How Fast Are the Ice Sheets Melting?

See allHide authors and affiliations

Science  24 Nov 2006:
Vol. 314, Issue 5803, pp. 1250-1252
DOI: 10.1126/science.1133325

If the ice sheets covering Greenland and Antarctica were to melt completely, they would raise sea level by about 65 m. But even a small loss of ice mass from the ice sheets would have a great impact on sea level, particularly on low-lying islands and coastal regions. New satellite observations, including those reported by Luthcke et al. on page 1286 of this issue (1), now allow estimates of the mass balances of the ice sheets and their evolution through time.

For the past 3000 years, global sea level has remained stable, but since the end of the 19th century, tide gauges have detected global sea-level rises [∼:1.8 mm/year on average over the past 50 years (2, 3)]. Satellite altimetry data document a rate of ∼:3 mm/year since 1993 (4). However, it remains unclear whether the recent rate increase reflects an acceleration in sea-level rise or a natural fluctuation on a decadal time scale.

Present-day sea-level rise has several causes. During the past decade, ocean warming has contributed roughly half of the observed rate of sea-level rise (5), leaving the other half for ocean-mass increase caused by water exchange with continents, glaciers, and ice sheets (6). The contribution of mountain glaciers and small ice caps to sea-level rise in the past decade is estimated to be ∼:0.8 mm/year (7). These figures constrain the contribution from ice sheets to less than 1 mm/year in the past decade.

Since the early 1990s, remote-sensing data based on airborne laser and satellite radar altimetry, as well as the space-borne Synthetic Aperture Radar Interferometry (InSAR) technique, have provided the first observations of ice sheet mass balance (813). These observations indicate accelerated ice-mass loss in recent years in the coastal regions of southern Greenland. In contrast, slight mass gain is reported in central high-elevation regions. Over Antarctica, remote sensing indicates accelerated mass loss in the western part of the continent (10), whereas the eastern part is gaining some mass as a result of increased precipitation (11, 12).

Because of these contrasting behaviors—mass loss in coastal regions and mass gain in elevated central regions—ice-sheet mass loss exceeds mass gain only slightly. Thus, according to the recent mass-balance estimates, the ice sheets presently contribute little to sea-level rise. However, great uncertainty remains, mainly because of incomplete coverage by remote-sensing surveys, spatial and temporal undersampling, measurement errors, and perturbation from unrelated signals. In addition, each technique has its biases. For example, radar altimetry misses narrow coastal glaciers because of inadequate ground resolution, and ice elevations measured by the radar are much less reliable over steep, undulated surfaces than over flat high-elevation surfaces. Another uncertainty arises because conversion of elevation change to mass change requires assumptions about the surface density of snow or ice to be made.

Since 2002, the NASA/DLR Gravity Recovery and Climate Experiment (GRACE) satellite mission has provided a new tool for precise measurements of ice-sheet mass balance, with nearly complete coverage of the high-latitude regions up to 89°N/S. GRACE measures the spatiotemporal change of Earth's gravity field. Over the ice sheets, this change can be converted into ice-mass change, assuming that the gravity change results from a change in surface mass.

Several studies have reported estimates of Greenland and Antarctica ice-mass change from GRACE (1419). The GRACE results confirm those from other remote-sensing techniques, that is, net ice-mass loss from Greenland and West Antarctica and a slight ice-mass gain over East Antarctica (see the figure). The GRACE results over Greenland also suggest accelerated ice-mass loss since 2002, in agreement with InSAR results (13).

Ice-sheet mass change estimated by different remote-sensing techniques.

(Left) Greenland. (Right) West and East Antarctica. The ice sheet mass change is given in gigatons per year. The numbers refer to different investigations as quoted in the reference list. Open bars correspond to GRACE results, filled bars to results from other techniques. The estimate from (15) is an average over the whole of Antarctica. On the right, positive values are for East Antarctica and negative values for West.

However, the GRACE-based mass-balance estimates are highly scattered (see the figure). One reason is the short time span of the analyses (2 to 4 years, depending on the study). Over Greenland, ice mass varies widely from year to year. Because the analyses do not overlap exactly in time, different trend estimates are to be expected.

Another cause of scatter is contamination from geodynamic processes related to Earth's response to ice melt from the last deglaciation. This effect, which depends on poorly known parameters, is mainly available from modeling, with important differences between models. Moreover, over Antarctica, this effect is of the same order of magnitude as present-day ice-mass change.

A third source of uncertainty is the coarse resolution (400 to 600 km) of most GRACE results (1419). As a result, the estimated ice-sheet mass change includes contributions not only from small isolated glaciers in the vicinity of the ice sheets, but also from other gravity signals (of oceanic, hydrologic, and tidal origin) from surrounding regions. These perturbing signals are still poorly known, and therefore difficult to be corrected for.

To improve the GRACE resolution, Luthcke et al. have applied a new approach over Greenland: They determined mass concentrations at a local scale from appropriate processing of the GRACE observations. This approach differs from the standard method, in which global solutions of the time-varying gravity field are computed, and a regional filter is then applied to extract the mass signal over the area of interest. The new approach minimizes the contamination from signals unrelated to the ice-sheet mass balance and provides results of finer resolution.

Luthcke et al. computed ice-mass change in six drainage basins of the Greenland ice sheet, ranging from coastal low-elevation to central elevated regions. They find ice-mass increase in high-elevation regions of northern Greenland, as suggested by satellite altimetry (11), and ice depletion at the margins of southern Greenland, in agreement with InSAR-based glacier discharge estimates (13). The results confirm accelerated ice flow in coastal regions of southeast Greenland. However, the trend is smaller than reported by some other recent GRACE-based studies (18, 19). Over the 2-year period of investigation, Luthcke et al.'s estimate of Greenland's contribution to sea-level rise amounts to ∼:0.3 mm/year.

However, further research is needed to improve estimates of Greenland and Antarctica mass balance (see the figure) and their contribution to sea level. Besides extending the time series of observations and reducing internal errors, it is important to reconcile estimates from different techniques and to eventually use them in synergy.

The greatest uncertainty in sea-level projections is the future behavior of the ice sheets. In recent years, the velocities of outlet glaciers in coastal regions of Greenland and Antarctica have accelerated, showing that a large fraction of ice-mass loss occurs through dynamical processes rather than surface melting (9, 10, 13). The dynamical response of the ice sheets to present-day climate forcing may thus play a much larger role than previously assumed. Future dynamical instabilities of the ice sheets is of major concern, given their potential impact on sea level (20), yet comprehensive modeling of such dynamical effects is in its infancy.

Improved mass-balance estimates from remote-sensing observations, such as those reported by Luthcke et al., will inform on the ongoing behavior of the ice sheets and help to validate models. This goal requires long time series of satellite observations, and hence continuity of space missions.

HyperNotes Related Resources on the World Wide Web

Ice Sheets

Ice Sheets, Greenland Ice Sheet, and Antarctic Ice Sheet Articles in Wikipedia.

Ice over the Poles A fact sheet from NASA's Earth Observatory.

Cryosphere Resources from the National Snow and Ice Data Center (NSIDC). Related science and data links are provided.

Land Ice Sheets A presentation by the Palaeoclimates Group, Department of Meteorology, University of Reading, UK.

Modern Glaciers and Ice Sheets Lecture notes by J. S. Aber for a course on Ice Age environments.

Climate Change—Breaking the Ice 24 March 2006 special issue of Science. Links to related previously published articles are provided.

Remote Sensing

Altimetry A presentation on the Aviso Web site.

Radar Altimeters and Light Detection and Ranging Instruments Introductions in the Earth Observation Handbook of the European Space Agency (ESA).

Geophysical Remote Sensing with Gravity, Radar and Microwave Remote Sensing, and Earth's Surface in 3D Sections in N. M. Short's Remote Sensing Tutorial, hosted by NASA's Goddard Space Flight Center.

InSAR Basics A presentation by the NPA Group.

ICESat Mission NASA satellite mission that includes study of ice sheet mass balance using GLAS (Geoscience Laser Altimeter System).

ERS-1 and ERS-2, ENVISAT, and CryoSat-2 ESA remote-sensing Earth observation satellite missions.

Mass Balance

Mass Balance A definition in the glossary provided by the Carbon Dioxide Information Analysis Center.

Ice Sheet Mass Balance Information from the Climate Change Institute, University of Maine.

“Mass Balance of Polar Ice Sheets” (8) Review by E. Rignot and R. H. Thomas in the 30 August 2002 issue of Science, a special issue titled “Trouble in Polar Paradise.”

Gravity Recovery and Climate Experiment

Gravity Recovery and Climate Experiment (GRACE) and GRACE Tellus Information from NASA's Jet Propulsion Laboratory.

GRACE Information from the University of Texas Center for Space Research. A fact sheet and a brochure are provided.

GRACE Mission Information from GeoForschungsZentrum, Potsdam, Germany.

“Greenland Mass Balance from GRACE” (14) 2005 Geophysical Research Letters article by I. Velicogna and J. Wahr, made available by I. Velicogna, Department of Physics, University of Colorado.

“Satellite Gravity Measurements Confirm Accelerated Melting of Greenland Ice Sheet” (18) Report by J. L. Chen, C. R. Wilson, and B. D. Tapley in the 10 August 2006 issue of Science.

Further Reading

Sea-Level Change 1990 report available from the National Academies Press.

“Ice-Sheet and Sea-Level Changes” (20) Review by R. B. Alley, P. U. Clark, P. Huybrechts, and I. Joughin in the 21 October 2005 issue of Science.

“Mass Changes of the Greenland and Antarctic Ice Sheets and Shelves and Contributions to Sea-Level Rise: 1992–2002” (11) December 2005 Journal of Glaciology article by H. J. Zwallyet al., available on the ICESat Web site.

“A Worrying Trend of Less Ice, Higher Seas” News Focus article by R. A. Kerr in the 24 March 2006 special issue of Science.

“Hitting the Ice Sheets Where It Hurts” Perspective by R. Bindschadler in the 24 March 2006 special issue of Science.

“Sea Level Rise During Past 40 Years Determined from Satellite and in Situ Observations” Report by C. Cabanes, A. Cazenave, and C. Le Provost in the 26 October 2001 issue of Science. “How fast are sea levels rising?” by J. A. Church is a related Perspective in the issue.

The Authors

Anny Cazenave is at the Laboratoire d'Etudes en Géophysique et Océanographie Spatiales, Observatoire Midi-Pyrénées, Toulouse, France.

S. B. Luthcke, D. D. Rowlands, R. D. Ray, and F. G. Lemoine are at the Planetary Geodynamics Laboratory, NASA Goddard Space Flight Center. H. J. Zwally and W. Abdalati are in the Cryospheric Sciences Branch, NASA Goddard Space Flight Center. R. S. Nerem is at the Colorado Center for Astrodynamics Research, Cooperative Institute for Research in Environmental Sciences, University of Colorado. J. J. McCarthy and D. S. Chinn are in the Science Division, SGT Inc., Greenbelt, MD.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
View Abstract

Stay Connected to Science

Navigate This Article