• Arctic Sea Ice
  • The Past 100 Years
  • Earth's Orbital Variations
  • Abrupt loss of Arctic sea ice
  • Polar ice cap gone by 2030?
  • Why is the Arctic sea ice is shrinking?
  • The Implications
• References
• Further Information
• Related Blogs

Figure 1. Annual average change in surface air temperature from land weather stations, relative to the average for 1961-1990, for the region from 60 to 90° north. Image credit: The Arctic Climate Impacts Assessment (ACIA).

The Arctic is a region particularly sensitive to climate change, since temperatures are, on average, near the freezing point of water. Slight shifts in the average temperature can greatly change the amount of ice and snow cover in the region. For example, as sea ice melts in response to rising temperatures, more of the dark ocean is exposed, allowing it to absorb more of the sun's energy. This further increases ocean temperatures, air temperatures, and ice melt in a process know as the "ice-albedo feedback" (albedo means how much sunlight a surface reflects). This is a positive "feedback", meaning that once the warming and melting starts, the conditions change in a way to amplify the warming and the melting. Summertime Arctic sea ice extent has fallen 39% between 1979 and 2007, and up to 50% since the 1950s. Have we reached the "tipping point" for Arctic sea ice? Has an irreversible ice-albedo feedback taken hold and will amplify until the Arctic Ocean is entirely ice-free later this century? To answer these questions, it is helpful to look at the past and see if similar warmth and ice melt have occurred.

The Past 100 Years

The Arctic Climate Impact Assessment (ACIA), published in November 2004, was a uniquely detailed study of Arctic climate compiled by 300 scientists over three years. The study found that while temperatures in the Arctic have increased significantly since 1980 (Figure 1), there was also a period in the 1930s and 1940s when temperatures were almost as warm. If one defines the Arctic as lying poleward of 62.5° north latitude (Polyakov, 2003), the 1930s and 1940s were the warmest period in the past 100 years. Looking at Figure 1, one cannot dismiss the possibility that temperatures in the Arctic oscillate in a 50-year period, and we are due for a cooling trend that will take temperatures below normal by 2030.

There is an important difference between the earlier warming and today. Since 1980, with the exception of Antarctica, the entire globe warmed. This was not true in the 1930s and 1940s. If regional warming is due to an atmospheric oscillation, it is likely that there will be other regions of cooling. If there are not regions of cooling, then it is more likely that the warming trend is due to global warming. Thus, the 1930s and 1940s warming in the Arctic is thought to be fundamentally different from the warming of the past 20 years. Furthermore, the past 20 consecutive years have all been above normal in temperature, whereas during the 1930s and 1940s there were a few cooler than average years interspersed with the very warm years. A detailed breakdown by month and region of the 100-year history of Arctic temperatures was performed by Overland et al. (2004). They found no evidence of a natural 50-year warming and cooling cycle in Arctic temperatures, and concluded that the warming since 1980 was unique. However, they stopped short of blaming the recent warming on human-emitted greenhouse gases (anthropogenic forcing). The ACIA, though, concluded that humans were likely to blame for the recent Arctic warming:

Figure 2. Summertime temperatures in the Arctic during the Mid-Holocene Warm Period (about 6,000 years ago), compared to today's temperatures. Image credit: NOAA.

"It is suggested strongly that whereas the earlier warming was natural internal climate-system variability, the recent surface air temperature changes are a response to anthropogenic forcing. There is still need for further study before it can be firmly concluded that the increase in Arctic temperatures over the past century and/or past few decades is due to anthropogenic forcing."

Critics of the ACIA report have pointed out that the Arctic was much warmer than today during the period 4,000-7,000 years ago. It is true that Arctic summers were warmer during the period 4,000-7,000 years ago (Figure 2). The mean July temperature along the northern coastline of Russia may have been 2.5 to 7.0° C warmer than present, and Scandinavian summer temperatures were 1.5 ° C higher than at present. The warming was caused by changes in the amount of sunlight the North Pole gets in summer due to variations in the Earth's orbit.

Earth's orbital variations

The Earth is tilted 23.5° on its axis, which is responsible for our seasons. Also, the Earth's orbit is not perfectly circular. We are closest to the sun in December and farthest in July. Thus, the Southern Hemisphere gets more sunlight in their summer (December-February) than the Northern Hemisphere does in their summer (June-August). The Earth slowly wobbles around its axis with a period of 23,000 years, which is called precession. You can demonstrate precession by studying the motion of a toy top.

Currently, the Earth is oriented so that the North Pole is pointed towards the Sun in the wintertime. Because of the precession, 4,000-7,000 years ago, the North Pole was pointed towards the Sun during the summertime. This resulted in warmer summer temperatures in the Arctic, because there was more intense summer sunlight. Also, winter temperatures were colder, since the Earth was farther from the sun during the Northern Hemisphere winter. The Arctic will again be warmer in summer 16,000 to 19,000 years from now when the cycle repeats. The ACIA examined the average temperature in the Arctic. The average remained about the same during the period 4,000-7,000 years ago; the summers warmed and the winters cooled. The warming since 1980 and the 1930s was observed over all seasons. The use of the Arctic temperatures 4,000-7,000 years ago as an analogue to evaluate with these more recent warm periods is not a consistent comparison.

Abrupt loss of Arctic sea ice

Figure 3. Average September Arctic sea ice coverage as observed by satellites between 1979 and 2007. Image credit: National Snow and Ice Data Center.

While very warm temperatures were also seen in the Arctic in the 1930s, the loss of Arctic sea ice observed in the past 20 years has no precedent in the historical record. For example, the fabled Northwest Passage opened in 2007, an event that has not occurred since at least 1497, and probably for a much longer span of time. Between 1979 (the year satellite imagery of the North Pole first became available) and 2006, Arctic sea ice extent shrunk by about 10% in winter (4% per decade) and 20% in summer (8% per decade). The loss of sea ice, when plotted on a graph (Figure 3), roughly followed a straight line over time. There were a few noisy ups and downs, reflecting colder and warmer years. A trend that approximately follows a straight line is called a "linear" trend. A continued linear summertime 8% per decade loss of sea ice would leave the summertime Arctic Ocean ice-free by 2100. The ocean would still partially freeze in winter, with about 50% of the ocean covered with ice.

In 2007, the Arctic sea ice did not follow the straight line, and dropped much more than suggested by the linear trend. Sea ice loss, which had declined 20% in summer from 1979 to 2006, doubled to 39% in 2007, according to the National Snow and Ice Data Center. In one year, as much ice was lost as in the previous 28 years! Sea ice extent for July, August, and September decreased 20%, 31%, and 39% compared to the averages for those months from 1979-2000. The minimum on September 15, 2007 was 23% less than the previous record set on September 20-21, 2005. Summertime Arctic sea ice extent in 2007 was 50% of values measured in the 1950s (Figure 4). One may look at the graph and wonder, but what about sea ice loss in other seasons? It hasn't been nearly so severe. The summer ice is, however, most important, because if the ice disappears in the summer when the Sun is shining, the ocean can absorb the summer heat. This sets up the conditions in the Arctic for less ice formation in the autumn, and hence, greater ice loss in future years. It's the initiation of the positive ice-albedo feedback.

Figure 4. Arctic sea ice extent since 1900, as estimated from satellite and ship reports compiled by Walsh and Chapman (2001). Image credit: University of Illinois cryosphere group.

The possibility of amplified sea ice loss was discussed in a December 2006 paper (Holland et al.), titled "Future abrupt reductions in the summer Arctic sea ice". The authors ran the Community Climate System Model, one of the top climate models used to formulate the "official word" on climate, the 2007 Intergovernmental Panel on Climate Change (IPCC) report. The model represented the years 1979-2006, and successfully predicted the 20% loss of summer sea ice during that period. The model then assumed that levels of greenhouse gases would continue to increase, until a doubling of CO2 levels occurred in 2100. This is considered a "middle-of-the-road" scenario, and assumes a sequence of events will unfold over the coming decades: humans will make some modest efforts to control greenhouse emissions, but not enough to prevent dangerous climate change. The model found that Arctic sea ice continued to decline linearly until about 2024, resulting in about 60% sea ice coverage in September (Figure 5). During this period, the vertical thickness of the sea ice declined from about four meters to one meter. Beginning in 2025, the rate of sea ice loss suddenly tripled, resulting in the total loss of the summertime polar sea ice by 2040. The authors theorized that once ice reaches a critical thickness--in this case, one meter--the processes that create open water suddenly become more efficient, resulting in a rapid disintegration of the remaining ice. The authors concluded that "abrupt changes in the summer Arctic sea ice cover are quite likely and can occur early in the 21st century, with the earliest event in approximately 2015".

Polar ice cap gone by 2030?

The prevailing view among climate scientists had been that an ice-free Arctic ocean would occur in the 2070-2100 time frame. The February 2007 report from the U.N.-sponsored Intergovernmental Panel on Climate Change (IPCC), warned that without drastic changes in greenhouse gas emissions, Arctic sea ice will "almost entirely" disappear by the end of the century. The recent observations and the Holland et al. model study suggest that it is conceivable that a complete loss of summer Arctic sea ice will occur far earlier. In a 2007 interview published in The Guardian, Dr. Mark Serreze, an Arctic ice expert with the National Snow and Ice Data Center, said: "If you asked me a couple of years ago when the Arctic could lose all of its ice, then I would have said 2100, or 2070 maybe. But now I think that 2030 is a reasonable estimate. It seems that the Arctic is going to be a very different place within our lifetimes, and certainly within our children's lifetimes." While natural fluctuations in wind, ocean circulation, and temperatures are partly to blame for this loss of sea ice, human-caused global warming is also to blame. In the words of Dr. Serreze: "The rules are starting to change and what's changing the rules is the input of greenhouse gases. This year puts the exclamation mark on a series of record lows that tell us something is happening."

Figure 5. September Arctic sea ice extent observed in 1979 (yellow line), and 2005 (white area). The predicted coverage of sea ice by Holland et al. (2006) is shown for 2015 (red line) and 2040 (green line). Their model predicts that sea ice in summer by 2040 will occur only in narrow bands along the Canadian Arctic coast. However, there will still be about 50% sea ice coverage in winter. Original image taken from NASA.

Some argue that the process of achieving both consensus and rigor in the IPCC report yields a "conservative" estimate of climate change. It is true that predictions which involve phase changes are among the most difficult for climate models. This is made even more challenging for sea ice, which sits in water and is subject to amplified melting by stirring in the water, and is also sensitive to the local salinity of the water. If there are to be surprises in the predictions of climate change, then they are likely to involve phase changes. In a warming climate, this would involve the transition of water from ice to liquid.

Why is the Arctic sea ice is shrinking?

Warmer air and water temperatures due to global warming
Annual average surface temperature has increased about 1° C since 1980 over the Arctic, which accounts for much of the sea ice melt. In addition, some melting has occurred from beneath the ice, due to warmer ocean waters. Global warming has heated up both the North Pacific and North Atlantic waters significantly over the past 30 years. Warmer waters have been brought into the Arctic Ocean from the Pacific via an ocean current flowing through the Bering Strait between Alaska and Russia, and from the Atlantic via an ocean current flowing northwards along the European coast.

Long-term natural temperature oscillations
Long-term natural temperature oscillations in the Arctic may be contributing to the current sea ice decline. Divine and Dick (2006) studied records of the Arctic sea ice edge from 1750 to present, taken from shipping records in the Nordic Seas north of Europe. They found that a natural cycle with a period of 60-80 years appeared to be responsible for an observed advance and retreat in the sea ice, and that this natural cycle might be near its minimum point in 2005. They also noted that the current sea ice extent in the Nordic Seas is the lowest since reliable records began in 1850.

Natural wind patterns
The Arctic Oscillation is an observed natural pattern of surface pressure variations in the Northern Hemisphere. The "positive index" of the AO is defined when the surface pressure is below normal at the North Pole and above normal at about 45° north latitude. Positive Arctic Oscillation conditions steer storms farther north, bringing stronger surface westerly winds in the North Atlantic and warmer and wetter than normal conditions to the Arctic and northern Europe. The winds and ocean currents during the positive Arctic Oscillation mode tend to drive sea ice from west to east along the north shore of Canada, then out of the Arctic Ocean through the channel of water to the east of Greenland (Fram Strait).

Figure 6. The Arctic Oscillation (AO) index from 1899 - 2006. The AO is a measure of the difference in surface pressure between the North Pole and about 45° north latitude. Image credit: Dave Thompson of Colorado State University.

When one looks at the wintertime pattern of the Arctic Oscillation (AO) over the past 100 years, a mostly random pattern of positive and negative AO modes is apparent (Figure 6). However, one anomalous period is very striking: a string of seven consecutive years with a positive AO, including two years (1989 and 1990) with the highest AO index ever observed. During this period, strong westerly winds rapidly flushed more than 80% of the oldest, thickest sea ice out of the Arctic Ocean, leaving most of the Arctic covered with ice less than three years old (Figure 7). Younger ice is much thinner, and melts much more readily. Rigor and Wallace (2004) estimate that at least half of the loss of sea ice in the Arctic since 1979 is due to these six years of strange weather with very low surface pressure over the Arctic. Did climate change cause this unusual pattern between 1989 and 1995? It is possible, but no one has published any papers showing how this might have occurred. For now, the assumption is that this major loss of Arctic sea ice due to wind patterns between 1989-1995 is natural.

Since the strange positive Arctic Oscillation years of 1989-1995, a number of years with negative AO have occurred. Normally, during negative AO years, ice extent and thickness increase in the Arctic. But instead, ice extent and thickness during these periods has continued to decline.

The monthly AO index also exhibited mostly positive values during September to November 2005 and March 2006 to March 2007, enhancing ice transport out of the Arctic. In September 2005, a pronounced atmospheric low pressure over the Barents Sea, in concert with a strong high pressure over the Canadian Basin, set up unusually strong northerly winds over Fram Strait, pushing ice more than a year old (perennial ice) that had accumulated north of Greenland out of the Arctic like a runaway train. Between March 2005 and March 2007, the amount of perennial ice fell by 23% (Nghiem et al., 2007), and perennial ice in the Arctic is now half of what it was in the 1950s. Seasonal ice (ice that is less than a year old), is thinner, more easily compressed, and can be pushed by wind out of the Arctic much more readily. It also melts more readily, since it is thinner. When an unusually strong high pressure system settled over the North Pole during the summer of 2007, the lower than average cloud cover that resulted allowed the sun to drastically melt the relatively thin ice that decades of gradual melting had created. The ice will re-freeze in winter, but this new seasonal ice reflects less sunlight than perennial ice. When summer returns, this will lead to even warmer temperatures in the Arctic. It is difficult to see how the Arctic will recover from 2007's drastic loss of sea ice, and an ice-free Arctic in summertime is possible by 2013, according to Arctic research scientist Dr. Wieslaw Maslowski of the Naval Postgraduate School in Monterey, California.

The implications

Direct effect on sea level rise
The melting of the Arctic sea ice will not raise ocean levels appreciably, since the ice is already floating in the ocean. Sea ice melting does slightly contribute to sea level rise since the fresh melt water is less dense than the salty ocean water is displaces. According to the sea level FAQ of Dr. Robert Grumbine of NOAA's sea ice group, if all the world's sea ice melted, it would contribute to about 4 millimeters of global sea level rise. This is a tiny figure compared to the 20 feet of sea level rise locked up in the ice of the Greenland ice sheet, which is on land.

Figure 7. The change in age and thickness of sea ice between 1987 and 2005. In 1987, most of the Arctic sea ice was old and thick, generally more than ten years old. A period of strong positive Arctic Oscillation conditions between 1989 and 1995 created winds and currents that flushed most of this old ice out of the Arctic Ocean, through Fram Strait to the east of Greenland. The new ice that replaced the old ice is much thinner. Image credit: Rigor, I. G., and J. M. Wallace (2004), "Variations in the age of Arctic sea-ice and summer sea-ice extent," Geophys. Res. Lett., 31, L09401, doi:10.1029/2004GL019492.

Indirect effect on sea level rise
The biggest concern about Arctic sea ice loss is the warmer average temperatures it will bring to the Arctic in coming years. Warmer temperatures will accelerate the melting of the Greenland ice sheet, which holds enough water to raise sea level 20 feet. 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 will probably need to be revised upwards in light of the unexpectedly high sea ice loss in the Arctic.

Effect on the jet stream and planetary weather patterns
Continued loss of Arctic sea ice may dramatically change global weather and precipitation patterns in the decades to come. The jet stream will probably move further north in response to warmer temperatures over the pole, which will bring more precipitation to the Arctic. More frequent and intense droughts over the U.S. and other regions of the mid-latitudes may result from this shift in the jet stream. Changes to the course of the jet stream affect weather patterns for the entire planet, and we can expect impacts on the strength of the monsoons and recurvature likelihood of hurricanes. Exactly what these effects might be and what regions would be most strongly affected are difficult to predict. In any case, the reduced Arctic sea ice should give most of the Northern Hemisphere a delayed start to winter this year, and for the forseeable future.

Coastal damage in the Arctic
More open water in the Arctic Ocean allows erosion due to wave action to affect the coast for longer periods, particularly during fall, when storms tend to be stronger with higher storm surges. The resulting destruction has already forced residents of the Alaskan town of Shishmaref to vote to abandon their village. More than half the residents of the nearby village of Kivalina were forced to evacuate on September 13 2007, when 25-40 mph winds drove a 3-4 foot high storm surge into the town. The U.S. Army Corps of Engineers built a $3 million sea wall to protect the town, but the wall has not been able to hold against recent storms. Over 100 feet of coastline has been lost in the past three years.

More open water also means more moisture and heat will be available to power storms. These stronger storms will bringer higher winds and higher storm surges to coastal areas in the Arctic over the remainder of the 21st century, resulting in increased erosion and flooding of low-lying areas.

A final note on model uncertainty
Global warming skeptics often criticize using computer model climate predictions as a basis for policy decisions, saying that these models are too uncertain. However, the uncertainty goes both way--sometimes the models will underestimate climate change. This was also the case when the ozone hole opened up--another case where the models failed to predict a major climate change. The atmosphere is not the well-behaved, predictable entity the models try to approximate it as. The atmosphere is wild, chaotic, incredibly complex, and prone to sudden unexpected shifts. By pumping large amounts of greenhouse gases into the air, the climate has been destabilized. We can expect many more surprises in the coming decades that the models have not predicted.

Figure 8. Coastal erosion due to less sea ice has had dramatic effects on the Alaskan village of Shishmaref. Image credit: American Association for the Advancement of Science.

References

Divine, D.V., and C. Dick, "Historical variability of sea ice edge position in the Nordic Seas", JGR Oceans 111, C01001, doi:10.1029/2004JC002851, 2006.

Holland, M.M., C.M. Bitz, B. Tremblay, 2006 "Future abrupt reductions in the summer Arctic sea ice", Geophysical Research Letters, 33, L23503, December 2006.

Holloway, G. and T. Sou, 2001, "Has Arctic Sea Ice Rapidly Thinned?", Journal of Climate 15, p1691-1701, 2001.

Nghiem, S. V., I. G. Rigor, D. K. Perovich, P. Clemente-Colón, J. W. Weatherly, and G. Neumann (2007), "Rapid reduction of Arctic perennial sea ice", Geophys. Res. Lett., 34, L19504, doi:10.1029/2007GL031138.

Overland, J.E, M.C. Spillane, D.B. Percival, M. Wang, H.O. Mofjeld (2004), "Seasonal and Regional Variation of Pan-Arctic Surface Air Temperature over the Instrumental Record", Journal of Climate, 17:17, pp3263-3282, September 2004.

Polyakov, V., et al. (2003), "Variability and Trends of Air Temperature and Pressure in the Maritime Arctic, 1875-2000", Journal of Climate, 16, 2067-2077.

Rigor, I. G., and J. M. Wallace (2004), "Variations in the age of Arctic sea-ice and summer sea-ice extent," Geophys. Res. Lett., 31, L09401, doi:10.1029/2004GL019492.

Rothrock D.A., Y. Yu, and G.A. Maykut, 1999: "Thinning of the Arctic sea ice cover." Geophys. Res. Lett., 26, 3469-3472.

Walsh, J.E and W.L.Chapman, 2001, "Twentieth-century sea ice variations from observational data", Annals of Glaciology, 33, Number 1, January 2001 , pp. 444-448.

Sources of Further Information

National Snow and Ice Data Center
The Cryosphere Today at the University of Illinois
Daily AMSR-E sea ice maps from the University of Bremen
RealClimate.org
NASA has posted a beautiful satellite image of the Arctic ice cap at the September 15 2007 minimum, showing the open water of the Northwest Passage.