Antarctic supraglacial lakes and ice-shelf collapse

What are supraglacial lakes?

Supraglacial lakes form on the surface of glaciers. Meltwater accumulates in depressions or hollows on the ice surface. We think of the East Antarctic Ice Sheet as being very cold and dry, with temperatures well below 0°C, but in some places, surface melting does occur in the summer melt season. In Antarctica, this occurs primarily from November to February1. Supraglacial lakes are critically important for the stability of ice sheets, glaciers and ice shelves, and can help drive ice-shelf collapse.

Settings of glacial lakes. Lakes can form in-front of glaciers and ice sheets (proglacial lakes), on top of the ice (supraglacial lakes), within the ice (englacial) and beneath glaciers and ice sheets (subglacial lakes).

Meltwater accumulates on the glacier or ice-shelf surface when firn (snow that has not yet been compressed into ice) becomes sufficiently saturated and can no longer absorb meltwater2,3.

Supraglacial lakes commonly refreeze at the end of the melt season, but some have been observed to drain into the ice6,14,16.

meltwater lakes
Supraglacial meltwater lakes on McMurdo Ice Shelf. Credit: Neil Glasser.

Why are supraglacial lakes important in Antarctica?

Supraglacial lakes are important because they can influence ice flow and mass balance in three ways. Firstly, because the ponded meltwater is darker than the surrounding white-grey ice, lakes have a lower albedo. More energy from the sun’s rays is absorbed rather than reflected,meaning the lakes can locally increase surface melting in a positive feedback effect4,5.

Secondly, supraglacial lakes can cause short-term ice flow accelerations when they drain through the ice sheet6. The meltwater can reach the bed of the ice sheet, and encourage basal sliding and thus faster ice flow.

Thirdly, supraglacial lakes have been linked to ice shelf collapse on the Antarctic Peninsula7,8. Ice shelves are the floating extensions of land ice; they receive ice from ice-flow from inland tributary glaciers and lose mass by calving icebergs. Ice shelf removal increases the flow of inland ice into the ocean and its contribution to sea-level rise9,10.

Schematic cartoon of a glacier flowing into an ice shelf, showing the grounding line and calving at the ice cliff at the edge of the ice shelf.

It is therefore important to understand where supraglacial lakes form and how they influence the behaviour and stability of the ice sheet. We need these data so that we can identify which parts of Antarctica may be more sensitive to climate warming and how the Antarctic Ice Sheet may respond under a future warming climate11.

Landsat-8 satellite image of supraglacial lakes on the Amery Ice Shelf, East Antarctica, taken on 17th January 2019. Image size is 185 km x 180 km. Lakes (bright blue), exposed rock outcrops called nunataks (black), snow (white) and bare ice (pale blue) are visible. Image credit: U.S. Geological Survey.

Where are supraglacial lakes found in Antarctica?

Satellite remote sensing since the 1970s has allowed researchers to map supraglacial lakes over the whole Antarctic continent12,13. This has shown that thousands of supraglacial lakes are widespread around the margins of Antarctica during the summer melt season, including the periphery of the East Antarctic Ice Sheet12-16. A recent assessment found lakes up to 500 km inland, and at up to >1500 m above sea level13.

This is not a new phenomenon; satellite imagery records supraglacial lakes persistently forming annually for decades on Antarctic ice shelves. For example, lakes have been recorded on the Amery Ice Shelf in East Antarctica since at least 1974, where the largest recorded supraglacial lake (~80 km long) lake has been observed12,16-20.

Locations around Antarctic where supraglacial lakes have been observed together with examples: (a) Larsen C Ice Shelf, (b) George VI Ice Shelf, (c) Riiser-Larsen Ice Shelf, (d) Langhovde Glacier, (e) Ross Archipelago, (f) McMurdo Ice Shelf, (g) Sørsdal Glacier, (h) Mawson Glacier. Green shaded regions in the central map represent the number of published studies reporting supraglacial lakes in that location. Note that the number of studies reporting lakes in a given location does necessarily correspond to the number of lakes forming, or how long lakes have been present in this location. Lakes mapped in January 2017 in a recent East Antarctic assessment by Stokes et al. (2019) are shown in purple. Images reproduced from: Martin Truffer, University of Alaska Fairbanks (a), Frithjof C. Küpper, University of Aberdeen (b), Matti Leppäranta, University of Helsinki (c), Takehiro Fukuda, Hokkaido University (d), NASA Operation IceBridge (e), Chris Larsen, NASA Operation IceBridge (f), Sarah Thompson University of Tasmania (g), and Richard Stanaway, Australian National University (h). Credit: Arthur et al. (2020).

This interactive Google Map below allows you to explore supraglacial lakes on the floating ice tongue of Sørsdal Glacier, Princess Elizabeth Land, East Antarctica. You can clearly see meltwater filling depressions and crevasses on the ice surface and lakes being fed by meltwater from upstream.

What controls supraglacial lake distributions in Antarctica?

Most supraglacial lakes in Antarctica form on ice shelves, where the long, flat profile of the floating ice shelf and low surface elevations encourage meltwater ponding13. Supraglacial lakes typically cluster a few kilometres down-ice from the grounding line12,13,21,31.

The figure below shows how supraglacial lake distributions around the continent are controlled by interactions between local and regional wind patterns, ice surface topography and albedo21.

Distribution of supraglacial lakes in relation to wind scour zones, exposed rock, blue ice, near-surface wind speed, surface melt rate and ice shelf vulnerability to surface-melt-induced collapse.
Distribution of supraglacial lakes in relation to wind scour zones, exposed rock, blue ice, near-surface wind speed, surface melt rate and ice shelf vulnerability to surface-melt-induced collapse. Surface melt fluxes are derived from QuickScat scatterometer (2000–2009 average, Trusel et al., 2013) and include meltwater that could refreeze in the snow/firn. Ice shelf vulnerability is derived from data from the QuikSCAT satellite and represents the relative concentration of refrozen meltwater in the firn (Alley et al., 2018). The grounding line and coastline are represented by dotted and solid black lines respectively. Arrows represent near-surface wind speed vectors (Lenaerts et al., 2017; Luckman et al., 2014). Supraglacial lake distributions are reproduced from Stokes et al. (2019). On Larsen C Ice shelf (A), supraglacial lake formation is restricted to ice shelf inlets (blue stars), driven by föhn-enhanced melting. On Roi Baudouin Ice Shelf (B), lakes clustered at the grounding zone are associated with katabatic wind-enhanced melting. The co-occurrence of lakes with rock outcrops and wind-scoured blue ice is prominent on other East Antarctic ice shelves, such as Shackleton (C) and Amery (D). Credit: Arthur et al. (2020).

Antarctic Peninsula

On the Antarctic Peninsula, such as the Larsen C Ice Shelf, strong down-slope winds called föhn winds cause intense melting (>400 mm w.e. yr –1) and localised ponding, even in Antarctic winter22-26.

Fohn winds are important for warming ice shelves on the eastern Antarctic Peninsula.
Fohn winds are important for warming ice shelves on the eastern Antarctic Peninsula.
Overview map of the Antarctic Peninsula.
Overview map of the Antarctic Peninsula. Note the large floating ice shelves (labelled in blue). From Davies et al., 2012 (Quaternary Science Reviews).

East Antarctica

In coastal East Antarctica, supraglacial lakes are clustered near ice-shelf grounding lines, but are often absent further downstream towards calving fronts because higher snowfall here means surface meltwater can be absorbed into the porous snowpack before it can pond on the ice surface15.

Persistent katabatic winds induce vertical air mixing andcan scour snow surfaces to expose lower-albedo blue ice, which intensifies melting and ponding15. Lakes often form close to exposed rock, which has a low albedo and locally increases surface melting12.

Supraglacial lakes are sensitive to katabatic winds, which can cause intense short-lived snowmelt events31.

Katabatic winds (meaning “descending winds”) are ‘drainage winds’, carrying high-density cold air from a high elevation down a slope under the force of gravity. The near-surface wind field over Antarctica is dominated by katabatic winds. From: Wikipedia.

Links to ice-shelf collapse

Supraglacial lakes can flex and fracture ice shelves when they grow and drain7,8,27. A good example is the near-synchronous drainage of over 2750 lakes (up to 6.8 metres deep) on Larsen B Ice Shelf in the days before it collapsed in February-March 200227.

Lakes are thought to have drained through a process called hydrofracturing, where meltwater triggers fracturing by filling and expanding vertical crevasses7, 27. This is thought to have triggered a ‘chain reaction’ that caused widespread fracturing and lake drainage8.

Supraglacial lakes have been observed on other ice shelves before they collapsed or partially disintegrated, including Larsen A and Prince Gustav in 19957,28. They could therefore be important indicators of future ice shelf instability and collapse29,30.

Examples of supraglacial lakes on Antarctic Peninsula ice shelves prior to their disintegration or major calving
Examples of supraglacial lakes on Antarctic Peninsula ice shelves prior to their disintegration or major calving. Prince Gustav Ice Shelf (A), which disintegrated in 1995; Larsen B ice Shelf (B and C), which disintegrated in 2002; Larsen C Ice Shelf (D); George VI Ice Shelf (E); and Wilkins Ice Shelf (F), which underwent major calving in 1998 and 2008. Imagery credits: (A) Rott et al. (1996), (B) Glasser and Scambos (2008), (C) National Snow and Ice Data Center, (D) Luckman et al. (2014), (E) Labarbara and MacAyeal (2011), (F) Scambos et al. (2009). Figure credit: Arthur et al. (2020).

The likelihood of lakes triggering ice shelf collapse in other locations is lower on thicker ice shelves confined within narrow embayments, such as the George VI Ice Shelf (between the mainland Antarctic Peninsula and Alexander Island), or those that are ‘pinned’ (locally in contact with the bedrock). This is because ice-shelf flow here is stabilised and so the ice shelf can support large numbers of lakes on its surface.

Glaciological structures on George VI Ice Shelt, Antarctic Peninsula, illustrating the dominance of meltwater channels, longitudinal structures and rifts towards the south ice front. From Holt et al., 2013.

Observing Antarctic supraglacial lakes from space

The inaccessibility of lakes in Antarctica makes obtaining field measurements difficult, but satellite remote sensing can be used to map supraglacial lakes at sub-metre resolution on weekly timescales, and to measure seasonal changes in lake depth and volume (16,31).

Improvements in satellite remote sensing over time

The figure below shows how the ability to resolve individual supraglacial lakes has vastly improved since coarser sensors like AVHRR and MODIS, which were unable to distinguish these features.

Higher resolution (>10m) satellites such as Sentinel-2 and Worldview-3 can resolve lake bathymetry and surface features such as partial lake ice coverage.

Examples of the evolution of satellite image sensor resolution and detection of supraglacial lakes on Beaver Lake, adjacent to Amery Ice Shelf, East Antarctica. Panels moving from top left to bottom right represent satellite sensors in order of increasing spatial resolution. Smaller insets show the same supraglacial lake to demonstrate the improvement in detail. Imagery: Scambos et al. (1996); United States Geological Survey; Google, Maxar Technologies. Figure credit: Arthur et al. (2020).

Recent assessments have used imagery from multiple optical satellite sensors (such as Landsat-8 and Sentinel-2) to maximise the amount of cloud-free imagery for mapping supraglacial lake distributions at the continent scale12,13.

Measuring changes in supraglacial lakes

Changes in the extent, area and volume of individual supraglacial lakes can be tracked through multiple melt seasons using satellites with a short revisit time (time elapsed between observations of the same point on earth)16,31.

For example, the 16-day revisit of Landsat-8 and the 5-day revisit of the Sentinel-2A/B satellite constellation have been used to calculate seasonal fluctuations in supraglacial lake area, depth and volume through multiple melt seasons16,31. This is important for estimating the volume of meltwater stored in lakes on the surface of the Antarctic Ice Sheet and the potential for hydrofracturing (where the lake helps to break up an ice shelf).

Satellite imagery recently recorded an extreme, potentially unprecedented, melt event in January 2020 around Antarctica during unprecedented warm air temperatures, which caused widespread surface melting and lake growth.

This post draws from the review of recent progress made in understanding Antarctic supraglacial lakes using satellite remote sensing by Arthur et al. (2020) (ref. 35).

About the Author

This article was contributed by Jennifer Arthur from Durham University. Jennifer is a PhD candidate specialising in Antarctic surface hydrology and ice shelf dynamics. She’s interested in the distribution and seasonal evolution of supraglacial lakes and the potential influence of surface meltwater on outlet glacier and ice shelf stability. Jennifer’s PhD research uses satellite remote sensing to investigate this on the East Antarctic Ice Sheet. You can follow Jennifer on Twitter @AntarcticJenny and she is part of the @EGU_CR Cryosphere blog team.

Jennifer Arthur


1. Echelmeyer, K, Clarke, TS, Harrison, WD (1991) Surficial glaciology of Jakobshavns Isbræ, West Greenland: Part I. Surface morphology. Journal of Glaciology 37(127): 368–382.

2. Alley, KE, Scambos, TA, Anderson, RS, et al. (2018) Quantifying vulnerability of Antarctic ice shelves to hydrofracture using microwave scattering properties. Remote Sensing of Environment 210: 297–306.

3. Hubbard, B, Luckman, A, Ashmore, D, et al. (2016) Massive subsurface ice formed by refreezing of ice-shelf melt ponds. Nature Communications 7(11897): 1–6.

4. Lüthje, M, Feltham, DL, Taylor, PD, et al. (2006) Modeling the summertime evolution of sea-ice melt ponds. Journal of Geophysical Research: Oceans 111: C02001.

5. Tedesco, M, Luthje, M, Steffen, K, et al. (2012) Measurement and modeling of ablation of the bottom of supraglacial lakes in western Greenland. Geophysical Research Letters 39(2): 1–5.

6. Tuckett, PA, Ely, JC, Sole, AJ, et al. (2019) Rapid accelerations of Antarctic Peninsula outlet glaciers driven by surface melt. Nature Communications 10(4311): 1–8.

7. Scambos, T, Hulbe, CM, Fahnestock, M (2003) Climate-induced ice shelf disintegration in the Antarctic Peninsula. Paleobiology and Paleoenvironments of Eoscene Rocks Antarctic Research Series 76: 335–347.

8. Banwell, AF, MacAyeal, DR, Sergienko, OV (2013) Breakup of the Larsen B Ice Shelf triggered by chain reaction drainage of supraglacial lakes. Geophysical Research Letters 40(22): 5872–5876.

9. Scambos, TA, Bohlander, JA, Shuman, CA, et al. (2004) Glacier acceleration and thinning after ice shelf collapse in the Larsen B embayment, Antarctica. Geophysical Research Letters 31(18): 2001–2004.

10. Rignot, E, Casassa, G, Gogineni, P, et al. (2004) Accelerated ice discharge from the Antarctic Peninsula following the collapse of Larsen B ice shelf. Geophysical Research Letters 31, L18401.

11. Bell, RE, Banwell, AF, Trusel, LD, et al. (2018) Antarctic surface hydrology and impacts on ice-sheet mass balance. Nature Climate Change 8: 1044–1052.

12. Kingslake, J, Ely, JC, Das, I, et al. (2017) Widespread movement of meltwater onto and across Antarctic ice shelves. Nature 544(7650): 349–352.

13. Stokes, CR, Sanderson, JE, Miles, BWJ, et al. (2019) Widespread development of supraglacial lakes around the margin of the East Antarctic Ice Sheet. Scientific Reports 9(1): 13823.

14. Langley, ES, Leeson, AA, Stokes, CR, et al. (2016) Seasonal evolution of supraglacial lakes on an East Antarctic outlet glacier. Geophysical Research Letters 43(16): 8563–8571.

15. Lenaerts, JTM, Lhermitte, S, Drews, R, et al. (2017) Meltwater produced by wind-albedo interaction stored in an East Antarctic ice shelf. Nature Climate Change 7: 58–63.

16. Moussavi, MS, Pope, A, Halberstadt, ARW, et al. (2020) Antarctic supraglacial lake detection using Landsat 8 and Sentinel-2 imagery: Towards continental generation of lake volumes. Remote Sensing 12(1): 1–19.

17. Hambrey, MJ, Dowdeswell, JA (1994) Flow regime of the Lambert Glacier-Amery Ice Shelf system, Antarctica: Structural evidence from Landsat imagery. Annals of Glaciology 20: 401–406.

18. Mellor, M, McKinnon, G (1960) The Amery Ice Shelf and its hinterland. Polar Record 10(64): 30–34.

19. Phillips, HA (1998) Surface meltstreams on the Amery Ice Shelf, East Antarctica. Annals of Glaciology 27: 177–181.

20. Swithinbank, C (1988) Satellite image atlas of glaciers of the world: Antarctica. United States Geological Survey Professional Paper 1386-B 36(122): 122–124.

21. Arthur et al. (2020) Recent understanding of Antarctic supraglacial lakes using satellite remote sensing. Progress in Physical Geography: Earth and Environment.

22. Datta, RT, Tedesco, M, Fettweis, X, et al. (2019) The effect of foehn-induced surface melt on firn evolution over the Northeast Antarctic Peninsula. Geophysical Research Letters 46: 1–10.

23. Luckman, A, Elvidge, A, Jansen, D, et al. (2014) Surface melt and ponding on Larsen C Ice Shelf and the impact of föhn winds. Antarctic Science 26(6): 625–635.

24. Kuipers Munneke, P, Luckman, AJ, Bevan, SL, et al. (2018) Intense winter surface melt on an Antarctic ice shelf. Geophysical Research Letters 45(15): 7615–7623.

25. Trusel, LD, Frey, KE, Das, SB, et al. (2013) Satellite-based estimates of Antarctic surface meltwater fluxes. Geophysical Research Letters 40(23): 6148–6153.

26. Wiesenekker, JM, Kuipers Munneke, P, van den Broeke, MR, et al. (2018) A multidecadal analysis of Föhn winds over Larsen C ice shelf from a combination of observations and modeling. Atmosphere 9(172): 1–13.

27. Glasser, NF, Scambos, TA (2008) A structural glaciological analysis of the 2002 Larsen B ice-shelf collapse. Journal of Glaciology 54(184): 3–16.

28. Rott, H, Skarvca, P, Nagler, T (1996) Rapid collapse of Northern Larsen Ice Shelf, Antarctica. Science 271(788): 788–792.

29. Bell, RE, Banwell, AF, Trusel, LD, et al. (2018) Antarctic surface hydrology and impacts on ice-sheet mass balance. Nature Climate Change 8: 1044–1052.

30. Robel, A, Banwell, AF (2019) A speed limit on ice shelf collapse through hydrofracture. Geophysical Research Letters 46(21): 12092–12100.

31. Arthur et al. (in review) Distribution and seasonal evolution of supraglacial lakes on Shackleton Ice Shelf, East Antarctica. The Cryosphere Discussions,

32. Labarbera, CH, MacAyeal, DR (2011) Traveling supraglacial lakes on George VI Ice Shelf, Antarctica. Geophysical Research Letters 38(24): L24501.

33. Scambos, T, Fricker, HA, Liu, C, et al. (2009) Ice shelf disintegration by plate bending and hydro-fracture: Satellite observations and model results of the 2008 Wilkins ice shelf break-ups. Earth and Planetary Science Letters 280(1–4): 51–60.

34. Scambos, T, Bohlander, J, Raup, B (1996) Images of Antarctic Ice Shelves [Amery Ice Shelf]. Boulder, Colorado USA: National Snow and Ice Data Center. Available at:

35. Arthur, J. F., Stokes, C., Jamieson, S. S. R., Carr, J. R., & Leeson, A. A. (2020). Recent understanding of Antarctic supraglacial lakes using satellite remote sensing. Progress in Physical Geography: Earth and Environment, 0309133320916114.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.