Greenland's greenhouse

Glaciers in southern Greenland are flowing 30% to 210% faster then they were ten years ago, and the overall amount of ice dumped into the sea from Greenland increased from 90 cubic km in 1996 to 224 cubic km in 2005, up 250%. As a result, Greenland's contribution to average annual sea rise increased from .23 mm/year in 1996 to .57 mm/year in 2005, and now accounts for between 20% and 38% of the observed yearly global sea level rise. Two-thirds of Greenland's contribution (.38 mm/yr) was due to glacier dynamics (chunks of ice breaking off and melting), and one-third (.19 mm/yr) from melting. These were the results of a paper called "Changes in the Velocity Structure of the Greenland Ice Sheet" published last Friday in Science magazine. NASA scientist Eric Rignot and University of Kansas researcher Pannir Kanagaratnam used ten years of satellite radar interferometry data to arrive at their conclusions.

The authors attributed the speedier glacier flow in southeast Greenland to climate warming, and noted that there had been a 3° C rise in temperature in the past 20 years at one station there. Widespread glacier acceleration affected just the southern tip of Greenland south of 66° north between 1996 and 2000, then spread rapidly northwards to 70° north by 2005 to cover the southern half of Greenland. The authors anticipated that as glacier acceleration continued to spread northward, Greenland's contribution to global sea level rise would continue to increase in coming years.

Greenland's increase in glacier speed and a corresponding rise in global sea level are reason for great concern, since Greenland holds enough ice to raise global sea level by over 20 feet (6.5 meters), should the ice cap disintegrate. However, the paper does not discuss many complicating factors, and it is uncertain if the paper's findings mean that Greenland's ice cap is in immediate danger. The most worrying aspect of the paper's findings is that we are told that the computer models used to estimate how long it will take Greenland's ice will melt are significantly in error--and in the wrong direction!


Figure 1. Change in sea level from 1993 to the end of 2004 shows a steady increase of about 3 mm/year. No acceleration of sea level rise due to increased input from Greenland or other causes is apparent. Image credit: University of Colorado.

Is the new Greenland melting evident in global sea level trends?
Sea level is a surprisingly difficult thing to measure. Tide gauges are very noisy, and only show sea level trend for the coastal areas they happen to be installed on. Global sea level trends from these gauges show a rise of between 1.5 and 2.1 mm/yr for the period 1950-2000. Satellite data from the TOPEX/POSEIDON and JASON satellites can give a better global picture, and show a rise of 2.9 mm/yr for the period 1993-2004 (Figure 1). This had increased to 3.4 mm/year for the period 1993-2007. The reason for the disagreement between the tide gauges and satellite data is unknown. There is a lot of variability in the data, due to changes in evaporation and precipitation related to such events as El Niño and La Niña. Indeed, sea level is not rising everywhere. In Scandinavia, the land is still rebounding from the Ice Age, and local sea level is receding. Sea level is also not increasing in the South Pacific's Vanuatu Islands, which Michael Crichton focuses on in his State of Fear book. This lack of sea level rise is not well understood, but may in part due to regional ocean current and precipitation patterns that reduce the amount of sea level rise one might expect. Mitrovica et.al. (2001) argue that as an ice sheet melts, the gravitational pull of the ice sheet on the surrounding ocean decreases, so that a substantial melting of the Greenland ice sheet would result in substantial drop in sea level over the North Atlantic, and a major sea level rise over the South Pacific with the maximum rise near the southern part of South America.

Given how noisy the global sea level data is, it should be no surprise that an increasing trend in sea level due to the increased contribution from Greenland is not apparent in Figure 1. According to Rignot and Kanagaratnam, Greenland's contribution to global sea level rise increased from .23 mm/year in 1996 to .57 mm/year in 2005, an increase of .34 mm/year. This is less than the error bounds of .4 mm/yr in the Figure 1 satellite data. It is also worth noting that while Rignot and Kanagaratnam's estimates to contributions to sea level rise due to glacier flow (.38 mm/yr) are not disputed by other studies, their estimate of the amount of melting Greenland is undergoing (.19 mm/yr) is in dispute. For example, Box et al. (2004) came up with a global sea level rise of 1.5 mm/yr due to Greenland's contributions, and Chylek et. al. (2004) say that the ice sheet may not be adding to sea level rise at all.



Figure 2. Average temperatures for the two stations in Greenland with a century-long record. Top: Godthab. Bottom: Angmagssalik. Image credit: NASA Goddard.

Is Greenland's Ice Cap in danger of disintegrating?
Greenland's ice cap is probably not in immediate danger of disintegrating, if temperatures stay at their current levels. Most of Greenland has been in a cooling trend over much of the last 60 years. It is only during the past ten years that we have seen a sharp upward jump in the temperatures at many (but not all) Greenland locations. However, temperatures as warm as Greenland is seeing now were also observed back in the Medieval Warm Period of 800-1300 A.D., and again in the 1930s. We can see the warm period of the 1930's reflected in the temperature records for two Greenland stations with records extending back over a century (Figure 2). Presumably, Greenland's glaciers at that time accelerated to speeds similar to what we are seeing today, without the ice cap suffering significant disintegration. I haven't looked for records of glacier flow and iceberg calving for that time period to check this hypothesis; I doubt reliable records exist.

Natural Variability
The temperature plot of Figure 2 demonstrates that Greenland is subject to large decades-long changes in its climate due to natural variation. The 2-4° C increase in temperature during the 1920s must have been primarily due to natural causes, since human-emitted greenhouse gases were relatively low then. Research results show that the climate of Greenland is dominated by a regional weather pattern called the North Atlantic Oscillation (NAO). The NAO oscillates unpredictably between a negative phase and a positive phase. If the wintertime NAO is negative, the persistent low-pressure area near Iceland called the Icelandic Low moves towards the southern tip of Greenland, bringing a sharp increase in precipitation and warmer temperatures to the island. During the positive phase of the NAO, the Icelandic Low moves back towards Iceland, allowing colder and drier conditions to prevail over Greenland. The wintertime NAO during 1950-2000 was primarily positive, which led to cooling over virtually all of Greenland--the opposite of the global warming trend of most of the rest of the world (Chylek et al., 2004). This cooling reduced the amount of glacier break-up and melting one would have expected due to global warming. To make things more complicated, increased precipitation during the wintertime negative NAO phase tends to add mass to the ice sheet in the interior, and may partially or totally offset the mass loss due to melting from that phase's increased temperatures (Johannessen, 2005). This is a very complicated system with many unknowns! The question--which was not discussed in Rignot and Kanagaratnam's paper--is, how will the expected rise in global temperatures of 1.5 to 4.5° C this century affect the NAO, and thus Greenland's temperature and precipitation? The current consensus from the computer models is that global warming should act to create a more positive NAO, which would keep Greenland cooler and drier.

When will Greenland's Ice Cap be gone?
The consensus view (Gregory et. al, 2004), using computer models that treat the Greenland ice sheet as a static hunk of ice, has been that the Greenland ice sheet will melt in about a thousand years, if atmospheric CO2 doubles. However, the doubling in glacier flow observed in the past ten years comes as a major shock. The models used to come up with the 1000 year estimate do not account for changes in glacier speed at all! The unexpected increase in glacier flow probably occurred in response to the lubrication effect of melt-water penetrating down to the glacial bed, as well as other poorly-understood processes. The paper concluded: "Current models used to project the contribution to sea level from the Greenland Ice Sheet in a changing climate do not include such physical processes and hence do not account for the effect of glacier dynamics." In other words, the models were wrong. Climate change skeptics are find of criticizing computer models, and cite their inadequacy as grounds for dismissing the threat of climate change. Well, it works both ways. Climate change models can be off in the wrong direction--as we also saw with the Antarctic ozone hole, which was completely missed by the models. These new results imply that if Greenland warms significantly (at least 3° degrees C), Greenland's ice could melt in a few centuries, not 1000 years. With 20 feet of sea level rise locked up in its ice, sea level rises well beyond the capability of humans to handle could occur later this century. The real test of the stability of the Greenland Ice Sheet will come when we reach temperatures not seen since before the last ice age, 125,000 years ago. Warm temperatures then caused the Greenland Ice Sheet to mostly disintegrate, leading to perhaps 14-17 feet (4.5-5 meters) of sea level rise (Cuffey and Marshall, 2000). The likelihood of this scenario is highly uncertain, though, given our lack of understanding of the system, the high amount of natural variability, and the limited amount of historical data we have to look at.

One interesting political note
Rignot works for NASA, which has recently been embroiled in controversy over whether political appointees there had tried to silence NASA climate scientist Jim Hansen from voicing his opinions. According to Time Magazine, when Rignot was asked if anyone at NASA had tried to shut him up, he said he had not been subjected to any such pressure.

The University of Colorado has a nice image showing where Greenland melted in 2005, and which areas melted for the first time.

Jeff Masters

References
Box, J.E., D.H. Bromwich, and L-S Bai, 2004. Greenland ice sheet surface mass balance 1991-2000: Application of Polar MM5 mesoscale model and in situ data. J. Geophys. Res., 109, D16105, doi:10.1029/2003JD004451.

Chylek, P, J.E. Box, and G. Lesins, "Global Warming and the Greenland Ice Sheet", Climatic Change 63:, 201-204, 2004.

Gregory, J.M., Huybrechtsm, P., and Sarah C. B. Raper, "Threatened loss of the Greenland ice-sheet", Nature 428, 616 (8 April 2004) | doi:10.1038/428616.

Johannessen, O.M., et. al, "Recent Ice-Sheet Growth in the Interior of Greenland", Science, 310: 1013-1016, 11 November 2005; published online 20 October 2005 [DOI: 10.1126/science.1115356]

Cuffey, K.M., and S.J. Marshall, "Substantial contribution to sea-level rise during the last interglacial from the Greenland ice sheet," Nature, 404, 591-594 (2000).

Mitrovica, J.X., Tamislea, M.E., Davis, J.L., and G.A. Milne, "Recent Mass Balance of Polar Ice Sheets Inferred from Patterns of Global Seas-Level Change", Nature 409, 1026-1029, 2001.