Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


The plumbing of Greenland's ice

Observations of the water pressure in drilled boreholes and natural moulins on the Greenland Ice Sheet show how its underlying plumbing system controls ice motion during the course of the summer melt season. See Letter p.80

The interface between the base of an ice sheet or glacier and its underlying bed is of fundamental importance in controlling the speed at which the ice flows1,2,3. Of particular significance is how friction at the ice–bed interface is affected by the routing of meltwaters across the ice-sheet or glacier bed. On page 80 of this issue, Andrews and colleagues4 make a fundamental advance in our understanding of the hydrology underlying the Greenland Ice Sheet and of how the evolution of the subglacial drainage system controls ice motion during the course of the summer melt season — when water is generated by melting of snow and ice at the ice-sheet surface. The authors demonstrate that, during the latter part of the melt season, variations in water pressure in subglacial channels control daily patterns of ice motion, but a longer term slowdown in ice flow is dependent on decreasing water pressures in areas away from the channels.

Studies of mountain glacier systems1,3,5, and more recently in Greenland6, have investigated how the subglacial drainage system evolves over the course of the melt season and how this evolution affects ice motion. At the onset of the melt season, meltwaters flow over the glacier or ice-sheet surface before draining into the ice through crevasses or moulins, large natural vertical pipes, which can route this meltwater rapidly to the glacier or ice-sheet bed7 (Fig. 1). These initial meltwaters, on reaching the glacier bed, encounter a subglacial drainage system that is incapable of transporting the meltwater easily along the ice–bed interface. As a result, the water pressure in the subglacial drainage system increases, which decreases the friction at the ice–bed interface and the ice accelerates; in effect, the pressurized water is helping to partially float the overlying ice, enabling it to slide downhill more easily. However, as the volume of surface meltwaters routed to the glacier bed increases with warming summer temperatures, the water flowing across the bed starts to create, through the melting of ice at the ice–bed interface, subglacial channels that are more hydraulically efficient8,9. These channels enable the water to drain out of the glacier efficiently, thereby lowering the subglacial water pressure, so that the glacier slows down due to the decreasing flotation effect.

Figure 1: A moulin on the Greenland Ice Sheet.

James Balog/Aurora Photos

By monitoring water levels in large natural vertical pipes, which route water from melting surface snow and ice to the ice-sheet base, Andrews et al.4 acquired evidence demonstrating the presence of hydraulically efficient subglacial channels fed by the surface meltwater.

To understand more fully how hydrology affects the dynamics of the Greenland Ice Sheet, Andrews and colleagues used a suite of methods to investigate the link between water pressure at the ice-sheet base and ice motion. They drilled holes, using a hot-water 'drill', through about 600 metres of ice to the ice-sheet bed and inserted pressure sensors into these boreholes to measure the subglacial water pressure at their base, while simultaneously monitoring ice motion at the surface using Global Positioning System (GPS) data. They also lowered pressure sensors into moulins located between about 0.3 and 1.6 kilometres from the boreholes to measure fluctuations in water level, and thus pressure, in the moulins.

Andrews et al. observed systematic differences between the water-pressure measurements in the moulins and the boreholes, concluding that the moulins were connected to an efficient channelized component of the drainage system, whereas the boreholes monitored a hydraulically inefficient system unconnected to the channels. Water-pressure variations in the moulins (and thus channels) were positively correlated with daily patterns of ice motion, whereas borehole water pressures were anti-correlated. Water-pressure variations in the subglacial channels are therefore capable of affecting the friction at the ice–bed interface over a large enough area of the bed to enable the ice sheet to accelerate and decelerate on diurnal timescales. However, during the latter half of the melt season, ice motion gradually decreased, but mean moulin water levels (and thus channel pressure) remained relatively constant. By contrast, the water pressure in the boreholes decreased, implying that the longer-term seasonal slowdown was driven by changes in the unconnected subglacial drainage system away from the large subglacial channels. These findings imply that, to understand the dynamic behaviour of the ice sheet, it is essential to understand the processes going on in areas distal to the subglacial channels as well as in the channels themselves.

Andrews and colleagues' findings corroborate many of the earlier detailed borehole observations from mountain glacier systems1,3,10,11,12, indicating that the processes controlling the interaction between the hydrology and dynamics of ice-sheet and smaller valley-glacier systems are similar. The authors' observations also demonstrate how difficult it is to drill directly into areas affected by pressure variations in subglacial channels, because these areas cover only a small fraction of the glacier bed, in contrast to the surrounding distributed drainage system3,8.

There remain considerable uncertainties regarding the processes linking the hydrology and dynamics of the Greenland Ice Sheet. The distance to which efficient subglacial channels extend into the ice sheet during the melt season remains unclear; tests that use artificial tracers to track the speed at which water is routed from moulins to the ice-sheet margin indicate that efficient channels extend at least tens of kilometres into the ice sheet9, but will such channels extend further in a warming climate under enhanced surface melting? Furthermore, it is not clear whether the observations made are transferable to the rapidly moving tidewater glaciers — rivers of ice which move at approximately 1–10 km per year and are responsible for about half of the ice-mass loss from Greenland through the calving of large icebergs into the ocean13. The structure of these subglacial drainage systems, especially where the glaciers are flowing fastest as they near the ocean, is unknown, but is likely to be important in sustaining the high subglacial water pressures that enable the ice to slide so rapidly. Nevertheless, it is through further studies such as those by Andrews and colleagues that the complexities of the hydrological system lurking deep under the thick ice of the Greenland Ice Sheet will be unravelled.


  1. 1

    Iken, A. & Bindschadler, R. J. Glaciol. 32, 101–119 (1986).

    ADS  Article  Google Scholar 

  2. 2

    Alley, R. B. et al. J. Geophys. Res. 92, 8921–8929 (1987).

    ADS  Article  Google Scholar 

  3. 3

    Fountain, A. G. & Walder, J. S. J. Rev. Geophys. 36, 299–328 (1998).

    ADS  Article  Google Scholar 

  4. 4

    Andrews, L. C. et al. Nature 514, 80–83 (2014).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Mair, D. et al. J. Geophys. Res. Solid Earth 107, B8, 2175 (2002).

    ADS  Article  Google Scholar 

  6. 6

    Bartholomew, I. et al. Nature Geosci. 3, 408–411 (2010).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Catania, G. A. & Neumann, T. A. Geophys. Res. Lett. 37, L02501 (2010).

    ADS  Article  Google Scholar 

  8. 8

    Nienow, P. et al. Earth Surf. Process. 23, 825–843 (1998).

    Article  Google Scholar 

  9. 9

    Chandler, D. et al. Nature Geosci. 6, 195–198 (2013).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Gordon, S. et al. Hydrol. Process. 12, 105–133 (1998).

    ADS  Article  Google Scholar 

  11. 11

    Hubbard, B. et al. J. Glaciol. 41, 572–583 (1995).

    ADS  Article  Google Scholar 

  12. 12

    Murray, T. & Clarke, G. K. C. J. Geophys. Res. Solid Earth 100, 10231–10245 (1995).

    Article  Google Scholar 

  13. 13

    Joughin, I. et al. Science 338, 1172–1176 (2012).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Peter Nienow.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Nienow, P. The plumbing of Greenland's ice. Nature 514, 38–39 (2014).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing