Air, wind, and fire

Small experiments help with a real big challenge

Imagine water flowing smoothly down a river. When it hits a rock, the flow of the water moves and bends, creating turbulence. Air moves in a similar fashion as it flows through a forest and encounters objects and other movement in its path — but we can’t see it. Unless something is present to help us see its movement — like smoke, fog, or fire.

“Fire is a part of the air. It has a chemical reaction that releases heat and light, which enables us to see the movement of the air,” says USDA Forest Service scientist Scott Goodrick.

Complex patterns of fire behavior driven by fire-atmosphere interactions on a prescribed fire in pine flatwoods at Eglin Air Force Base in Florida. USDA Forest Service photo by Joseph O’Brien.

Goodrick has been studying fire-atmosphere interactions for years. He recently coauthored a paper in Nature that sheds light on how fire influences the winds, and vice versa, through data collected in a small-scale, real-world fire experiment.

“When wind moves through trees in a forest, it creates turbulence,” says Goodrick. “Turbulence increases the supply of oxygen, which affects fire behavior. Turbulence generated in one part of the fire can move through and affect other areas of the fire.”

This study is the first experiment investigating fire turbulence conducted outside rather than in a laboratory setting. For this study, Goodrick and colleagues from the University of California, Irvine scattered pine needles across the ground in a controlled area to create a heterogeneous environment — so the fire would act more like it would in a real landscape. Then they lit a fire and took images using high frequency Particle Image Velocimetry (PIV) to look at the air flow in and around the fire. PIV is an optical method used to visualize the fluid flow. In this case, the fluid is air, and PIV allows them to measure the flow around a fire.

“The PIV captured 30 images every second,” says Goodrick. “In these series, we can see how flames move from frame to frame; how smoke moves; how ash or a piece of grass moves. Then we use a bunch of fancy mathematics to describe the movement.”

Taking the images with the PIV camera relies on being able to track differences between frames of a video and translating these differences into motions. These differences could be from bits of ash being blown by the wind, wisps of smoke, or even ,subtle optical variations caused by the fact that hot air has a slight difference in density than cooler air. The camera can detect this even if our eyes can’t see it clearly — like the heat from a barbeque.

When the fire is lit, hot air and flame move upwards and cool air fills in underneath to replace the warmer air. As the fire grows, the movements of the air and the flames become more chaotic and complex. There are pressure differences between the inner, burned-out circle and outer, unburned fuel that cause interesting fire dynamics.

Illustration of flow patterns associated with spreading point ignition, with areas of updrafts over the fire and areas of downdrafts both inside and outside the fire.

“There are counter-rotating vortex pairs,” says Goodrick. “This vertical circular motion pushes hot gas outwards, preheating and drying out the fuels. That can result in fire bursts.”

Understanding how fire influences air movement in small-scale but real-life settings give researchers insight into wind patterns in larger fires. Wildfires are increasing in numbers and intensity across the globe. Prescribed burning is needed to restore the natural fire regime in many forest systems. With climate change and excessive fuel loading, we need models that can accurately predict fire behavior. Researchers will use data from this study to test the accuracy of fire models and improve them to better predict what will happen during prescribed and wildland fires.

Goodrick was instrumental in developing QUIC-fire, the first fire-atmosphere model that can be run on a laptop computer to generate fire predictions rather than a supercomputer. “Getting modeling tools into the hands of practitioners is a main goal,” says Goodrick. “We’re also working with partners to study previous fires and then adjust the models to better predict what future fires will do.”

The Forest Service’s 10 Year Strategy for Confronting the Wildfire Crisis aims to significantly increase acres burned with prescribed and managed fire. It is essential for managers to have fire models that accurately predict fire behavior under a variety of climatic and environmental conditions so they can make the safest decisions possible.

Read the full article in Nature.

For more information, email Scott Goodrick at

Find the latest publications by SRS scientists.