Forest ThreatNet

If forest threats were judged solely on the basis of their prominence in agency spending, wildland fire would have no equal. Each year, the Federal government spends more than $2 billion on wildland fire prevention, preparedness, suppression, and recovery. These funds are distributed among five Federal agencies—Forest Service, National Park Service, U.S. Fish and Wildlife Service, Bureau of Indian Affairs, and Bureau of Land Management. As the extent and severity of wildland fires has increased in recent years due to multiple reasons, so has the demand for more funding. Increased funding has led to greater scrutiny by the General Accounting Office and others regarding the Federal fire program’s effectiveness.

The Fire Program Analysis (FPA) is a special effort created specifically to respond to these concerns. FPA is a multi-agency effort that develops common interagency decision support tools for wildland fire planning and budgeting. This effort enables Federal wildland fire managers to jointly plan for funding allocations that cost-effectively accomplish interagency objectives. FPA also encourages State and local wildland fire partners to participate.

Above right: Lightning-ignited wildfire burning during day, Blue Mountains. (Photo credit: Dave Powell, USDA Forest Service, www.bugwood.org)

Since August 2006, EFETAC Director Danny C. Lee has co-led the Interagency Science Team (IST), commissioned to provide FPA with scientific support. The original IST—which included 13 scientists from the Forest Service, the U.S. Geological Survey, and academia—designed an analytical system to meet FPA needs.

"The IST design was used by the FPA development team to create a prototype system, which should be fully operational this June," explains Lee. The system will analyze budget alternatives locally and nationally; determine relative costs and benefits for the full scope of fire management activities; and identify cost-effective mixes of personnel and assets to implement these activities within a specified budget. IST scientists continue to work closely with the FPA team on this effort.

The IST design incorporated a combination of simulation models, GIS analyses, and decision support tools to model potential costs and consequences of alternative investment strategies. Two separate simulation models play important and complementary roles. The initial response model simulates local responses to random wildland fire ignitions. Fires that escape initial response or are allowed to burn for resource benefits are simulated using a large fire model. Both models are based on tools previously developed by IST members and Forest Service researchers, Jeremy Fried (Pacific Northwest Research Station) and Mark Finney (Rocky Mountain Research Station). The FPA system uses results from the simulation models in a larger decision support framework that allows rigorous and systematic analysis of tradeoffs among alternatives.

The large fire model has been challenging to design and implement. Large fires typically exceed 300 acres and can grow to 500,000 acres or more. In much of the U.S.—especially western states—large fires are a small percentage of those started, but they torch the bulk of acres burned and account for the most suppression costs incurred. Even advanced computers are challenged to simulate the numerous fire possibilities that could burn under multiple alternatives, especially considering the short time window allowed for each FPA analysis. To overcome this hurdle, Lee, Finney, and FPA team members created a statistical approach that reasonably approximates the large fire model’s results. This approach uses simulation results to produce a "model of the model" that can very quickly analyze multiple alternatives and produce spatially explicit estimates of burn probabilities and the flame intensities expected across a landscape.

Please visit FPA’s Web site at http://www.fpa.nifc.gov for additional information.

Above left: Georgia fires raged in 2007. (Photo credit: National Interagency Fire Center Archive, www.bugwood.org)