MISSOULA, Mont. -- Forty-eight hours into the costliest wildfire in Colorado's history, Mike Tombolato turned to his computer for answers on what to do next.
The wind-driven blaze was fast approaching the city of Boulder. As the operations section chief charged with suppressing the fire, he had to make the expensive decisions: when -- and if -- to evacuate those homes and where to send the helicopters carrying a sticky, red, fire-inhibiting slurry. If he chose well, fire creeping too close to homes could be stifled.
That's where fire models developed at the U.S. Forest Service's Missoula Fire Sciences Laboratory came in. From his office in Fort Collins, Colo., Tombolato typed in data about where the fire started, the environment around the burn and how wet it was outside. Minutes later, a computer spat out information that helped him plot his course.
"The system is not perfect yet and can't account for all winds effects," he said, "but it was giving us a timeline of fire advancement given the topography and the weather conditions." That hour-by-hour forecast helped him decide that he would wait on evacuating the next batch of residents unless the fire progressed to a certain point by that evening.
Not too many years ago, he and his firefighters would have been on their own.
At the Forest Service's 50-year-old lab in Missoula, researchers developed the basic science underpinning the agency's knowledge of how fires spread. They built the prediction models that emergency responders like Tombolato look to when they are facing difficult decisions. Even the wildfire risk ratings listed in front of local fire departments -- couched next to Smokey the Bear signs -- came from the lab.
Recalibrating for climate change
In recent years, its scientists have also added climate change to the lab's portfolio. Teams of researchers have started exploring how warming temperatures and all the natural phenomena that accompany it -- like the explosion in numbers of bark beetles that kill trees -- will affect future fires.
Broadly, they hope to create a clearer picture of how climate and fire will interact and to arm land managers with hard data they can use. They are working on the basic research that could eventually help answer burning questions such as: What type of trees should replace those that have gone up in smoke? What steps could be taken to revive a species?
But in some cases, they are literally racing wildfires to gather crucial information before wildfires wipe out the evidence.
"Everything the fire lab works on is climate change work, even if it is not labeled that way," said Colin Hardy, the program manager for the lab. "Wildfires will be undeniably impacted by climate change," he said. A hotter, drier future, he said, will likely affect how often fires happen, as well as their size and intensity.
So his researchers are fanning out into their respective fields, gathering clues in the lab and on the ground that they hope to weave into a clearer picture of future fire conditions.
Just as firefighters attack flames by air, in vehicles and on foot, scientists at the lab are tackling research questions with different scientific approaches.
Wei Min Hao looks to satellite data: Four times a day, he receives a real-time feed on the fire scene outside. If a steady wildfire is nearby, his team hops into a souped-up white van parked out back and races to catch it. At the scene, they break out a specialized laser and shoot its beam into the fire's smoke, all to better visualize how it billows through the air.
Answers on how that smoke travels, according to Hao, will help shed light on how black carbon wafts from Russia to the Arctic, settling on sea ice and speeding up melts. Hao's team also scoops up smoke samples in metal canisters and brings them back to the lab. There, Hao cracks them open and employs smoke-sniffing equipment to better quantify the greenhouse gases and other air pollutants released from a wildfire.
Filling the gaps in future scenarios
Mark Finney, by contrast, works within a 40-foot cement room, building fires of his own -- on purpose. Within what the Forest Service bills as a one-of-a-kind combustion chamber, Finney is trying to fundamentally alter the scientific understanding of how fire travels from one particle to the next.
He is responsible for about half a dozen of the hallmark fire prediction models used on the ground. Now he is rewriting the calculations that drive most fire prediction models. The current program, developed in the 1970s, was fine for its day, he said, but "it is mostly empirical and doesn't explain why it works."
The equation limits a prediction model's effectiveness because, while it can forecast how fast and far fire might spread, it assumes that spread will happen; it won't tell you if it's likely to spread at all, he said.
Research ecologist Rachel Loehman camps in the woods for months at a time, doing painstaking data collection on how many trees are in a stand, what kind of understory vegetation surrounds them, and how much potential fuel is on the ground.
Back at the lab, she uses that data to build three-dimensional models of forestland. With that textured perspective, she constructs different scenarios of past and future fires -- hoping to better clarify how climate changes will likely affect a given region.
Though the Intergovernmental Panel on Climate Change laid out different forecasts of climate change's effects on the planet, the way it affects the local forests she studies will likely fall somewhere within that spectrum, she said. "Instead of using canned scenarios, I fill in all these gaps with these constructed combinations of temperature and precipitation change," she said. For land managers, knowing the details of what climate trajectory a region is on could make big differences in how they plan.
If Glacier National Park, for example, is poised to change from forest to grasslands in hotter, drier conditions -- resulting in less carbon sequestration, less suitable habitat for wildfire species and less water security for humans -- that is something land managers would want to know as soon as possible, Loehman said.
Revisiting fires of the past
Key components of understanding the future outlook, however, involve delving into the details forests can tell of the past.
To get some of those answers, Loehman, who is head of the climate-fire team at the lab, and researcher Emily Heyerdahl work in the woods of central Oregon, boring holes in trees with telltale charring. "Written records of fire are really short. They only go back to the 1970s," said Heyerdahl. Trees, however, have a much longer memory enmeshed in their rings -- assuming you get to them before wildfires engulf the tree and destroy those clues.
When fire laps up against a part of a tree, that stunts tree ring growth at that particular spot. But trees produce resin to act as a bandage over the vulnerable area and grow new rings over the injury. At a microscopic level, the ring variation reveals what environmental conditions were at the time of the burn. Researchers can tap the thousands of other tree samples stored in the lab to compare rings and then pin down the year -- and even season -- of a past fire.
Loehman then works this past fire data into her model simulations, projecting conditions perhaps 500 years down the road.
The tree data also provides more immediate outlook information, explained Heyerdahl. The Forest Service was able to predict that this year's wildfire season would be so mild thanks in part of the details provided from the tree rings, she said.
Heyerdahl can identify certain trends in the rings that indicate El Niño or La Niña years, and "if you see the same pattern in tree rings enough times, you can get statistical confidence," she said.
Figuring out how much of these complicated data can be fused together is a difficult process -- even for an advanced computer. Longer still is the time it takes to gather the field data for the computer to chew on. But Loehman hopes to have initial answers on future climate projections for central Oregon within the next year.
Research is more than a funnel where science is added at one end and a management tool drops out of the bottom, explained Loehman. "Most of the questions that we're exploring at the fire lab are very fundamental -- landscape resilience, tipping points, climate drivers, physical fire effects, etc. -- that ultimately can be of use to managers -- but we have to do the science first."
"The climate change stuff the lab is working on hasn't trickled down yet," affirmed Laura Ward, a district fire management officer for the Lolo National Forest. She wants that information so she can better plan for what do after flames die down or gauge what the emissions benefit would be of mitigating large fires, she said. For long-term planning, she said, it would just be good to know: "What happens if we do nothing?"