Computers Help Plan Fire Management

This winter a team of people at the Bureau of Land Management Alaska Fire Service is playing a computer game, but with a very serious purpose. The game is fire management, and at stake is the future cost of fighting forest fires in Alaska.

Headed by Fire Management Officer Dave Uebersbach, the team includes Mark Jones, Mike Silva, Joe Ribar, Dan Burrows, Vince Mazzier, Brian Fox, and Dave Dash. They're working with a computer model that knows the cost of the elements of fighting fires, and it's been given historically accurate fire and weather data for Alaska from 1980 to 1986. Team members try different responses to the imaginary but realistic fires the model sets for them, trying to find the most economical and effective effort.

"Our job is to find the point of least acreage burned for least cost," said Liebersbach. "The model generates a fictitious fire, and we tell it what resources it has to fight the fire. Based on such things as the historic average weather conditions, the number of smokejumpers or crews, their response time, and their rate of fireline construction, the computer predicts how large the fire will get and what the cost will be to put it out."

How much a fire costs is a complicated question. In the computer model, cost includes the budget for initial attack readiness, cost of actual suppression, and change in value of the land because of the fire. (It doesn't consider the cost of human life or property, though protecting these are top priority.)

The more ready resources like smokejumpers and retardant bomber planes at hand for the initial attack on a wildfire, the less the cost of putting it out--up to a point. Fires caught early are less expensive to put out, because they're smaller. As the budget for the immediate response increases, eventually most fires are quenched when they're first discovered. But after that, increasing the budget won't lead to better results.

In picking that perfect point of efficacy and economy, the planners can experiment with new techniques for making the vital first assault on a fire--all within the computerized game. "For example," team member Mark Jones said, "a faster aircraft would get smokejumpers to a fire more quickly, but might cost more to operate. Would this change lower or increase the overall firefighting cost? What if we reduced the air tanker fleet by one and hired more smokejumpers with the money saved--would that be a change for the better or worse? The computer model can answer this kind of question and let us find the most cost-effective means of accomplishing the goal."

Team head Liebersbach is also part of a six-person national BLM task force formed in 1985 to develop a procedure for planning the activity of fire management. It's not a simple process: fire management plans must meet requirements set in resource management documents written by federal, state, and Native land managers, who decide the fire protection priorities on the lands they manage. Nor is it an abstract concern: the fire plan justifies the annual average funding level for fire management.

The Alaska team's job is complicated because the BLM Alaska Fire Service is mandated to manage fires on all Department of the Interior and Native corporation lands in the state, but it actually fights fires only in the northern half of Alaska. Through an agreement with the state Division of Forestry, state fire fighters protect the southern part. The budget must take that into account.

Although based on an earlier U.S. Forest Service program, the actual computer model used nationally for BLM planning was developed in Fairbanks by fire service employee Lloyd Eggan. His development of the fire management analysis planning system software and Liebersbach's presence on the national task force have given Alaska a leading role in the Bureau of Land Management's nationwide fire management planning.