ARTICLESPUBLISHED ONLINE: XX MONTH XXXX | DOI: 10.1038/NGEO1481

An aerial view of 80 years of climate-relatedglacier fluctuations in southeast GreenlandAnders A. Bjørk1*, Kurt H. Kjær1, Niels J. Korsgaard1, Shfaqat A. Khan2, Kristian K. Kjeldsen1,Camilla S. Andresen3, Jason E. Box4,5, Nicolaj K. Larsen6 and Svend Funder1

Widespread retreat of glaciers has been observed along the southeastern margin of Greenland. This retreat has beenassociated with increased air and ocean temperatures. However, most observations are from the satellite era; presatelliteobservations of Greenlandic glaciers are rare. Here we present a unique record that documents the frontal positions for132 southeast Greenlandic glaciers from rediscovered historical aerial imagery beginning in the early 1930s. We combinethe historical aerial images with both early and modern satellite imagery to extract frontal variations of marine- andland-terminating outlet glaciers, as well as local glaciers and ice caps, over the past 80 years. The images reveal a regionalresponse to external forcing regardless of glacier type, terminal environment and size. Furthermore, the recent retreat wasmatched in its vigour during a period of warming in the 1930s with comparable increases in air temperature. We show thatmany land-terminating glaciers underwent a more rapid retreat in the 1930s than in the 2000s, whereas marine-terminatingglaciers retreated more rapidly during the recent warming.

Since the mid 1990s, sections of the Greenland Ice SheetQ1 1

margin have been retreating significantly1,2, leading to2

glacier speedups and dynamic ice loss3–9. Notable are3

the major changes in southeast Greenland. Evidence from the4

Gravity Recovery and Climate Experiment mission10–12, radar5

interferometry13 and a global positioning system14 show significant6

increases in the mass-loss rate and glacier velocities in this region7

since late 2003. This acceleration is partly attributed to increased8

dynamic thinning triggered by rising air and ocean temperatures9,15.9

However, although the mass loss of the entire ice sheet has steadily10

increased16, glaciers along the southeastern margin decelerated and11

advanced in 2006 (refs 3,5,17), suggesting episodic rather than12

steady changes in frontal positions.13

In situ mass-balance measurements in southeast Greenland are14

rare and exist only for a single glacier over a longer timescale18.15

Consequently, frontal area and length changes of outlet glaciers16

have been used implicitly to derive changes in regional mass17

balance19–21. Several studies have investigated short-term frontal18

changes of glaciers on the southeast coast of Greenland and almost19

all have been limited to satellite imagery19–22. Although a few studies20

have investigated sporadic aerial imagery and frontal changes in the21

region on a longer timescale23,24, none have so far investigated the22

entire time span from 1931 to present.23

In the early twentieth century several expeditions were launched24

to east Greenland to produce new geographical maps and in many25

instances to map the coastlines for the first time. Using state-of-26

the-art photographic and aircraft technology, the seventh Thule27

Expedition25,26conducted a systematic survey of the southeast coast28

of Greenland in 1932–1933. After the initial use the images were29

classified and not generally accessible. Later, they were archived30

in a citadel outside Copenhagen, and knowledge of their location31

forgotten. Our rediscovery of the early twentieth-century images32

1Centre for GeoGenetics, Natural History Museum of Denmark, University of Copenhagen, Copenhagen, Denmark, 2DTU Space — National Space Institute,Technical University of Denmark, Department of Geodesy, Copenhagen, Denmark, 3Geological Survey of Denmark and Greenland, Department of MarineGeology and Glaciology, Copenhagen, Denmark, 4Department of Geography, The Ohio State University, Columbus, Ohio, USA, 5Byrd Polar ResearchCenter, The Ohio State University, Columbus, Ohio, USA, 6Department of Geoscience, Aarhus University, Aarhus, Denmark. *e-mail: andersb@snm.ku.dk.

Figure 1 | Historical aerial photographs from the seventh ThuleExpedition, 1933. The oblique photographs were recorded from a HeinkelMKII hydroplane and show the Helheim Glacier (SEGL014). Images (no.20.907–20.910) were recorded from left to right at 4,000 m elevation.

along with air photos obtained by the USmilitary in theWorldWar 33

II era provide us with a unique opportunity to observe changes in 34

Greenlandic glaciers at high spatial resolution from a period where 35

quantitative observations and measurements are absent (Fig. 1 and 36

Supplementary Information). Furthermore, the recent fluctuations 37

in theGreenland Ice Sheetmass balance underlines that longer-term 38

records are needed to explain a causal relationship between climate 39

and glacier variability in this region.

Q2

40

Unravelling historical imagery 41

We analyse the frontal behaviour of 132 glaciers along more than 42

600 km of the southeast Greenland coastline, between 61.5◦N 43

and 66.5◦N latitude. Combining aerial photographs (1931–1933, 44

NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience 1

ARTICLES NATURE GEOSCIENCE DOI: 10.1038/NGEO1481

Table 1 | Images and sources used in this study.

Year Date Data type Source Ground resolution Number of images used

1931 20 July Oblique aerial BAARE <60 m 361932 August Oblique aerial Seventh Thule <60 m 91933 August Oblique aerial Seventh Thule <60 m 2121933 Late July Terrestrial Seventh Thule n.a. 41943? End of melt season Vertical aerial USAAF/USN ∼1.7 m 1241965 Late July Satellite Corona 0.61 m 281972 Early September–early October Satellite Landsat 1 MSS 60 m 91981/1985 Late July Vertical aerial KMS 4 m (ortho) 3112000 August Satellite Landsat 7 ETM+ 15 m (pan) 82010 Late July/August Satellite Landsat 7 ETM+ 15 m (pan) 9

KMS, National Danish Survey and Cadastre; USAAF/USN, United States Army Air Force/United States Navy; n.a., not applicable. The ground resolution of the oblique aerial images vary within the frameof the image, only images within a ground resolution of 60 m or better have been used owing to the relative higher uncertainties in rectifying the oblique images. 1981/1985 vertical aerial images fromKMS were scanned at a resolution of ∼2 m. The Landsat 7 scenes have a spatial resolution of 30 m and were pan-sharpened to a resolution of 15 m using the panchromatic band 8. The exact time ofrecording remains uncertain for the USAAF/USN images (see Supplementary Information), based on the snow conditions images must have been recorded at the end of the melt season.

Gre

enla

nd Ic

e Sh

eet

Atla

ntic

Oce

an

38° W

66° N

62° N

1933¬1943 1943¬1965 1965¬1972 1972¬1981* 1981*¬2000 2000*¬2010

0¬1010¬2525¬5050¬100100¬200200¬420

0¬1010¬25

25¬5050¬100100¬200

200¬500500¬888

Advance (m yr–1)

Retreat (m yr–1)

km

N

0 50 100

Figure 2 | Frontal changes of southeast Greenland glaciers. Changes are shown as rates during six observation periods from 1933 to 2010. Red circlesshow retreat and green circles advance (m yr−1). A total of 132 glaciers are included, originating both from the ice sheet as well as local glaciers and icecaps. The data gap in the central fjord region in the second and third period is a result of a missing 1965 image. Note, the 17 southernmost aerial imageswere recorded in 1985 and not in 1981, and frontal change rates reflect 1985 position. Background image is an ASTER GDEM (product of METI and theNational Aeronautics and Space Administration) with land and ice-sheet masks added. Error bars reflect the measuring uncertainty for eachobservational period.

1943, 1981–1985), terrestrial photographs (1933) and satellite1

images (1965, 1972, 2000, 2010) we determine frontal changes2

over a period of 80 years (Table 1). A semi-automated reference-3

point method, modified from the box method20,21,27, is used to4

measure frontal fluctuations (Supplementary Information). We5

differentiate between local glaciers and ice caps peripheral to the6

ice sheet, and outlets of the Greenland Ice Sheet. The surveyed7

region covers a wide range of different topographic settings8

and ice marginal environments (Supplementary Information).9

Of 36 main (>2 km wide) marine-terminating glaciers, 30 are10

represented in the data set. The remaining six large outlet11

glaciers could not be digitized with adequate accuracy from the12

historical imagery.13

Throughout the 80 years of observation, the southeast Greenland14

glaciers have undergone widespread retreat, with only a few15

glaciers advancing (5%) or remaining stationary (<2%; Fig. 2). The16

observational period can be divided into three distinct periods of17

frontal change. Two recessional events stand out, one during the18

1930s (1933–1943) at the onset of our observations and one during 19

the 2000s (2000–2010). The two periods of retreat are interrupted 20

by a period of widespread advance from 1943 to 1972. 21

During our first observation period (1933–1943) most of the 22

glaciers retreated (Fig. 2) with typical rates of 10–50m yr−1 and 23

a maximum rate of 374m yr−1. This retreat occurred just after a 24

warming period with abnormally high temperatures and during 25

a period with slightly decreasing temperatures and higher-than- 26

usual accumulation rates28. The retreat pattern is non-uniform 27

as a number of glaciers in the central fjord region, primarily 28

marine-terminating glaciers originating from ice sheets, advanced 29

within this period (Fig. 2). The retreat in the 2000s is more uniform 30

with only 5% of the 132 glaciers advancing. The highest retreat 31

rates observed in the study occurred in this latest period. All 32

glaciers retreating at rates >200 m yr−1 are large (>2 km wide) 33

marine-terminating outlets from the ice sheet, with Midgaard 34

Glacier (in the northern part of the study area) showing the largest 35

retreat of 887m yr−1. 36

2 NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience

NATURE GEOSCIENCE DOI: 10.1038/NGEO1481 ARTICLES

1920 1940 1960 1980 2000

Ave

rage

ret

reat

rat

e (m

yr–1

)SS

T a

nom

aly

(°C

)

1910 1930 1950 1970 1990 2010

ObservationsAerial photoSatellite

1

0

Tasiilaq m

ean annual air tem

perature (°C

)

–1

–2

–3

b

0

–10

–20

–30

–40

–50

1.5

1.0

0.5

0

–0.5

–1.0

a

c

Year

Figure 3 | Average frontal changes of glaciers and temperature records.a, SST anomaly at two locations (see map to the right) during 1912–2011from Met Office Hadley Centre observations data sets (green and purplecurve) (Supplementary Information). The SST anomaly is relative to a1961–2000 baseline. b, The air temperature record from Tasiilaq (bluecurve). Periods of combined forcing from air and SST warming and coolingare shaded red and blue. c, Frontal change rates of many glaciers areaveraged, hence the mid-century cooling and related advance observed inFig. 2 is seen as a slowdown or slight retreat. Error bars reflect themeasuring uncertainty for each observational period.

Glacial response to warming1

Air temperatures in southeast Greenland have reached record2

heights since continuous observations began in 1895 (ref. 29).3

The temperature rise has been steady at 1.3 ◦C per decade since4

1993. However, the recent warming is not unique for southeast5

Greenland (Fig. 3). An earlier warming period of 2.0 ◦C per6

decade30,31 occurred from 1919 to 1932 (early twentieth-century7

warming, ECW) and was primarily dominated by spring warming,8

leading to a prolonged melt season31. The warming has been9

attributed to unusual circulation patterns in the atmosphere10

over the North Atlantic Ocean32 and lower insolation due to11

aerosol presence33. However, recorded annual temperatures are12

0.45 ◦C warmer in the recent decade than during the peak of13

the ECW. Both warming periods share similar North Atlantic14

Oscillation phases and follow the interdecadal variations and15

overall increase in sea surface temperature (SST) (Fig. 3). We16

therefore use the ECW as an analogue to the present and17

a means to better understand glacier response and sensitivity18

to external forcing.19

The 1930s and 2000s retreat events were associated with both20

rising air- and SSTs well above the mean (Fig. 3). These two21

retreat periods are also seen as the peak calving periods in a22

study of fjord sediments from Sermilik Fjord where the Helheim23

Glacier terminates and here it was documented that decadal scale24

variability in frontal positions is related to decadal scale changes25

in both air and sea temperatures34. Notably, our findings show26

that decadal scale climate-forced glacier changes are not restricted27

Retreat rates

1933¬1943 < 2000¬20101933¬1943 = ∗2000¬20101933¬1943 > 2000¬2010

Num

ber

of o

bser

vatio

ns

¬600 ¬400 ¬200 0

(1933¬1943¬2000¬2010)

Retreat rate difference (m yr¬1)200 400

a b

c 16

12

8

4

0

Figure 4 | Retreat-rate difference between the large retreat periods1933–1943 and 2000–2010. Yellow circles indicate glaciers experiencingfaster or equal change in the 1933–1943 period in comparison with the2000–2010 period; blue is the opposite; green circles indicate equal frontalretreat rates (within the measuring uncertainty of±7 m yr−1).a, Land-terminating glaciers. b, Marine-terminating glaciers. c, Histogramdepicting the difference in retreat rates between the two periods.

to the large Helheim Glacier, but that most of the glaciers along 28

the southeast coast followed a similar retreat pattern (Fig. 4). 29

Of all glaciers, 35% show higher retreat rates during the 1930s 30

and 20% show similar retreat rates. For land-terminating glaciers 31

42% were faster during the 1930s and 33% similar. However, 32

more glaciers retreated during the 2000s. Of the seven glaciers 33

undergoing high retreat rates during the 2000s (>200m yr−1 34

over a ten-year period), all are relatively large marine-terminating 35

outlet glaciers that have proved especially sensitive to changes in 36

ocean temperature9,15. 37

Divergence in frontal behaviour between the two warm events 38

is observed in the central fjord region (Figs 2 and 5). Here, a large 39

number of primarily marine-terminating glaciers originating from 40

the ice sheet advanced during the early observation period. This 41

region is more mountainous than others and hosts glaciers that 42

reach the fjords through long alpine valleys. Of all the advancing 43

marine-terminating glaciers from the ice sheet, nine out of ten 44

have abundant proglacial ice mélange. In this particular region, 45

the fjord ice breaks up later in the season than in the rest of 46

the study area, suggesting that the late breakup combined with 47

the presence of calf ice restrain rapid glacial retreat35. It is also 48

possible that the higher-than-normal accumulation in the ECW 49

(ref. 28) is dominating over the effect of increased temperatures 50

on melt rates. Moreover, some response to the early ECW may 51

already have occurred before the beginning of our observations 52

(Supplementary Information). 53

Mid-century cooling 54

As the ECW faded, a period of cooling began in the mid 1950s that 55

lasted until the early 1970s. The glacier response to the cooling was 56

an advance, primarily among marine-terminating glaciers. Surpris- 57

ingly, a large portion (60%) of our glaciers advanced, many already 58

at the beginning of the period (Fig. 5), whereas another 12% were 59

stationary during the period. This shows that the glaciers in this 60

region are more sensitive to climate change and react more rapidly 61

to cooling than indicated both by earlier studies36,37 and by more 62

NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience 3

ARTICLES NATURE GEOSCIENCE DOI: 10.1038/NGEO1481

38° W

38° W

66° N

66° N

Green

land

Ice S

heet

Gre

enla

nd Ic

e Sh

eet

Atla

ntic

Oce

an

62° N

62° N

1933¬1943 1943¬1965 1965¬1972 1972¬1981* 1981*¬2000 2000¬2010

1933¬1943 1943¬1965 1965¬1972 1972¬1981* 1981*¬2000 2000¬2010

dz (

m)

1933

¬20

07

land

-ter

min

atin

g m

argi

ns

0 100 200 300

Distance north to south (km)

400 500 600

Atla

ntic

Oce

an

0¬1010¬25

25¬5050¬100100¬200200¬420

0¬1010¬25

25¬5050¬100100¬200

200¬420500¬888

Advance (m yr–1)

Retreat (m yr –1)

0 50km

N

100

0

–20

Ave

rage

ret

reat

rat

e (m

yr–1

)

Ave

rage

ret

reat

rat

e (m

yr–1

)

Ave

rage

ret

reat

rat

e (m

yr–1

)

–40

–60

–80

0

–20

–40

–60

–80

0

–20

–40

–60

–80

0

–20

Ave

rage

ret

reat

rat

e (m

yr–1

)

–40

–60

–80Marine terminating Land-terminating Ice-sheet glaciers Local glaciers

a

b

c

d

0

100

200

300

1940 1960 1980 2000 1940 1960 1980 2000 1940 1960 1980 2000 1940 1960 1980 20001930 1950 1970 1990 2010 1930 1950 1970 1990 2010 1930 1950 1970 1990 2010 1930 1950 1970 1990 2010

Year Year Year Year

Figure 5 | A closer look at marine- and land-terminating glaciers. Frontal changes are shown as rates during six observation periods from 1933 to 2010.Red circles show retreat and green circles advance (m yr−1). a, Land-terminating glaciers. b, Marine-terminating glaciers. c, Aggregated frontal retreatrates, with measuring uncertainty for each period. Error bars reflect the measuring uncertainty for each observational period. d, Changes inland-terminating margin elevation for the period 1933–2007. Filled circles are ice-sheet glaciers, open circles are local glaciers and ice caps.

recent studies investigating marine-terminating outlets from 19721

onwards21. Exploring SST data we find that a period of colder-than-2

usual temperatures reached the coast of southeast Greenland in the3

late 1960s (Supplementary Information). However, the influence of4

cold water on the terminus of marine-terminating outlets does not5

explain the entire advance as numerous land-terminating glaciers6

advance in the period 1943–1981 (Fig. 5). Hence, a shorter melt7

season38 and increased accumulation28 resulted in glacier growth8

after the ECW. Considering the temporal resolution of our obser- 9

vations, it is possible that advance may have occurred and not been 10

documented in this data set. Such is the case with the Mittivakkat 11

Glacier, which advanced between 1946 and 1956 (ref. 39), and mi- 12

nor advances that are not detected in this data set have also been re- 13

ported in 2006 in east and southeast Greenland17,19. These findings 14

emphasize the fact that marine-terminating glaciers may respond 15

quickly to bothwarmer and colder ocean temperatures9,15,40. 16

4 NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience

NATURE GEOSCIENCE DOI: 10.1038/NGEO1481 ARTICLESIce-marginal sensitivity to external forcing1

A comparison between retreat rates for local and outlet glaciers2

from the Greenland Ice Sheet shows that local glaciers show3

similar retreat rates during the two warming periods, whereas4

ice-sheet glacier retreat has accelerated during the latest warming5

(Fig. 5). Land-terminating glaciers, most of which are of local6

origin, did however retreat more rapidly after the ECW. As most7

land-terminating glaciers are situated in valleys, the ablation area8

becomes smaller as the terminus retreats to higher ground. From9

1933 to the present (2007) some glacier fronts have been elevated10

more than 250m (Fig. 5d). This is more pronounced for local11

glaciers and ice caps. The elevation change probably explains the12

differences observed in retreat rate between the ECW and the recent13

warming, as the initial conditions for the two retreat periods have14

changed through time.15

It has been argued that the east coast can be divided into two16

distinct regions with regards to ice–climate response at 69◦N,17

with the southern region being more dynamic as a response to18

the influence of warm Atlantic water17. Our results show that19

glaciers in the entire southeast region south of 67◦N have reacted20

synchronously to the warming events regardless of type, size21

and terminal environment, indicating that the retreat observed22

during the past decade5,6,14,17,19,20 is not confined to large marine-23

terminating glaciers from the ice sheet.24

The acceleration of glacial retreat observed during the past25

decade is expected to continue with rising temperatures20,21.26

Climate models project continued warming through the twenty-27

first century and beyond41, and records of ocean heat content28

indicate strong increases in temperature in the past few decades42.29

With Greenland temperatures deviating at present from the30

ongoing temperature rise in the Northern Hemisphere, it has31

been proposed that Greenland mean temperatures must warm32

1.0–1.5 ◦C above the 2007 level to be in synchronicity with33

the Northern Hemisphere31. This has been supported by the34

observed temperature rise in the subsequent four years (2008–35

2011) and reinforces the expectation that continued warming36

will occur in the coming years. However, the recent high rate37

of retreat may come to a slowdown when retreating marine-38

terminating glaciers reach their grounding line and become less39

sensitive to the influence of ocean temperature19,20, or through40

positive or negative feedback mechanisms relating to the cold EastQ3 41

Greenland Coastal Current9.42

Our analysis of historical aerial imagery from the past 80 years43

shows that the glaciers of southeast Greenland have responded44

vigorously on a decadal scale to both warming and cooling. Major45

retreat has occurred not only during the recent warming18–20, but46

also during and after the ECW, when air and SSTs were similar47

to the present or slightly lower. However, the recent retreat rate48

is higher for marine-terminating glaciers whereas it was higher49

for land-terminating glaciers in the early twentieth century. The50

asymmetric response between glacier types suggests that although51

the retreat ceased and shifted to advance in response to the mid-52

twentieth century cooling, the lower retreat rate of land-terminating53

glaciers in the recent warming period may be due to their smaller54

overall extent and elevated terminus altitude. These findings are an55

important contribution to the sparse knowledge of local glaciers56

and ice caps in Greenland, as only a few studies so far have carried57

out investigations on a regional scale43,44 and none on the southeast58

coast of Greenland. Although research shows a less pronounced59

retreat north of the study area7,8,17,44, the recent record-setting60

high temperatures have resulted in an unprecedented extent of61

glacial retreat in southeast Greenland, which should continue with62

a persevering temperature increase. Our results have implications63

for future estimations of sea-level rise as retreat rates for marine-64

terminating glaciers are likely to increase as temperature rises until65

glacier fronts reach the grounding line, or when cold ocean currents66

re-establish, whereas retreat rates for land-terminating glaciers are 67

not likely to rise in the same order of magnitude. 68

Methods 69

Glacier length data were obtained from terrestrial, aerial oblique and vertical 70

photos and satellite images. All images were rectified in digital form according 71

to 1981/1985-orthorectified mosaics generated from a digital elevation model 72

(DEM) created from the 1981 and 1985 aerial images using BAES SocetSet 5.5.0. Q4 73

Images from 1981 and 1985 were scanned on a photogrammetric precision scanner 74

from negatives at a resolution of 15 µm (2m ground resolution), images from 75

1931, 1932, 1933 and 1943 were retrieved from paper copies at a resolution of 76

600 dpi. The main source of historical imagery is the seventh Thule Expedition 77

(1932–1933), but material from the British Arctic Air Route Expedition (BAARE; 78

1931) has been used to fill gaps not covered by the former. Rectification of images 79

covering each glacier was carried out using tie-points surrounding the glacier 80

front and subsequently tie-points were identified in 1981/1985 orthomosaics for 81

coregistration. Tie-points were placed at a similar elevation to that of the glacier 82

front, to minimize distortion within the rectified image. The type of transformation 83

varied according to number of available tie-points. Uncertainties in digitization 84

were found empirically by using known stationary landforms within the rectified 85

image (for example, trim lines and tidewater marks on bedrock) and thereby 86

calculating the standard deviation using the 1981/1985 orthomosaics as ground 87

truth. Equivalent rectification process and uncertainty estimation was carried 88

out with all digital satellite images. After the digitization of glacier fronts, glacier 89

lengths were measured using a glacier length tool developed for ESRI’s ArcGIS 10.0 Q5 90

that divides each glacier front into points with a separation of 15 m. From a static 91

measuring point placed at an arbitrary position up-glacier, the mean distances to 92

points were calculated (see Supplementary Information). Where 1933 images were 93

available, 1931 and 1932 images were used and frontal changes were calculated 94

to the 1933 date by linear regression. For all calculations involving average values 95

of frontal change rates, Midgaard Glacier (SEGL011, 66◦26′ N, 36◦44′W) was 96

discarded from the calculations owing to its extremely high retreat resulting 97

in an obvious skewing of the averages. Elevation changes of land-terminating 98

fronts were found by extracting the elevation values of the digitized 1933 fronts 99

from the 1981/1985-DEM, whereas recent elevations were extracted from a 100

2007 SPOT Spirit DEM. Q6 101

Received 21 January 2012; accepted 20 April 2012; 102

published online XXMonth XXXX 103

References 104

1. Joughin, I., Abdalati, W. & Fahnestock, M. Large fluctuations in speed on 105

Greenland’s Jakobshavn Isbrae Glacier. Nature 432, 608–610 (2004). 106

2. Howat, I. M., Joughin, I., Tulaczyk, S. & Gogineni, S. Rapid retreat and 107

acceleration of Helheim Glacier, east Greenland. Geophys. Res. Lett. 32, 108

L22502 (2005). 109

3. Nick, F. M., Vieli, A., Howat, I. M. & Joughin, I. Large-scale changes in 110

Greenland outlet glacier dynamics triggered at the terminus. Nature Geosci. 2, 111

110–114 (2009). 112

4. Vieli, A. & Nick, F. M. Understanding and modelling rapid dynamic changes 113

of tidewater outlet glaciers: Issues and implications. Surv. Geophys. 32, 114

437–458 (2011). 115

5. Howat, I. M., Joughin, I. & Scambos, T. A. Rapid changes in ice discharge from 116

Greenland outlet glaciers. Science 315, 1559–1561 (2007). 117

6. Luckman, A., Murray, T., de Lange, R. & Hanna, E. Rapid and synchronous 118

ice-dynamic changes in east Greenland. Geophys. Res. Lett. 33, L03503 (2006). 119

7. Rignot, E. & Kanagaratnam, P. Changes in the velocity structure of the 120

Greenland Ice Sheet. Science 311, 986–990 (2006). 121

8. Joughin, I., Smith, B. E., Howat, I. M., Scambos, T. &Moon, T. Greenland flow 122

variability from ice-sheet-wide velocity mapping. J. Glaciol. 56, 415–430 (2010). 123

9. Murray, T. et al. Ocean regulation hypothesis for glacier dynamics in southeast 124

Greenland and implications for ice sheet mass changes. J. Geophys. Res. 115, 125

F03026 (2010). 126

10. Velicogna, I. &Wahr, J. Acceleration of Greenland ice mass loss in spring 2004. 127

Nature 443, 329–331 (2006). 128

11. Chen, J. L., Wilson, C. R. & Tapley, B. D. Satellite gravity measurements 129

confirm accelerated melting of Greenland Ice Sheet. Science 313, 130

1958–1960 (2006). 131

12. Luthcke, S. B. et al. Recent Greenland ice mass loss by drainage system from 132

satellite gravity observations. Science 314, 1286–1289 (2006). 133

13. Rignot, E., Box, J. E., Burgess, E. & Hanna, E. Mass balance of the Greenland 134

Ice Sheet from 1958 to 2007. Geophys. Res. Lett. 35, L20502 (2008). 135

14. Khan, S. A. et al. Elastic uplift in southeast Greenland due to rapid ice mass loss. 136

Geophys. Res. Lett. 34, L21701 (2007). 137

15. Straneo, F. et al. Rapid circulation of warm subtropical waters in a major glacial 138

fjord in east Greenland. Nature Geosci. 3, 182–186 (2010). 139

16. Rignot, E., Velicogna, I., van den Broeke, M. R., Monaghan, A. & Lenaerts, J. 140

Acceleration of the contribution of the Greenland and Antarctic ice sheets to 141

sea level rise. Geophys. Res. Lett. 38, L05503 (2011). 142

NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience 5

ARTICLES NATURE GEOSCIENCE DOI: 10.1038/NGEO1481

17. Seale, A., Christoffersen, P., Mugford, R. I. & O’Leary, M. Ocean forcing of1

the Greenland Ice Sheet: Calving fronts and patterns of retreat identified by2

automatic satellite monitoring of eastern outlet glaciers. J. Geophys. Res. 116,3

F03013 (2011).4

18. Mernild, S. H. et al. Increasing mass loss from Greenland’s Mittivakkat5

Gletscher. Cryosphere 5, 341–348 (2011).6

19. Howat, I. M., Joughin, I., Fahnestock, M., Smith, B. E. & Scambos, T. A.7

Synchronous retreat and acceleration of southeast Greenland outlet8

glaciers 2000–06: Ice dynamics and coupling to climate. J. Glaciol. 54,9

646–660 (2008).10

20. Moon, T. & Joughin, I. Changes in ice front position on Greenland’s outlet11

glaciers from 1992 to 2007. J. Geophys. Res. 113, F02022 (2008).12

21. Howat, I. M. & Eddy, A. Multi-decadal retreat of Greenland’ s13

marine-terminating glaciers. J. Glaciol. 57, 1–8 (2011).14

22. Dwyer, J. Mapping tide-water glacier dynamics in east Greenland using Landsat15

data. J. Glaciol. 41, 584–595 (1995).16

23. Weidick, A. Satellite Image Atlas of Glaciers of the World GREENLAND USGS17

Professional Paper 1386-C (United States Government Printing Office, 1995).18

24. Weidick, A. Historical fluctuations of calving glaciers in south and west19

Greenland. Rapp. Geol. Surv. Greenland 161, 73–79 (1994).20

25. Gabel-Jørgensen, Dr. Knud Rasmussen’s contribution to the exploration of21

the south-east coast of Greenland, 1931–1933. Geogr. J. 86, 32–49 (1935).Q7 22

26. Gabel-Jørgensen, C. A. A. Report on the Expedition - 6. og 7. Thule-Expedition23

til Sydøstgrønland 1931–1933.Meddelelser om Grønland 106 (1940).Q8Q9

24

27. Howat, I. M., Box, J. E., Ahn, Y., Herrington, A. & Mcfadden, E. M. Seasonal25

variability in the dynamics of marine-terminating outlet glaciers in Greenland.26

J. Glaciol. 56, 601–613 (2010).27

28. Wake, L. M. et al. Surface mass-balance changes of the Greenland Ice Sheet28

since 1866. Ann. Glaciol. 50, 178–184 (2009).29

29. Box, J. E. et al. Greenland [in ‘‘State of the Climate in 2010’’]. Bull. Am.30

Meteorol. Soc. 92, 161–171 (2011).31

30. Chylek, P., Dubey, M. K. & Lesins, G. Greenland warming of 1920–1930 and32

1995–2005. Geophys. Res. Lett. 33, L11707 (2006).33

31. Box, J. E., Yang, L., Bromwich, D. H. & Bai, L. Greenland Ice Sheet surface air34

temperature variability: 1840–2007. J. Clim. 22, 4029–4049 (2009).35

32. Wood, K. R. & Overland, J. E. Early 20th century Arctic warming in retrospect.36

Int. J. Climatol. 30, 1269–1279 (2009).37

33. Wild, M. et al. From dimming to brightening: Decadal changes in solar38

radiation at Earth’s surface. Science 308, 847–850 (2005).39

34. Andresen, C. S. et al. Rapid response of Helheim Glacier in Greenland to40

climate variability over the past century. Nature Geosci. 5, 37–41 (2012).41

35. Amundson, J. M. et al. Ice mélange dynamics and implications for terminus42

stability, Jakobshavn Isbræ, Greenland. J. Geophys. Res. 115, F01005 (2010).43

36. Weidick, A. Present-day expansion of the southern part of the inland ice.44

Rapp. Geol. Surv. Greenland 152 (1991).45

37. Meier, M. F. & Post, A. Fast tidewater glaciers. J. Geophys. Res. 92, 46

9051–9058 (1987). 47

38. Box, J. E. Survey of Greenland instrumental temperature records: 1873–2001. 48

Int. J. Climatol. 22, 1829–1847 (2002). 49

39. Larsen, H. V. Runoff studies from the Mitdluagkat Gletcher in SE-Greenland 50

during the late summer 1958. Danish J. Geogr. 58, 54–65 (1959). 51

40. Straneo, F. et al. Impact of fjord dynamics and glacial runoff on the circulation 52

near Helheim Glacier. Nature Geosci. 4, 1–6 (2011). 53

41. IPCC Climate Change 2007: The Physical Science Basis (Cambridge Univ. Press, 54

2007). 55

42. Levitus, S. et al. Global ocean heat content 1955–2008 in light of recently 56

revealed instrumentation problems. Geophys. Res. Lett. 36, L07608 (2009). 57

43. Yde, J. C. & Knudsen, N. T. 20th-century glacier fluctuations on Disko Island 58

(Qeqertarsuaq), Greenland. Ann. Glaciol. 46, 209–214 (2007). 59

44. Jiskoot, H., Juhlin, D., Pierre, H. S. T. & Citterio, M. Tidewater glacier 60

fluctuations in central east Greenland coastal and fjord regions (1980s–2005). 61

Ann. Glaciol. 53, 35–44 (2012). 62

Acknowledgements 63

This study could not have been possible without the aid of The National Survey and 64

Cadastre (KMS, Denmark) who gave access to the historical photographs from the 65

seventh Thule Expedition. We are also grateful to the Scott Polar Research Institute (UK), 66

and the Arctic Institute (Copenhagen, Denmark) who also supplied access to historical 67

images. A. Pedersen (MapWorks, Denmark) wrote the script for the glacier length tool. 68

SPOT Spirit DEM was obtained from B. Csatho and S. Nagarajan, Geology Department, 69

University at Buffalo, USA. The ASTER GDEM data were obtained through the online Q10 70

data pool at the National Aeronautics and Space Administration Land Processes 71

Distributed Active Archive Center, United States Geological Survey /Earth Resources 72

Observation and Science Center, Sioux Falls, South Dakota. We thank K. L. Bird and E. 73

Willerslev who edited the manuscript. This work is a part of the RinkProject financially 74

supported by the Danish Research Council (FNU) no. 272-08-0415 and the Commission 75

for Scientific Research in Greenland (KVUG). Q11 76

Author contributions 77

A.A.B. and K.H.K. designed and conducted the study, N.J.K. conducted photogrammetry, 78

K.K.K. undertook the geographic information system analysis, J.E.B., C.S.A. and S.A.K. 79

carried out climate and SST analysis. All authors contributed to the discussion and 80

writing of the manuscript. 81

Additional information 82

The authors declare no competing financial interests. Supplementary information 83

accompanies this paper on www.nature.com/naturegeoscience. Reprints and permissions 84

information is available online at www.nature.com/reprints. Correspondence and 85

requests for materials should be addressed to A.A.B. 86

6 NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience

Page 1

Query 1:Please note that the first paragraph has been

edited according to style.Query 2:

Please provide post codes for all addresses.

Page 5

Query 3:Text amended to ‘positive or negative’ here.

OK?Query 4:

Please define ‘BAES’ here.Query 5:

Please define ‘ESRI’ here.Query 6:

Please define ‘SPOT’ here.

Page 6

Query 7:Please provide initials for the author name in ref.

25Query 8:

Please provide page range or article id of refs 26,36Query 9:

Please check ref. 26 ok as given.Query 10:

Please define ‘ASTER GDEM’ here.Query 11:

Thanks to anonymous reviewers is against style sohas been removed from acknowledgement section.

General Queries

Query 12:Please note that reference numbers are formatted

according to style in the text, so that any referencenumbers following a symbol or acronym are givenas ‘ref. XX’ on the line, whereas all other referencenumbers are given as superscripts.

Query 13:Please define ‘METI’ and ‘ASTER GDEM’ in Fig.

2 caption.

Query 14:

Please define ‘*’ in Fig. 2.