New Approaches to Understanding Wildfires

Forecasting fire-weather interactions, assessing emissions, and understanding social impacts

Persistent drought and hot weather are expected to make the 2007 wildland fire season unusually severe. Spring fires caused extensive damage in the Southeast and in southern California, and experts warn that much of the West is vulnerable to large blazes this summer. These conditions follow a record-breaking 2006 season which, by early October, had seen an unprecedented 9.93 million acres burned, almost twice the average for the past decade.

In 2006, Science reported that warmer summer temperatures appear to be increasing the duration and intensity of the wildfire season in the western United States through longer summers, more droughts, and earlier snowmelt. Compared to the years 1970–86, major fires have increased fourfold since then, and total burn area has increased sixfold.

Wildfires have major impacts on regional and global pollution and can affect climate and weather. They could even complicate efforts to regulate greenhouse gases.

"Climate, vegetation structure, and fire regimes interact with one another and with human activities," explains Guy Brasseur, director of NCAR's Earth and Sun Systems Lab. "Integrated approaches will be required to adapt to altered fire activities resulting from climate change, especially in terms of social and economic practices. NCAR wants to be at the forefront of the research on these issues."

NCAR scientists study a wide range of wildfire topics, their work ranging from forecasting fire and weather to modeling emissions, assessing mercury, and analyzing social impacts. Here's a look at recent developments in wildfire research at NCAR.

NCAR scientist Janice Coen is approaching wildfire as a high-impact weather forecasting challenge.

"People can think of this approach as a weather forecast model that exchanges feedbacks with another model, such as a land surface, ocean, or sea ice model. But in this case, the fire model takes wind and humidity information from the atmospheric model, calculates how the fire would behave and grow, and provides releases of heat, water vapor, and smoke back into the atmospheric model."

In conjunction with colleague Ned Patton, Coen is working on WRF-Fire, a module of the Weather Research and Forecasting model (WRF) that will allow the university community to study fire-weather interactions and fire impacts on air quality. She's trying to improve short-range weather forecasting techniques and develop methods to verify forecasts. Along with university colleagues, she's also developing techniques for assimilating fire mapping and weather data.

Coen has a special interest in how multiple fires interact with each other by modifying their surrounding atmosphere, a research question that is especially applicable to backfires. A backfire is a small fire set intentionally ahead of a major wildfire to prevent the wildfire's advance by consuming fuels in its path.

"If the backfire is too intense or set too far away, it becomes another uncontrollable fire," Coen says. "Backfires sometimes escape and are blamed for large losses." She cites a case in which more than one hundred homeowners filed suit against the Forest Service, claiming that a backfire set during Montana's Bitterroot fires of 2000 was responsible for destroying their homes.

New techniques developed by NCAR's Gabriele Pfister and Christine Wiedinmyer are giving scientists a better grasp on making more accurate regional estimates of wildfire emissions.

Last spring, Wiedinmyer introduced the first version of a model for regional gas and particulate emissions from wildfires in North America. The model uses a combination of satellite and ground-based data to refine estimates of fuel loading (the amount of burnable material in a given area).

She's currently working on state-by-state comparisons between this summer's fire emissions and emissions from human-generated sources such as industry and automobiles. She's found that wildfires in Idaho and Washington in August emitted much more carbon monoxide (a toxic gas and ozone precursor) and particulate matter than human sources in those states. By the end of September, wildfires in the west had accounted for a substantial fraction of the United States' annual fine particulate matter and carbon monoxide emissions.

"There are a lot of uncertainties in these estimates, but even within a factor of two, they are still significant amounts," Wiedinmyer says.

Wildfire emissions on such a large scale could complicate efforts to regulate greenhouse gases. Wiedinmyer has found that fires in California this summer emitted a significant fraction of the state's annual energy-related emissions of carbon dioxide. Governor Arnold Schwarzenegger recently signed landmark legislation to cap the state's greenhouse gas emissions. "Big fire seasons could potentially have an impact on California's new policy," Wiedinmyer says.

The next step will be to connect her model with WRF-Chem, a WRF module that simulates trace gases and aerosols, to determine the effect that fire emissions have on atmospheric chemistry and, by extension, human health.

Pfister has focused on modeling emissions from severe wildfires that burned in Alaska and Canada in 2004. Using a different modeling technique than Wiedinmyer's, she's shown that these fires emitted about as much carbon monoxide as did human-related activities in the continental United States during the same time period. The fires also increased ground-level ozone by 25% or more in parts of the northern continental United States and by 10% as far away as Europe.

The social dimensions of wildfire Kathleen Miller. (Photo by Carlye Calvin, UCAR.) NCAR economist Kathleen Miller is collaborating with a graduate student at CU-Boulder on a project that examines the effects of information on the decisions homeowners make in wildfire-prone areas. Their goal is to document what people understand about wildfires, identify the significance of uncertainty about wildfire risk, and assess the value of various types of information for decisions regarding land development and wildfire risk mitigation. In particular, they hope to determine the extent to which people's decisions to build homes in fire-prone areas and/or mitigate risks are sensitive to differences in how they perceive the probability of wildfires. "We take the perspective of looking at the relevant decisions as multi-faceted investment and consumption decisions made under uncertainty," Miller explains. The team recently completed a survey of residents in fire-prone areas of Clear Creek County, Colorado, to document their understanding of wildfire risks and how they make decisions to mitigate that risk. Preliminary results suggest that most residents understand their vulnerability to wildfires, and have made investments to reduce their exposure. In addition, the majority of homeowners consciously choose to accept the risks in exchange for the lifestyle benefits of living in a mountain community. The next step is to analyze how homeowners' decisions to take steps to mitigate fire hazard are affected by property values.

"It's important to see how the influence of these fires can reach large parts of the atmosphere, perhaps even the entire Northern Hemisphere," Pfister says. "This has significant implications as societies take steps to improve air quality."

Pfister's next step is to look at the impact that greenhouse gases and aerosols from wildfires have on the change in radiation (heat) entering or leaving the climate system, also called radiative forcing.

Colleague Steve Massie is using satellite data to study wildfires in Siberia as part of a broader study of Asian air pollution. "There's a stew of aerosols that comes out of Asia, and some of it gets transported across the Pacific Ocean," Massie says. "We're using satellite data to unravel this."

Emissions from Siberian wildfires are one of the three main components of the mixture, along with urban coal emissions and dust from the Gobi Desert. Steve and colleagues hope to use the satellite data to determine how much smoke and gas the fires inject into the lower stratosphere, about 10 miles (16 kilometers) above Earth's surface.

Mercury storage in vegetation

During a wildfire, mercury stored in foliage and ground litter is released into the atmosphere. In a field study in Prince Albert National Park in Saskatchewan, Hans Friedli and colleagues have taken a close look at where and how mercury is distributed within the plants and soil of a boreal forest. They've found that more than 90% of the total mercury resides in the soil. This mercury is subject to release during severe wildfires.

The discovery has significant implications for research on mercury releases under drought and climate change scenarios. "As a result of global warming, permafrost melts and bogs can burn, and the large stores of mercury in them get released," Friedli says. "With global warming occurring in boreal areas, this is a real concern."

On a similar note, James Greenberg is looking at emissions of volatile organic compounds (VOCs) from vegetation that has been heated in advance of an oncoming fire or that smolders afterward, but that hasn't reached its burning point.

Using a Proton Transfer Reaction Mass Spectrometer, Greenberg heated leaves and measured their emissions at different temperatures as they approached combustion. "We found that most of the oxygenated compounds that come out of vegetation during this process come out during the pre-flaming state," Greenberg says. "We also showed how individual organic compounds evolve during different states of heating."

The research will help scientists determine how much pollution comes from a fire itself and how much comes from pre-fire release of VOCs, giving modelers a better sense of how fire emissions evolve over time. "If we can break a fire into its different components, we can model the emissions much better," Greenberg says.

Summer 2007 (adapted from Staff Notes Monthly, October 2006)