An Actuarial Approach to Modeling Wildfire Risk

Alan Ager and Mark Finney

USDA Forest Service, Pacific Northwest Research Station and Rocky Mountain Research StationA number of wildfire risk systems have been developed in recent years to provide land managers with tools to examine potential wildfire impacts.  However, few of these efforts use well-established concepts and definitions of risk from the actuarial sciences, and none are sufficiently detailed for watershed-scale fuels treatment planning.  In the context of wildfire, risk is the expected loss from a fire, calculated as the product of (1) probability of a fire at a specific intensity and location, and (2) the resulting financial or ecological damage.  The process of wildfire risk assessment is concerned with changes in expected loss in response to fuel treatments, suppression,  structure improvements, and assumptions about fire weather.  We developed a wildfire risk model based on the expected loss concept and tested it on 16,000 ha wildland-urban interface in Eastern Oregon.  The modeling approach was designed to be compatible with ongoing wildfire risk mitigation projects on federal lands in terms of resolution and data requirements.  Conditional wildfire probabilities were calculated by simulating large numbers of wildfires via  a mechanistic wildfire spread model.  Fire weather was simulated using Monte Carlo methods and historical weather data.  A financial loss function used flame length at each pixel to determine the fire effects on standing timber volumes and residential structures.  We also considered the positive value of low intensity wildfire in terms of reducing fuel loads and future wildfire intensity.  We simulated a range of fuel treatment alternatives and examined the net change in expected loss. The results demonstrated the importance of considering both the probability of a disturbance and the potential loss in a risk assessment, especially for spatially heterogeneous disturbances like fire.  Future work will include the development of loss functions for other resources of concern.  Our work advances the application of actuarial science to wildfire risk management and fuels treatment planning on Federally-managed lands.

Wednesday Morning Plenary

corresponding author:

Alan Ager
USDA Forest Service
Pacific Northwest Research Station
1401 Gekeler Lane
La Grande, OR 97850-3368
541-278-3740
aager@fs.fed.us