News & Views | Published:

Climate change

Greenland's ice on the scales

  • A Correction to this article was published on 25 October 2006

Satellite measurements of changes in Earth's gravity field reveal ice loss from Greenland's ice sheet. Over the past four years, this melt has contributed to global sea-level rise at an accelerating rate.

The volume of the ice sheet that covers most of Greenland is so large that, were it to melt completely, sea levels across the world would rise by about 7 metres. Furthermore, an increase in its delivery of fresh water to the oceans could weaken or disrupt the 'thermohaline' circulation of oceanic salt water1, profoundly altering the climate of the Northern Hemisphere.

Such doomsday scenarios are well rehearsed, but — expressed in this way — not necessarily accurate. If cold areas such as the centre of Greenland warm up, it might actually snow more. That would, in turn, thicken the ice sheet and remove water from the global oceans. The very different densities of snow, ice and water mean that measuring the volume of the Greenland ice sheet does not provide the complete answer as to whether it is growing or shrinking. The ideal method is to measure how the mass of the ice sheet is changing with time.

In two complementary studies, Velicogna and Wahr (on page 329 of this issue)2 and Chen et al. (published online in Science)3 do just that. They show that the Greenland ice sheet lost between 192 million and 258 million tonnes of ice each year between April 2002 and April 2006 (equivalent to a volume of 212–284 km3). This rate of ice loss is equivalent to a rise in sea level of 0.5±0.1 mm yr−1, which is higher than many previous estimates. Both studies also show that the rate at which ice was being lost increased dramatically in the course of the study: the loss rate in the period 2004–06 was 2.5 times higher than that between 2002 and 2004 (ref. 2).

Both studies used data from the Gravity Recovery and Climate Experiment (GRACE), funded by NASA and the German Aerospace Center (DLR), which measures Earth's gravity field from space. GRACE consists of two satellites orbiting Earth. These satellites are separated by a distance of around 220 km that varies slightly as the satellites pass over anomalies in the gravity field. Since their launch in March 2002, the GRACE satellites have mapped the global gravity field every 30 days (Fig. 1). Over time, that field should show evidence of changes in the ice-sheet mass.

Figure 1: Amazing GRACE.

Anomaly map of Earth's gravity field, as measured by the NASA/DLR GRACE satellite.

But calculated ice-mass changes are only as accurate as the models used to remove other mass-change signals — those caused by tidal and non-tidal changes in the oceans, and by changes in the atmosphere and in Earth's mantle as it rebounds after the last ice age. At high latitudes, these models are not without error. The GRACE studies2,3 attempt to account for these other variations and their uncertainties, leaving a residual signal that results from the net loss of glacier ice into the oceans alone.

In particular, the high density of mantle rocks means that the gravity signal is very sensitive to even small errors in the model of rebound. But Velicogna and Wahr's estimate of uncertainty in the rebound rate2 would have to be increased by a factor of ten to change their conclusion of an overall loss of ice-sheet mass to an overall gain. And as this error would be constant over the timescale of the GRACE measurements, the change in the rate of mass loss is a highly stable result.

The GRACE data can also be used to indicate where the ice is being lost. Most of the loss is from south or southeast Greenland2,3, with Chen et al. reporting an additional area of loss in the northeast3. Airborne and satellite altimetry over the ice sheet shows further detail of the spatial pattern, albeit over different periods. The central portions of the ice sheet, with elevations above 1,500 m, are indeed thickening, fed by increased snowfall4. The margins, in contrast, are thinning5. The outlet glaciers that feed ice from the centre of Greenland to the ocean are also depleting rapidly (Fig. 2, overleaf). This is occurring particularly in the southeast, where thinning rates can be more than 10 m yr−1 (ref. 6). This depletion coincides with mass loss identified in the southeast using GRACE2,3. The mass loss in the northeast2 is possibly caused by thinning of the northeast Greenland ice stream4. Chen et al. postulate that changes in glaciers in the Svalbard archipelago, which lies northeast of Greenland between Norway and the North Pole, might be part of the explanation3.

Figure 2: Accelerated slippage.

A 2005 image of Kangerdlugssuaq, an outlet glacier of the Greenland ice sheet, from a Disaster Monitoring Constellation satellite. The glacier's calving speed doubled during 2004; the yellow line shows marginal position in 2000, about 4 km in front of imaged position.

The short period over which the GRACE observations were made means that the measured changes in the rate of mass loss could simply be the result of variations in snowfall or summer melt. Indeed, 2002–03, which preceded the mass-loss acceleration, was a year of unexpectedly high snowfall in southeast Greenland6, and 2005, which immediately followed it, was a year of record melt7. But the difference between snow accumulation and meltwater run-off accounts for only around a third of the mass loss from Greenland8. The remainder is lost through the calving of icebergs at the margins of fast-flowing outlet glaciers. The flow rate of many of Greenland's outlet glaciers increased between 1996 and 2000, and again in the period to 2005, especially in the south8. In spring 2004 — at the same time as the increase in mass loss recorded by GRACE2 — significant accelerations in the flow and calving rate of two major outlet glaciers occurred9.

The agreement of the GRACE results2,3 with measurements of glacier dynamics8 on the scale and the timing of the mass loss suggests that the accelerating contribution to sea-level rise — which in 2004–06 was equal to almost 0.7 mm yr−1 (ref. 2) — results from changes in the dynamics of outlet glaciers. Current model predictions from the Intergovernmental Panel on Climate Change suggest a sea-level rise of 0.5±0.4 m during the 21st century. But these models contain only a small component of the dynamic response of glaciers10, and the GRACE results indicate that more rapid changes are occurring than the models predict. The GRACE results can thus help us to re-evaluate the rates of loss from the ice sheet that we should expect through climate warming.

It is clear that there is much we don't understand about the current response of the Greenland ice sheet. Records over short periods have to be treated with caution, and we cannot be certain that changes represent a profound alteration in the behaviour of the sheet. But several independent sources now confirm overall mass loss from the Greenland ice sheet, together with unexpected and rapidly changing behaviour. Uncertainties remain, but the GRACE results provide one of the best estimates of overall mass balance of the ice sheet.

They do not, however, reveal the detailed pattern, at least not yet. It is vital that we use a variety of instruments and techniques to make continued observations of the ice sheet's response, and complement these with studies aimed at understanding the processes that are driving the observed changes. Such a programme will allow us to improve our predictive models of the Greenland ice sheet, and assess the timing and extent of its future contribution to sea-level rise.

Change history

  • 26 October 2006

    The figures given in mass units for Greenland's ice loss are 1,000 times too small.


  1. 1

    Fichefet, T. et al. Geophys. Res. Lett. 30, 1911 (2003).

  2. 2

    Velicogna, I. & Wahr, J. Nature 443, 329–331 (2006).

  3. 3

    Chen, J. L., Wilson, C. R. & Tapley, B. D. Science doi:10.1126/science.1129007 (2006).

  4. 4

    Johannessen, O. M., Khvorostovsky, K., Miles, M. W. & Bobylev, L. P. Science 310, 1013–1016 (2005).

  5. 5

    Thomas, R., Frederick, E., Krabill, W., Manizade, S. & Martin, C. Geophys. Res. Lett. 33, L10503 (2006).

  6. 6

    Krabill, W. et al. Geophys. Res. Lett. 31, L24402 (2004).

  7. 7

  8. 8

    Rignot, E. & Kanagaratnam, P. Science 311, 986–990 (2006).

  9. 9

    Luckman, A., Murray, T., de Lange, R. & Hanna, E. Geophys. Res. Lett. 33, L03503 (2006).

  10. 10

    Intergovernmental Panel on Climate Change The Science of Climate Change (Cambridge Univ. Press, 2001).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.