Climate Change and Arctic Sea Ice

Introduction

The Arctic ice pack is melting. A large body of recent scientific evidence now verifies what was once science fiction speculation set in an indefinite future.

Covering an average of 14 million square kilometers at its greatest extent in February and about half this size at its smallest extent in the northern autumn,[1] the Arctic's sea ice is a major driver of global weather systems. The light surface of the ice (which in scientific terms has a high "albedo") reflects solar energy away from the Earth and acts as a natural refrigerator for the planet. Ice and melt water from the Arctic Ocean have profound affects on ocean circulation patterns on the North Atlantic, and from there to ocean and other climate systems over the entire planet.

The Arctic's sea ice is home to a wide variety of wildlife, including polar bears, arctic foxes, seals, walruses, and whales, fish species such as Arctic cod and char, and sea birds such as guillemots, auks, and eiders. The sea ice is also used as an important transportation route by caribou and musk ox and a traditional hunting ground for the Inuit, that remarkable indigenous culture of the far north.

Declining ice extent and area

Detailed data on the extent of Arctic sea ice is available from ship and other scientific observations since at least 1953, and from weekly satellite observations since 1972. (The ice extent is the area of the ice pack north of the ice edge, including small areas of ice- enclosed open water, and is slightly larger than the actual ice area.) A U.S. study concluded that minimum ice extent in the northern autumn has declined by 3.6 percent per decade since 1961.[2] The Nansen Environmental and Remote Sensing Centre in Norway found a 4.6 percent decline in ice extent and a 5.8 percent decline in actual ice area between 1978 and 1994.[3] Tentative results suggest that this decline accelerated between 1987 and 1994.[4]

Most of the sea ice decline so far has occurred in the Barents, Kara and East Siberian seas north of Russia, as well as in the Sea of Okhotsk, an icy enclave of the Pacific Ocean northwest of Japan. In sharp contrast, an increasing amount of sea ice has developed in the Davis Strait and Labrador Sea between Canada and Greenland.[5]

Key factors

Two key factors appear to be influencing these changes in ice distribution. The first factor is an intense warming of more than 0.5 degrees Celsius per decade that has occurred over the past thirty years over much of the landmasses of Siberia, Alaska and Western Canada, combined with a weaker cooling trend in western Greenland and the Labrador Sea.[6]

The second factor is a major change in Arctic air circulation patterns. Average sea level pressure has dropped over the central Arctic Ocean,[7] and there has been an increase in high latitude storms.[8] As a result of these changes, relatively warm spring and summer air masses from much of the Arctic coast are able to penetrate far over the Arctic Ocean, melting some sea ice and driving much of the remainder away from the shore and towards the central ice pack.

Lengthening melt season

One of the major results of these changes has been a lengthening warm season during which sea ice can melt. The melt season has varied between 55 and 75 days between 1979 and 1996, and has lengthened at a rate of 5.3 days (8 percent) per decade during that time.[9] Once areas of open water, called "leads", have opened in the ice, these darker areas of ocean (technically called areas of lower "albedo") reflect less sunlight, warming up and thus melting still more ice.

Melting in the Beaufort Sea

The most recent evidence from Ice Station SHEBA suggests that sea ice decline may be spreading into the Beaufort Sea as well.

Ice Station SHEBA (Surface Heat Budget of the Arctic Ocean) consists of a number of scientific research stations surrounding a Canadian Coast Guard icebreaker, the Des Groseilliers, frozen into the ice pack in September 1997 in the Beaufort Sea. SHEBA is in part a follow-up of the Arctic Ice Dynamics Joint Experiment (AIDJEX) which took place in the same region in 1975-76, and thus makes it possible to monitor detailed changes in the Beaufort Sea over the last twenty years.

In a preliminary paper released on the Internet, SHEBA scientist Miles McPhee and colleagues commented "During the icebreaker transit into the perspective SHEBA site this past September [1997], we were struck by the lack of thick ice. Where we expected the mean thickness to be between 2 and 3 m, we were hard pressed to find floes more than 1.5 m thick. Even on the comparatively thick floe chosen for the SHEBA site, the thickest un-ridged ice rarely exceeded 1.8 m. Taken by itself, this might be attributed to selective sampling (icebreakers often seek out thinner, more navigable ice), or natural variation in regional ice thickness. However, with the first measurements of ocean temperature and salinity at the SHEBA site in early October, we found the upper ocean to be less saline and warmer than we had expected, and surmised that this indicated excessive melting."[10]

By carefully examining the salinity and temperature of layers of ocean water up to 500 meters deep, the SHEBA scientists were able to construct a history of the 1997 melt season in the Beaufort Sea. They concluded that the equivalent of more than two meters of ice must have melted in the Beaufort Sea that year to explain the unusually low levels of salt. This is an astonishingly high figure given that the AIDJEX expedition had found that the mean ice thickness was between two and three meters. AIDJEX had found that about 0.8 meters of ice melted during the 1975 summer, so the SHEBA figures were more than twice as high.

SHEBA measurements and calculations suggest that most, if not all, of the fresh water added to the Beaufort Sea came from melting Beaufort sea ice in the summer of 1997, and imply that much of the ice of the Beaufort Sea was much thinner (perhaps more than a meter thinner) at the end of the melt season in 1997 than it was during the end of the melt season in 1975.

The 1997 results are admittedly based in part on complex mathematical models and not entirely on direct measurement of ice thickness. The SHEBA scientists will remain frozen into the ice during the entire 1998 melt season, and will have an unusual opportunity to directly monitor the response of the Beaufort ice pack in great detail.

Warming in the Lincoln Sea

Signs of climate warming have even been observed in the Lincoln Sea, the most northerly sea in the Arctic Ocean, between Greenland and Ellesmere Island. Annual measurements taken during 1989-94 show a clear warming trend, with average temperatures 75 meters below the surface warming about 0.3 degrees Celsius during this period.[11] Although these measurements are not sufficient by themselves to establish a long term trend, the ongoing disintegration of the Milne, Ward Hunt and Ayles floating ice shelves of Ellesmere Island over the last few decades suggest that this trend may be real.[12]

Greenhouse gases and Arctic warming

Although direct systematic observations of Arctic sea ice have only been available for a few decades, much longer term surrogate temperature records from the surrounding land masses, including ice cores, tree rings, and lake bed pollen samples, suggest that the Arctic land area is now warmer than it has been in at least 400 years. A landmark survey published in November 1997 concluded that although this warming has likely been influenced by natural forces, including decreased volcanic activity, increased solar irradiance and natural variability in the climate system itself, the best explanation for most of the warming after 1920 is increasing levels of greenhouse gases.[13]

Future warming

It is likely that greenhouse-gas-induced Arctic warming is one of the major factors for the significant decline in sea ice area and thickness observed in many Arctic seas over the past few decades. General circulation computer models of the atmosphere project that greenhouse warming will occur more intensely over the Arctic in the future than any other part of the planet, largely because melting snow and ice will replace lighter surfaces with darker tundra and ocean surfaces, lowering the albedo, decreasing the sunlight reflected from the Arctic, and so accelerating the warming trend. The warming observed so far over much of the Arctic land masses is only a small fraction of the intense 8-16 degree Celsius increase in winter temperatures projected by computer models if greenhouse gas levels double over the next few decades.[14]

What effect would a greenhouse gas doubling have on the Arctic ice pack? This is a difficult question because Arctic warming will have complex effects on air and ocean circulation, clouds and precipitation. Nevertheless, most computer models project a dramatic decline in Arctic sea ice.[15]

Perhaps the most sophisticated analysis published to date is from Warren M. Washington and Gerald A. Meehl at the National Center for Atmospheric Research (NCAR) in Boulder, Colorado. The NCAR model includes coupled simulations of both the atmosphere and the oceans, with special attention paid to modeling sea ice and clouds. The model projects that warming the Arctic would create low hanging clouds that would tend to cool down the region and mitigate the warming to some degree. Nevertheless, sea ice thickness is reduced to less than half a meter in most Arctic seas during the winter, and no ice remains by the end of the summer melt season with the exception of an ice accumulation, less than a meter thick and perhaps four times the size of Iceland, floating off the Laptev Sea.[16]

The Arctic Without the Permanent Sea Ice - Local Implications

If current trends continue, and computer models are accurate in their broad details, Arctic winter ice cover will be much thinner in a greenhouse future, and almost non-existent in the summer. These dramatic changes would have enormous implications for Arctic plants, animals, and inhabitants.

Most Arctic marine species depend upon the presence of sea ice. The Arctic marine food chain begins with ice algae that cling to the underside of the dark ice pack all winter and creates a dense mat under the ice with the end of the long darkness in spring. About six weeks later, a phytoplankton bloom develops in the water beneath the ice. As the ice begins to break-up, the bloom spreads into a wide 20-80 kilometer belt surrounding the ice edge. This highly productive ice-edge ecosystem is home to numerous crustaceans and other invertebrates. These in turn are eaten by fish species such as Arctic cod. Organic material released from the ice algae mat and the phytoplankton bloom enriches the floor of the vast Arctic continental shelves, supporting a benthic (sea bottom) community of shellfish and other invertebrates. Unique among the world's ecosystems, the ice-edge zone moves thousands of kilometers each year, north in spring and south in fall. Walrus, numerous species of seals and cetaceans such as belugas and narwhals all follow the ice-edge, taking advantage of the ready access to food and (for the walrus and seals) the availability of ice to haul-out on for sunning, mating and raising pups. Seals are in turn preyed on by polar bears, humans and Arctic foxes.[17]

The almost complete elimination of multiyear ice in the Arctic Ocean is likely to be immensely disruptive to ice-dependent microorganisms, which will lack a permanent habitat. Preliminary results from the SHEBA ice camp in the Beaufort Sea suggest that ice algae and other micro-organisms may have already been profoundly affected by warming over the last 20 years. Samples taken at the SHEBA site indicate that most of the larger marine algae have died out, and have been replaced with a much less productive community of microorganisms more usually associated with freshwater ecosystems. The nutrient level in the water under the ice has decreased significantly.[18]

A recent study has suggested that increased concentration of greenhouse gases in the atmosphere, in addition to warming the lower atmosphere, will prevent heat from rising into the upper atmosphere. In particular, lower stratospheric temperatures over the Arctic will accelerate ozone depletion and delay any recovery.[19] Increased rates of spring ultraviolet radiation may have significant impacts on Arctic ice-dependent micro-organisms.

Perhaps the first regions to be affected will be ice-dependent seas near but outside the Arctic Ocean proper, including the Bering Sea, the Sea of Okhotsk, and Hudson Bay. These areas are currently covered in seasonal winter ice, which could vanish altogether with continued warming.

Walruses, which travel long distances on floating sea ice that allows them to feed over a wide area may be particularly vulnerable.[20] In a greenhouse future, sea ice will likely melt rapidly in the spring, shrinking quickly over continental shelf areas and withdrawing to the deep ocean of the central Arctic. This could be devastating to walrus, which use sea ice as a platform from which to feed primarily on shellfish on the bottom of shallow continental shelf areas.[21]

Many species of seal are ice-dependent, including the spotted seal, which in the Bering Sea breeds exclusively at the ice edge in spring; the harp seal, which lives at the ice edge all year; the ringed seal, which give birth to and nurse their pups on sea ice; the ribbon seal and the bearded seal.[22]

Polar bears would be threatened by any decline in ringed seal populations, their main food source. Moreover, polar bears are dependent on sea ice for hunting and transportation. In Hudson and James bays, all polar bears already fast during the 4 ice-free summer months they are forced to spend ashore, and pregnant females fast for 8 months.[23] Any extension in this ice-free period would increase starvation rates and decrease birth rates.

The summer and fall of 1990 may be an example of future conditions. Strong winds in May carried unusually warm air masses from Siberia out over the Arctic Ocean. Continued warmth in June promoted early breakup and consistent winds in July pushed ice away from the Siberian coast and towards the North Pole. By August, ice cover was 21 percent below normal. The low ice conditions persisted into September, which was 19 percent below normal. Almost all of this reduction occurred in the East Siberian Sea, with lesser reductions occurring in the adjacent Chukchi and Laptev seas.[24]

Even in October, there was no significant sea ice within several hundred kilometers of Wrangel Island, stranding walrus colonies and several hundred polar bears on the island. This Russian island, which marks the boundary between the East Siberian and Chukchi seas, is the largest polar bear denning area in the world.

According to a BBC report, the bears had no food supply beyond the walrus colonies, and only fully grown and experienced male polar bears are able to kill walrus. The other bears had to scavenge the remains of dead walrus. Arctic researcher Nikita Ovsyanikov said "with global warming, these conditions of climate and ice might become a regular occurrence. If that happens, the polar bear - now thought to be recovering in numbers since 1973 ... could be in trouble again."[25]

Earlier springs and late winter rain may cause polar bear snow dens to collapse. Two cubs and their mother bear were killed in such an incident near the Beaufort Sea coast in 1991.[26]

Arctic scientist Vera Alexander warns that "it seems likely that overall biological productivity of the arctic seas would be severely reduced by a temperature increase. Essentially all the distinctive Arctic animals would disappear."[27]

Sea ice changes are also affecting the Inuit. In a speech prepared for a meeting of the signatories to the Framework Convention on Climate Change in 1996, Rosemarie Kuptana, from the Beaufort Sea community of Sachs Harbour and then President of the Inuit Circumpolar Conference, noted:

"Climate change has not gone unnoticed at the community level in the Arctic. Today, our hunters are noticing changes in our homeland -- such as discolorations and thinning of sea ice, changes in the leads and open water areas, and the presence of animals not previously found in our region ... Highly experienced and knowledgeable hunters have had experiences falling through areas of sea ice they have previously known to be safe."[28]

The Arctic Without the Permanent Sea Ice - Global Implications

The significance of the physical and chemical processes taking place in the Arctic region extend far outside it. The polar area has been described as a "refrigerator in the equator to pole transport of energy".[29] The NCAR model projects a 3.8 degree Celsius global temperature increase for greenhouse gas doubling, near the high end of typical projections, in part because of shrinking Arctic sea ice.[30]

As well as being an area where nutrients are recycled and released into the water, the Polar Front region in the North Atlantic plays a fundamental role in the driving of ocean currents. At the front near the Greenland, Iceland and Norwegian (GIN) seas and the Labrador Sea, warm salty water from the North Atlantic is cooled by Arctic waters and by intense heat loss to the atmosphere; it becomes more dense and sinks to deeper layers of the ocean. Salt rejected as sea ice forms also increases the density and contributes to the process. Although a slow process, this sinking takes place over a wide area and each winter several million cubic kilometers of water sink and begin moving slowly south along the bottom of the Atlantic Ocean. It is known as thermohaline circulation because it is driven in part by temperature and partly by salinity differences.

The dense, cooled water becomes part of what is termed the Ocean Conveyor and the water eventually returns to the surface in the Indian and Pacific Oceans. As warm water returns to the Atlantic, the current moves pole-wards as the Gulf Stream and North Atlantic Drift, warming northwestern Europe substantially.[31] In addition, the formation of deep-water also dissolves carbon dioxide from the atmosphere and effectively removes it. This is of significance in the global cycling of carbon. The Arctic region, therefore, plays a fundamental role in ocean circulation patterns, which in turn determine climate patterns over the rest of the globe.

In 1968, a large area of unusually cold, fresh water appeared off the west coast of Greenland. Subsequent analysis suggests that this water was derived from melting sea ice which had broken off the Arctic ice pack and drifted south. This area of water, now called the Great Salinity Anomaly (GSA), significantly reduced deep water formation in the North Atlantic.[32] Fresher water is less dense than saltier water, and the GSA made the GIN and Labrador seas more buoyant, reducing the deep water formation that drives global ocean circulation.

Unusually cold, fresh water has been increasing in the Labrador Sea in recent years. There are a number of possible expectations for this, including the GSA, which drifted into the Labrador Sea in the late sixties and early seventies, and the possible disappearance of sea ice bridges further north, allowing drifting and melting sea ice to enter the Davis Strait and the Labrador Sea from the north.[33] In any case, a persistent freshening trend threatens the continued functioning of the Labrador Sea deep water formation.

Growing awareness of the vulnerability of thermohaline circulation to climate change has sparked a large amount of scientific research over the last few decades. The results are not encouraging.

The Ocean Conveyor appears to have operated fairly reliably over the past several thousand years. However, an examination of ice cores from both Greenland and Antarctica shows that this has not always been the case in the more distant past, and abrupt climatic changes associated with large amounts of sea ice in the North Atlantic and rapid changes in thermohaline circulation may have occurred repeatedly in the past. [34]

Mathematical modeling suggests that a continuous freshwater source about twice the size of the GSA would be enough to permanently prevent deep water formation in both the GIN and Labrador Seas. Moreover, a continuous freshwater source only half the size of the GSA would be enough to shut down the Labrador deep water formation.[35]

But where could such a continuous source of freshwater come from? Certainly not directly from melting sea ice. If enough of the 20 or 30 trillion cubic meters of Arctic sea ice were to melt each year to create a source twice the size of the GSA, the Arctic sea ice would be exhausted after 10-15 years.[36] The Greenland ice sheet is about 100 times larger than the floating sea ice, and so it could contribute for a much longer period.[37]

However, the most likely source for increased freshwater in the far North Atlantic is increased precipitation. As the climate warms and the sea ice melts, scientists expect that more rain and snow will fall on the Arctic Ocean and the North Atlantic, reducing the saltiness and density of the water. But would this be enough to shut off thermohaline circulation?

Computer modeling suggests yes - if enough warming happens fast enough. Modelers Syukuro Manabe and Ronald J. Stouffer find that doubling greenhouse gas levels creates enough precipitation to cut thermohaline circulation in half - and then circulation recovers back to normal levels over several centuries. However, if levels quadruple, possible if emissions increase unabated over the next century, circulation stops permanently.[38]

Using a simpler model than Manabe and Stouffer, Thomas F. Stocker and Andreas Schmittner find that a fast doubling of levels of greenhouse gases would shut the thermohaline circulation down permanently, while a slower doubling would only reduce the rate of circulation. Perhaps the most disturbing aspect of this study is that Stocker and Schmittner's "fast" scenario corresponds to current rates of greenhouse gas emission.[39]

Changes in precipitation are already beginning to occur. In Alaska, west of latitude 141, precipitation has increased by 30% between 1968 and 1990.[40] Elsewhere precipitation in high latitudes has increased by 15% over the last forty years.[41]

Not enough is known yet to estimate the impact of a thermohaline shutdown on northern Europe. Global warming may counterbalance the cooling caused by the shutdown of the Gulf Stream, resulting in only moderate temperature change. On the other hand, ocean circulation may halt much more quickly than the planet warms, dropping Europe quickly into a deep freeze.[42]

In any case, switching off the ocean's main global circulation system would have profound effects on global marine life and fisheries, during a century in which enormous pressure is already likely to be placed on ocean ecosystems and human food supplies. Moreover, shutting down deep water formation in the North Atlantic would eliminate important sinks for greenhouse gases, increasing their level in the atmosphere and the rate of climate change.[43]

Conclusion

Evidence for melting Arctic sea ice is available from many different sources. Warming Arctic landmasses; declining sea ice area, extent and thickness; decreasing salinity; and major changes in Arctic and North Atlantic air and ocean circulation all form part of the picture. Impacts have already been observed on many scales: to Arctic ice algae and other micro-organisms, to walrus and polar bear populations and to Arctic human inhabitants, such as the Inuit. Long term climate records suggest that most of this warming, especially after 1920, is driven by increasing levels of human-created greenhouse gases in the atmosphere.

Computer modeling suggests that, if warming and levels of greenhouse gases continue to increase, most of the permanent ice pack is likely to melt and be replaced by seasonal winter ice. This Arctic meltdown would threaten the productivity of the Arctic Ocean and the continued existence of many Arctic animals, including walrus, many seal species, and polar bears. It would also threaten the traditional lifestyle of the Inuit, the indigenous inhabitants of the Arctic coast.

The accelerated Arctic warming that would result from the removal of the permanent ice pack would significantly increase precipitation over the Arctic Ocean and far North Atlantic. This precipitation, combined with melt water from sea ice and the Greenland ice sheet, would reduce the salinity of the North Atlantic. Computer models suggest that these changes in salinity, especially if they happen quickly, may severely reduce or completely switch off the North Atlantic Conveyor, which is the major driving force for the Gulf Stream and global ocean circulation. This may significantly cool the climate of northern Europe, and is likely to severely disrupt global marine life and fisheries, as well as reducing the ocean's ability to remove greenhouse gases from the atmosphere.

[1] W.L. Chapman and J.E. Walsh, 1993. "Recent variations of sea ice and air temperatures in high latitudes", Bulletin of the American Meteorological Society 74:33-47

[2] Chapman and Walsh 1993, op.cit

[3] O. M. Johannessen, E. Bjorgo, and M.W. Miles, 1996. "Global Warming and the Arctic" (Letter), Science 271:129

[4] O. M. Johannessen, et.al., 13 July 1995. "The Arctic's shrinking sea ice", Nature 326:126-7.

[5] Claire L. Parkinson, 1992."Spatial patterns of increases and decreases in the length of the sea ice season in the north polar region, 1979-1986", J. Geophys. Res., v. 97, n. C9, pp. 14,377-14,388; and James A. Maslanik, Mark C. Serreze and Roger G. Barry, 1996. "Recent decreases in Arctic summer ice cover and linkages to atmosphere circulation anomalies", Geophys. Res. Lett. 23(13):1677-1680.

[6] Chapman and Walsh 1993, op.cit.

[7] J.E. Walsh, William L. Chapman and Timothy L. Shy, 1996. "Recent decrease of sea level pressure in the central Arctic", J. Climate, 9:480-486.

[8] Annette Varani, 1996. "At the Edge", in Distributed Active Archive Centres (DAACs): Supporting Earth Observing Science in 1995, National Snow and Ice Data Center, published on the Internet at http://www-nsidc.colorado.edu/NASA/YEARBOOK_1995/edge.html

[9] Douglas M. Smith, 1998. "Recent increase in the length of the melt season of perennial Arctic sea ice", Geophysical Research Letters 25(5):655-658.

[10] Miles G. McPhee, Tim Stanton, Jamie Morison, and Douglas Martinson, 1997. "Freshening of the upper ocean in the Central Arctic: Is perennial sea ice disappearing?", available on the Internet at http://sheba.apl.washington.edu/whatshot/whatshot.html

[11] John L. Newton and Barbara J. Sotirin, "Boundary undercurrent and water mass changes in the Lincoln Sea", J. Geophys. Res. V. 102, n. C2, pp. 3393-3403, 1997.

[12] Martin O. Jeffries, "Arctic ice shelves and ice islands: Origin, growth and disintegration, physical characteristics, structural stratigraphic variability and dynamics", Reviews of Geophysics 30(3):245-267.

[13] J. Overpeck, K. Hughen, D. Hardy, R. Bradley, R. Case, M. Douglas, B. Finney, K. Gajewski, G. Jacoby, A. Jennings, S. Lamoureaux, A. Lasca, G. MacDonald, J. Moore, M. Retelle, S. Smith, A. Wolfe, G. Zielinski, 1997. "Arctic environmental change of the last four centuries", Science 278:1251-1256

[14] Chapman and Walsh 1993, op.cit.

[15] Chapman and Walsh 1993, ibid.

[16] Warren M. Washington and Gerald A. Meehl, 1996. "High-latitude climate change in a global coupled ocean-atmosphere-sea ice model with increased atmospheric CO2", J. Geophys. Res., v. 101, n. D8, pp. 12,795-12,801.

[17] Cynthia T. Tynan and Douglas P. DeMaster, 1997. "Observations and predictions of Arctic climate change: potential effects on marine mammals", Arctic 50(4):308-322.

[18] I. Melnikov, "Preliminary biological and chemical oceanographic evidence for a long-term warming trend in the Arctic Ocean (current materials of the SHEBA Ice Camp, Beaufort Sea)", abstract, in James Morison, Knut Aagaard, and Michael Steele (eds.), Draft Report on Study of the Arctic Change Workshop, November 10-12, 1997, Polar Science Center, Applied Physics Laboratory, University of Washington, Seattle, Washington, Appendix D, pp. 21-22

[19] D.T. Shindell, D. Rind, and P. Lonergan, 1998. "Increased polar stratospheric ozone losses and delayed eventual recovery owing to increasing greenhouse-gas concentrations", Nature 392:589-592.

[20] Vera Alexander, 1992. "Arctic Marine Ecosystems", in Robert L. Peters and Thomas E. Lovejoy, Global Warming and Biological Diversity, Yale University Press, New Haven, pp. 221-232.

[21] Tynan and DeMaster 1997, p. 316

[22] Vera Alexander 1992, and Ian Stirling and Andrew E. Derocher, Possible impacts of climatic warming on polar bears, Arctic 46(3):240-245, 1993.

[23] Stirling and Derocher 1993

[24] Mark C. Serreze, James A. Maslanik, Jeffrey R. Key, and Raymond F. Kokaly, "Diagnosis of the record minimum in Arctic sea ice during 1990 and associated snow cover extremes", Geophys. Res. Let. 22(16):2183-2186, 1995.

[25] BBC Wildlife, "Ursus ubiquitous", December 1992

[26] Stirling and Derocher, 1993

[27] Alexander 1992

[28] Rosemarie Kuptana, 1996. Speech written for the Second Conference of the Parties to the U.N. Framework Convention on Climate Change, Geneva, Switzerland, July 16-19.

[29] Arctic Monitoring and Assessment Program (AMAP), 1997. Arctic Pollution Issues: A State of the Environment Report, Oslo.

[30] Washington and Meehl 1996, op.cit., p. 12,796

[31] C. Bernes 1996. The Nordic Arctic Environment - Unspoilt, Exploited, Polluted? Report Nord 1996: 26, Nordic Council of Ministers, Copenhagen; AMAP 1997; A.J. Weaver 1993. "The oceans and global warming", Nature 364:192-193

[32] Cecilie Mauritzen and Sirpa H�kkinen, 1997. "Influence of sea ice on thermohaline circulation in the Arctic-North Atlantic Ocean", Geophys. Res. Lett. 24(24):3257-3260.

[33] For a layman's examination of the cooling and freshening of the Labrador Sea and its implications for global climate, see Myron Arms, Riddle of the Ice: A Scientific Adventure into the Arctic, Anchor Books, Doubleday, New York, 1998.

[34] Wallace S. Broecker, "Thermohaline circulation: the Achilles Heel of our climate system: Will man-made CO2 upset the current balance?", Science 278:1582-1588, November 1997

[35] Andrew J. Weaver, 1995. "Driving the ocean conveyor", Nature 378:135-136; Stefan Rahmstorf, 1995."Bifurcations of the Atlantic thermohaline circulation in response to changes in the hydrological cycle", Nature 378:145-149.

[36] Rahmstorf 1995

[37] R.A. Warrick, et.al., 1996. "Changes in sea level", Table 7.1, p. 372, in J.T. Houghton, et.al. (eds.), Climate Change 1995: The Science of Climate Change, IPCC Second Assessment Report, Cambridge University Press.

[38] Syukuro Manabe and Ronald .J. Stouffer, 1993. "Century-scale effects of increased atmospheric CO2 on the ocean-atmosphere system", Nature 364: 215-218.

[39] Stefan Rahmstorf, 1997. "Risk of sea-change in the Atlantic", Nature 388:825-826; Thomas F. Stocker and Andreas Schmittner, 1997. "Influence of CO2 emission rates on the stability of the thermohaline circulation", Nature 388:862-865.

[40] Bering Sea Impact Study (BESIS) 1997. The Impacts Of Global Climate Change In The Bering Sea Region:

An Assessment Conducted By The International Arctic Science Committee Under Its

Bering Sea Impacts Study (BESIS), Girdwood, Alaska, 18-21 September 1996

[41] AMAP 1997

[42] Rahmstorf 1997, ibid.

[43] Ibid.