It's wildfire season in the western U.S., and one fire in particular has taken the spotlight this week. Burning thousands of acres, the High Park Fire
has already killed one woman and destroyed over 100 structures near Fort Collins, CO.
Two summers ago, the Fourmile Canyon Fire became the most devastating wildfire in my home state of Colorado's history, generating $217 million in insurance claims. Shortly thereafter, I wrote an article for my school magazine about groups of physicists, IT specialists, and climate scientists who use computer models to predict and track wildfires. Although wildfire modeling has developed significantly over time, its roots still exert great influence thanks to the field's "father": Dick Rothermel.
A view of the High Park Fire from space. Image Credit: NASA
An aeronautical engineer by training, Rothermel transitioned into fire modeling for the USDA in the 1960's and 1970's. By 1972, he became the first person to develop a set of equations to model the spread of wildfire, and his legacy pervades the variety of modeling software available today.
Rothermel wanted to measure how quickly a wildfire would advance given certain environmental conditions. Primarily, wind velocities, the type of vegetation and the local topography determine the spread of a wildfire. By plugging in these initial conditions to the Rothermel equation, modelers can get a rough estimate of a fire's speed, enabling more strategic firefighting.
The original Rothermel equation. The variables on the right side of the equation correspond to various wind, fuel and topography factors that determine wildfire spreading.
Although Rothermel's equations gave modelers a good starting point, wildfires are a complicated beast. Variable weather conditions can be extremely difficult to model, and a number of wildfire phenomena can cause erratic spreading.
For instance, the hottest wildfires can cause the tops of trees to ignite, shooting debris in all directions. When these flaming branches land – sometimes as far as a mile away – new hotspots emerge. In addition, wind speeds and directions can vary significantly over time in small areas, adding another layer of complexity that the models have to capture.
But refinements of the original Rothermel equation have made strides. Scientists working on wild fires have tried to account for as many variables as possible, and the advent of supercomputers has allowed them to predict the spread of wildfires as they develop. In fact, the U.S. Forest Service has started to incorporate modeling into their real-time firefighting.
When I was writing my article on wildfires two summers ago, I interviewed a modeling specialist at IBM, a researcher with a government agency, and even researchers who analyzed the social media response to natural disasters. As computing power has increased and social media have exerted greater influence, research into wildfire modeling has burgeoned. Hopefully, models will become even more useful in the future.
For a very thorough overview of Dick Rothermel and his contributions to fire science, see this article
(PDF) from Fire Science Digest.
My old article, which includes some more information about wildfire modeling, can be found here
If you want to keep up with Hyperspace, AKA Brian, you can follow him on Twitter