CHAPTER
8
Glacier fluctuations and dynamics around
the margin of the Greenland Ice Sheet
Leigh A. Stearns and Hester Jiskoot
ABSTRACT
Greenland’s ice cover has undergone remarkable
changes in the last two decades as a response to
forcing from the atmosphere and ocean. This period
coincides with the evolution of remote-sensing platforms towards higher spatiotemporal resolutions.
In this chapter, we give an overview of some of
the key glaciological findings emerging from Greenland in the past two decades, and describe two case
studies in which GLIMS data are used to provide
new insights into regional changes in Greenland’s
glaciers.
8.1
GREENLAND GLACIOLOGY
Greenland—the largest relic of Pleistocene glaciation in the Northern Hemisphere—contains enough
ice ( 2:9 10 6 km 3 ) to raise global mean sea level
by 7 m if it melted completely (Bamber et al. 2001).
Peripheral to the ice sheet thousands of smaller
local ice masses exist, including ice caps, valley
glaciers, mountain glaciers, ice fields, and glacierets.
Local ice caps and local glaciers are difficult to
delineate, especially since many are not detached
ice bodies, but connected to the ice sheet, separated
by ice flow divides (Yde 2011). Best estimates of the
total ice cover on Greenland are between 1,801,000
to 1,824,000 km 2 , of which 88,000 to 90,000 km 2
are local ice masses (Kargel et al. 2012, Rastner et
al. 2012, Citterio and Ahlstrøm 2013). Ice discharge
through ice sheet outlet glaciers accounts for 70%
of the mass discharge in Greenland (Rignot and
Kanagaratnam 2006). Many of these outlet glaciers
are undergoing rapid changes in dynamics, driven
by changes in atmospheric and ocean forcings.
The mechanisms driving the large changes in
Greenland glacier dynamics are not fully understood. The near-coincident timing of the changes
observed on Kangerdlugssuaq and Helheim glaciers
in East Greenland (Stearns and Hamilton 2007),
several smaller outlet glaciers in the southeast
(Rignot and Kanagaratnam 2006), and Jakobshavn
Isbræ in the west (Joughin et al. 2004), suggests a
common trigger such as climate warming may be
responsible. These observations highlight the sensitivity of large outlet glaciers to climate-related perturbations and imply that current estimates for the
predicted sea level contribution from the Greenland
Ice Sheet need to be reevaluated to account for
rapid changes in ice dynamics.
Over the past 30 years, Arctic surface temperatures have increased 0.5 C per decade (Gillett
et al. 2008). This warming has been accompanied
by other changes: sea ice extent has decreased
8.6 2.9% (Serreze et al. 2007) and Northern
Hemisphere ocean temperatures have warmed by
0.19 0.13 C (Polyakov et al. 2004) in the past
decade. In Greenland alone, surface observations
184
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
indicate recent increases in temperature (1.5 C per
decade; Box et al. 2006), seasonal ablation (16% per
decade; Fettweis et al. 2011, Mernild et al. 2011),
surface meltwater runoff (19% per decade; Hanna
et al. 2005, Box et al. 2006), mass flux due to outlet
glacier acceleration (140% from 2000 to 2005;
Rignot and Kanagaratnam 2006), and overall net
mass loss. Current models (even those excluding the
dynamic response of outlet glaciers) predict collapse
of the Greenland Ice Sheet if surface temperatures
increase by 3 C (Church et al. 2001, Gregory et al.
2004). When enhanced surface melt and subsequent
glacier accelerations are included in models, the
Greenland Ice Sheet becomes more sensitive to
warming temperatures (Parizek and Alley 2004).
8.1.1 Ice sheet mass changes
Repeat satellite observations show that the Greenland Ice Sheet (GIS) lost mass at an accelerating
rate during the late 1990s through the mid-2000s
(e.g., Rignot and Kanagaratnam 2006, Velicogna
2009). Mass is added to the ice sheet by precipitation, and lost by surface and basal melting, and ice
flux across the grounding line (usually resulting in
iceberg calving). Recent studies report that Greenland lost approximately 290 Gt yr1 in 2009, an
acceleration of 30 Gt yr2 over the period
2002–2009 (Velicogna 2009). At least half of this
increase in mass loss is due to the acceleration of
several large marine-terminating outlet glaciers,
some of which are discharging two to four times
more ice into the ocean than they were in the early
1990s (Rignot and Kanagaratnam 2006, van den
Broeke et al. 2009).
The mass balance of the GIS can be determined
by the mass budget approach or geodetic
approaches (using geodetic estimates of volume
change to infer mass change). The mass budget
approach compares catchment-wide accumulation
rates with ice flux across a prescribed gate at or near
the grounding line. It requires knowledge of the
surface mass balance (snowfall minus surface
ablation), which is reconstructed from atmospheric
circulation models and in situ records such as ice
cores. The discharge flux requires estimates of outlet glacier velocity and ice thickness (cross-sectional
area of the gate). Studies employing this technique
predominantly rely on interferometric synthetic
aperture radar data (InSAR) ice velocities because
of their extensive spatial coverage, but catchmentspecific studies use optical imagery (e.g., Stearns et
al. 2005) or GPS data (Hamilton and Whillans
2000, Thomas et al. 2001) to determine ice flux.
Snow accumulation is one of the most difficult
parameters to constrain (e.g., Eisen et al. 2008,
Ettema et al. 2009). There are approximately 500
in situ surface mass balance measurements in
Greenland, covering a range of time periods and
of varying quality (Ettema et al. 2009). Atmospheric reconstructions, constrained by in situ
measurements are used to give spatial coverage of
ice sheet surface mass balance.
The main geodetic approaches used in determining mass change over GIS come from the timevariable gravity data collected by the Gravity
Recovery and Climate Experiment (GRACE) and
airborne or satellite altimeters. The GRACE satellite pair have generated gravity estimates of Greenland approximately every month since early 2002.
Ice mass change is one of the variables that impacts
the gravity signal over Greenland. In order to
isolate this component, corrections must be applied
for other variables that impact the gravity field:
redistribution of mass in the atmosphere, the ocean,
the crust and mantle—glacial isostatic adjustment
(GIA)—and water/snow/ice stored on land (but not
connected to the ice sheet) (Velicogna et al. 2005).
Combined, these variables are referred to as mass
‘‘leakage’’. Deriving GIS mass change estimates
from GRACE measurements is complicated by
the limited spatial resolution and nonrandom noise
inherent in the gravity data. There are fundamental
differences in the approach of various studies for
deriving GIS mass change from GRACE gravity
data and accounting for leakage errors (Velicogna
et al. 2005, Chen et al. 2006, Luthcke et al. 2006,
Ramillien et al. 2006, Wouters et al. 2008). These
processing differences yield a range in mass loss
estimates that are greater than the associated errors.
However, the regional trend of mass change
appears consistent.
To provide further constraints on GIS mass loss
estimates, GRACE solutions are combined with
surface elevation data (ICESat; Slobbe et al.
2009, Ewert et al. 2011), the mass budget approach
(Rignot et al. 2011), or geodetic uplift rates (Khan
et al. 2010). Smaller sized basins are also being
investigated (Wouters et al. 2008, Chen et al.
2011, Schrama and Wouters 2011), so that mass
changes can be more easily connected to changes
in forcing mechanisms.
Laser altimeters measure precise surface elevation (h) along aircraft or satellite ground tracks.
Repeat altimeter measurements provide an estimate
of the rate of elevation change with time (dh=dt)
Greenland glaciology 185
which can be used to estimate mass balance on the
scale of the ice sheet (e.g., Krabill et al. 2000) or
basin. Elevation changes are converted to volume
and mass changes by correcting for coincident
changes in surface mass balance, firn compaction,
and crustal motions (e.g., Zwally and Li 2002,
Thomas et al. 2006, Wingham et al. 2006).
The geodetic and mass budget approaches both
reveal an acceleration of mass loss in GIS of 19 Gt
yr2 from 1992 to 2010 (30 Gt yr2 for the period
2002–2010 covered by GRACE; Rignot et al. 2011).
Between 2000 and 2005, the ice loss was largest in
southeast Greenland, largely due to the increase in
ice discharge from Helheim and Kangerdlugssuaq
glaciers (e.g., Rignot and Kanagaratnam 2006,
Howat et al. 2007, Stearns and Hamilton 2007).
Since 2006, ice loss in southeast Greenland has
decelerated, and ice loss in northwest Greenland
has accelerated, according to both the GRACE
record (Chen et al. 2011, Rignot et al. 2011,
Schrama and Wouters 2011), and GPS observations
of elastic rebound (Khan et al. 2010).
8.1.1.1
Surface mass balance
Estimates of GIS mass balance require knowledge
of surface mass balance (SMB)—the annual sum
of mass accumulation (snowfall, rain) and ablation
(sublimation, runoff ). In the past two decades, the
SMB in Greenland has decreased rapidly—driven
by increases in surface melting and runoff (e.g.,
Ettema et al. 2009). Surface melt runoff now
accounts for roughly half the annual mass loss from
Greenland (the other half is by iceberg calving)
(e.g., Zwally and Giovinetto 2001, Rignot and
Kanagaratnam 2006, Hanna et al. 2008, van den
Broeke et al. 2009).
Precipitation on GIS falls predominantly as snow
(94%), with some rain (6%) in coastal areas in the
south (Ettema et al. 2009). Zones of high accumulation exist where low-pressure systems extend over
the ice sheet in southeast Greenland (the Icelandic
Low), and migrate north along the west coast
(Ettema et al. 2009). Generally, accumulation
increases from north to south, and is higher in
southern coastal areas (Bales et al. 2009). Improvements in annual accumulation estimates from
shallow ice cores (e.g., Banta and McConnell
2007), snow pits and weather stations (e.g., Bales
et al. 2009) illuminate details of these spatial accumulation patterns. There is no significant long-term
trend in either regional or ice sheet–wide accumula-
tion rates in the past 50 years (Hanna et al. 2006),
according to climate reanalysis models.
Ablation (non–iceberg calving mass loss) in
Greenland is dominated by runoff (90%), with
the remainder coming from evaporation and sublimation (Ettema et al. 2009). The increase in runoff
coincides with the steady lengthening of the melt
season. Using a spatially distributed meteorological
snow and ice model, Mernild et al. (2011) determined that the length of the melt season on GIS
has increased at a rate of 2 days per year since 1972,
yielding an extended melt period of 70 days between
1972 and 2010. A consequence of the extended melt
season, combined with the increased surface temperatures in GIS, is the observed increase in surface
melting (both volume and area) in the past two
decades. Surface melt trends are quantified through
satellite-based observations (e.g., Mote 2007,
Tedesco 2007, Hall et al. 2008, Wouters et al.
2008, Bhattacharya et al. 2009) and model reconstructions (e.g., Fettweis et al. 2007, Hanna et al.
2008, Ettema et al. 2009, van den Broeke et al. 2009)
and generally show a near-linear increase in melt
area and volume, punctuated by a decrease in melt
area in 1995 (Bhattacharya et al. 2009) and
increases in 2002 (Steffen 2004), 2007 (Mote
2007), and 2010 (Tedesco et al. 2011). At higher
elevations, meltwater refreezes in the winter snowpack, so an increase in melt extent does not necessarily coincide with an increase in runoff.
The annual average SMB for GIS for 1958–2007,
derived from high-resolution climate modeling, is
an estimated 469 41 Gt yr1 —the sum of accumulation (697 Gt yr1 from snow accumulation; 46
Gt yr1 from rainfall) and ablation (248 Gt yr1
from runoff and 26 Gt yr1 from evaporation/
sublimation) (Ettema et al. 2009).
8.1.1.2
Ice discharge
In the past decade, over half of Greenland’s net
mass loss has been due to the acceleration and
retreat of outlet glaciers (van den Broeke et al.
2009). Between 1992 and 2009, ice discharge
increased steadily by 9.0 1 Gt yr2 (Rignot et
al. 2011). This steady mass loss trend is punctuated
by the acceleration of several large marineterminating outlet glaciers; between 1998 and
2005, Jakobshavn Isbræ in West Greenland, and
Helheim and Kangerdlugssuaq glaciers in East
Greenland underwent dynamic changes in flow that
nearly doubled annual mass loss (Joughin et al.
2004, Stearns and Hamilton 2007). Current trends
186
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
in dynamic mass loss may be episodic or short lived,
so caution is necessary when extrapolating these
trends for sea level rise predictions (Price et al.
2011). Understanding the physical processes that
control ice discharge is crucial in constraining
current and future mass balance estimates.
Synchronized changes in ice discharge from tidewater outlet glaciers, despite being located several
hundred kilometers apart, suggests sensitivity to
environmental (atmospheric or ocean) forcing.
Atmospheric warming increases the amount of thinning in the ablation zone, and provides additional
surface water (which can penetrate to the glacier
bed, causing acceleration via enhanced lubrication).
Tidewater glaciers are subject to large changes in
flow dynamics on short timescales ranging from
days to decades, as a result of forcings as diverse
as ocean tides (e.g., de Juan et al. 2010), ocean
circulation (Holland et al. 2008, Straneo et al.
2010, 2011), mélange extent (Amundson et al.
2010, Howat et al. 2011), and air temperature
(Shepherd et al. 2001, Thomas et al. 2003, Andersen
et al. 2010); this range of variability and an incomplete understanding of relevant processes make it
difficult to predict tidewater glacier response in a
warming climate and contributes one of the largest
sources of uncertainty in sea level forecasts (Bindoff
and Willebrand 2007, Lemke 2007).
8.2
CASE STUDY 1: CENTRAL EAST
GREENLAND MARGIN
FLUCTUATIONS AND CLIMATE
SENSITIVITY FROM A GLIMS
GLACIER INVENTORY AND
ASTER GDEM
8.2.1 Introduction
More than half of the glaciers peripheral to the
Greenland Ice Sheet, 50,000 km 2 , are located in
central East Greenland (67–72 N) and drain into
Scoresby Sund, Kangerdlugssuaq Fjord, and the
Blosseville Kyst (Blosseville Coast) (Jiskoot et al.
2003). This region of extreme topography contains
a variety of glacier types, many of which are tidewater terminating. The largest local ice cap, Geikie
Plateau, is 300 to 500 m thick (Christensen et al.
2000), has 2–3 m annual snow accumulation, and
frequent melt events even at higher elevations (Dall
et al. 2001). The largest tidewater outlet glacier,
Kong Christian IV Gletscher, drains partly from
the Geikie Plateau and partly from the Greenland
Ice Sheet and has a bed that is more than 600 m
below sea level within 20 km of the glacier terminus
(Thomas et al. 2009). This is an area where there has
been very limited glaciological research and only
few recent quantitative remote-sensing studies
(e.g., Dwyer 1995, Jiskoot et al. 2001, 2003, Luckman et al. 2003, Pritchard et al. 2003, 2005, Jiskoot
and Juhlin 2009, Thomas et al. 2009). No mass
balance or other monitoring programs exist
(Weidick 1995), although a new initiative, the Programme for Monitoring of the Greenland Ice Sheet
(PROMICE), has installed an automatic mass balance station in the region (Ahlstrøm and the PROMICE Project Team 2008). The regional runoff
from East Greenland to the North Atlantic is
important in global thermohaline circulation,
salinity, and sea ice dynamics (Mernild et al.
2008), hence it is important to (i) establish the
glacierized area, so that ice volume extrapolations
can be made, and (ii) establish the advance/retreat
rates of the glaciers, and especially of those with
tidewater margins.
Since the neoglacial period, the majority of landterminating glaciers in the Geikie Plateau region
have receded up to a few kilometers (Weidick
1995). There were no noticeable changes in the ice
margin positions of most calving fronts between the
late 19th century and the mid-1980s (Weidick 1995).
Between 1978 and 1991, major tidewater glacier
termini along the northern coast of Kangerdlugssuaq Fjord and the southern Blosseville Kyst
showed little change or a slight loss (0.1–0.5 km 2 )
in the glacier tongue area (Dwyer 1995). However,
the majority of these glaciers are of surge type
(Jiskoot et al. 2003), and it is unclear whether these
losses were responses to dynamic or climatic causes
(Dwyer 1995). Extreme glaciodynamic terminus
fluctuations occurred in Sortebræ, a surge-type
tidewater glacier, which experienced a retreat of
1.5 km between 1933 and 1943, a surge advance
of 10 km by 1950, a retreat of 8 km between1950
and 1981 followed by stagnation between 1981 and
1992 and a second surge advance of >5 km between
1992 and 1994 (Jiskoot et al. 2001). Other recent
examples of surge-related glaciodynamic terminus
fluctuations include the land-terminating Sermeq
Peqippoq Glacier with a retreat of 0.5 km between
1987 and 2000, and an advance of 2.8 km between
2000 and 2007 (Jiskoot and Juhlin 2009), and a
tributary of Bredegletscher, which changed from
being land-terminating to tidewater-terminating
during its 1 km advance some time between
1987 and 2000 (Jiskoot et al. 2012). Neighboring
Case Study 1: Central East Greenland margin fluctuations and climate sensitivity
187
Figure 8.1. Location map of Geikie Plateau region with glacier outlines from the new glacier inventory, and DEM
shading (light is highest elevation, while dark is lowest) according to ASTER Global DEM Version 1 data. Red
circles indicate regions where large errors in the DEM occur.
glaciers did not change their terminus positions
significantly during the same period. In the late
1970s, the regional transient snow line was 700 m
asl in southern parts of central East Greenland and
1,000–1 300 m asl in northern parts (Weidick 1995).
It had risen to 1,000–1,500 m asl for at least half the
glaciers in 1999–2000 (Jiskoot et al. 2001, 2003),
which agrees with the range of 1,200–1,500 m asl
suggested for modern Scoresby Sund glaciers (Lie
and Paasche 2006).
Predicted changes in the central East Greenland
region are highly sensitive to climate change
because of the specific character of its glaciers
and regional climate:
1. About two thirds of the total area drains through
tidewater glaciers (Jiskoot 2002, Jiskoot et al.
2003), which will have a direct dynamic response
to warming ocean currents and rising sea level
(Murray et al. 2010, Christoffersen et al. 2011).
The three largest tidewater glaciers have sustained flow velocities of 6.5–14 m day1 and
drain 25% of the total central East Greenland
glacier area (Luckman et al. 2003). Other large
tidewater glaciers have flow velocities of 0.1–2.5
m day1 with summer velocities of 1–3 m day1
(Dwyer 1995).
2. Between 30 and 70% of central East Greenland
glaciers are of surge type (Weidick 1988, Jiskoot
et al. 2003) and can suddenly reach speeds of 5 m
day1 up to 21 m day1 (Pritchard et al. 2005,
Jiskoot and Juhlin 2009) and advance several
kilometers (Jiskoot et al. 2001, Jiskoot and
Juhlin 2009). Some surges cause extreme
calving events. For example, Sortebræ’s calving
flux of 3.5–5.5 km 3 yr1 during its 1992–1994
surge, which is 10–20% of Jakobshavn Isbræ’s
calving flux (Jiskoot et al. 2001, Pritchard et al.
2003).
3. Many of the Scoresby Sund glaciers are polythermal (Kirchner 1963, Citterio et al. 2009, Jiskoot and Juhlin 2009), hence changes in local
temperature and precipitation rates might affect
their thermal regime and ice dynamic behavior
over time.
4. Climate models using IPCC emissions scenario
A1B predict 2.8–4.3 C temperature increase for
Greenland over the next century and radiative
forcing models show central East Greenland as a
‘‘hotspot’’ (Solomon et al. 2007).
5. The timing of breakup of sea ice in this region is
positively correlated with increased surface melt,
especially early breakup in the month of July
(Rennermalm et al. 2009).
188
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
In order to assess glacier characteristics, recent
changes, and sensitivity to future climate change
in central East Greenland, a detailed glacier inventory was compiled of the Geikie Plateau region,
using semiautomated digitization from ASTER
and Landsat 7 imagery. Preliminary results of this
inventory are presented in this chapter, and
include tidewater margin fluctuations between
1995 and 2004/2005, and climate sensitivity of landterminating glaciers related to snow line elevations
and glacier hypsometry. More detailed studies of
this region can be found in Kargel et al. (2012)
and Jiskoot et al. (2012).
8.2.2 Methods
Glacier outlines were extracted from 68 ASTER
L1B scenes with a spatial resolution of 15 m and
six orthorectified Landsat 7 ETMþ pan-sharpened
images with a resolution of 14.5 m. The vast
majority of ASTER images were from late-summer
2004/2005, while images with dates between May
and August 2000 were used to fill in gaps. Landsat 7
images were in two south–north strips of three
images each, where the easternmost strip, covering
the Blosseville Kyst, was from July 12, 2000 and the
inland strip from August 28, 2001. ASTER images
were co-registered with Landsat 7 images, and in
two regions where there was an absence of (cloudfree) Landsat 7 images, ASTER to ASTER image
co-registration was applied. Image-to-image coregistering was accomplished through matching of
20 ground control points distributed in a pattern
closely following that in Fig. 8.2. RMS errors were
within 3.0 m for each co-registered ASTER image.
Through co-registration, the spatial resolution of
the ASTER images was upsampled slightly to
14.5 m.
Glaciers were identified from mosaics of two to
four ASTER scenes of the same date (and gain
setting; Raup et al. 2000), using supervised distance
classification. Training regions, each containing at
least 1,500 pixels, for different surface classes were
selected from different areas within each mosaic.
The training areas were grouped in a binary class
(glacier/nonglacier), in which the glacier class comprised snow and exposed glacier ice, and the nonglacier class contained sediment, bedrock, water,
vegetation, shadow, and seawater. Shaded ice was
included in the glacier class, and shaded bedrock/
sediment in the nonglacier class, but this shadowfiltering technique was only moderately successful.
Systematic comparison of supervised classification
Figure 8.2. Pattern of 20 ground control points or ‘‘tie
points’’ for georeferencing ASTER images to orthorectified Landsat 7 images.
tools in ENVI 4.3 revealed that the maximum likelihood classification and Mahalanobis distance
classification performed best in identifying the
glacier class, and the end results were quite similar.
However, the Mahalanobis distance classification
runs approximately three times faster because it
assumes all class covariances to be equal, and was
therefore applied.
The resulting raster glacier masks were exported
to ArcGIS 9.2, in which small polygons and irregularities were removed using the Enhanced Lee
Filter, with a threshold of 3 pixels. The filtered
raster images were converted to polygons and
concatenated into one large ArcGIS shapefile.
However, since the entire mosaicked polygon file
was well over 4 GB in size, the polygons had to
be cleaned using the generalized function with a
minimum distance of 100 m, which introduced
some inaccuracies (see Fig. 8.3d). The cleaned polygons were converted into one large mosaic, and
overlaid onto the Landsat 7 mosaic. Morainecovered termini, shadow zones, nunataks, and ice
divides were then manually digitized. A glacier
threshold of 2.0 km 2 was applied, for reasons
including difficulties in distinguishing seasonal
snow from glaciers at this scale, the small proportion of area relative to the overall glacierized area in
this region, the absence of calving margins in this
size class, and for time-management reasons. Fig.
8.3 shows examples of each of these four steps in the
semiautomated glacier classification process.
Case Study 1: Central East Greenland margin fluctuations and climate sensitivity
189
Figure 8.3. Steps in semiautomated glacier extraction: (a) Mahalanobis distance classification of glacier surface
(blue) and nonglacier surface (yellow); (b) removal of small polygons and irregularities using the Enhanced Lee
Filter; (c) vectorized version of Fig. 3b superimposed on ASTER image, (d) glacier outlines after manual correction
and cleaning with the generalized function. This version was used for the mosaic. The pregeneralized version was
retained for analysis of individual glaciers. See Online Supplement 8.1 for higher resolution version.
Calving glacier margins were manually extracted
by retracing margin lines and snapping these to
glacier margin polygons derived from the semiautomated method. Calving margins from a glacier
inventory based on InSAR images and topographic
maps of the mid-1980s to mid-1990s (Jiskoot 2002,
Jiskoot et al. 2003) were also manually retraced.
Through overlay of these two margins any differences in area (positive or negative; retreat and
advance) were summed, so that for each glacier a
calving margin area change between 1985/1995 and
2004/2005 was obtained. For four case studies, calving margins from all available images were
extracted, where the maximum number of margins
could be three and between the following periods:
1985 or 1995, 2000/2001, and 2004/2005.
Glacier hypsometry for individual glaciers and
for the total land-terminating glacier cover was
calculated and plotted by counting pixels (which
have constant area), and seasonal snow lines were
extracted from interactively stretched ASTER
scenes from July/August 2003–2005 (cf. Jiskoot et
al. 2009). Snow lines could be determined for 60%
of the glaciers (180 glaciers), but are mostly lacking
in the southern half of the region.
Errors in georeferencing were between 0 and 4
pixels with an average of 2 pixels (29 m) and manual
digitization errors of calving margins were within
1 pixel (14.5 m). Hence, the total areal error for the
ASTER-to-ASTER or Landsat 7-to-ASTER calving margin was 33 m 2 . However, the shift in the
calving margin from 1987 to 1995 (Jiskoot et al.
2003) was systematic to the east, and was between
75 and 255 m, but manual shift correction through
overlaying of coastal bedrock control points
reduced this average error to <150 m. The resulting
error in linear change of tidewater glacier margins
termini is 154 m (square root of the sum of the
squares of component errors) for the period
between 1985/1995 through 2000/2005. The maximum error in area change is thus 154 m times the
width of the glacier terminus.
8.2.3 Results
8.2.3.1
Glacier inventory
The glacier inventory contains 332 glaciers, with a
total glacierized area of 41,591 km 2 . Glaciers range
in size from 2 km 2 (thresholded) to 11,079 km 2 for
190
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Kong Christian IV, which partly drains the Greenland Ice Sheet (e.g., Thomas et al. 2009). Glacier
area is lognormally distributed, and the minimum
glacier area cutoff of 2 km 2 only underestimates the
total glacierized area by 0.5–1%. Glacier types
include ice caps, snowfields, and mountain, valley,
and outlet glacier systems, of which many are tidewater terminating. By spatially joining the glacier
inventory of Jiskoot et al. (2003) with this new
inventory, the four-category surge-type glacier
classification was transferred, with class 0 ¼ normal
glaciers (no morphological evidence for surge
behavior), 1 ¼ possibly surge type (one or two
equivocal morphological features of past surge
behavior), 2 ¼ probably surge type (two or more
unequivocal), 3 ¼ glaciers with an observed surge
(only three in our region: Sortebræ complex (Jiskoot et al. 2001), Sermeq Peqippoq (Jiskoot and
Juhlin 2009), and a tributary of Bredegletscher
Glacier (Jiskoot et al. 2012). About 30 glaciers were
added to the original glacier inventory of Jiskoot et
al. (2003) in this region. Given that 56 glaciers are
class 2 or 3, 17% of glaciers in the region are of
surge type with certainty. Another 23% (75 class 1)
are possibly of surge type, hence the total estimated
percentage of surge type glaciers in the Geikie Plateau region is between 17 and 40%.
Most glaciers drain from the Geikie Plateau (with
a highest elevation of 2,895 m asl) to sea level and
end in tidewater margins. The total length of calving fronts in 1995 was 235 km, while in 2000 it was
196 km, and roughly 90% of the total glacierized
area drains through 120 tidewater glaciers (Fig.
8.4). The widths of calving margins range from
0.1 to 13 km, and the length–frequency is lognormally distributed. Evidently, the role of calving
and potential influence of ocean currents, sea ice,
and sea level change in this region are important.
8.2.3.2
Calving margin dynamics
Subtracting the 2004/2005 margins from those constructed from 1985 maps and 1995 InSAR imagery
(Jiskoot et al. 2003) resulted in an overall calving
margin area loss (through retreat) of 74 km 2 , an
area gain (through advance) of 4.3 km 2 , and a
net calving area loss of 70 36 km 2 . This translates
to a mean overall annual retreat rate between 3 km
and 7 km (many smaller tidewater margins were
digitized from 1985 topographic maps, but most of
the larger margins were derived directly from 1995
InSAR imagery; Jiskoot et al. 2003). Some glaciers
(e.g., Borggraven, a non-surge-type glacier) show
quite large interannual calving margin fluctuations
which are possibly related to the presence of sea ice
or fast ice, which is visible in some of the earlysummer ASTER images.
Typical patterns are shown in four case studies
(Fig. 8.5): (a) Sortebræ, a surge-type glacier with
an observed surge between 1992 and 1995 (Jiskoot
et al. 2001, Pritchard et al. 2005), has lost 11.5 km 2
since 1995, (b) a relatively small unnamed surgetype glacier (unofficially nominated ‘‘Ryberg
Glacier’’), has lost 2.5 km 2 since 1995, (c) Kong
Figure 8.4. Glacier inventory map with tidewater-terminating glaciers in dark gray, tidewater margins in red, and
land-terminating glaciers in light gray. The abbreviated glacier names correspond to the four case studies presented
in Fig. 8.5. Figure can also be viewed as Online Supplement 8.2.
Case Study 1: Central East Greenland margin fluctuations and climate sensitivity
191
Figure 8.5. Case studies of tidewater margin changes. Two surge-type glaciers in their quiescent phase: (a)
Sortebræ and (b) Ryberg Glacier. Two non-surge-type fast-flowing glaciers: (c) Kong Christian IV Gletscher, and
(d) Magga Dan Gletscher. See Fig. 8.4 for locations and text for details. Figure can also be viewed as Online
Supplement 8.3.
Christian IV, a non-surge-type outlet glacier partly
draining the Greenland ice sheet, shows no significant change in the position of its tidewater margin
between 1995 and 2004/2005 (areal loss of 0.97
km 2 ), and (d) Magga Dan, a non-surge-type glacier
draining into Scoresby Sund, was also virtually
unchanged between 1995 and 2004/2005 (areal
change 0.02 km 2 ). The last two glaciers have extremely fast ice flow of the order of 6.5–14.0 m
day1 (2.4–5.11 km yr1 ) at the margin (Luckman
et al. 2003), suggesting the tongues may be close to
flotation.
The stationarity of the calving margin of Kong
Christian IV and Magga Dan Glaciers may reflect
their relatively constant fast flow, limited thinning
rate, and shallow submarine shoals (Luckman and
Murray 2005, Joughin et al. 2010, Jiskoot et al.
2012). Laser altimeter surveys, with NASA’s
Airborne Topographic Mapper (ATM) of Kong
Christian IV Glacier between 1993 and 2006
showed that the terminal 20 km thinned between
0.5 and 0.7 m yr1 in that period, and that the
thinning rate appears to be decreasing over time.
Farther inland, the glacier was roughly in balance
between 1993 and 1998, but thinned by 0.5 m yr1
between 1998 and 2006 (Thomas et al. 2009).
Kangerdlugssuaq Glacier, a fast-flowing outlet gla-
cier from the Greenland Ice Sheet immediately
south of the Geikie Plateau region, retreated rapidly
in 2004/2005 and this event coincided with possible
warming of water masses in Kangerdlugssuaq
Fjord in 2004 (Christoffersen et al. 2011). Moreover, retreat occurred in a region where ice surface
elevations had been only about 10 m above flotation levels, suggesting that retreat was probably
caused by slow thinning after 2001 that allowed
the ice to float free from its bed and almost immediately break up into icebergs (Thomas et al. 2009).
Analysis of possible patterns in the tidewater
margin fluctuations of local glaciers (Kong Christian IV was removed from this analysis as it partly
drains the Greenland Ice Sheet) resulted in the
following:
1. There is no correlation between calving width
and terminus retreat/advance rate.
2. There is no north–south or east–west spatial
pattern in the terminus retreat/advance rate
(i.e., glaciers along the Kangerdlugssuaq Fjord,
Blosseville Kyst and those draining north into
Scoresby Sund did not have significant differences in retreat/advance rates).
3. There appears to be some correlation between
surge-type glaciers and increased calving. For
192
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
non-surge-type glaciers (class 0 and 1) the calving
area change between 1985 and 2001/2004 was
36 km 2 , which is 0.2% of the total area of
17,527.5 km 2 of these glaciers. For surge-type
glaciers (class 2 and 3) the calving area change
for the same period was 42 km 2 , which is
0.45% of the total area of 9345.8 km 2 of these
glaciers. If the classification of ‘‘surge-type’’ were
to include class 1 glaciers, then surge-type
glaciers show a tidewater margin retreat of 62
km 2 since 1995 (a handful have been measured
since 1985) and non-surge-type glaciers a retreat
of only 8 km 2 since 1985. It therefore appears
that some of the calving margin retreat could be
related to surge dynamics rather than just to
changes in mass balance or ocean water temperature (i.e., a dynamically amplified response to
climatic drivers).
8.2.3.3
Mass balance sensitivity of
land-terminating glaciers from
hypsometric analysis
We use the new ASTER Global DEM to calculate
standard glacier inventory data (aspect, elevation,
and slope; Paul et al. 2009), and generate glacier
hypsometries in order to assess snow line characteristics, and sensitivity of glaciers to projected climate
change. Glacier hypsometry is the area–elevation
distribution of either individual glaciers, or a
glacierized region as a whole. In combination with
mass balance curves, or individual snow lines, this
measure is important in assessing individual or
regional climate sensitivity of a group of glaciers.
Hypsometry is also a factor in the response time of
glaciers to a change in regional climate (e.g., Furbish and Andrews 1984, Raper and Braithwaite
2009; see also this book’s Chapter 33 by Kargel
et al.). Unfortunately we discovered that the
ASTER GDEM has extensive areas with large
vertical errors (>1,000 m), so-called ‘‘mushroom
regions’’ or ‘‘mole hills’’, due to cloud cover on
the images from which the ASTER GDEM was
derived. One of these regions can clearly be seen
in the accumulation zone of Kong Christian IV in
Fig. 8.1. At this latitude there are no automated
techniques available to fill these regions and correct
the error with reasonable accuracy. Because of these
errors, and because land-terminating glaciers are
more sensitive to surface mass balance changes than
tidewater-terminating glaciers (which are influenced
by calving dynamics and therefore also linked to
ocean temperature and sea ice dynamics), hypso-
Figure 8.6. Normalized hypsometric curve of 180
land-terminating glaciers, annotated with the average
late-summer snow line in 2003–2005 (1,050 m asl)
and a corresponding AAR of 56%. A rise in ELA
(approximated by snow line) of 200 and 400 m corresponds to a change of AAR to 42 and 29%, respectively.
See text for details.
metric analysis was only performed using landterminating glaciers. The hypsometry of all landterminating glaciers combined (180 glaciers representing approximately 9% of the total glacierized
area of the Geikie Plateau region) is shown in Fig.
8.6, and depicts a near equidimensional type of
area–elevation distribution (Furbish and Andrews
1984).
The average snow line elevation of glaciers for
which snow lines could be established (60% of the
298 glaciers in the Geikie Plateau region) was 1,092
m asl, which is a rise of 82 m over the average snow
line elevation of 1,010 m asl derived from latesummer Landsat images of 1999/2000 (Jiskoot et
al. 2003). Of these 60% that are land terminating
the average snow line was at 1,050 m asl. Combining this snow line elevation, which we take as a
good approximation of equilibrium line altitude
(ELA), with the hypsometry for the 180 landterminating glaciers reveals that the accumulation
area ratio (AAR; a measure of glacier health; Paterson 1994) was 56% in 2003–2005 (see Fig. 8.6). If
the predicted regional increase of 2.8–4.3 C over the
next century is roughly translated into a rise in snow
line, it would rise between 200 and 400 m. The
resulting AAR for these glaciers would be reduced
Case Study 2: A comparison of high-rate GPS and ASTER-derived measurements on Helheim Glacier 193
to 42% for a rise in snow line of 200 m, and reduced
to only 29% for a rise in snow line of 400 m (Fig.
8.6). In this simple analysis it is assumed that the
overall shape of the hypsometric curve will not
change significantly over this period. Thus, predicted regional warming (Solomon et al. 2007),
combined with the sensitivity of surface melt to
earlier breakup of sea ice (Rennermalm et al.
2009), land-terminating glaciers will have low
viability for survival under predicted climate
conditions.
8.3
CASE STUDY 2: A COMPARISON
OF HIGH-RATE GPS AND ASTERDERIVED MEASUREMENTS ON
HELHEIM GLACIER
8.3.1 Introduction
Satellite remote sensing has revolutionized polar
glaciology by providing frequent coverage over
large spatial regions that are difficult to access by
field-based programs. Sequential observations can
span decades, longer than most traditional field
methods. Spaceborne measurements of surface
elevation and flow speed are of particular relevance
to studies of ice dynamics. Radar and laser altimetry is the most common method of obtaining
surface elevations (e.g., Krabill et al. 2004, Thomas
et al. 2006, Wingham et al. 2006), but elevations can
also be extracted from optical imagery using photoclinometry (e.g., Scambos and Fahnestock 1998)
and stereo imaging (e.g., Berthier et al. 2005,
Stearns and Hamilton 2007). Glacier velocities
can be derived from interferometric analysis (e.g.,
Joughin et al. 1999) or speckle tracking on radar
images (e.g., Wuite 2006), or from feature tracking
on visible band images (e.g., Scambos et al. 1992,
Howat et al. 2005, Stearns et al. 2005). Each technique has its advantages and limitations.
In this case study, ice velocities are derived from
optical satellite imagery by tracking the displacement of surface features in sequential images. Feature tracking can be performed at varying levels of
complexity ranging from manual (e.g., Lucchitta
and Ferguson 1986), to semiautomatic (e.g., Ferrigno et al. 1993), to automatic (e.g., Scambos et
al. 1992, Whillans and Tseng 1995), with each technique producing a progressively larger number of
matches. Here, we assess the accuracy of a widely
used automatic feature-tracking technique.
Validating remote-sensing observations with
ground-based measurements is necessary to verify
that information extracted from the satellite data
accurately characterizes geophysical processes. In
this study, we use ground-based GPS elevation
and velocity measurements to assess the accuracy
of ASTER-derived data products. ASTER imagery
has been used extensively in glaciology to map
changes in glacier geometry (e.g., De Angelis
2003), surface elevation and volume change (e.g.,
Kääb et al. 2002, Vignon et al. 2003, Paul et al.
2004, Howat et al. 2005, Stearns and Hamilton
2007), and ice velocity (e.g., Kääb et al. 2002,
Howat et al. 2005, Stearns et al. 2005), although
few of these studies have been validated with field
measurements.
Helheim Glacier (66.5 N, 38 W), located in East
Greenland, has undergone rapid changes in ice
dynamics in the past few years (e.g., Howat et al.
2005, Luckman et al. 2006, Stearns and Hamilton
2007) including rapid flow acceleration (by 40%
at the glacier front, Fig. 8.7), thinning (up to 60 13
m in one year), and terminus retreat (5 km
between 2003 and 2005) (Stearns and Hamilton
2007). These events and changes were largely quantified using repeat ASTER images of Helheim
Glacier.
GPS instruments were deployed on Helheim
Glacier to collect frequent velocity measurements
during the 2006 summer. The ASTER sensor, on
board the Terra satellite, obtained two usable
images of Helheim Glacier while the GPS units were
operating. This overlap provides a rare opportunity
to compare satellite-derived ice velocity and surface
elevation measurements with in situ field data.
8.3.2 Data
8.3.2.1
Ground-based GPS surveys
We installed a network of 19 GPS receivers on and
around Helheim Glacier in late June, 2006. Sixteen
receivers were installed on the glacier, in a configuration including stations both on the glacier
centerline, and offset from the centerline (Fig.
8.8).Three GPS receivers were installed at rock sites
(sites 1–3) surrounding the on-ice network to help
define a stable geodetic reference frame (Fig. 8.8).
Nine of the stations spanned an upglacier distance
of 20 km from a point 15 km behind the calving
front. The stations operated for 60 days, and
recorded data at a rate of one sample every five
seconds. In addition to the long-term GPS network,
we deployed four receivers (sites 14–19), for 2 to 5
194
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Figure 8.7. (A) An ASTER image of the trunk of Helheim Glacier from August 3, 2004. (B) Ice velocity along the
profile in panel A, derived from Landsat ETMþ and ASTER image pairs (diamonds), and overlapping GPS
measurements (stars) along and adjacent to the profile.
Figure 8.8. A DEM of Helheim Glacier, derived from an ASTER image taken on August 30, 2006. The red dots
represent GPS instruments which operated between August 25–30, 2006.
Case Study 2: A comparison of high-rate GPS and ASTER-derived measurements on Helheim Glacier 195
days each, at locations just behind the calving front.
In this study, we are only interested in the GPS data
that overlap with satellite images: the period from
August 25 to August 30, 2006.
The GPS data were processed using the GIPSY
software package (Lichten and Border 1987) and
high-precision kinematic data-processing methods
(e.g., Elósegui et al. 1996, 2006) to estimate the
time-dependent positions of GPS sites on the glacier
relative to the static antennas on nearby bedrock.
Processing incorporated precise satellite orbits from
the International GNSS Service (IGS), with no
further orbit improvement. A second-order quadratic was fit to the 5 s position data to obtain daily
velocities. A linear fit was applied to the sites at the
calving front (sites 14–17), which were occupied for
a shorter time interval. Uncertainties in velocity are
less than 0.1 m day1 (de Juan et al. 2010).
8.3.2.2
Satellite remote sensing
Two DEMs were generated using ASTER images
taken at midday on August 25 and August 30, 2006.
Processing of the stereo bands to epipolar geometry, and parallax matching was done automatically
using commercial software developed by the
Japanese ASTER Science Team and described by
Fujisada et al. (2005). Products generated using
identical procedures can now be ordered from the
NASA/USGS Land Processes Distributed Active
Archive (LP DAAC) at http://edcimmswww.cr.usgs.
gov/pub/imswelcome The commercial software
produces DEMs with a post-spacing of 30 m, which
are subsequently interpolated to 15 m to match the
resolution of the VNIR bands.
Geolocation of the ASTER DEMs is entirely on
the basis of the satellite ephemeris information contained in the image header file, which is considered
to be better than 50 m (Fujisada et al. 2005). DEM
uncertainties are a combination of systematic
errors, and random errors due to satellite positioning, image acquisition geometry, and atmospheric
conditions. We detect a systematic bias in the
vertical of 17.79 m between the two DEM scenes
based on relative elevation differences of static bedrock surfaces. Once this bias is removed, calculated
random errors contribute to a root mean square
error of 7.1 m for the image pair, based on a
comparison of elevation differences in static regions
(Stearns and Hamilton 2007). This error is consistent with Fujisada et al. (2005), who report a DEM
vertical accuracy of 20 m with 95% confidence (2).
Rivera et al. (2005) report an RMS error of 17 m,
based on a comparison of ASTER DEMs and
photogrammetrically produced DEMs. In a similar
study, Kääb (2002) compared ASTER DEMs with
DEMs produced by photogrammetry for mountain
regions in the Swiss Alps and New Zealand. In such
cases, uncertainties in absolute elevations can be
quite large (60 m RMS) because of rugged topography (Kääb 2002). The uncertainties in relative
elevations, important for surface elevation change
and volume loss estimates, are usually much
smaller. Stevens et al. (2004) note that, in the
absence of appreciable atmospheric water vapor,
RMS uncertainties for relative ASTER DEMs are
less than 10 m for moderately rugged terrain.
8.3.2.3
ASTER-derived velocity data
Velocities are derived from automatic tracking of
surface features on sequential ASTER images using
a cross-correlation technique implemented in the
IMCORR software package (Scambos et al.
1992). The software was originally developed for
mapping displacements on low-slope ice streams
in West Antarctica (e.g., Bindschadler and Scambos
1991, Scambos et al. 1992) from Landsat imagery,
but has been adapted to steep, fast-moving outlet
glaciers and other image types (Whillans and Tseng
1995, Wuite 2006, Ahn and Howat 2011).
IMCORR tracks the displacement of surface features (e.g., crevasses, seracs) in two co-registered and
orthorectified images. The program uses a normalized cross-covariance correlation method to match
the surface features in each image pair.
IMCORR uses small subscenes (‘‘chips’’) from
each image to track displacements. A ‘‘reference
chip’’ from the older image moves in a grid-like
pattern through a ‘‘search chip’’ in the newer image
(Fig. 8.9). For each position, a correlation coefficient is calculated, creating a 2D correlation function that has a peak shape. The reported match is
the location with the maximum correlation value.
The shape of the correlation function is an important indicator of measurement accuracy: the
sharper the peak, the more accurate the match
(Wuite 2006). If no match is found, a null vector
is output.
IMCORR allows the user to control the size and
offset of the chips to adjust for the time separation
of the images, the speed of the glacier, the size of
trackable features, and the direction of flow.
Because glacier flow speeds can range from centimeters to kilometers per year within a single image,
the strength of the correlation will vary. Optimizing
196
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Figure 8.9. Two ASTER scenes of Helheim Glacier illustrating the IMCORR technique. A reference window in the
2005 scene is compared with a larger search window in the 2006 scene. Once a match is found, the displacement is
calculated from the midpoint of the two chips. Boxes are enlarged in the figure, for clarity.
the correlation strength can be done by manually
adjusting the chip sizes, or implementing an automatic adjustment in the code (Wuite 2006).
Sequential images used for cross-correlation
must be largely cloud free, and are required to have
similar illumination characteristics (Scambos et al.
1992). The ASTER instrument’s cross-track, offnadir scene acquisition capability (24 ) in the
VNIR introduces an image geometry change that
must be considered during scene selection. Over
regions of rugged relief, such as in East Greenland,
we find that the pointing angles of sequential images
need to be within 3 to maintain similar geometric
characteristics. If the pointing angle difference is
greater than 3 , panoramic distortion inhibits
cross-correlation. A further consideration is the
time interval between sequential image acquisition.
The time separation must be long enough for features to be displaced more than the measurement
uncertainty, but not so long that features are distorted beyond recognition.
The measured displacements of surface features
have several sources of uncertainty originating
from image orthorectification, co-registration, and
application of the feature-matching technique.
Orthorectification using the ASTER DEM translates DEM errors onto the orthorectified image.
Kääb (2002) reports a 10 m ground position error
for rough terrain and a 3 m error for moderate
terrain based on a similar analysis in the Swiss Alps.
Overall, resampling errors during orthoprojection
translate to positional errors that are at the subpixel
(<15 m) level.
Uncertainty associated with the image crosscorrelation technique is also smaller than the pixel
size of 15 m. Matches with uncertainties larger than
1 pixel are discarded. Because uncertainty in the
acquisition times of the imagery is negligible, velocity uncertainty is inversely proportional to the time
separation of the image pairs. In this study, because
our two images are only 5 days apart, the cumulative velocity error is relatively large (4.24 m day1 ,
or 21% of total velocity).
8.3.3 Results
8.3.3.1
Elevation
The DEM software outputs elevations in the
EGM96 geoid, at an interpolated post-spacing of
15 m. To permit comparison with GPS ellipsoidal
heights, the DEM heights were converted to the
WGS-84 ellipsoid using parameters found at
http://earth-info.nga.mil /GandG/wgs84/gravitymod/
egm96/intpt.html The geoid–ellipsoid difference is
50 m at Helheim Glacier. The results of the comparison are shown in Fig. 8.10A and Table 8.1. The
August 30 DEM has a systematic bias of 17.79 m,
ASTER data for GLIMS: STARS, DARs, gain settings, and image seasons
197
Figure 8.10. Elevation results from GPS and two ASTER-derived DEMS. (A) Elevation differences between GPS
data and ASTER DEMs. ’’Raw’’ elevations are the values for pixels nominally containing each GPS site. The asterisk
indicates that better elevation matches were obtained from neighboring pixels (within 50 m). The August 25/30,
2006 comparison is done with bias-corrected values for the August 30, 2006 DEM. ASTER elevation uncertainties
are 10 m, as described in the text. (B) GPS results, and ASTER elevations from a DEM with 15 m pixel postspacing.
and consistently underestimates the GPS elevations
with an RMS difference of 33.71 m. The RMS
difference between the August 25 DEM and the
GPS elevations is 12.56 m. These comparisons are
based on DEMs with 15 m post-spacings and the
nominal pixel coordinates of each GPS site. GPS
sites were installed on the peaks of nunataks (rock
sites 1–3) and on relatively high locations on the ice
(sites 4–17) to improve satellite visibility, which
partly explains why ASTER DEMs systematically
underestimate GPS measurements. Some of the
remaining difference might be due to image geolocation. The geolocational uncertainty of ASTER
DEMs (50 m) (Fujisada et al. 2005) means that
the best elevation match is not always at the pixel
location prescribed by the coordinates of the GPS
site. GPS to DEM elevation differences can be minimized by searching for better elevation matches in
pixels within 50 m of the original pixel (i.e., within
the range of geolocational uncertainty of the GPS
site on the image) (noted with an asterisk in Fig.
8.10). The results show an RMS difference of 9.56 m
for the August 25 DEM, and 25.52 m for the
August 30 DEM (Table 8.1).
When comparing glacier surface elevations over
time, the accuracy of each DEM relative to other
DEMs is more important than the absolute accuracy of an individual DEM. We detect a systematic
bias between the two DEMs, in which the August
30 DEM yields elevations consistently lower than
the August 25 DEM by an RMS of 17.79 m. The
bias was quantified by measuring the elevation
difference over static bedrock regions on the two
images, in 5 5 km boxes. The bias accounts for
most of the RMS difference of 22.94 m between
the two DEMs (at each GPS location). When the
August 30 DEM is corrected for the bias, the RMS
difference drops to 6.52 m. Most of the remaining
198
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Table 8.1. Summary of errors for absolute and relative
DEMs of different post-spacings. The asterisk indicates
that elevations from neighboring pixels (within 50 m)
were considered. The absolute DEM RMS quantifies
the difference between the GPS elevations and the
elevations derived from ASTER DEMs on August 25
and August 30. The relative DEM describes the RMS
difference between the two DEMs at each site.
DEM
description
August 25
absolute
DEM
August 30
absolute
DEM
Relative
DEM
15 m
12.56 m
33.71 m
17.31 m
15 m*
9.56 m
25.52 m
22.94 m
150 m*
9.37 m
16.86 m
10.53 m
difference is due to ‘‘noise’’. The 15 m DEMs are
‘‘noisy’’ products, as a result of the crevassed and
rugged surface of Helheim Glacier. Stearns and
Hamilton (2007) eliminated much of this noise by
resampling the DEMs to 150 m using a bicubic
spline. By carrying out a similar smoothing of the
August 25 and August 30 DEMs, we obtain RMS
differences of 10.53 m (Table 8.1).
8.3.3.2
Velocity
Velocities were derived from the two ASTER scenes
using several different IMCORR input parameters
and grid sizes. The grid interval determines the
spacing of the reference chips, and therefore the
number of velocity vectors. A small grid interval
will increase the number of vectors, but can lead
to oversampling because the individual vectors are
not statistically independent (Wuite 2006). The
results, shown in Fig. 8.11, suggest that different
grid spacings produce slightly different velocities.
An IMCORR grid spacing of 5 grid cells, which
results in a post-spacing of 75 m (because the image
resolution is 15 m), generates slightly faster
velocities, probably because of the oversampling
issue mentioned above.
Overall, ASTER-derived velocities are consistent
with GPS measurements (Fig. 8.12). The RMS difference between the GPS and the ASTER velocities
(gridded to 150 m) is 0.89 m day1 , well within the
errors assigned to the ASTER results. The offset is
higher using velocities gridded to 375 m (2.46 m
day1 ) and 75 m (1.53 m day1 ). For Helheim
Glacier, a fast-flowing glacier, gridding the featuretracking results to 150 m generates flow speeds
Figure 8.11. The influence of different IMCORR grid spacings on derived velocities.
Discussion and conclusion 199
Figure 8.12. Velocity results from GPS and ASTER-derived velocity measurements.
which best match GPS observations. Depending on
the gridding routine, individual point measurements have an RMS difference of between 0.89 m
day1 and 2.46 m day1 , or 6–17% of the flow
speed.
The IMCORR software yields the displacements
in X and Y components, which are used to determine the direction of flow (Fig. 8.13). ASTER
flow azimuths are compared with GPS results in
Fig. 8.13. The RMS difference of the azimuths is
8.60 . This small difference shows that the featuretracking results duplicate both the magnitude and
direction of flow, even with a very short time
separation between the image pairs.
8.4
Figure 8.13. The velocity vectors of GPS (red) and
ASTER-derived (black) data. ASTER velocities for sites
4–6 and 17 are not available. Sites 1–3 are rock sites.
DISCUSSION AND CONCLUSION
A direct comparison of satellite data and terrestrial
measurements shows that ASTER imagery is well
suited for applications in glacier dynamics.Velocity
measurements derived from ASTER images capture the magnitude and direction of ice flow. In a
new glacier inventory we have documented 41,591
km 2 of glaciers, mainly detached from the Greenland Ice Sheet, in central East Greenland. Multiple
200
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
repeat ASTER and Landsat 7 ETMþ images of the
Blosseville Kyst show tidewater glacier changes,
including an overwhelming dominant pattern of
rapid but time-variable rates of retreat. The RMS
difference between ASTER DEMs and GPS elevations ranges from 9.37 to 12.56 m (for DEMs with
no systematic bias), depending on whether
smoothed or unsmoothed DEMs are used. Most
of the elevation errors arise from the geolocation
error of individual images, with the remaining
difference probably being due to GPS sites being
placed on locally high terrain to improve satellite
visibility. The two ASTER DEMs demonstrate
good repeatability over the glacier surface,
especially at a grid spacing of 150 m, and after
biases are removed.
Finally, while ASTER imagery usually generates
good DEMs and velocity maps, images should be
scrutinized before use. Images with clouds, low Sun
elevations, high off-nadir pointing angles, or inappropriate gain settings will not produce good
results. Systematic biases between DEMs do occur,
and DEMs should be validated over static surfaces
(e.g., bedrock) to test their relative geolocation and
elevations. This is especially true when using
ASTER DEMs to detect changes in elevation on
glaciers. Given the error budget, ASTER DEMs are
probably only valid when glacier elevation changes
are greater than 25 m over 5 years.
8.5
ACKNOWLEDGMENTS
L.A.S. partially carried out her work at the
University of Maine and thanks Gordon S.
Hamilton for collaborative input. Research
support for L.A.S. was provided by NASA grant
NNX08AD38G awarded to G.S. Hamilton. H.J.
thanks her student Dan Junlin for assistance with
glacier inventory development and analysis.
Research support to H.J. was through NSERC Discovery and NSERC UFA grants. ASTER data
courtesy of NASA/GSFC/METI/Japan Space Systems, the U.S./Japan ASTER Science Team, and
the GLIMS project.
8.6
REFERENCES
Ahlstrøm, A.P., and the PROMICE project team (2008)
A new programme for monitoring the mass loss of the
Greenland ice sheet. Geological Survey of Denmark and
Greenland Bulletin, 15.
Ahn, Y., and Howat, I.M. (2010) Efficient, Automated
Glacier Surface Velocity Measurement from Repeat
Images Using Multi-Image/Multi-Chip (MIMC) and
Null Exclusion Feature Tracking (Tech. Rep. TGRS2010-00396.R2), Institute of Electrical and Electronics
Engineers, Washington, D.C.
Ahn, Y., and Howat, I.M. (2011) Efficient, automated
glacier surface velocity measurement from repeat
images using multi-image/multi-chip (MIMC) and null
exclusion feature tracking. IEEE Trans. on Geoscience
and Remote Sensing, 49(8), 2838–2846.
Amundson, J.M., Fahnestock, M., Truffer, M., Brown,
J., Lüthi, M.P., and Motyka, R.J. (2010) Ice mélange
dynamics and implications for terminus stability,
Jakobshavn
Isbræ,
Greenland.
Journal
of
Geophysical Research, 115(F1), F01005, doi: 10.1029/
2009JF001405.
Andersen, M.L., Larsen, T.B., Nettles, M., Elósegui, P.,
van As, D., Hamilton, G.S., Stearns, L.A., Davis, J.L.,
Ahlstrøm, A.P., de Juan, J. et al. (2010) Spatial and
temporal melt variability at Helheim Glacier, East
Greenland, and its effect on ice dynamics. Journal of
Geophysical Research, 115(F4), F04041, doi: 10.1029/
2010JF001760.
Bales, R.C., Guo, Q., Shen, D., McConnell, J.R., Du, G.,
Burkhart, J.F., Spikes, V.B., Hanna, E., and Cappelen,
J. (2009) Annual accumulation for Greenland updated
using ice core data developed during 2000–2006 and
analysis of daily coastal meteorological data. Journal
of Geophysical Research, 114(D6), 1–14.
Bamber, J.L., Layberry, R.L., and Gogineni, S.P. (2001)
A new ice thickness and bedrock data set for the
Greenland Ice Sheet, 1: Measurement, data reduction,
and errors. Journal of Geophysical Research, 106(D4),
33773–33780.
Banta, J.R., and McConnell, J.R. (2007) Annual accumulation over recent centuries at four sites in central
Greenland. Journal of Geophysical Research,
112(D10114).
Berthier, E., Vadon, H., Baratoux, D., Arnaud, Y.,
Vincent, C., Feigl, K., Remy, F., and Legresy, B.
(2005) Surface motion of mountain glaciers derived
from satellite optical imagery. Remote Sensing of
Environment, 95(1), 14–28.
Bhattacharya, I., Jezek, K.C., Wang, L., and Liu, H.
(2009) Surface melt area variability of the Greenland
ice sheet: 1979–2008. Geophysical Research Letters,
36(L20502), doi: 10.1029/2009GL039,798.
Bindoff, N., and Willebrand, J. (2007) Observations:
Oceanic climate change and sea level, Climate Change
2007: Working Group I: The Physical Science Basis,
Intergovernmental Panel on Climate Change (IPCC),
Washington, D.C., p. 48.
Bindschadler, R.A., and Scambos, T. (1991) Satelliteimage-derived velocity field of an Antarctic Ice Stream.
Science, 252, 242–246.
Box, J.E., Bromwich, D.H., Veenhuis, B.A., Bai, L.S.,
Stroeve, J.C., Rogers, J.C., Steffen, K., Haran, T.,
and Wang, S.H. (2006) Greenland ice sheet surface
mass balance variability (1988–2004) from calibrated
References 201
Polar MM5 output. Journal of Climate, 19(12), 2783–
2800.
Chen, J., Wilson, C., and Tapley, B. (2006) Satellite
gravity measurements confirm accelerated melting of
Greenland Ice Sheet. Science, 313(5795), 1958–1960.
Chen, J.L., Wilson, C.R., and Tapley, B.D. (2011) Interannual variability of Greenland ice losses from satellite
gravimetry. Journal of Geophysical Research, 116(B7).
Christensen, E.L., Reeh, N., Forsberg, R., Jorgensen,
J.H., Skou, N., and Woelders, K. (2000) A low-cost
glacier-mapping system. Journal of Glaciology,
46(154), 531–537, doi: 10.3189/172756500781833142.
Christoffersen, P., Mugford, R.I., Heywood, K.J.,
Joughin, I., Dowdeswell, J.A., Syvitski, J.P.M., Luckman, A., and Benham, T.J. (2011) Warming of waters
in an East Greenland fjord prior to glacier retreat:
Mechanisms and connection to large-scale atmospheric conditions. The Cryosphere, 5, 701–714.
Church, J.A., Gregory, J.M., Huybrechts, P., Kuhn, M.,
Lambeck, K., Nhuan, M.T., Qin, D., Woodworth,
P.L., Anisimov, O.A. et al. (2001) Changes in sea level.
In: J.T. Houghton, Y. Ding, D.J. Griggs, M. Noguer,
P.J. van der Linden, X. Dai, K. Maskell, and C.A.
Johnson (Eds.), Climate Change 2001: The Scientific
Basis. Contribution of Working Group 1 to the Third
Assessment Report of the Intergovernmental Panel on
Climate Change, Cambridge University Press, Cambridge, U.K.
Citterio, M., and Ahlstrøm, A.P. (2013) (Brief communication) The aerophotogrammetric map of Greenland
ice masses. The Cryosphere, 7(2), 445–449.
Citterio, M., Mottram, R., Larsen, S.H., and Ahlstrøm,
A. (2009) Glaciological investigations at the Malmbjerg mining prospect, central East Greenland. Geological Survey of Denmark and Greenland Bulletin, 17.
Dall, J., Madsen, S.N., Keller, K., and Forsberg, R.
(2001) Topography and penetration of the Greenland
Ice Sheet measured with airborne SAR interferometry.
Geophysical Research Letters, 28(9), doi: 10.1029/
2000GL011,787.
De Angelis, H. (2003) Glacier surge after ice shelf collapse. Science, 299(5612), 1560–1562.
de Juan, J., Elósegui, P., Nettles, M., Larsen, T.B., Davis,
J.L., Hamilton, G.S., Stearns, L.A., Andersen, M.L.,
Ekström, G., Ahlstrøm, A.P. et al. (2010) Sudden
increase in tidal response linked to calving and acceleration at a large Greenland outlet glacier. Geophysical
Research Letters, 37(12), L12,501.
Dwyer, J. (1995) Mapping tide-water glacier dynamics in
East Greenland using Landsat data. Journal of Glaciology, 41, 584–595.
Eisen, O., Frezzotti, M., Genthon, C., Isaksson, E.,
Magand, O., van den Broeke, M.R., Dixon, D.A.,
Ekaykin, A., Holmlund, P., Kameda, T. et al. (2008)
Ground-based measurements of spatial and temporal
variability of snow accumulation in East Antarctica.
Reviews of Geophysics, 46(RG2001), 139.
Elósegui, P., Davis, J.L., Johansson, J.M., and Shapiro, I.
(1996) Detection of transient motions with the Global
Positioning System. Journal of Geophysical Research,
101(B5), 11249–11262.
Elósegui, P., Davis, J.L., Oberlander, D., Baena, R., and
Ekström, G. (2006) Accuracy of high-rate GPS for
seismology. Geophysical Research Letters, 33(L11308),
doi: 10.1029/2006GL026065.
Ettema, J., van den Broeke, M.R., Van Meijgaard, E.,
Van De Berg, W.J., Bamber, J.L., Box, J.E., and Bales,
R.C. (2009) Higher surface mass balance of the Greenland ice sheet revealed by high-resolution climate
modeling. Geophysical Research Letters, 36(12), 1–5.
Ewert, H., Groh, A., and Dietrich, R. (2011) Volume and
mass changes of the Greenland ice sheet inferred from
ICESat and GRACE. Journal of Geodynamics, doi:
10.1016/j.jog.2011.06.003.
Ferrigno, J.G., Lucchitta, B.K., Mullins, K.F., Allison,
A.L., Alen, R.J., and Gould, W.G. (1993) Velocity
measurements and changes in position of Thwaites
Glacier/iceberg tongue from aerial photography,
Landsat images and NOAA AVHRR data. Annals of
Glaciology, 17, 239–244.
Fettweis, X., Van Ypersele, J.-P., Gallée, H., Lefebre, F.,
and Lefebvre, W. (2007) The 1979–2005 Greenland ice
sheet melt extent from passive microwave data using
an improved version of the melt retrieval XPGR algorithm. Geophysical Research Letters, 34(5), 5.
Fettweis, X., Tedesco, M., van den Broeke, M., and
Ettema, J. (2011) Melting trends over the Greenland
ice sheet (1958–2009) from spaceborne microwave data
and regional climate models. The Cryosphere, 5, 359–
375.
Fujisada, H., Bailey, G., Kelly, G., Hara, S., and Abrams,
M. (2005) ASTER DEM performance. IEEE Transactions on Geoscience and Remote Sensing, 43(12), 2707–
2714.
Furbish, D.J., and Andrews, J.T. (1984) The use of hypsometry to indicate long-term stability and response of
valley glaciers to changes in mass transfer. Journal of
Glaciology, 41(139), 199–211.
Gillett, N.P., Stone, D.A., Stott, P.A., Nozawa, T., Karpechko, A.Y., Hegerl, G.C., Wehner, M.F., and Jones,
P.D. (2008) Attribution of polar warming to human
influence. Nature Geoscience, 1(11), 750–754.
Gregory, J., Huybrechts, P., and Raper, S. (2004)
Threatened loss of the Greenland Ice Sheet. Nature—
News&Views, 428, 616.
Hall, D.K., Williams, J.R.S., Luthcke, S.B., and
Digirolamo, N.E. (2008) Greenland ice sheet surface
temperature, melt and mass loss: 2000–06. Journal of
Glaciology, 54(184), 81–93.
Hamilton, G., and Whillans, I. (2000) Point measurements of mass balance of the Greenland Ice Sheet using
precision vertical Global Positioning System (GPS)
survey. Journal of Geophysical Research, 105(B7),
16295–16301.
Hanna, E., Huybrechts, P., Janssens, I., Cappelen, J., and
Stephens, A. (2005) Runoff and mass balance of the
Greenland ice sheet: 1958–2003. Journal of Geophysical
Research, 110(D13108).
202
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Hanna, E., McConnell, J., Das, S., Cappelen, J., and
Stephens, A.G. (2006) Observed and modeled Greenland Ice Sheet snow accumulation, 1958–2003, and
links with regional climate forcing. Journal of Climate,
19, 344–358.
Hanna, E., Huybrechts, P., Steffen, K., Cappelen, J.,
Huff, R., Shuman, C., Irvine-Fynn, T., Wise, S., and
Griffiths, M. (2008) Increased runoff from melt from
the Greenland Ice Sheet: A response to global warming. Journal of Climate, 21(2), 331–341.
Holland, D., Thomas, R., de Young, B., and Ribergaard,
M. (2008) Acceleration of Jakobshavn Isbrae triggered
by warm subsurface ocean waters. Nature Geoscience,
1, 659–664.
Howat, I.M., Joughin, I., Tulaczyk, S., and Gogineni, S.
(2005) Rapid retreat and acceleration of Helheim
Glacier, East Greenland. Geophysical Research Letters,
32(L22502), doi: 10.1029/2005GL024,737.
Howat, I., Joughin, I., and Scambos, T. (2007) Rapid
changes in ice discharge from Greenland outlet
glaciers. Science, 607, doi: 10.1126/science.1138478.
Howat, I.M., Box, J.E., and Ahn, Y. (2011) Seasonal
variability in the dynamics of marine-terminating
outlet glaciers in Greenland. Journal of Glaciology,
56(198), 113.
Jiskoot, H. (2002) Central East Greenland GLIMS Glacier
Database, Digital Media, National Snow and Ice Data
Center/World Data Center for Glaciology.
Jiskoot, H., and Juhlin, D.T. (2009) Surge of a small East
Greenland glacier, 2001–2007, suggests Svalbard-type
surge mechanism. Journal of Glaciology, 55(181), 567–
570.
Jiskoot, H., Pedersen, A.K., and Murray, T. (2001)
Multi-model photogrammetric analysis of the 1990s
surge of Sortebræ, East Greenland. Journal of Glaciology, 47(159), 677–687.
Jiskoot, H., Murray, T., and Luckman, A. (2003) Surge
potential and drainage basin characteristics in East
Greenland. Annals of Glaciology, 36, 142–148.
Jiskoot, H., Curran, C.M., Tessler, D.L., and Shenton,
L.R. (2009) Changes in Clemenceau Icefield and
Chaba Group glaciers, Canada, related to hypsometry,
tributary detachment, length-slope and area-aspect
relations. Annals of Glaciology, 50(53), 133–143.
Jiskoot, H., Juhlin, D., Citterio, M., and St. Pierre, H.
(2012) Tidewater glacier fluctuations in central East
Greenland coastal and fjord regions (1980s–2005).
Annals of Glaciology, 53(60), 35–44.
Joughin, I., Gray, L., Bindschadler, R., Price, S., Morse,
D., Hulbe, C., Mattar, K., and Werner, C. (1999)
Tributaries of West Antarctic ice streams revealed by
RADARSAT interferometry. Science, 286(5438), 283–
286.
Joughin, I., Abdalati, W., and Fahnestock, M. (2004)
Large fluctuations in speed on Greenland’s Jakobshavn Isbræ glacier. Nature, 432, 608–610.
Joughin, I., Smith, B.E., Howat, I.M., Scambos, T., and
Moon, T. (2010) Greenland flow variability from
ice-sheet-wide velocity mapping. Journal of Glaciology,
56(197), 415–430.
Kääb, A. (2002) Monitoring high-mountain terrain
deformation from air- and spaceborne optical data:
Examples using digital aerial imagery and ASTER
data. ISPRS Journal of Photogrammetry and Remote
Sensing, 57(1/2), 39–52.
Kääb, A., Huggel, C., Paul, F., Wessels, R., Raup, B.,
Kieffer, H., and Kargel, J. (2002) Glacier monitoring
from ASTER imagery: Accuracy and applications.
Paper presented at Proceedings of EARSeL LISSIG
Workshop: Observing Our Cryosphere from Space,
p. 43.
Kargel, J.S., Ahlstrøm, A.P., Alley, R.B., Bamber, J.L.,
Benham, T.J., Box, J.E., Chen, C., Christoffersen, P.,
Citterio, M., Cogley, J.G. et al. (2012) (Brief communication) Greenland’s shrinking ice cover: Fast times but
not that fast. The Cryosphere, 6, 533–537.
Khan, S.A., Wahr, J., Bevis, M., Velicogna, I., and Kendrick, E. (2010) Spread of ice mass loss into northwest
Greenland observed by GRACE and GPS. Geophysical Research Letters, 37(6), L06,501.
Kirchner, G. (1963) Observations at bore holes sunk
through the Schuchert Gletscher in north-east Greenland. Journal of Glaciology, 4(36), 817–818.
Krabill, W., Abdalati, W., Frederick, E., Manizade, S.,
Sonntag, J., Swift, R., Wright, W., and Yungel, J.
(2000) Greenland ice sheet: High elevation balance
and peripheral thinning. Science, 289(5478), 428–430.
Krabill, W., Hanna, E., Huybrechts, P., Abdalati, W.,
Cappelen, J., Csatho, B., Frederick, E., Manizade,
S., Martin, C., and Sonntag, J. (2004) Greenland Ice
Sheet: Increased coastal thinning. Geophysical
Research Letters, 31(L24402).
Lemke, P. (2007) Changes in snow, ice, and frozen
ground, Climate Change 2007: Working Group I: The
Physical Science Basis, Intergovernmental Panel on
Climate Change (IPCC), Washington, D.C., p. 48.
Lichten, S.M., and Border, J.S. (1987) Strategies for highprecision Global Positioning System orbit determination. Journal of Geophysical Research, 92, 12751–
12762.
Lie, Ø., and Paasche, Ø. (2006) How extreme was northern hemisphere seasonality during the Younger Dryas?
Quaternary Science Reviews, 25, 404–407.
Lucchitta, B., and Ferguson, H. (1986) Antarctica:
Measuring glacier velocity from satellite images.
Science, 234, 1105–1108.
Luckman, A., and Murray, T. (2005) Seasonal variation
in velocity before retreat of Jakobshavn Isbræ, Greenland. Geophysical Research Letters, 32(L08501).
Luckman, A., Murray, T., Jiskoot, H., Pritchard, H., and
Strozzi, T. (2003) Automatic feature-tracing measurement of outlet glacier velocities on a regional scale in
East Greenland. Annals of Glaciology, 36, 129–134.
Luckman, A., Murray, T., de Lange, R., and Hanna, E.
(2006) Rapid and synchronous ice-dynamic changes in
East Greenland. Geophysical Research Letters,
33(L03503).
References 203
Luthcke, S., Zwally, H., Abdalati, W., Rowlands, D.,
Ray, R., Nerem, R., Lemoine, F., McCarthy, J., and
Chinn, D. (2006) Recent Greenland ice mass loss by
drainage system from satellite gravity observations.
Science, 314(1286), 5.
Mernild, S.H., Liston, G.E., and Hasholt, B. (2008)
East Greenland freshwater runoff to the Greenland–
Iceland–Norwegian Seas 1999–2004 and 2071–2100.
Hydrological Processes, 22(23), 4571–4586.
Mernild, S.H., Mote, T.L., and Liston, G.E. (2011)
Greenland ice sheet surface melt extent and trends:
1960–2010. Journal of Glaciology, 57(204), 621–628.
Mote, T.L. (2007) Greenland surface melt trends 1973–
2007: Evidence of a large increase in 2007. Geophysical
Research Letters, 34(L22507).
Murray, T., Sharrer, K., James, T.D., Dye, S.R., Hanna,
E., Selmes, N., Luckman, A., Hughes, A.L.C., Cook,
S., and Huybrechts, P. (2010) Ocean regulation
hypothesis for glacier dynamics in southeast Greenland and implications for ice sheet mass changes.
Journal of Geophysical Research, 115(F03026), doi:
10.1029/2009JF001522.
Parizek, B., and Alley, R. (2004) Implications of
increased Greenland surface melt under globalwarming scenarios: ice-sheet simulations. Quaternary
Science Reviews, 23(9/10), 1013–1027.
Paterson, W. (1994) The Physics of Glaciers (Third Edition), Pergamon Press, Oxford, U.K.
Paul, F., Huggel, C., and Kääb, A. (2004) Combining
satellite multispectral image data and a digital elevation model for mapping debris-covered glaciers.
Remote Sensing of Environment, 89, 510–518.
Paul, F., Barry, R.G., Cogley, J.G., Frey, H., Haeberli,
W., Ohmura, A., Ommanney, C.S.L., Raup, B.,
Rivera, A., and Zemp, M. (2009) Recommendations
for the compilation of glacier inventory data from
digital sources. Annals of Glaciology, 50(53), 119–126.
Polyakov, I., Alekseev, G., Timokhov, L., Bhatt, U.,
Simmons, H., Walsh, D., Walsh, J., and Zakharov,
V. (2004) Variability of the Intermediate Atlantic
Water of the Arctic Ocean over the last 100 years.
Journal of Climate, 17, 4485 –4497.
Price, S., Payne, A.J., Howat, I.M., and Smith, B. (2011)
Committed sea-level rise for the next century from
Greenland ice sheet dynamics during the past decade.
Proceedings National Academy Sciences U.S.A., 108,
1–6.
Pritchard, H., Murray, T., Strozzi, T., Barr, S., and
Luckman, A. (2003) Glacier surge-related topographic
change derived from Synthetic Aperture Radar interferometry. Journal of Glaciology, 49(166), 381–390.
Pritchard, H.D., Murray, T., Luckman, A., Strozzi, T.,
and Barr, S. (2005) Glacier surge dynamics of
Sortebræ, East Greenland, from synthetic aperture
radar feature tracking. Journal of Geophysical
Research, 110(F03005).
Ramillien, G., Lombard, A., Cazenave, A., Ivins, E.,
Llubes, M., Remy, F., and Biancale, R. (2006) Interannual variations of the mass balance of the Antarctica
and Greenland ice sheets from GRACE. Global and
Planetary Change, 53(3), 198–208.
Raper, S.C.B., and Braithwaite, R.J. (2009) Glacier
volume response time and its links to climate and
topography based on a conceptual model of glacier
hypsometry. The Cryosphere, 3, 183–194.
Rastner, P., Bolch, T., Mölg, N., Machguth, H., Le Bris,
R., and Paul, F. (2012) The first complete inventory of
the local glaciers and ice caps on Greenland. The Cryosphere, 6(6), 1483–1495.
Raup, B.H., Kieffer, H.H., Hare, T.M., and Kargel, J.S.
(2000) Generation of data acquisition requests for the
ASTER satellite instrument for monitoring a globally
distributed target: Glaciers. IEEE Transactions on
Geoscience and Remote Sensing, 38(2), 1105–1112.
Rennermalm, A.K., Smith, L.C., Stroeve, J.C., and Chu,
V.W. (2009) Does sea ice influence Greenland ice
sheet surface-melt? Environmental Research Letters,
4(024011), doi: 10.1088/17489326/4/2/024,011.
Rignot, E., and Kanagaratnam, P. (2006) Changes in the
velocity structure of the Greenland Ice Sheet. Science,
311, 986–990.
Rignot, E., Velicogna, I., van den Broeke, M.R.,
Monaghan, A., and Lenaerts J. (2011) Acceleration
of the contribution of the Greenland and Antarctic
ice sheets to sea level rise. Geophysical Research
Letters, 38(5), L05,503, doi: 10.1029/2011GL046,583.
Rivera, A., Casassa, G., Bamber, J., and Kääb, A. (2005)
Ice-elevation changes of Glaciar Chico, southern Patagonia, using ASTER DEMs, aerial photographs and
GPS data. Journal of Glaciology, 51(172), 105–112.
Scambos, T., and Fahnestock, M.A. (1998) Improving
digital elevation models over ice sheets using
AVHRR-based photoclinometry. Journal of Glaciology, 44(146), 97–103.
Scambos, T., Dutkiewicz, M., Wilson, J., and Bindschadler, R. (1992) Application of image crosscorrelation to the measurement of glacier velocity
using satellite image data. Remote Sensing of Environment, 42(3), 177–186.
Schrama, E.J.O., and Wouters, B. (2011) Revisiting
Greenland ice sheet mass loss observed by GRACE.
Journal of Geophysical Research, 116(B2), B02,407.
Serreze, M., Holland, M., and Stroeve, J. (2007) Perspectives on the Arctic’s shrinking sea-ice cover. Science—
Perspectives, 315, 1533–1536.
Shepherd, A., Wingham, D., Mansley, J., and Corr, H.
(2001) Inland thinning of Pine Island Glacier, West
Antarctica. Science, 291(5505), 862–864.
Slobbe, D.C., Ditmar, P., and Lindenbergh, R.C. (2009)
Estimating the rates of mass change, ice volume change
and snow volume change in Greenland from ICESat
and GRACE data. Geophysical Journal International,
176(1), 95–106.
Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis,
M., Averyt, K.B., Tignor, M., and Miller, H. (Eds.)
(2007) Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on
204
Glacier fluctuations and dynamics around the margin of the Greenland Ice Sheet
Climate Change, 2007, Cambridge University Press,
Cambridge, U.K.
Stearns, L., Hamilton, G., and Reeh, N. (2005) Multidecadal record of ice dynamics on Daugaard Jensen
Gletscher, East Greenland, from satellite imagery and
terrestrial measurements. Annals of Glaciology, 42(1),
53–58.
Stearns, L.A., and Hamilton, G.S. (2007) Rapid volume
loss from two East Greenland outlet glaciers
quantified using repeat stereo satellite imagery. Geophysical Research Letters, 34(L05503), doi: 10.1029/
2006GL028982.
Steffen, K. (2004) The melt anomaly of 2002 on the
Greenland Ice Sheet from active and passive microwave satellite observations. Geophysical Research Letters, 31(L20402).
Stevens, N., Garbeil, H., and Mouginis-Mark, P. (2004)
NASA EOS Terra ASTER: Volcanic topographic
mapping and capability. Remote Sensing of Environment, 90(3), 405–414.
Straneo, F., Hamilton, G.S., Sutherland, D.A., Stearns,
L.A., Davidson, F., Hammill, M.O., Stenson, G.B.,
and Rosing-Asvid, A. (2010) Rapid circulation of
warm subtropical waters in a major glacial fjord in
East Greenland. Nature Geoscience, 3(3), 1–5.
Straneo, F., Curry, R., Sutherland, D., Hamilton, G.S.,
Cenedese, C., Våge, K., and Stearns, L.A. (2011)
Impact of fjord dynamics and glacial runoff on the
circulation near Helheim Glacier. Nature Geoscience,
4, 322–327.
Tedesco, M. (2007) Snowmelt detection over the Greenland ice sheet from SSM/I brightness temperature daily
variations. Geophysical Research Letters, 34(2), 6.
Tedesco, M., Fettweis, X., van den Broeke, M.R., Van De
Wal, R.S.W., Smeets, C.J.P.P., Van De Berg, W.J.,
Serreze, M.C., and Box J.E. (2011) The role of albedo
and accumulation in the 2010 melting record in Greenland. Environmental Research Letters, 6(1), 014,005.
Thomas, R., Csatho, B., Davis, C., Kim, C., Krabill, W.,
Manizade, S., McConnell, J., and Sonntag, J. (2001)
Mass balance of higher-elevation parts of the Greenland ice sheet. Journal of Geophysical Research,
106(D24), 33707–33716.
Thomas, R., Abdalati, E., Frederick, E., Krabill, W.,
Manizade, S., and Steffen, K. (2003) Investigation of
surface melting and dynamic thinning on Jakobshavn
Isbrae, Greenland. Journal of Glaciology, 49(165), 231–
239.
Thomas, R., Fredrick, E., Krabill, W., Manizade, S., and
Martin, C. (2006) Progressive increase in ice loss from
Greenland. Geophysical Research Letters, 33(L10503),
4.
Thomas, R., Frederick, E., Krabill, W., Manizade, S.,
and Martin, C. (2009) Recent changes on Greenland
outlet glaciers. Journal of Glaciology, 55(189), 147–162.
van den Broeke, M., Bamber, J., Ettema, J., Rignot, E.,
Schrama, E., Van De Berg, W.J., Van Meijgaard, E.,
Velicogna, I., and Wouters, B. (2009) Partitioning
recent Greenland mass loss. Science, 326(5955), 984–
986.
Velicogna, I. (2009) Increasing rates of ice mass loss
from the Greenland and Antarctic ice sheets revealed
by GRACE. Geophysical Research Letters, 36(19),
L19,503.
Velicogna, I., Wahr, J., Hanna, E., and Huybrechts, P.
(2005) Short term mass variability in Greenland, from
GRACE. Geophysical Research Letters, 32(L05501).
Vignon, F., Arnaud, Y., and Kaser, G. (2003) Quantification of glacier volume change using topographic and
ASTER DEMs. Paper presented at Geoscience and
Remote Sensing Symposium, IGARSS’03. Proceedings.
2003 IEEE International, 4, 2605–2607.
Weidick, A. (1988) Surging Glaciers in Greenland: A
Status Report (Tech. Rep. 140), Rapport Grønlands
Geologiske Undersøgels.
Weidick, A. (1995) Satellite Image Atlas of Glaciers of
the World: Greenland (Tech. Rep. 1386-C), U.S. Geological Survey Professional Paper.
Whillans, I.M., and Tseng, Y.H. (1995) Automatic tracking of crevasses on satellite images. Cold Regions
Science and Technology, 23, 201–214.
Wingham, D., Shepherd, A., Muir, A., and Marshall, G.
(2006) Mass balance of the Antarctic ice sheet. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 364(1844),
1627–1635.
Wouters, B., Chambers, D., and Schrama, E.J.O. (2008)
GRACE observes small-scale mass loss in Greenland.
Geophysical Research Letters, 35(L20501).
Wuite, J. (2006) Spatial and temporal dynamics of three
East Antarctic outlet glaciers and their floating ice
tongues, Ph.D. thesis, The Ohio State University.
Yde, J. (2011) Greenland glaciers outside the ice sheet. In:
V.P. Singh, P. Singh, and U.K. Haritashya (Eds.),
Encyclopedia of Snow, Ice and Glaciers (Encyclopedia
of Earth Sciences Series), Springer-Verlag, Berlin.
Zwally, H., and Giovinetto, M. (2001) Balance mass flux
and ice velocity across the equilibrium line in drainage
systems of Greenland. Journal of Geophysical
Research, 106(D24), 33717–33728.
Zwally, H.J., and Li, J. (2002) Seasonal and interannual
variations of firn densification and ice-sheet surface
elevation at Greenland summit. Journal of
Glaciology, 48(161), 199–207.