The Arctic heats up

NCAR researchers study climate change in Earth’s northern realms

The Arctic is on thin ice—literally as well as figuratively. New research shows that tempera­tures were warmer there in the 1990s than any decade in at least 2,000 years. Arctic sea ice has been dwindling for the past few decades, with 2009 levels well below normal. Greenland's melting ice sheets have the potential to dramatically raise sea levels, and thawing permafrost threatens to release massive amounts of methane into the atmosphere.


Sea ice near Barrow, Alaska.Sea ice on the Chukchi Sea near Barrow, Alaska.

Although many UCAR/NCAR labs and programs touch on the Arctic in their research, the core of the organization's Arctic research is in ESSL/CGD. Researchers there are observing, measuring, and modeling the Arctic's changing snow, ice, land, and water, in hopes of understanding what the warm 21st century has in store for the top of the world.

Arctic sea ice outlook

Atmospheric circulation patterns in August helped spread out sea ice, slowing ice loss in most regions of the Arctic, according to the National Snow and Ice Data Center. Even so, sea ice levels for 2009 were well below average once more. The two years with the greatest summer loss of Arctic ice in the 30-year-long satellite record are 2007 (first place) and 2008 (second place). By early September of this year, sea ice extent had dropped below the minimum for 2005 (which was previously in third place), and melting was expected to continue until late in the month.

With the Arctic Ocean becoming more accessible, there is increasing interest from the shipping indus­try, oil and gas companies, northern governments, and the polar community in general in short-term predictions of sea ice levels.

In CGD, Marika Holland is researching how to most effectively forecast sea ice on a short-term, seasonal timescale in the Arctic's changing environment. "There are seasonal forecasting methods in place for Arctic sea ice, but it's questionable whether they will continue to work in the Arctic regime we're now experiencing," Marika says.

These seasonal forecasting methods have tra­ditionally been based on statistical methods that look at historical trends. In a new study, Marika and colleagues used NCAR's Community Climate System Model (CCSM) to assess the inherent predictability of sea ice, and whether this predictability will change as the Arctic warms. "We wanted to take a step back," Marika says. "Should we even expect that, given knowledge of January conditions, we can predict sum­mertime conditions?"

Their results suggest that climate change does influence predictability and that forecasters will likely need to modify the tools they use for seasonal fore­casts, relying more on physically-based models rather than statistical methods.

Sea ice loss and future climate

Also in CGD, a team led by Clara Deser has used an atmospheric general circulation model to explore how projected losses in Arctic sea ice may affect climate. The results were published in the Journal of Climate in September. 

Polar bear tracks in snow.Polar bear tracks in crystalline structures known as frost flowers near Barrow, Alaska, during the OASIS field project in March 2009.

The goal of the study was to isolate the direct impacts of Arctic sea ice loss without feedbacks from other components of the climate system, notably the oceans. Scientists in the United Kingdom and Norway have undertaken similar efforts, but the CGD study is the first to apply a high-resolution, integrated model to this research question.

Running experiments with the CCSM coupled to the Community Land Model (CAM), the researchers found that loss of summer Arctic sea ice will result in considerable (more than 9°F, or 5°C) warming by 2100 over adjacent land regions north of 65° latitude, particularly during autumn and winter. The modeling experiment also forecasts increased snowfall over land, especially over Siberia and northern Canada, as ice loss sends moisture from the ocean into the atmosphere.

Another important finding is that although sea ice loss peaks in September, the atmosphere's full response is delayed until November or December, according to the model. "This is because the ocean is more efficient at losing heat to the atmosphere in winter when air temperatures are lowest and winds strongest," Clara explains.

The researchers' next step is to incorporate an interactive ocean component into its atmospheric model to study how ocean temperatures, salinity, and currents respond to sea ice loss. This new version of the model will be used for the next IPCC assessment, due out in 2013.

Greenland's ice sheets and sea level rise

It's not just sea ice that's melting. New research led by Aixue Hu in CGD looks at Greenland's melting ice sheets, focusing on how they may raise sea levels.

Aixue Hu. Aixue Hu.

Aixue and his team used the CCSM to study differ­ent scenarios that depend on how fast the island's ice melts in the coming century. Their results, published in Geophysical Research Letters last May, conclude that the island's melting ice sheets may drive more water than previously thought this century toward the northeastern coasts of the United States and Canada, threatening coastal cities such as Halifax, Boston, and New York.

"If the Greenland melt continues to accelerate, we could see significant impacts this century from the resulting sea level rise," Aixue says. "Major northeast­ern cities are directly in the path of the greatest rise."

Map showing sea level rise from melting Greenlandic ice sheets.This image, based on new computer modeling, shows that sea level rise may be an additional 4 inches (10 centimeters) higher by populated areas in northeastern North America than previously thought. Extreme northeastern North America and Greenland may experience even higher sea level rise. (Image courtesy Geophysical Research Letters, modified by UCAR.) [ENLARGE]

According to the study, if Greenland's ice melts at moderate to high rates, ocean circulation by 2100 may shift and cause sea levels off the northeast coast of North America to rise by about 12 to 20 inches (about 30 to 50 centimeters) more than in other coastal areas. The research builds on recent reports that have found that sea level rise associated with global warming could adversely affect North America, and its findings suggest that the situation is more threatening than previously believed.

  Unlike water in a bathtub, water in the oceans does not spread out evenly. The northeast coast of North America is especially vulnerable to the effects of Greenland's ice melt because of circulations that transport warm water from the tropical Atlantic Ocean northward, where it cools and descends to cre­ate a dense layer of cold water. As a result, sea level is currently about 28 inches (71 cm) lower in the North Atlantic than the North Pacific, which lacks such a dense layer. (See the debut issue of UCAR Magazine for more on regional variations in sea-level rise.)

A 2,000-year Arctic climate record

Temperatures in the Arctic were warmer in the 1990s than any decade in at least 2,000 years, accord­ing to research by CGD's Dave Schneider published in Science in early September. The study, which incor­porates geologic records and computer simulations, provides new evidence that the Arctic would be cool­ing were it not for greenhouse gas emissions that are overpowering natural climate patterns.

The study was led by Northern Arizona Univer­sity's Darrell Kaufman, who collaborated with CGD's paleoclimate group, where Dave recently finished his postdoctoral appointment, to develop a syn­thesis of Arctic climate records and compare them with model simulations. The researchers used data from sediments in Arctic lakes, glacial ice, and tree rings to reconstruct temperatures for the last 2,000 years, extending far beyond the 400 years of Arctic temperature records previously available at that level of detail.

Graph showing Arctic temperatures.Recent research from CGD’s paleoclimate group shows that the Arctic reversed a long-term summer cooling trend and began warming rapidly in recent decades. The blue line shows estimates of summertime Arctic land temperatures over the last 2,000 years, based on proxy records from lake sediments, ice cores, and tree rings. The green line shows the long-term cooling trend. The red line shows the recent warming based on actual observations. (Image courtesy Science, modified by UCAR.) [ENLARGE]

The study is the first to quantify a pervasive cool­ing across the Arctic on a decade-by-decade basis related to a cyclic wobble in Earth's tilt relative to the Sun. Over the last 7,000 years, the timing of Earth's closest pass by the Sun has shifted from September to January. This has gradually reduced the intensity of sunlight reaching the Arctic in summertime, as Earth is farther from the Sun.

The analysis shows that summer temperatures in the Arctic cooled at an average rate of about 0.36°F (0.2°C) per thousand years, in step with the reduced energy from the Sun. The orbital cycle that produced the cooling continues today, but was overwhelmed in the 20th century by human-induced warming, resulting in summer temperatures in the Arctic by the year 2000 that were about 2.5°F (1.4°C) higher than would have been expected from the continued cyclic cooling.

David SchneiderDavid Schneider.

"This study provides us with a long-term record that reveals how greenhouse gases from human ac­tivities are overwhelming the Arctic's natural climate system," Dave says. "It's particularly important be­cause the Arctic, perhaps more than any other region on Earth, is facing dramatic impacts from climate change."

The impermanence of permafrost

A study published in June 2008 by CGD's Dave Lawrence, working with colleagues at the National Snow and Ice Data Center, showed that the rate of climate warming over northern Alaska, Canada, and Russia could more than triple during periods of rapid sea ice loss. The research raised concerns about thawing permafrost (permanently frozen soil), which releases carbon as well as methane, a greenhouse gas that is 23 times more effective than carbon dioxide at trapping heat in the atmosphere.

Dave and postdoctoral researcher Sean Swenson are currently laying groundwork to study in more depth how the carbon cycle will respond to Arctic warming and permafrost thaw. They're developing a dynamic wetlands scheme to model how changes driven by terrain, climate, and permafrost thaw con­trol the distribution and extent of wetlands.

Such a model is critical because one of the keys to predicting the carbon cycle's response to thawing permafrost is predicting how the changing climate and thawing permafrost will affect the Arctic's hydrol­ogy. When dead plant materials break down under wet soil conditions, methane is produced, whereas in dry conditions, enhanced decomposition tends to pro­duce carbon dioxide.

"Since methane is a much stronger greenhouse gas, you will see a much different climate impact if the Arctic land surface becomes wetter, either due to increased precipitation or changing landforms due to permafrost thaw, than if it becomes drier," Dave says. Mapping these regional variations in precipitation and permafrost thaw will be critical to predicting the future of the Arctic's climate cycle.


Polar amplification:
Why the Arctic warms quicker

The reason that Earth's far northern latitudes are so vulner­able to climate change is a phenomenon called polar amplifica­tion, which causes the polar regions, particularly the North Pole, to experience enhanced warming compared to other parts of the globe. By some estimates, the Arctic is warming twice as fast as the rest of the world.

A number of factors contribute to polar amplification, one of the most important being the feedback effect caused by ice albedo. Polar ice has a high albedo, meaning that it reflects more incoming solar radiation than darker land or water surfaces. As ice melts in the Arctic, the exposed land and water absorb radiation, in turn amplifying the warming effect. Other influences, such as natural variability in Arctic temperatures and the transport of heat to the Arctic via the ocean, may also play a role in polar amplification.