Ponds on ice: a cause of ice shelf break-up?

Melt ponds, which are exactly what the name implies – ponds of melted water formed on ice – have been observed in recent years on the Antarctic Peninsula, the most northern part of Antarctica. Previously this area had been too cold for melting to occur but, due to a 2.5 degrees Centigrade warming trend on the Peninsula over the last 50 years (several times the global average), this has begun to change.[1]

Melt ponds on Arctic sea ice such as those initially studied by CPOM researchers, copyright NASA Godard Space Flight Center. Copyright NASA Goddard Space Flight Center, obtained from Creative Commons licensed content on Flickr.

The Larsen B ice shelf on the Antarctic Peninsula hit the headlines in 2002 when it spectacularly collapsed and an area of ice larger than Rhode Island was lost in a matter of weeks.[2] This collapse occurred during record warm air temperatures when both the length and extent of the melt season reached a new high. Areas of the ice shelf with melt ponds collapsed whereas adjacent areas with few or no melt ponds did not. This suggests that melt ponds may be playing a role in triggering ice shelf collapse. However, there are many other processes – such as rising ocean temperatures – which may also play a role and so a greater understanding of ice-shelves and the mechanisms affecting them is essential in predicting their future.

A common misconception is that when ice shelves collapse they contribute to sea level rise. This is not strictly true; ice shelves are already floating on the water and in the same way that when ice cubes in your glass of water melt the water level in the glass doesn’t go up, ice shelf break-up (and subsequent melting) won’t add to global sea level. By Archimedes’ Principle (that’s the one from the ‘jumping out of the bath shouting “Eureka”’ story) the floating ice has already displaced a volume of water that is virtually the same as the amount that would be added to the sea if the ice shelf were to melt. Technically, a very small amount of sea level rise would occur on melting as ice shelves are made of fresh water and sea water is salty (and therefore more dense than fresh water) so slightly less water is displaced than the total melted amount would add but the effect is so small it is negligible: the equivalent of less than 4cm of sea level rise if all floating ice (ice shelves and sea ice) were to melt.

However, the fact that ice shelf break-up doesn’t contribute to sea level directly doesn’t mean that it doesn’t have an effect on it. When Larsen B broke up it was noted that the glaciers which fed into the ice sheet accelerated, possibly because there was no longer any force from the ice sheet impeding their flow. The ice coming from these glaciers is coming from the land and going into the sea so does contribute to sea level rise. This acceleration can continue for several years after the ice shelf has collapsed,[3] so the effect of a large collapse can actually be quite significant. On top of this, there is also the loss of habitat and the effect that large amounts of cold fresh water from the melted ice shelves have on ocean circulation and chemical make-up to be considered.

So what is it about these ponds makes them so key in understanding what will happen to the ice shelves? Firstly, it’s important to note that ‘pond’ is quite a deceptive term – on the Antarctic Peninsula we’re talking about something that can cover several square kilometers as oppose to something that would be found in your back garden, so these ponds are actually a fairly significant feature of the ice shelf.

One of the crucial things that melt ponds affect is albedo. Albedo is the proportion of incoming radiation that a surface reflects back. Ice, being white, has a high albedo whereas a melt pond is darker and has a lower albedo: it will reflect less energy and thus absorb more energy. Absorbing more energy means more heat is absorbed, more ice melts, ponds grow, albedo is further decreased, more energy is absorbed and so on in a positive feedback loop that results in greater and greater amounts of ice melting.

Meltwater is also thought to influence ice shelf break up when it fills crevasses in the ice. The forces that act on crevasses that open in the ice sheet tend to close them but meltwater filling existing surface crevasses provides an outward pressure that can allow them to remain open. Additionally, meltwater can cause the crevasses to penetrate further through the ice shelf, possibly far enough for them to reach the ice shelf’s base and impact on its structural integrity.[1]

Researchers at the Centre for Polar Observation and Modelling (CPOM) [4] at the University of Reading are looking at improving climate models by incorporating a better understanding of the processes affecting melt pond formation into the models. Initially, work was done on modelling melt ponds on Arctic sea ice [5]; one of the greatest uncertainties in predicting future global temperature changes is sea ice level fluctuations and this is something the level of melt ponds will play a role in. Now they have successfully incorporated a physically based melt pond scheme into the CICE sea ice model in order to simulate the stages of melt pond development.[6]

Currently a physical model of Arctic sea ice melt ponds is being adapted to look at the Antarctic, focussing on the Larsen C ice shelf as melt ponds have appeared there and it is thought to be a potential candidate for future collapse. The ponds found on Antarctica are much larger and over a greater scale than the Arctic sea ice ones and, unlike sea ice, the ice-shelves do not contain salt, adding an extra challenge in modifying previous work. If successful this model will help to provide new insights into the role of meltwater in the collapse of ice shelves and help to improve future versions of ice sheet models and climate predictions.


  1. T. A. Scambos, C. Hulbe, M. Fahnestock, and J. Bohlander, The link between climate warming and break-up of ice shelves in the Antarctic Peninsula, J. Glaciol. 46, pp. 516–530, 2000.
  2. National Snow & Ice Data Center- Larsen B Ice Shelf Collapses in Antarctica, 19 July 2013.
  3. T. Scambos, J. Bohlander, C. Shuman, and P. Skvarca, Glacier acceleration and thinning after ice shelf collapse in the Larsen B embayment Antarctica, Geophys. Res. Lett. 31, p. L18402, doi: 10.1029/2006GL020670, 2004.
  4. Centre for Polar Observation and Modelling- Current Research Programmes, 19 July 2013.
  5. F. Scott and D. Feltham, A model of the three-dimensional evolution of Arctic melt ponds on first-year and multiyear sea ice, J. Geophys. Res. 115, p. C02064, doi: 10.1029/2010JC006156, 2010.
  6. D. Flocco, D. Feltham, and A. Turner, Incorporation of a physically based melt pond scheme into the sea ice component of a climate model, J. Geophys. Res. 115, p. C08012, doi: 10.1029/2009JC005568, 2010.