New Approaches to Cold-Season Impacts
Scientists at NCAR and their colleagues elsewhere are at work on warning and detection systems to help keep the wheels of transport moving, even when snow and ice threaten. Meanwhile, basic research is painting a clearer picture of how weather’s usual ebb and flow crystallizes into patterns that can persist for much of a winter—and how those might evolve in a changing climate.
Globe-spanning insight on weather cycles
British mathematician Sir Gilbert Walker was the first to connect the Southern Oscillation (or SO, a see-saw in atmospheric pressure between the western and eastern tropical Pacific) to cyclic weather and climate patterns far afield.
AT THE UNIVERSITIES
A year-round perspective on Indian monsoons
Few places offer as dramatic a seasonal shift as does India, where a bone-dry winter is typically followed by torrential summer rain. Peter Webster (Georgia Institute of Technology) believes the more placid Indian winters get short shrift.
Occasional outbreaks of cold air, especially in the Delhi region, can kill large numbers of people who may be ill-clothed or poorly sheltered. “It’s an intriguing thing,” says Webster. “If you ask meteorologists to look at the weather map, they’ll go straight to where it’s raining. It’s rare to think of looking for impacts where it’s not raining.” At Georgia Tech, Webster’s group carries out forecasts on a range of time scales for several clients in Bangladesh and India. Their work brings added value to modeling from NOAA and the European Centre for Medium-range Weather Forecasts. “We take those numerical models, join them with our hydrological models, and then use a fair amount of statistics.” Along with 1- to 10-day forecasts and seasonal outlooks, Webster’s group delves into the 20- to 30-day period, long considered beyond the pale in weather prediction due to the rapid growth of errors over time. Webster stresses the need for communication in improving forecast products: “Making a forecast and sticking it out on the loading dock, hoping somebody will pick that forecast up, is a useless endeavor. You really have to work with the user community.”
Knowledge tailored for decision makers
How do we characterize the uncertainty within even the best forecasts? How do we prepare for extreme weather and climate events—including their effects on health? Linda Mearns is looking at these and other questions as the director of NCAR’s Weather and Climate Impact Assessment Science Initiative.
“Uncertainty is one of the main themes,” says Mearns, an expert in formulating climate-change scenarios and studying the agricultural impacts of climate change. The initiative focuses on critical scientific gaps in the weather and climate arenas that have proved particularly challenging for decision makers and scientists alike. The goal is to help scientists generate knowledge in forms designed to have real value to the people who shape our societies. In 2004 the initiative brought a wide-ranging array of experts to Boulder for a colloquium on climate and health. The participants explored atmospheric links, societal responses, and research directions for health issues that ranged from the West Nile virus to heat-related illness. Overall, Mearns sees the initiative as “a tremendous opportunity to really integrate the physical science being done at NCAR with the environmental and social sciences.
We gird ourselves with flannel, fleece, and multivitamins. We splurge on all-season tires, super-insulating windows, and high-efficiency heaters. Despite all we do to soften the annual journey through winter, the season still takes a toll that is often brutal. Many hundreds of Americans are killed each year in motor-vehicle accidents linked to icy or snowy roads. A mere hour’s delay from a snow squall at a major airport can cost airlines hundreds of thousands of dollars. And an unexpectedly strong cold wave can throw energy companies into a scramble to keep up with soaring demand.
These setbacks needn’t be as costly—in lives or dollars—as they are now. Atmospheric scientists are making continual gains in their ability to detect and predict winter’s worst hits. Such breakthroughs can take years to filter into practice, though. At NCAR, a wide range of work is geared toward accelerating this process and helping society better cope with winter impacts. Transportation, whether by air, train, or motor vehicle, is a particular focus.
Along with these efforts, climate scientists at NCAR are using powerful software to tease out the processes that give each winter its character: unusually snowy in one region, pleasantly mild in the next, or distressingly dry in still another. The triggers can be as distant as a slight shift in the warm ocean water that girds the Equator.
Even in a world that most atmospheric researchers expect to get steadily warmer, winter and its frigid fury won’t be vanishing anytime soon.
Slipping science into the art of road treatment
The task seems easy enough: keep ice from coating highways, and plow snow as soon as it falls. But while drivers worry only about the roads they navigate, maintenance crews have to deal with a vast network of highways all at once. Applying too much chemical or the wrong one is expensive and can degrade the roadside environment. But pavement can ice or thaw in minutes over small areas. Undertreating roads, or missing a flash-frozen stretch, could result in a fatal accident.
As they straddle this icy line, most of the nation’s highway maintenance crews rely on spotty weather information, sometimes hours out of date. Time-tested rules of thumb for treatment may work well in some situations but not in others.
“Highway agencies around the world are anxious to obtain better weather information and use it more effectively,” says NCAR scientist William Mahoney. But, he adds, many of these agencies have little in-house meteorological expertise, “so they’re often working in a vacuum.”
The Web-based Maintenance Decision Support System (MDSS), developed by a multiagency team led by NCAR and sponsored by the Federal Highway Administration’s Road Weather Management Program, promises to save lives, cut costs, and help keep drivers on the move. It does so by giving highway crews more than a generic outlook, such as “snow likely tonight,” according to project manager Mahoney. “We need weather and road condition forecasts that are more specific, more timely, and tailored for decision makers who are not meteorologists.”
During the winters of 2003 and 2004, state road crews in Iowa put the MDSS to work during 15 episodes of winter weather. Decades of experience made some managers dubious at first, according to Iowa winter operations administrator Dennis Burkheimer. Their concern, in his words: “You’re going to try to replace a supervisor with a computer box?” However, he adds, “We were all very open to evaluating the product. Working with the scientists and giving them feedback, it just got better and better over time.”
While on their usual routes, the crews gathered key data along 16 segments of interstate and U.S. highway near Des Moines and Ames, Iowa. Some of these roads were studded with sensors that relayed pavement temperature, a critical factor in freeze-thaw cycles. Above ground, the wintry weather was sampled by automated weather stations, satellites, and radar.
The reports from Iowa fed into MDSS software at NCAR’s Boulder headquarters. The system balanced advice from a roundtable of computer models, then relayed road-specific instructions to the maintenance crews: whether or not to plow, how much deicing chemical to apply per mile, and when to reapply treatment. Also factored in were such possibilities as whether a surface might refreeze after a brief warm-up.
While far from infallible, the MDSS produced encouraging results in its Iowa evaluation. Wilfrid Nixon (University of Iowa), an engineer who tracks best practices in winter road treatment around the world, gave the MDSS high marks. “It’s an intelligent computer system that captures the state of the art and presents it in a very useful and accessible format.” Moreover, says Nixon, the project served to build bridges between road crews and meteorologists. “You’ve got two communities that don’t really speak the same language. It highlighted the need for much better communication.”
Its major test now complete, the MDSS will soon be in the marketplace. Over 100 specialists from commercial weather firms, universities, and federal labs—as well as nearly half of U.S. state transportation departments—met in July 2004 to discuss how the private sector can tailor the MDSS to specific states and regions. “It’s not a plug-and-play system,” notes Mahoney, “so the commercial rollout will take some time.”
Meanwhile, NCAR and partners continue with smaller-scale testing to improve the system. “Models often miss shallow or thin cloud layers that greatly impact road temperature forecasts,” says Mahoney. Thin cirrus clouds, for example, allow sunlight to filter through by day, but help trap radiation from Earth by night. In both cases, the effect is to keep temperatures higher and potentially keep roads from freezing. The MDSS team hopes to leverage the strong points of each computer model in their array and address the gray areas that remain. The ultimate goal: capturing threats as fine grained as a single ice-covered bridge lying in wait along a cold, wet highway.
The dance of ice and water
Whirling inside the cloud banks of a winter storm are millions of snowflakes, ice pellets, and water droplets. The mix is different for each storm, and the dancers can shift identities in a flash: snowflakes collide and grow, raindrops congeal into sleet.
One might easily overlook the tiny agents known as supercooled water droplets. Lacking a mote of dust, salt, or pollution at their core, these droplets stay liquid even as temperatures fall well below freezing. When they strike a surface—such as an airplane wing—they freeze on contact. An unprotected plane flying through clouds that are dense with such droplets is sheathed in a quick-forming layer of ice that roughens wings, produces drag, and may ultimately bring down the aircraft.
Useful as it is, Doppler weather radar cannot spot in-cloud icing very well. Strong radar echoes from larger raindrops, ice crystals, and snowflakes tend to obscure the supercooled droplets, which are often less than 50 microns (0.002 inch) in diameter. But NCAR’s aviation-weather specialists have attacked the icing problem on several fronts. Two online services now tell pilots where the worst risk of icing aloft is and how it should evolve. And a new hybrid radar system called S-Polka may soon be able to measure pockets of supercooled water.
“Icing outlooks have been around for years, but until recently they were very subjective tools,” says NCAR’s Marcia Politovich. She heads the in-flight icing component of an aviation weather research effort supported by the Federal Aviation Administration at NCAR since 1989. Over the past few years, her team has developed two breakthrough Web services.
As their names imply, the Current Icing Product and Forecast Icing Product—available freely on the Internet—show pilots the present and future states of icing risk at various locations and altitudes. Both products use a 0-to-100 scale that gives a sense of relative threat. Produced hourly, the products draw on surface and pilot reports, satellite and radar data, and a fast-updating forecast model. The CIP and FIP are especially useful for commuter planes and other propeller-driven aircraft. These typically cruise at lower, ice-prone altitudes, and they often lack the front-of-the-wing, ice-fighting heaters that are standard on jets.
Thousands of visitors each week vouch for the appeal and usefulness of the CIP and FIP Web sites. But researchers crave more—they want to actually see the supercooled particles that hide so skillfully within clouds. In 2004 NCAR unveiled a two-in-one radar that may provide just that view.
It’s called S-Polka, a name that blends two elements. One is S-Pol, a workhorse Doppler radar maintained by NCAR and employed by university scientists for nearly a decade. It sends signals in the S-band frequency (3,000 MHz) that are vertically and horizontally polarized—thus, the “pol.” Its new partner is another, much smaller and faster polarized radar that operates in the Ka-band (35,000 MHz).
Like other Doppler radars, S-Pol detects precipitation and infers wind speed, but it can’t detect very small objects, including tiny supercooled droplets. “When it comes to cloud particles, we can’t yet interpret the standard radar echo,” explains NCAR’s Jothiram Vivekanandan, the project’s lead scientist.
The Ka-band radar picks up signals from smaller droplets than S-Pol and covers a smaller area. By mounting the two radars side by side on a single pedestal so they can operate in unison—sending pulses within microseconds of each other—NCAR researchers hope to be able to identify the supercooled droplets that threaten aircraft.
Although the Ka-band radar had a few glitches in the 2004 tests, the basic S-Polka technique appears sound. Next on tap: translating the radar returns into maps and data that show where supercooled droplets lurk. If all goes well, systems modeled on S-Polka could make their way to airports across the country over the next few years. S-Polka is also being tapped for other studies of cloud microphysics, such as the 2005 Rain in Cumulus over the Ocean Experiment.
Ice, frost, and tomorrow’s climate
Why does one winter bring a barrage of snow and ice storms, while another stays largely dry and mild? El Niño and La Niña make up one explanation. They’re one of the biggest sources of variability in North American winters. When the eastern tropical Pacific is unusually warm (El Niño), U.S. temperature contrasts are typically diminished and heavy precipitation is focused along the Sun Belt from California to Florida. La Niña’s cooler waters tend to foster highly variable U.S. winters.
Looking over the last half-century, NCAR’s Grant Branstator is exploring how a slight shift west or east in the location of El Niño’s oceanic warming can make a very big difference. When the warming is focused slightly farther west than usual, the typical global impacts become far stronger, says Branstator. This takes shape through a sequence of ridges and troughs in the jet stream dubbed the Circumglobal Waveguide Pattern. If the warming is focused slightly eastward, the worldwide effects are weaker, but the U.S. West Coast feels the brunt of El Niño. These effects seem to result from changes in the subtropical jet stream, around 30°N. “These aren’t just subtle effects—in some years they jump out at you,” Branstator says.
In collaboration with Frank Selten (KNMI, the Royal Netherlands Meteorological Institute), Branstator is examining output from a large set of global climate simulations to see how El Niño impacts might evolve over the next century. Thus far, the model’s signature of human-induced climate change favors the Circumglobal Waveguide Pattern and its effects—but not El Niño per se. According to Selten, “El Niño itself hardly changes.”
Another clue to a winter’s character plays out with the help of a jet stream far to the north. Strong upper-level winds typically encircle the Arctic Ocean each winter, but they sometimes dip southward with a frigid air mass. This ebb and flow has been dubbed the Arctic Oscillation (or the Northern Annular Mode). Since the 1980s, the oscillation has tended to stay in a positive phase, keeping the polar jet contracted and high pressure bottled up in the Arctic for weeks at a time. This mode often leads to long spells of mild weather across the central and western United States; parades of warm, wet storms across northern Europe; and persistent dryness around the Mediterranean Sea.
What causes the Arctic Oscillation to favor its positive phase? NCAR’s James Hurrell and NOAA’s Martin Hoerling have found a link to warming in the tropical Indian Ocean. Showers and thunderstorms that pump heat skyward across this vast area appear to be rearranging atmospheric features across the Northern Hemisphere. Further warming of these tropical waters might act to reinforce the regional patterns seen in U.S. and European winters over the last few years.
As for the North Atlantic—smaller and cooler than its tropical counterpart—most scientists see it as a slave to wintertime trends in the atmosphere more than a driver. To put this notion to the test, Clara Deser and her NCAR colleagues joined Gudrun Magnusdottir (University of California, Irvine). They jump-started the NCAR Community Climate System Model with recent trends in North Atlantic water temperature and ice cover. The trends were magnified in order to bring out possible interplay with the atmosphere more clearly.
“The atmosphere’s response to the ice was much stronger than to the ocean temperatures. That was quite a surprise,” says Deser. What the models can’t yet explain is the intriguing difference between the Greenland Sea, where the peak extent of winter ice is decreasing year by year, and the Labrador Sea, where it’s increasing. “We think these changes are a result of the Arctic Oscillation,” says Deser, “but it’s hard to disentangle the interactions in the model.”
Whatever its cause, the flavor of recent U.S. and European winters may become standard fare by later in this century, according to NCAR’s Gerald Meehl. He and colleagues Claudia Tebaldi and Douglas Nychka simulated the period 2080–2099, based on a global climate model supported by the U.S. Department of Energy and NCAR. They found that frost days (when the nighttime low temperature dips below freezing) should become less common across most of the United States and Europe. This tendency would be strongest across the western parts of North America and Europe, where relatively mild ocean air should sweep in more often—a trend already observed in both areas. Most of the reduction in frost days would occur in springtime rather than autumn, another trend that’s already unfolding.
“I was surprised that the model was able to simulate the pattern and seasonality of the observed trends in decreasing frost days over the United States,” says Meehl. “Nobody in the climate science community had looked at day-to-day extremes in frost days before.” The study, along with a companion work on heat waves, will inform the next Intergovernmental Panel on Climate Change report, due in 2007 (see Highlights article, The Century After Tomorrow).
Meehl and colleagues also looked at how changes in frost days affect the length of the growing season. Although there’s an influence, says Meehl, the relationship between changes in frost days and the growing season isn’t clear-cut. With many millions of dollars in agriculture at stake, however, there’s plenty of societal motivation for further research on this topic, as well as on many other aspects of winter’s annual arsenal.