Top climate story of 2008: Arctic sea ice loss

The top climate story of 2008, as it was in 2007, was the extraordinary summertime sea ice retreat in the Arctic. For the second consecutive year, we experienced the opening of the fabled Northwest Passage through the Canadian Arctic waters. Explorers have been attempting to sail the Northwest Passage since 1497, and 2007 and 2008 are the only known years the passage has been ice-free. In addition, 2008 saw the simultaneous opening of the Northeast Passage along the coast of Russia. This meant that for the first time in recorded history, the Arctic ice cap was an island--one could completely circumnavigate the Arctic Ocean in ice-free waters. Although the summer ice extent in 2008 finished 9% higher than 2007's record minimum, it was still an extraordinary 34% below average, according to the National Snow and Ice Data Center. Furthermore, the ice was thinner at the September 2008 minimum compared to 2007, so the total ice volume (thickness times area) was probably at its lowest point in recorded history in 2008.


Figure 1. Daily arctic sea ice extent for September 12, 2008. The date of the 2008 minimum (white) is overlaid on September 16, 2007--last year's minimum extent (dark gray). Light gray shading indicates the region where ice occurred in both 2007 and 008. Image credit: National Snow and Ice Data Center.

The Arctic "perfect storm" of summer weather in 2007 did not repeat in 2008
The summer of 2007 saw a "perfect storm" of weather conditions favorable for ice loss. Unusually strong high pressure over the Arctic led to clear skies and plenty of sunshine. Arctic winds, which usually blow in a circular fashion around the Pole, instead blew from the south, injecting large amounts of warm air into the Arctic. How unusual were these conditions? Well, at last month's meeting of the American Geophysical Union, the world's largest scientific conference on climate change, J.E. Kay of the National Center for Atmospheric Research showed that Arctic surface pressure in the summer of 2007 was the fourth highest since 1948. Cloud cover at Barrow, Alaska was the sixth lowest. This suggests that once every 10-20 years a "perfect storm" of weather conditions highly favorable for ice loss invades the Arctic. The last two times such conditions existed was 1977 and 1987.

The 2008 melting season began in March with slightly greater ice extent than had been measured in previous years, thanks to a relatively cold winter during 2007-2008. However, since so much ice had melted during the summer of 2007, most of the March 2008 ice was thin first-year ice, which extended all the way to the North Pole. The total ice volume in the Arctic in March 2008 was lower than what the record-breaking year of 2007 had seen. This led to speculation that a new record minimum would be set in 2008, and Santa's Workshop would plunge into the ocean as ice melted at the North Pole. However, the "perfect storm" of summertime weather conditions did not materialize in 2008. From May through July, cooler temperatures and winds less favorable to ice loss occurred. When very warm temperatures moved into the Arctic in August, the ice loss rate accelerated to levels higher than in 2007. However, with sunlight waning, ice loss was not able to reach the levels seen in 2007. Arctic temperatures in the summer of 2008 were up to 4°C cooler along the Siberian coast than in 2007 (Figure 2).


Figure 2. Difference in surface temperature (°C) between the summer of 2008 and the summer of 2007. Blues and purples indicate areas where is was cooler in 2008. The biggest change was over the Bering Sea between Alaska and Russia, where exceptionally sunny weather with southerly winds in 2007 caused record-breaking warmth. Image credit: NOAA/ESRL.

The future of arctic sea ice
Climate models have done a poor job predicting the recent record loss of arctic sea ice (Figure 3). None of the models used to formulate the official word on climate, the 2007 United Nations IPCC report, foresaw the shocking drop of 2007-2008. At the December 2008 AGU meeting, Wieslaw Maslowski of the Navy Postgraduate School hypothesized that the reason for this was the models' improper handling of ocean currents and how they transport heat. He blamed 60% of the melting during the past decade on heat brought in by ocean currents, and projected that summertime arctic sea ice would completely disappear by 2016. Dr. Jim Overland of NOAA's Pacific Marine Environmental Laboratory was more conservative, projecting a 2030 demise of arctic sea ice. He thought we would be "hanging around where we are for a while", and thought it would take two more unusual summers like the "perfect storm" of 2007 to push the system to an ice-free state. He further noted that while summertime air temperatures have been near record levels the past few years in the Arctic, there has been one period of comparable warmth, in the 1930s and 1940s. The year 1941 still ranks as the warmest year in the Arctic, though 2007 was virtually tied with it. However, the warmth of the 1930s and 1940s was different than the current warming, and was caused by the Siberian High moving unusually far east over Europe, driving warm, southerly winds over Greenland. The warmth in the past decade, in contrast, is associated with a warming of the entire planet, and is not due to an unusual pressure pattern driving warm air into the region. This means that the current warming is accompanied by much warmer ocean waters, which have helped caused much of the arctic sea ice loss the past two years by melting the ice from beneath.


Figure 3. Arctic sea ice extent from observations (thick orange line) and 13 model forecasts used to formulate the 2007 IPCC report (light lines). The thick black line is the multi-model ensemble mean, with the standard deviation plotted as a dashed black line. Image has been updated to include the observed 2007 and 2008 measurements. Image credit: Arctic sea ice decline: Faster than forecast by Stroeve et al., 2007.

The impact on the jet stream
The unprecedented melting of arctic sea ice the past two summers has undoubtedly had a significant impact on the early winter weather over the Northern Hemisphere. Several modeling studies presented at the December AGU meeting showed that sea ice melt on this scale is capable of injecting enough heat into the atmosphere to result in a major shift in the jet stream. Dr. Overland remarked that the early cold winter over North America this winter, and the exceptionally cold and snowy early winter in China last winter, were likely related to arctic sea ice loss. The sea ice loss induced a strong poleward flow of warm air over eastern Siberia, and a return flow of cold air from the Pole developed to compensate. Thus regions on either side of eastern Siberia--China and North America--have gotten unusually cold and snowy winters as a result.

The impact on sea level rise
The loss of arctic sea ice will have little impact on sea level rise over the next few decades. Since the ice is already floating in the ocean, melting it does not change sea level much--just like when ice melting in a glass of water will not change the level of liquid in the glass. In the case of sea ice, there is a slight sea level rise, since the fresh melt water is less dense than the salty ocean water it displaces. If all the world's sea ice melted, it would raise global sea level by only 4 mm. This is a tiny figure compared to the 20 feet of sea level rise that would occur from complete melting of the Greenland ice sheet--which is on land.

The impact on melting of the Greenland Ice Sheet
The big concern with arctic sea ice melt is the warmer temperatures it will bring to the Arctic, which will bring about an accelerated melting of the Greenland Ice Sheet. As the sea ice melts, the resulting warmer average temperatures will increase the amount of dark, sunlight-absorbing water at the pole, leading to further increases in temperature and more melting of sea ice, in a positive feedback loop. As temperatures warm, partial melting of the Greenland Ice Sheet will raise global sea levels. While no one is expecting 20 feet of sea level rise from the total melting of the Greenland Ice Sheet for many centuries, even one meter (3.3 feet) of sea level rise due to the partial melting of the Greenland Ice Sheet can cause a lot of trouble. The official word on climate, the 2007 IPCC report, predicted only a 0.6-1.9 foot sea level rise by 2100, due to melting of the Greenland ice sheet and other factors. These estimates did not include detailed models of ice flow dynamics of glaciers, on the grounds that understanding of the relevant processes was too limited for reliable model estimates. The IPCC estimates were also made before the shocking and unexpected loss of arctic sea ice of the past two summers. In light of these factors, a large number of climate scientists now believe the IPCC estimates of sea level rise this century are much too low. The most recent major paper on sea level rise, published this month by Grinsted et al., concluded that there was a "low probability" that sea level rise would be in the range forecast by the IPCC, and predicted a 0.9 - 1.3 meter (3 - 4.3 feet) rise by 2100. Pfeffer et al. last month concluded that a "most likely" range of sea level rise by 2100 is 2.6 - 6.6 feet (0.8 - 2.0 meters). Their estimates came from a detailed analysis of the processes the IPCC said were understood too poorly to model--the ice flow dynamics of glaciers in Greenland and Antarctica. The authors caution that "substantial uncertainties" exist in their estimates, and that the cost of building higher levees to protect against sea level rise is not trivial. Other recent estimates of sea level rise include 1.6 - 4.6 feet (0.5 - 1.4 meters) by Rahmstorf (2007).

What would 3 feet of sea level rise mean?
Rising sea levels will lead to permanent and intermittent flooding in low-lying coastal areas across the world. A global sea level rise of .9 meters (3 feet) would affect 100 million people worldwide, mostly in Asia. The impact of hurricane storm surges will significantly increase as a result of sea level rise. Given a 3 foot rise in sea level, Hurricane Ike's storm surge would have overwhelmed the levees in Port Arthur, Texas, flooding the city and its important oil refineries. Galveston's sea wall would have been overtopped and possibly destroyed, allowing destruction of large portions of Galveston. Levees in New Orleans would have been overtopped, resulting in widespread flooding there, as well. I'll have a full analysis of who's at risk, and what the risks are, in a series of forthcoming blog posts this year.

What can we do?
One reasonable suggestion, presented by Trish Quinn of NOAA at the December 2008 AGU meeting, would be to limit the amount of crop residue burning that goes on in Eastern Europe and Asia each year. These fires generate large amounts of black soot that blows into the Arctic. These black particles on the white ice leads to a significant amount of warming during the summer months, when the black particles absorb sunlight.

For more information
The wunderground sea level rise page has detailed background info on sea level rise.
The wunderground Northwest Passage page is also a good reference.
realclimate.org has a nice post summarizing the recent sea level research.

I'll have a new blog post Wednesday or Thursday.

Jeff Masters