Forest Service scientists with the Southern Research Station (SRS) Center for Forest Disturbance Science (CFDS) recently tested Rabbit Rules, a simplified model they developed that can be used to quickly calculate fire and smoke behavior from prescribed burns.
The impacts of smoke on air quality are particularly important in the South, where managers use prescribed fire to treat 6 to 8 million acres of forest and agricultural lands each year. Almost anywhere they burn will be close enough to cities, towns, and suburbs for smoke to potentially impact air quality and human health. Complex terrain and fuels and variable weather contribute to the difficulty of determining precisely how a prescribed fire will spread or exactly where smoke will end up.
“We originally designed the model, Rabbit Rules, to provide data to Daysmoke, a model that simulates smoke plumes from prescribed burns,” says Gary Achtemeier, SRS research meteorologist with the CFDS atmosphere science team. “The model quickly showed that it could also be used to simulate fire spread and fire behavior.”
Rabbit Rules is a fire spread model based on the principle that very complex equations can be replaced by simple rules for computing purposes. The “rabbit” is an autonomous agent, a software entity that can act in its own interest in response to other elements—for instance, other rabbits and food—in its virtual environment.
“The purpose of using an autonomous agent is to reduce the complexity of fire modeling while keeping explanatory power,” says Achtemeier. “The rabbit seemed like a good icon to show how fire ‘hops’ from fuel element to fuel element as it spreads. Using a cellular grid, you can reduce the fire spread problem to finding out when a rabbit will jump, how far it will jump, and how long it will live.”
Rabbits are converted back to fire by “training” them to hop according to observed fire spread rates. When the model is run, the outcomes reveal complex distributions of fire, fire behavior, fire spread and fire weather.
Achtemeier compared simulations from Rabbit Rules with data from an experimental grass fire conducted in 2006 along the coastal plain in Texas. In an article published in September in the International Journal of Wildland Fire, he reported that the model accurately simulated wind patterns and observed fire behavior at the site. “It took five minutes on a laptop to run Rabbit Rules on this fire,” says Achtemeier. “Other available models would take three to four hours on a supercomputer. Rabbit Rules provides a place to start in determining if more complex analysis is needed.”
In 2011, Achtemeier and fellow researchers Scott Goodrick and Yongqiang Liu made what may be the first attempt ever to model prescribed burn set by aerial ignition. Rabbit Rules simulated 6,000 ignitions over a 50-minute period on a 1650-acre tract as part of an RxCADRE experiment at Eglin Air Force Base in Florida. “Rabbit Rules simulated the key variables such as the number of updraft core smoke plumes generated by the fire,” says Achtemeier. “Daysmoke, which simulates wildland fire plumes, requires the updraft core number as an input before it can be initialized. Managers can use Rabbit Rules for this important step when trying to determine whether conditions are right for burning.”
For more information, email Gary Achtemeier at email@example.com .