| 十大美女 | 好山好水 | 著名建筑 | 创意图片 | 名画名字 | 风云人物 | 最高/最大/最长/最... | 更快/更好/更高/更全面/更深刻/更... |
您当前的位置:首页 > 真善美好 > 更快/更好/更高/更全面/更深刻/更...

Computing a Better Fire Forecast

时间:2013-08-10 23:45:50  来源:  作者:

Science 9 August 2013:
Vol. 341 no. 6146 pp. 609-611
DOI: 10.1126/science.341.6146.609
News Focus
Wildfire Science
Wildfire Science
Computing a Better Fire Forecast
Eli Kintisch
Scientists and firefighters ponder new ways to predict the spread of wildfire as the U.S. West faces ever more potent blazes.


View larger version:
In this page In a new window
Download PowerPoint Slide for Teaching
Firestorm. The 2006 Esperanza wildfire burned 16,000 hectares of chaparral and killed five firefighters in southern California.

CREDIT: ZUMAPRESS/NEWSCOMFirefighters who were there that deadly morning would later describe trees with glowing orange bark and flames that leapt hundreds of meters in moments. An arsonist in the southern California town of Cabazon started the October 2006 blaze, known as the Esperanza fire. Havoc reigned at dawn as residents evacuated and fire trucks snaked up smoky mountain roads.

At about 6:30 a.m., a five-man crew from Engine 57 set up a water pump above one edge of the fire, hoping to save an octagon-shaped house. But the blaze swept up an adjacent gully and over the structure in less than 10 seconds. None of the crew escaped.

Now, 7 years after the Esperanza tragedy, U.S. wildfire specialists are still debating exactly what happened—and whether improved computer models might help scientists, firefighters, and disaster planners avert such deadly surprises. Relatively crude existing simulations already help the U.S. Forest Service and other agencies predict the course of hundreds of fires each year, mostly in the western United States. But researchers agree that these models do a poor job of simulating extreme events like Esperanza, when wildfire can seemingly act capriciously and erratically.

"The current models have been stretched to their limits," says Kelly Close, a fire behavior analyst in Fort Collins, Colorado. They will be stretched further if a warming climate makes extreme fires more common, raising the odds of additional Esperanza-like calamities, such as the Yarnell, Arizona, blaze that killed 19 elite firefighters in June. More sophisticated models that can simulate both the extreme behavior of individual fires and the risk of fire across entire landscapes are on the horizon. But a variety of factors, including limited funds, technical disagreements, and the sometimes clashing cultures of academic scientists and wildland firefighters, is complicating the search for alternatives.

Model behavior

View larger version:
In this page In a new window
Download PowerPoint Slide for Teaching
Enlightenment. These days, meteorologist Janice Coen (above) is working to develop fire-atmosphere simulations that improve pioneering models devised by Richard Rothermel (left, behind window) in the 1970s.

CREDITS (TOP TO BOTTOM): PHOTO BY CARLYE CALVIN/© UCAR; COURTESY OF THE JOINT FIRE SCIENCE PROGRAMWildfire modeling is a relatively young endeavor. In the early 1970s, when the U.S. Forest Service sought to analyze fire risk over large landscapes, it turned to its research laboratory in Missoula, Montana. There, fire scientist Richard Rothermel led a team that observed fires in field tests and the laboratory. They went on to write a set of equations, which for the first time related the speed and direction of a model fire's spread to terrain, the vegetation available for fuel, and nearby wind patterns.

Forty years later, Rothermel's formulas are the foundation for two varieties of operational fire models. Strategic modeling tools allow managers to forecast, well beforehand, how fires might spread or visualize where future fires might break out; tactical tools allow modeling of specific fires on shorter timescales, sometimes with an eye toward figuring out whether to confront them or let them burn. Perhaps the most widely used model is a tactical model called FARSITE, which uses terrain, fuel, and weather data to project a fire's path over a 2D map.

Fire analysts use it tens of thousands of times a year in their work, says FARSITE developer Mark Finney of the Missoula lab. And in the hands of an analyst who understands its limitations, FARSITE can be a valuable tool. In 1994, for example, officials in Glacier National Park were confronted with a pair of wildfires. They faced "a lot of tense decision-making," Finney says, over whether the flames might harm settlements, and whether they should put firefighters in harm's way. But FARSITE predicted that the fire would stay within a safe area, Finney says, and it did. In other instances, it has given planners a reliable 3-to 5-day forecast of a fire's spread, enabling them to more efficiently—and safely—deploy crews.

Still, Rothermel-based tools like FARSITE have some glaring weaknesses. They assume that fuel is evenly distributed, for instance, and that it burns uniformly. They don't simulate the movement of particles and gases that can affect a fire's path, or the complex relationships with the atmosphere that can enable some blazes to create their own wind and weather, even generating so-called pyrocumulus clouds. The models also have a hard time with complex phenomena like rotating air masses known as fire whirls, fires spread by embers, or explosive blasts of flame that can shoot out from a fire's flanks. A Rothermel-based model "describes very well a fire burning in a field of wheat," one of the researcher's colleagues once told Fire Science Digest. "As you get further away from that uniformity, the less accurate it becomes."

来顶一下
返回首页
返回首页
发表评论 共有条评论
用户名: 密码:
验证码: 匿名发表
推荐资讯
相关文章
栏目更新
栏目热门