A family video making the rounds from Glacier National Park’s Reynolds Creek fire shows hints of a revolutionary breakthrough in fire science — if you know where to look.
And it turns out, we’ve been literally looking in the wrong place to understand how wildfires spread.
According to new research developed in part by scientists at the Missoula Fire Sciences Laboratory, we’ve been dazzled by the flames in the treetops when we really should be concentrating on forces near the ground.
Missouri teenager Lakota Duncan’s backseat video catches one opening in the trees where the wall of flames can be seen about 200 yards away. Tongues of fire are shooting from the treetops, racing the Duncans’ car as it speeds along Glacier's Going-to-the-Sun Road.
Back in Missoula’s U.S. Forest Service lab, research forester Mark Finney has been trying to explain how that spread occurs. Firefighters’ lives depend on outwitting such firestorms.
But for years, what Finney saw in the lab didn’t really answer the question.
“We wanted to know how does the flame front advance?” Finney said. “It’s a big fire management concern, so we’re looking for tools to mitigate fire spread. And we’ve been producing practical tools that do a good job for what they’re designed to do. But they don’t explain the physical process.”
For decades, most fire science assumed radiant heat and landscape were the driving forces in fire spread. The inferno grew according to how the winds pushed it, how drier or wetter the available fuels became, and how the terrain sloped up or down.
Understanding that should explain how the Reynolds Creek fire went from what Lakota Duncan thought was a small car fire on Tuesday afternoon to a 2,000-acre firestorm threatening Rising Sun Motor Inn by that evening. The fire was estimated at 3,166 acres on Saturday.
But research upended the model. It’s convection, not radiation that governs fire spread. Radiant heat is the warmth you feel standing next to a campfire. Convective heat is the burn you get when an ember lands on your foot.
In the case of a wildfire, it’s the superheated gases released by combustion that ignite more fuel.
But hot gas rises, right? How does a fire move forward if all the heat should be rising up and away from the trees?
“Radiation wasn’t igniting the fuels,” fellow research physical scientist Jack Cohen said. “That left us with a big problem. How do you keep those hot gases down so they can come in contact with the fuel?”
Furthermore, convection both heats and cools. Hot soup burns your tongue when it touches, but you can chill it by blowing on the spoon and convecting some of that heat away.
And it’s not the big logs but the little pine needles that really fuel a moving fire. Those fine fuels are fickle: They absorb heat quickly but they lose it just as fast. So fast, in fact, that a blast of wind from a fire front can actually cool fine fuels so much they won’t ignite by radiant heat.
If fire science were a car, Cohen said, we’ve become really good drivers but poor mechanics. We know pushing the gas pedal or brake makes the car go faster or slower but never looked under the hood to learn how the engine and transmission actually work.
“It’s the difference between experiencing something and understanding something,” Cohen said. “If you don’t ask questions, you don’t make discoveries.”
In a way that makes basic research seem almost like magic, the question they asked produced a discovery none of the scientists expected.
The problem involved a burn chamber used to study the spread of fire through different fuels — grass, pine needles, wood blocks, etc. In experiments, the researchers lay a layer of fuel on a platform, blow wind across it, light it, and film the results. This measures how fast flames move from one piece of fuel to another in different conditions — something firefighters see all the time.
But try as they might, they could never get a smooth burn. Some spot or another flared faster or slower. Scientists want to either control all variables in an experiment or eliminate them so they can be sure which caused what. They assumed the flare-ups were due to some batch of needles being extra-pitchy or oddly clumped. So they started using strips of cardboard with precisely cut ridges, like the teeth of a comb.
The goal was seeing how long and short ridges burned differently, which might give an idea how different fuel types spread fire. Instead, they exposed how fire actually works.
“We stumbled upon the principle that explains fire behavior,” Finney said. “It was a lucky thing.”
“Almost accidental,” added Fire Lab mechanical engineer Jason Forthofer. “We ended up seeing something we had never seen before.”
Once the oddities of natural fuel beds were removed, the researchers could see the basic mechanics of how flames move from one spot to another. And it turns out, we’ve always been looking in the wrong place.
“In a forest fire, you see peaks of flame with troughs in between,” Cohen said. “We’ve always been looking at the peaks of the flames, but that’s not where the action is.”
As a wall of advancing wildfire moves forward, the saw blade of peaks and troughs pulsates from below — not above. Between every two flaming peaks flows a pair of corkscrew vortexes that spin the superheated gas down and outward — into the easily ignitable grass and pine needles below.
The movement follows a classical — but weird — physics pattern that can be seen with the naked eye if you know what to look for.
“This is a well-understood principle in other forms of fluid dynamics,” Finney said. “We see this behavior in air ducts and sand dunes and beaches. But in fire, we’re just catching up to that.”
“We started looking at videos of actual fires, and they’re everywhere,” Cohen added. He pulled up news footage of the June 30 forest fire that burned 30 homes in Wenatchee, Washington. A helicopter shot clearly shows the peaks and troughs of fire, with visible swirls of smoke and flame running at ground level.
The Missoula Fire Science Lab scientists teamed with colleagues at the University of Maryland and University of Kentucky to flesh out the new observations.
They conducted outdoor experiments at Camp Swift, Texas, burning whole fields of grass and stands of trees. They proved the behavior they saw in little lab tests scaled up identically in real field conditions.
The whole thing has now been published in the Proceedings of the National Academy of Sciences, titled “Role of Buoyant Flame Dynamics in Wildfire Spread.” It’s coauthored by Michael Gollner at the University of Maryland and Kozo Saito at the University of Kentucky.
Like in a car, knowing how transmission gears transfer power from engine to wheels doesn’t help you get from Point A to Point B any quicker. But practical knowledge of the physics of fire behavior may still save both money and lives.
“One very common place where lots of people die in firefighting is a ‘chimney fire,’ where the fire is growing along a steep canyon wall,” Forthofer said. “As it’s burning along one side of the wall, it’s drafting a small amount of air. As it grows, it needs more air to draft, but the canyon wall restricts the air flow. Then suddenly you get this rapid transition and it jumps to both sides of the canyon. That might be the most common transition that kills firefighters.”
Firefighters are trained to recognize the kinds of fuels, weather conditions and topography where those flash transitions occur, Finney said, but they’re not taught the physics of how it happens.
The same principles are at work in a prescribed burn, where lines of workers with drip torches ignite vegetation to clear slash. Each worker’s fire is essentially a peak, and the spacing between them is the trough. The physics determines how effectively those little fires will grow together and clear out the burn zone.
“It seems like we should have understood this a long time before we did,” Finney said. “We spent all this time producing practical tools rather than stepping back and looking at the process.”