Science Briefs

Getting a Grip on Future Ice Sheet Changes

The current rate of sea level rise is about 3 mm/year, but how much will it be over the course of the next century? Human-caused, or anthropogenic, climate change is making temperatures over Greenland warmer. Will sea level rise be faster in the future because of melting of the Greenland ice sheet?

Map of ice melt in Greenland

Figure 1, at right Extent of ice melt from the Greenland ice sheet in 1992 and 2005. Pink areas are where at least 1 day of melting occurred in 1992, while red pink plus red areas are where melting occurred in 2005. Click for larger image. (Image: Huff and Steffen, CIRES, Univ. Col.)

If Greenland were to completely melt, global sea levels would rise by around 6.5 meters. And while that isn't likely to happen anytime soon, Greenland did lose 2-3 meters of ice the last time that the Arctic was as warm as we expect it to be by the year 2100. It is challenging to determine how fast the Greenland ice sheet might melt because of the short observational record and the lack of adequate computer models for predicting future ice sheet decay and the difficulty in seeing what is happening inside an ice sheet.

The Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report report indicated that there would be up to 59 cm of sea level rise over the next century, primarily a result of the expansion of the oceans due to warming. They stated that if the observed contributions from the Greenland and Antarctic ice sheets between 1992 and 2003 were to increase in direct parallel with global average temperature change, the upper ranges of sea level rise would increase by 10 to 20 cm. This prediction, however, is based on data collected in a very short period of time — mostly from the last decade — and is not enough to give a clearer idea about what might happen to the Greenland ice sheet.

Map of Laurentide ice sheet thickness

Figure 2, at left: Topography of Laurentide ice sheet (9 kyr NP) used as input to climate model simulation. Units are meters of ice thickness. (Image: Carlson et al. See references.)

The second challenge is that ice sheet modeling is still in its infancy, in part due to the lack of observations of ice sheet decay, and therefore cannot accurately predict what the melt might look like. The current state-of-the-art ice sheet models suggest a slow ice sheet response to global warming, but they cannot accurately reproduce ice age to warm age cycles of the past. This limits their use in future predictions of ice sheet changes until they improve. It is thought that part of the problem with these models is their lack of small-scale processes that speed up the flow and retreat of ice sheets.

To overcome these challenges, one approach for examining the potential for future changes to Greenland is to explore the last example of an ice sheet disappearance around 9000 years ago. At that time, the Laurentide ice sheet, a remnant of the last ice age, covered much of central to eastern Canada and the Hudson Bay. At that time, it had roughly twice the volume of ice that Greenland does today. Milankovitch cycles, "wobbles" in Earth's orbit that change the timing and amount of sunlight reaching the surface, are thought to be the underlying cause for the decay of this ice sheet, although climate feedbacks likely sped up its retreat.

The geologic evidence of the retreat of this ancient ice sheet indicates that it increased global sea levels by around 1.3 cm/year, significantly faster than the current sea level rise projections from IPCC (that haven't yet fully accounted for ice sheet melt).

Map mof model simulated ice sheet mass balance Figure 3, at right: Model-simulated annual mass balance predicted (units: mm/day), with the gray lines denoting the equilibrium line. (Image: Carlson et al. See references.)

But were the "stresses" on this ancient ice sheet really similar to those of today or the near future? To understand the details of these stresses requires the assistance of a global climate model. We used the NASA GISS ModelE-R to study the climate conditions of 9,000 years ago, including changing the incoming sunlight (or solar radiation) and greenhouse gases and adding in the Laurentide ice sheet.

Importantly, ModelE-R used special water isotope "tracers" that are recorded in many of the records of past climate change. This allowed us to evaluate the skill of the climate model by making direct comparisons between its results and the geologic evidence of past climate. The comparison revealed that it did a very good job at simulating the climate causing the disappearance of this ancient ice sheet, thereby improving our confidence in the model.

The driving factor in melting the Laurentide ice sheet was indeed the increased solar radiation caused by a change in the earth's orbit, which in turn increased summer temperatures. Given current warming trends linked to human activities, we expect that similar temperature rises may occur over Greenland by 2100.

The new analysis suggests that the pressures — particularly the temperature increase — on the Laurentide ice sheet 9000 years ago were similar to those that Greenland will face by the year 2100. This implies a potential for melting substantially greater than the IPCC projections (up to a factor of 10 by the end of the century).

Our study did not consider "dynamic" ice effects. An example of this occurs when water melts on the surface and drains to the bottom of the ice, making the land surface more slippery and thus speeding the rate at which ice flows. These effects are likely to increase the rate of ice sheet disappearance.

We are continuing to address this uncertainty in the fate of the Greenland ice sheet model by more careful examination of the Laurentide. New ice sheet models are under active development at GISS and elsewhere, and we hope that the "skill test" of the paleoclimate proxy tracers used in this study will improve the ice sheet model as well.


Carlson, A.E., A.N. LeGrande, D.W. Oppo, R.E. Came, G.A. Schmidt, F.S. Anslow, J.M. Licciardi, and E.A. Obbink, 2008: Rapid early Holocene deglaciation of the Laurentide ice sheet. Nature Geosci., 1, 620-624, doi:10.1038/ngeo285.


Please address all inquiries about this research to Dr. Allegra LeGrande.