer (1999–2002) and of the St. Patrick
Bay ice caps (observed only 2001) was dirty, bare
glacier ice characterized by accumulations of
wind-blown dust in well-developed cryoconite
holes – all suggesting net ablation over an extended
period of time.
Ice-cap area changes
All four Hazen Plateau ice caps experienced con-
siderable marginal recession since at least 1959
(Figs 2, 3). Marginal recession was greatest (up to
c. 700 m) for the flat, low-lying parts of the ice caps
and less along the steeper and sheltered sections of
the ice margins. This presumably was due to local
increases in snow accumulation related to wind
drifting. This retreat of the ice margins led to de-
creases in overall ice-cap area amounting to be-
tween 30 and 47% since 1959 (Table 2). The mar-
gin of Murray Ice Cap retreated 10–30 m each year
in 1999–2001, resulting in an annual area reduction
of c. 2.5% (Table 2). The margins of all four ice
caps were visually thinning and rapidly disintegrat-
ing over the course of each summer. This was vis-
ibly illustrated by one section of the Simmons Ice
Cap margin which retreated c. 10 m over 15 days
in late July 2001. In addition, two small holes (c.
200 m
2
) developed in the SW-lobe of Simmons Ice
Cap at c. 1030 m a.s.l. during July/August 2001
(i.e. ice-free area), which are likely to accelerate ice
margin disintegration and retreat in the coming
years. We were unable to conduct quantitative area
measurements in 2002, but field observations indi-
cated a continued recession of Murray Ice Cap of
c. 40 m at its terminus in this year. The SW lobe of
Simmons Ice Cap in 2002 was almost completely
separated from the main ice cap at an elevation of
c. 1030 m a.s.l. (Fig. 2; see also:
www.geo.umass.edu/climate/hazen/sic_99_02.html).
Discussion
Our new data, in combination with previously pub-
lished work (Tables 1, 2) allow a generalized re-
construction of the Hazen Plateau ice caps’ mass
balance history for the last four decades (Fig. 4a).
The ice caps experienced net ablation and shrink-
age for an extended period of time until some time
in the early to mid-1960s (Hattersley-Smith and
Serson 1973). This was followed by a phase of net
accumulation and ice-cap growth until the early to
mid-1970s (Hattersley-Smith and Serson 1973;
Bradley and England 1977). Since that time, the ice
caps have again experienced overall net mass loss
and marginal recession. There is evidence for some
inter-annual variations in mass balance superim-
posed on the general trend (e.g. 1982/83), as well
as for spatial variability across the Hazen Plateau
(e.g. 1976) (Table 2).
This general temporal pattern was also exhibited
by other glaciers studied in the Canadian High Arc-
tic, with generally positive mass balances from the
mid-1960s to the mid-1970s, followed by generally
negative mass balances thereafter (e.g. Fig. 4b).
This documented increase in glacierization across
much of the High Arctic coincided with a period of
F
i
g.
4
.
(
a
)
G
enera
li
ze
d
mass
b
a
l
-
ance history of the Hazen Plateau
ice caps. (b) Annual net mass bal-
ance of the White Glacier (WG)
(cf. Fig. 1). (c) July freezing level
heights at Alert (1951–2001), with
mean of 1150 m a.s.l. shown as a
dotted line (1993 data missing).
Gray shading indicates the eleva-
tion range of the Hazen Plateau ice
caps (c. 800–1100 m). Mass bal-
ance data for White Glacier are
from www.trentu.ca/geography/
glaxmbal.htm; White Glacier is
considered a representative exam-
ple of Canadian High Arctic gla-
ciers (Cogley et al. 1996). Vertical
lines through all three plots repre-
sent the period of reduced summer
melt and increased glacierization
in the High Arctic from c. 1964 to
1976.
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Geografiska Annaler · 86 A (2004) · 1
50
reduced summer melt conditions and increased an-
nual precipitation (Bradley and Miller 1972; Bra-
dley 1973; Bradley and England 1978; Alt 1987).
Corroborating this are upper air sounding data (Fig.
4c) from the closest official Canadian weather sta-
tion at Alert (Fig. 1), which show a decrease in July
freezing level height of c. 100 m between 1964 and
1976 relative to the long-term (1951–2001) mean
of 1150 m a.s.l. July freezing level heights have
also been shown to be highly correlated with gla-
cier ELAs and mass balance (Bradley 1975; Brad-
ley and England 1978) in the Canadian High Arctic.
This comparatively small shift in climate and low-
ering of the ELA was evidently sufficient for the
Hazen Plateau ice caps to experience predominant-
ly positive mass balance years and expansion dur-
ing this period from the mid-1960s to the mid-
1970s (cf. Bradley 1975). However, the regional
ELA appears to be located, on average, above the
summits of the ice caps for at least the last c. 50
years (Fig. 4c). This suggests that the contempo-
rary climatic conditions on the Hazen Plateau are
not severe enough to sustain permanent ice cover
on the plateau (cf. Ohmura et al. 1992). These find-
ings support the interpretation by Bradley and Ser-
reze (1987a) that the Hazen Plateau ice caps are out
of equilibrium with current climate.
The overall cumulative mass balance of the
Hazen Plateau ice caps and other Canadian High
Arctic glaciers (Koerner 1995; Dowdeswell et al.
1997; Serreze et al. 2000) has been negative for the
last c. 40 years, with a turn towards even more neg-
ative values during the 1990s. This mass loss has led
to a retreat of the ice margins and resulted in shrink-
age of the Hazen Plateau ice caps since at least 1959
(Fig. 5). There is additional evidence from snow-pit
and firn-core studies (Hattersley-Smith 1963) for
elevated summer temperatures and increased melt-
ing in the period from c. 1925 to 1961, suggesting
overall more negative glacier mass balances in the
Canadian High Arctic during the first part of the
20th century, compared to the last 40 years of direct
measurements (Koerner 1995).
The Hazen Plateau ice caps are to some extent
end-members of the full glacier–climate response
spectrum in the sense that their response to a given
climatic perturbation is relatively more extensive
and rapid than in the case of larger, more dynamic
High Arctic ice caps or glaciers. These ice caps ap-
pear to have formed relatively recently (Koerner
1989) during the so-called ‘Little Ice Age’ (LIA, c.
1600–1850). They probably maintained their max-
imum Neoglacial extents as late as c. 1925, similar
to other small glaciers and ice caps on northern
Ellesmere Island (Hattersley-Smith 1969). Evi-
dence for an increased extent of ice and/or peren-
nial snow on the Hazen Plateau at some point in the
recent past, probably the LIA, is also provided by
a well-defined lichen trim line around Murray Ice
Cap (c. 9.6 km
2
) and on the plateau between Mur-
ray and Simmons Ice Caps (Fig. 2). A similar li-
chen trim line is evident around Simmons Ice Cap
and the St. Patrick Bay ice caps, but has not yet been
mapped in detail. Additional plateau surfaces of
comparable elevation in this region probably also
supported small ice caps or perennial snowfields at
that time. In this context the period of overall pos-
itive glacier mass balances from the mid-1960s to
mid-1970s may provide a useful analogue for the
reduced summer melt conditions in the High Arctic
during the LIA (cf. Alt 1987).
Conclusions
The Hazen Plateau ice caps have experienced con-
siderable marginal recession and significant overall
mass loss since at least 1959. The sensitivity of
these ice caps to changes in climate is enhanced by
(1) the low amounts of winter snow accumulation,
(2) the absence of iceflow, and (3) the small vertical
relief. They are out of equilibrium with modern cli-
mate and considered to be relicts of past climatic
conditions with reduced summer melt and/or in-
Fig. 5. Area reduction of the Hazen Plateau ice caps 1959–2001,
with a generalized linear projection suggesting complete disap-
pearance of the ice caps by mid-21st century or soon thereafter.
MASS BALANCE AND AREA CHANGES OF FOUR HIGH ARCTIC PLATEAU ICE CAPS, 1959–2002
Geografiska Annaler · 86 A (2004) · 1
51
creased snowfall although winter snowfall varia-
tions appear to be largely inconsequential in terms
of annual mass balance today. The ice caps proba-
bly formed during the LIA and will continue to lose
mass and retreat under current climatic conditions.
They are likely to disappear over the next few dec-
ades, unless climatic conditions deteriorate as they
did in the mid-1960s. The decay of these ice caps
is likely to accelerate in the near future due to feed-
back processes such as ice-margin disintegration.
This study demonstrates that even infrequent mass
balance and ice-area measurements can be useful in
assessing the general mass balance regime of High
Arctic glaciers, especially if some additional his-
torical information is available.
Acknowledgements
Research was supported by a US National Science
Foundation Grant (OPP-9819362) to the Universi-
ty of Massachusetts. The Polar Continental Shelf
Project (Energy, Mines, and Resources Canada)
provided superb logistical and generous equipment
support. We also thank Parks Canada (Nunavut
Field Unit) for continuous support. M. Serreze gen-
erously shared his original field data and F. Keimig
provided invaluable help with the data analysis.
The thoughtful and detailed comments and sugges-
tions of two reviewers greatly improved an earlier
version of this manuscript.
Carsten Braun, Douglas R. Hardy, Raymond S.
Bradley, Climate System Research Center, Depart-
ment of Geosiences, Morrill Science Center, 611
North Pleasant Street, University of Massachu-
setts, Amherst, MA 01003, USA.
E-mail: carsten@geo.umass.edu
References
Alt, B.T., 1987: Developing Synoptic Analogs for Extreme Mass
Balance Conditions on Queen Elizabeth Island Ice Caps. J.
Clim. Appl. Meteorol., 26(11): 1605–1623.
Bradley, R.S., 1973: Recent Freezing Level Changes and Climatic
Deterioration in the Canadian Arctic Archipelago. Nature,
243: 398–400.
Bradley, R.S., 1975: Equilibrium-line altitudes, mass balance, and
July freezing-level heights in the Canadian High Arctic. Jour-
nal of Glaciology, 14(71): 267–274.
Bradley, R.S. and England, J., 1977: The Simmonds Ice Cap. In:
Bradley, R.S. and England, J., (eds): Past glacial activity in the
High Arctic. Contribution No. 31. Department of Geology
and Geography, University of Massachusetts. 177–182.
Bradley, R.S. and England, J., 1978: Recent climatic fluctuations
of the Canadian High Arctic and their significance for glaci-
ology. Arctic and Alpine Research, 10(4): 715–731.
Bradley, R.S. and Miller, G.H., 1972: Recent Climatic Change
and Increased Glacierization in the Eastern Canadian Arctic.
Nature, 237: 385–387.
Bradley, R.S. and Serreze, M.C., 1987a: Mass balance of two
High Arctic plateau ice caps. Journal of Glaciology, 33(113):
123–128.
Bradley, R.S. and Serreze, M.C., 1987b: Topoclimatic studies of
a High Arctic plateau ice cap. Journal of Glaciology, 33(114):
149–158.
Braun, C., Hardy, D.R. and Bradley, R.S., 2001: Recent recession
of a small plateau ice cap, Ellesmere Island, Canada. Journal
of Glaciology, 47(15): 154.
Cogley, J.G. and Adams, W.P., 1998: Mass balance of glaciers
other than ice sheets. Journal of Glaciology, 44(147): 548–
563.
Cogley, J.G., Adams, W.P., Ecclestone, M.A., Jung-Rothenhaeu-
sler, F. and Ommanney, C.S.L., 1996: Mass balance of White
Glacier, Axel Heiberg Island, N.W.T., Canada, 1960–91.
Journal of Glaciology, 42(142): 548–563.
Dowdeswell, J.A. et al., 1997: The Mass Balance of Circum-Arc-
tic Glaciers and Recent Climate Change. Quaternary Re-
search, 48: 1–14.
Grudd, H., 1990: Small glaciers as sensitive indicators of climatic
fluctuations. Geografiska Annaler, 72A(1): 119–123.
Hattersley-Smith, G., 1963: Climatic inferences from firn studies
in northern Ellesmere Island. Geografiska Annaler, 45(2–3):
139–151.
Hattersley-Smith, G., 1969: Glacial features of Tanquary Fiord
and adjoining areas of northern Ellesmere Island, N.W.T.
Journal of Glaciology, 8(52): 23–50.
Hattersley-Smith, G. and Serson, H., 1973: Reconnaissance of a
small ice cap near St Patrick Bay, Robseson Channel, North-
ern Ellesmere Island, Canada. Journal of Glaciology, 12(66):
417–421.
Jansson, P., 1999: Effect of uncertainties in measured variables
on the calculated mass balance of Storglaciaren. Geografiska
Annaler, 81A(4): 633–642.
Koerner, R.M., 1979: Accumulation, ablation and oxygen isotope
variations on the Queen Elizabeth Islands ice caps, Canada.
Journal of Glaciology, 22(86): 25–41.
Koerner, R.M., 1989: Queen Elisabeth Islands Glaciers. In: Ful-
ton, R.J., (ed.): Quaternary Geology of Canada and Green-
land. Geology of Canada, (No. 1). Geological Survey of Can-
ada. 464–477.
Koerner, R.M., 1995: Canadian Arctic. In: Jania, J. and Hagen,
J.O., (eds): Report on mass balance of Arctic glaciers. Sos-
nowiec/Oslo, International Arctic Science Committee. Work-
ing Group on Arctic Glaciology. 5–8.
Miller, G.H., Bradley, R.S. and Andrews, J.T., 1975: The glacia-
tion level and lowest equilibrium line altitude in the High Ca-
nadian Arctic: Maps and climatic interpretation. Arctic and
Alpine Research, 7(2): 155–168.
Ohmura, A., Kasser, P. and Funk, M., 1992: Climate at the equi-
librium line of glaciers. Journal of Glaciology, 38(130): 397–
411.
Ommanney, C.S.L., 1969. A study in glacier inventory: The ice
masses of Axel Heiberg Island, Canadian Arctic Archipelago.
Axel Heiberg Island Research Reports, Glaciology 3. McGill
University, Montreal. 105 p.
Ommanney, C.S.L., 1977: Quadrennial report to the Permanent
Service on the Fluctuations of Glaciers on Canadian glacier
variations and mass balance changes. Fisheries and Environ-
ment Canada. Inland Waters Directorate, Glaciology Divi-
sion. Ottawa.
Østrem, G. and Brugman, M., 1991: Glacier mass-balance meas-
urements. A manual for field and office work. National Hy-
CARSTEN BRAUN, D.R. HARDY AND R.S. BRADLEY
Geografiska Annaler · 86 A (2004) · 1
52
drology Research Institute Science Report 4. Environment
Canada, Saskatoon. 224 p.
Paterson, W.S.B., 1969: The Meighen Ice Cap, Arctic Canada:
Accumulation, Ablation and Flow. Journal of Glaciology,
8(54): 341–352.
Rosqvist, G. and Østrem, G., 1989: The sensitivity of a small ice
cap to climatic fluctuations. Geografiska Annaler, 71A(1):
99–103.
Serreze, M.C. and Bradley, R.S., 1987: Radiation and cloud ob-
servations on a High Arctic plateau ice cap. Journal of Gla-
ciology, 33(114): 162–168.
Serreze, M.C. et al., 2000: Observational evidence of recent
change in the Northern High-Latitude environment. Climatic
Change, 46: 159–207.
Manuscript received September 2002, revised and accepted May
2003.