WWETAC Projects

Project Title: Wildfire Risk Analysis at the National Forest Scale

Status: Ongoing

Principal Investigator: Alan Ager, Western Wildland Environmental Threat Assessment Center  

Collaborators: Mark Finney, Carol Miller and Mark Parisien, Rocky Mountain Research Station, Missoula, MT; Deschutes National Forest, Bend, OR; Nicole Vaillant, Western Wildland Environmental Threat Assessment Center

E-mail Contact: aager[at]fs.fed.us

Key Issue: Federal land management agencies struggle with the preservation of critical habitat and other highly valued resources on fire-prone landscapes in the western US.  Wildfire continues to erode existing habitat networks for Federally-listed species. The policy debate over active versus passive management has generated many landscape management questions that remain to be resolved among agencies.  The problem is particularly acute for species like the Northern spotted owl that inhabit the eastside Cascade National Forests in Oregon and Washington.

Goals and Objectives: In this study, we demonstrate the application of risk assessment tools to examine wildfire risk among and within an array of conservation reserves and other land management designations on the Deschutes National Forest.  In particular, we posed three questions (1) What is the relative wildfire risk among the array of management designations on the Forest; (2) Are specific conservation reserves contributing to wildfire risk and to the loss of other highly valued resources?  (3) How do fuel treatments reduce wildfire risk when conservation reserves are excluded from treatment?

Description of Work: Burn probabilities at specific intensity classes were estimated for each 90 x 90 m pixel by simulating 50,000 wildfires under burn conditions that replicated recent severe wildfire events on the Forest.  For each polygon within each management area and conservation reserve we calculated: (1) the average burn probability, (2) the expected flame length, and (3) the average size of fire generated by an ignition within the polygon. 

Deschutes NF Management Objectives Map


Management Areas on the Deschutes NF
Reserves                                        Area (Ha)
Owl Critical Habitat                   38,421
Owl Home Range                     50,138
LSR                                        89,108
Aquatic Buffers                        15,705
Bald Eagle Habitat                     8,598
Old Growth                              16,984
Research Natural Areas              3,502
Intensive Recreation Areas       26,648
Scenic View Retention               76,857
Special Interest Areas                8,200
Wildland Urban Interface          82,023
Managed Forest                     256,331

The land allocations and conservation reserves on the Deschutes NF create a mosaic of management objectives unrecognized by wildfire.  General Forest where fuel treatments are concentrated is shown in light green.   

burn probability and flame length graph


The plot to the right shows the average burn probability and expected flame length for each polygon in each management area.  Increasing burn probability indicates higher likelihood of burning given a random ignition on the Forest.  Increasing expected flame length measures the fire intensity given a wildfire occurs.  Risk is a combination of likelihood and loss (i.e. intensity), hence data points to the upper right indicate highest wildfire risk areas.  The plot shows high variability among the polygons in each management area.  Old growth and visual areas show the highest likelihood of burning.  Fish habitat reserves show some of the highest flame lengths.  The owl reserves show relative high flame lengths and low probability of burning.

fire size to burn probability ration map

Wildfire sources and sinks were examined by calculating the ratio of burn probability to the fire size generated by an ignition.  The ratio is a relative measure of how much a management area contributes versus receives fire in the surrounding landscape.  Areas that have high burn probability but do not generate large fires from an ignition are a wildfire sink.  Conversely, locations that generate large fires when an ignition occurs and have low burn probability are wildfire sources.  

Map showing ratio of fire size to burn probability for a portion of the Deschutes NF.  High values (red) are wildfire sources, and low values (blue) are sinks.

The plot below shows the average burn probability and fire size for individual polygons within each management area.  Specific polygons within each of the management areas can be identified as wildfire sources (high fire size).  In particular, riparian buffers and owl home ranges show high fire size relative to burn probability.

average burn probability and fire size graph

burn probability and fire size graphA fuel treatment priority map for the Forest was used to simulate fuel treatments and examine change in wildfire risk.  The treatment scenario called for 64,000 ha of treatments in the general forest management areas.  The ratio of the burn probability after and before the treatments was used to examine change in wildfire likelihood.  The analysis suggested large reductions in burn probability to conservation and other reserves.  For instance, the likelihood of a fire in the old growth reserves was 30% of the pre-treatment conditions. The effect of treatments on both burn probability and fire size for specific reserves like old growth show large reductions post treatment.

relative risk post-treatment graph

Summary of Results: This work demonstrated the application of risk assessment to analyze the relative risk to human and ecological values at the national forest scale.  The analyses revealed spatial variation in wildfire risk that is useful in prioritizing fuels treatments and guiding other wildfire mitigation activities.  The work also illuminated the conflict between biodiversity conservation efforts on federally-managed lands and the high wildfire risk on fire-prone landscapes.

Information from these analyses can guide future conservation efforts for TES habitat affected by wildland fire.  For instance, if there are optimal patterns of fuel treatments to slow the spread of fire, then it follows that there are optimal patterns of conservation reserves to avoid fire.  The arrangement of fuels, the proportion of the reserve relative to the matrix, the shape of the reserves and the arrangement of the reserves relative to dominant fire weather and the topology of fire are all important factors.  Spatial patterns of wildfire risk will be a key factor to design adaptive reserves and sustain habitat. 

The analysis shown here can be easily extended to other national forests to prioritize fuel treatments and analyze the manifold effects of conservation and other reserves on wildfire risk to ecological and human values.

Deliverables and Citations:

Ager, A.A., Finney M.A., Vaillant, N. (in prep).  Wildfire risk among conservation strategies on fire prone landscapes.

Ager, A.A., Finney, M.A. 2009.  Application of wildfire simulation models for risk analysis.  Geophysical Research Abstracts, Vol. 11, EGU2009-5489, 2009. 

Parisien, M.-A., C. Miller, A.A. Ager, and M.A. Finney.  2010. Use of artificial landscapes to isolate controls on burn probability. Landscape Ecology 25:79-93.

Miller, C., M.-A. Parisien, A.A. Ager, and M.A. Finney. 2008. Evaluating spatially-explicit burn probabilities for strategic fire management planning. WIT Transactions on Ecology and the Environment 119: 245-252.

Seli, R.C., Ager, A.A., Crookston, N.L., Finney, M.A., Bahro,  B., Agee, .J.K., McHugh, C.W.  2008.  Linking a Spatially Explicit Fire Model to the Forest Vegetation Simulator. Proceedings of the third FVS conference, Fort Collins, CO. February 13-15, 2007.  RMRS-P-54 p27-39.

Finney, M.A., Seli, R.C., McHugh, C., Ager, A.Bahro, B., Agee, J.K. 2007.  Simulation of long-term landscape-level fuel treatment effects on large wildfires.  International Journal of Wildland Fire 16:712–727

Kerns, B., Ager, A. 2007.  Risk assessment for biodiversity in Pacific Northwest forests.  Forest Ecology and Management 246:38-44

Project ID: FY10AA69