Runaway heat? A darkening Arctic may accelerate warming trends there.

Slide into the black vinyl seat of a car that s been parked for hours on a sunny summer day, and your exposed skin will be assaulted by heat. On an 80[degrees]F (27[degrees]C) day, that dark vinyl can reach a scorching 180[degrees]F (82[degrees]C). If the material were white instead of black, it would reflect more of the sunlight, and the seat would be tens of degrees cooler. Now, imagine painting an entire region of the world in blacks instead of whites, and you'll understand climate-change researchers' concern that the dark-vinyl effect may be taking hold in the Arctic.

Increasing concentrations of greenhouse gases such as carbon dioxide in the atmosphere are making the world warmer, on average, most scientists say. However, new research suggests that in some localities an even greater effect stems from changes in the percentage of light that's reflected from Earth's surface--a parameter called albedo.

Because the average temperature at many Arctic locales hovers at or near the freezing point of water, even a small increase in local warming can have big consequences. Among the effects that scientists have recorded in the Arctic are earlier snowmelt in the spring, the spread of shrubs into tundra areas once hospitable only for grasses, and the dwindling of sea-ice coverage in the summer. Each of these changes makes the Arctic a darker place. As the darker materials absorb more solar radiation than lighter ones do, they increase local warming, so these changes in albedo tend to be self-amplifying.

Such cycles of change have occurred in the Arctic after past ice ages--around a dozen during the past 1 million years. But the trends appear to be moving much faster today than they did in previous eras. They're also occurring beneath an atmosphere that holds more planet-warming carbon dioxide than formerly.

This conspiracy of circumstances might elicit changes in the Arctic that some scientists fear could go well beyond those seen in recent interglacial periods and drive the region into a widespread meltdown by the end of this century. Such an ice-free state hasn't occurred in the Arctic for millions of years.

GROWING HEAT Average temperatures worldwide have risen about 0.6[degrees]C in the past 100 years. In some regions, temperatures have warmed even faster. For example, in the Arctic portions of Alaska and western Canada, average summer temperatures have increased about 1.4[degrees]C just since 1961. Other Arctic regions, including Greenland, Scandinavia, and northern Russia, show similar accelerated warming trends.

Albedo changes appear to be the primary cause of the recent warming, at least in Alaska. For one thing, a pronounced summertime warming in that state's interior doesn't seem to be correlated with changes in weather patterns that steer storm systems, says Matthew Sturm, a geophysicist at the U.S. Army Cold Regions Research and Engineering Laboratory in Fort Wainwright, Alaska. Although short-term variations in climate caused by the El Nino--Southern Oscillation and the longer-term influences from the Pacific Decadal Oscillation affect Alaska's winter temperatures, the summertime effect of those phenomena is weak at best.

A decrease in summer sea ice north of Alaska in recent decades can't explain the warming measured in Alaska's interior. Warming trends caused by declines in summer sea ice would also show up in autumn and winter, seasons when the heat gained in summer would De released to the atmosphere. But temperatures during those seasons haven't warmed significantly, says Sturm.

Alaska's summer warming is best explained by an increase in the length of the state's snowfree season, Sturm and his colleagues report in the Oct. 28 Science. Since the 1960s, the end of the snowmelt in Barrow, on the coast of the Arctic Ocean, has come 1.3 days earlier, on average, with each passing decade. During the same period, snowmelt inland in the northern foothills of the Brooks Range has ended about 3.6 days earlier each decade. Overall, snowmelt in the Alaskan Arctic has advanced about 2.5 days per decade since the 1960s.

These earlier snowmelts have increased the amount of sunlight absorbed by the ground. It's gone up by about 1.5 percent per decade, says Sturm. So, with each decade, each square meter of ground surface takes in about 3.3 watts more energy than it did the decade before. That rate of heat increase is several times what scientists project to result from emissions from automobiles and industrial activity.

Local changes in albedo caused by earlier snowmelt account for about 95 percent of the heating observed in arctic Alaska, Sturm and his coworkers estimate. The other 5 percent results from changes in vegetation, with 3 percent coming from the expansion of forests and 2 percent from new acreage covered by shrubs.

Although shrubs are now the smallest players in albedo change, their significance will probably overtake that of forests, says Sturm. That's because boosting summertime temperatures 1[degrees]C or so typically triggers increased shrub growth within a decade, whereas conversion of tundra to forest occurs more slowly.

SPREADING LIKE WILDFIRE Since 1950, the acreage of shrub-dominated tundra on Alaska's North Slope has increased about 1.2 percent per decade. Once shrubs take hold in an area previously covered by grasses, they trigger ecosystem changes that perpetuate, and even accelerate, the shrubs' growth and expansion.

The nitrogen content of the soil plays a major role in stimulating shrub growth. Warmer temperatures enhance activity of nitrogen-producing bacteria in soil. Furthermore, as shrubs prosper, they produce more leaves, which fall to the ground each autumn and decompose, releasing nitrogen into the soil.

Snow accumulates more thickly beneath shrubs than it does beneath tundra grasses, and it better insulates the ground during frigid winter months, says Stunn. Wintertime soil temperatures beneath shrubs are 3[degrees]C to 10[degrees]C higher than they'd be in shrubfree terrain--a difference that boosts the soil's nitrogen supply about 25 percent, thereby making conditions even better for further shrub growth, says Sturm.

Field experiments by Sturm and three coworkers confirm that the growth and expansion of shrubs can have a significant effect on local albedo. The scientists conducted their experiments during the winters of 2001 and 2002 at five sites near Council, Alaska. Because the test plots were within a few kilometers of each other, their temperatures and the amounts of precipitation and sunshine that they received were similar, says Sturm. Only their plant cover differed, ranging from shrubfree tundra at the least-vegetated site to forest at the site with the most vegetation.

By suspending instruments on a 50-meter-long cable across each test plot, the researchers could measure the albedo of the terrain during winter and spring without disturbing the plots' vegetation or the surface of the snow as it melted. The scientists found that as little as 10 centimeters of snow cover above the tops of the plants could block sunlight from reaching the plants' dark bark and branches. In such cases, between 70 and 90 percent of the light that fell on the snow was reflected, a range similar to that reflected from snow atop a glacier or sea ice.

Snow depth was at its greatest at the Alaskan test sites in April 2001. At sites where grasses and short, supple shrubs were prevalent, the vegetation was bent to the ground with the weight of the overlying snow, and the albedo was about 85 percent. At the site where tall, stiff shrubs predominated--and where only 1 percent of the shrubs stuck up through the snow--the albedo was equally high, says Sturm. But at the woodland site, where shrubbery still above the snow covered about 10 percent of the test plot, the albedo was only 60 percent.

As the snow melted in the spring and more vegetation became exposed, the albedos of all the test sites dropped to their summertime values, between 11 and 19 percent. Sturm and his colleagues report their findings in the Sept. 28 Journal of Geophysical Research (Biogeosciences).

The type of vegetation at a site greatly affected not only overall albedo but also how quickly albedo changed during the snowmelt. As the depth of snow over buried shrubs declined, more and more sunlight reached the plants' dark branches. The branches absorbed some of the radiation and became warmer--"like little solar ovens," says Sturm. That process melted cavities within the snowdrift, and shrubs popped up through the surface. As a result, the local albedo dropped more rapidly than it did at sites where grasses were prevalent.

Given this effect, if shrubs replace extensive areas of grass-covered tundra, the overall Arctic climate could undergo a major transformation, Sturm and his colleagues suggest. At the latitude of the Arctic Circle, such a transition could boost average radiation absorption by 20 watts per square meter.

Computer simulations by some scientists suggest that if shrubs covered all the tundra regions of Alaska's Arctic Slope, average temperatures in July would rise between 1.5[degrees]C and 3[degrees]C. Other research teams' regional models indicate that shrubs exposed above the snow would boost the temperature about 6[degrees]C.

Broad regions of the Arctic are poised for a transition to shrub dominance, says Sturm. The shrubs "won't need to migrate. In many spots, they're already there, just in a stunted form," he notes. In Alaska, tall shrubs now occupy about 20 percent of the North Slope's tundra but eventually may cover the entire area. That ground is now dominated by a mix of tundra grasses and small shrubs.

Across the Arctic, the tall shrubs could eventually replace grasses over about 4 million square kilometers.

Overall, arctic soils such as peat bogs and tundra hold more than a quarter of the carbon stored in soils worldwide since the end of the last ice age about 11,500 years ago (SN: 1/17/04, p. 37). Field tests suggest that increased shrub coverage in the Arctic would cause the soil to lose carbon. This loss would be greater than the gain in carbon stored in plant material aboveground. Therefore, says Sturm, changing from grasses to shrubs could actually increase the atmospheric concentration of carbon dioxide, causing the world to warm even quicker.

MELTING CAP A transition taking place in the ice-covered ocean that sits atop the world may prove far more significant for the Arctic's albedo than is the shrub expansion on land. The sea ice is in danger of melting, argue Jonathan T. Overpeck of the University of Arizona in Tucson and his colleagues in the Aug. 23 Eos.

The two other major repositories of Arctic ice face less of a threat. "It would take centuries, if not millennia, to melt the Greenland ice sheet," says Overpeck. Likewise, the deep permafrost that underlies many Arctic lands probably will not melt for centuries.

Between the late 1970s, when satellites first started monitoring sea-ice coverage, and 2001, the area that's covered by sea ice has been decreasing about 6.5 percent per decade.

Now, thanks in part to the record global warmth of the past 4 years--all of which rank in the 5 hottest years since 1861--the extent of end-of-summer sea ice has reached record lows, says Ted A. Scambos, a climatologist at the National Snow and Ice Data Center in Boulder, Colo. Late last summer, sea ice covered only 5.32 million k[m.sup.2], about two-thirds that measured during 1970s summertimes.

While dry snow atop sea ice reflects between 70 and 90 percent of the radiation that strikes it, open water reflects only 7 to 10 percent, says Scambos. So, when ice melts and albedo takes a nosedive, the upper layers of the ocean absorb more sunlight, and sea-surface temperatures rise. That extra warmth melts yet more sea ice during the summer and slows the formation of sea ice in the autumn. That reduces sea-ice coverage the following spring, giving summer melting even more of a head start.

At some point in the sea ice's decline--maybe even a point that's already been passed--changes in albedo begin to accelerate the rate of melting. Taking into account Arctic sea-ice declines of the past 4 years and those measured since the late 1970s, the ice coverage has been diminishing at a rate of 8 percent per decade.

At that rate of decline, the Arctic Ocean by the end of this century could be icefree late in each summer, Overpeck and his colleagues estimate. Although data from ice cores drilled from Greenland suggest that the Arctic was warmer about 125,000 years ago than it is today, there's no evidence that the Arctic Ocean in the past 1 million years or so has been icefree, he notes.

Over that period, atmospheric concentrations of carbon dioxide fluctuated between 200 and 280 parts per million (ppm), says Overpeck. But today, the average global concentration of that greenhouse gas sits above 375 ppm, and it's rising each year. So, melting may progress much farther than it did during earlier interglacial periods, when that warming influence wasn't in place.

Considering the rates of sea-ice decline experienced in the past 4 years, Scambos is concerned. "It's easy to conclude that we're on a rapid and accelerating downward trend," he notes.

Sturm is a bit more conservative regarding the runaway melting that's poised to plague the Arctic. "We're on the top of the slide now" he says. But he notes that this scenario may not last long: "The [increasing rates of melting] seem to have a lot of boosters and not so many brakes."

Brakes or not, the Arctic Ocean late this century should still receive at least a thin glaze of sea ice during the perpetual night of winter, says Scambos. But that's cold comfort, because come summertime, when sunlight lasts 24 hours a day, each square meter of icefree sea will absorb about 100 watts more energy than it would if it were covered with snow-topped ice.

That's not quite as bad as upholstering the Arctic with black vinyl, but the long-term effects on the planet's climate may be just as uncomfortable.

HOT STUFF--Dark shrubs absorb more sunlight than snow does, so when bushes poke above the drifts, the surrounding air warms and snow melts faster.

[ILLUSTRATION OMITTED]

THEN AND NOW -- Shrub coverage in northern Alaska has increased substantially in recent decades, as shown in these aerial photos of the same site taken in July 1950 (top, in black and white) and August 2002 (bottom).

[ILLUSTRATION OMITTED]