Assessing Risks to Multiple Resources Affected by Wildfire and Fuels Treatment using an Integrated Probabilistic Framework

Steven P. Norman, Sandra Jacobson, Christine Damiani, and Danny C. Lee

USDA Forest Service Eastern Forest Environmental Threat Assessment Center (1,4) and Pacific Southwest Research Station (2-3)

The tradeoffs that surround forest management are inherently complex, often involving multiple temporal and spatial scales. For example, conflicts may result when fuel treatments are designed to mediate long-term fuel hazards, but activities could impair sensitive aquatic habitat or degrade wildlife habitat in the short term. This complexity makes it hard for managers to describe and communicate the conditional nature of risk and to justify planned activities to stakeholders. In addition, our understanding of how proposed activities will affect resources of concern is often limited due to informational shortcomings and imprecise models. To be robust and transparent, a risk assessment framework needs to reveal these limitations while quantifying the probable outcomes of project effects to multiple resources of concern. In this paper, we describe the effects of fuel treatments using such a planning framework, called CRAFT (Comparative Risk Assessment Framework and Tools). CRAFT provides a platform from which diverse ancillary models and other relevant information can be transparently integrated and evaluated.

We conducted our case study in a portion of the Hayfork Adaptive Management Area of the Shasta-Trinity National Forest, California. As is typical of other mixed-conifer forests of California, this area has experienced decades of fire suppression and severe fire effects are increasingly likely. With Forest managers, we identified a range of measurable objectives involving the Wildland Urban Interface, fire behavior, fire effects and sensitive wildlife. We then developed a conceptual model describing how components of the system inter-relate. From this, we developed a probabilistic framework, using Bayesian Belief Networks, in which we employed existing fire and vegetation models to address the effects of fuel treatments on various resources for different fire weather and management scenarios. Our model provides decision makers and stakeholders with insight into the conditional probability that efforts conducted to reduce fire-related risks will be successful.

Wednesday Morning Plenary

corresponding author:

Steven P. Norman
USDA Forest Service
Pacific Southwest Research Station
1700 Bayview Drive
Arcata, CA 95521-6013
stevenorman@fs.fed.us