Behavior Modification: Tempering Fire at the Landscape Level

Joint Fire Science Program, February 27, 2008

With a history of management choices that have suppressed fire in the West, ecosystems in which fire would play a vital role have developed tremendous fuel loads. As a result, conditions are prime for fires to grow large, escape attack measures, and become catastrophic conflagrations that damage watersheds, forest resources, and homes. With a quiver of treatment options, land managers have successfully used prescribed burning and thinning to modify landscapes at the stand level. But planning treatments to modify fuel build up on a patch of forest is vastly different than planning treatments that could modify fire’s spread over larger landscapes.

Using information specific to a site, such as fuels, topography, and weather, simulations are run to identify the pathways fire would likely follow, the elements that would cause a fire to grow from moderate to severe, and the treatment options that would best modify the fuel load present. The simulations identify the best placement of treatment units and number of units on a landscape. Little is known about how long treatments will last, but studies suggest the benefits are limited to 10 to 15 years. To achieve desired effects in tempering fire’s behavior, land managers must apply optimally placed treatments at a rate of 1% to 2% per year.

Key Findings
• The pattern of fuel treatment units on a landscape is critical. Fuel treatment patterns placed optimally on a landscape (along a fire corridor) are roughly twice as efficient at changing large fire growth as random arrangements.

• When arranged in an optimal pattern, fuel treatment must occur at a rate of 1% to 2% per year to achieve reductions in large fire sizes or growth rates. The rate of treatment must produce treated area faster than the rate of plant regrowth and new fuel accumulation.

• Using sufficient treatment rates, the benefits of a fuel treatment program take about 1 to 2 decades to achieve.

• Long-term programs of fuel treatment involve maintenance of previously treated areas as well as implementation of new treatment units. The location of the treatment areas as they relate to the major corridors for fire spread are the most important factor in determining whether to maintain them.

• Variation in treatment unit sizes has the least impact on modifying large fires compared to treatment pattern and rate of treatment.

With a history of management that suppressed fire, lands in western North America bear conditions that foster the growth of large or “problem fires”—conflagrations that escape initial attack and spread far from where they start. Fire hazards, as a result, are greater, and threaten values—the safety of our homes and communities, the protection of our watersheds, and the beauty of our natural lands. Adding to the hazard, homes are built in ever greater numbers at the wildland-urban interface (WUI). In recent years, catastrophic fires have grabbed public attention, stimulating renewed interest in fuel treatments and prompting new research studies. Mark Finney, fire science researcher with the USDA Forest Service’s Missoula Fire Sciences Laboratory, has invested much of his time in providing answers and offering strategic planning tools to modify fire behavior.

Complexities of a conflagrant nature
“Fuel,” Finney offers, “is the only element of fire behavior that is manageable, since weather and topography are beyond human control.” Understanding how weather, topography, and fuels play a role in allowing fire to blow up is vital, but changing the character of fuels on the landscape and the arrangement of fuels offers the only possible means to resist a fire’s ability to turn into a disaster. To conduct experiments that haven’t been possible on actual, large-scale landscapes, Finney has developed a method to map how fires develop. But before he models fire behavior on a particular landscape, Finney looks at the real world concerns of stakeholders and those charged with managing the land. Equipped with a quiver of treatment options, having assembled data on weather patterns and plant species that comprise the fuel load on a given landscape, Finney places himself in the land manager’s role, with questions on predicting where fire will go, what fire will do, and how to curb fire’s energy so it can’t get out of hand.

Trial by (facsimile of) fire
A landscape of ponderosa pine forest located near Flagstaff, Arizona, historically burned at short-intervals, removing fuels on the ground, allowing plants to carry out their life histories. As management choices in the past continually prevented fire, this ponderosa pine forest now experiences, under extreme conditions, crown fires as the dominant fire pattern.

A California landscape in the Stanislaus National Forest in the heart of the central Sierra Nevada contains a mix of vegetation and ownership—its western edges representative of the wildland-urban intermix of the foothills. With fire excluded for the past century, surface and crown fuels now make a relatively continuous fuel complex with the potential for large and severe fires under extreme conditions.  The foothills of the central and northern Sierra Nevada have already experienced these kinds of fires, catastrophic events that have resulted in losses and costs in the hundreds of millions of dollars.

What is the best management plan to reduce fuels on each of these landscapes? Where should treatment sites be situated on the landscape to obstruct fire growth? What rate should treatments be applied to produce cumulative change in fire behavior? How fast does plant matter regrow? How do areas excluded from treatment possibilities—due to private ownership or wilderness designation, for example—affect treatment benefits? To study the California, western Montana, and eastern Washington sites that represent different forest conditions in the western United States, Finney set them afire, by simulation that is.

Using the Forest Vegetation Simulator (FVS) with its Fire and Fuels Extension (FFE), Finney and his collaborators were able to plug in critical variables: species types, tree diameters, stand height, canopy cover, canopy base height, canopy bulk density, fuel pools, treatment history, vegetation growth rates, topography, historical weather conditions that have produced fire spread directions and rates, wind speeds, and moisture levels. Then they ignited a spot and watched fire burn the land. Next, in another simulation, they selected treatment options and directed them at the modeled landscape to see which treatment or treatment combinations could weaken a fire’s ability to grow into a problem conflagration.

To test options and patterns, Finney ran comparisons that placed treatment units using optimal (along fire corridors, for example) versus random placements. The simulation models were able to calculate the impacts of treatments in terms of how fast a fire spread, how large a fire grew, and how likely an area would burn once fire grew large or escaped initial attack. The difficulty for planning managers in placing units and determining unit sizes, Finney found, is that actual landscapes, as opposed to unit sizes developed for models, contain complex variations in fuels, topography, wind direction, and fuel moisture. “Under complex conditions, the size and orientation of a given treatment is only efficient in the context of other possible units encountered immediately before and after the fire moving across the landscape,” Finney notes. “Each unit modifies the path of fire into succeeding units.”

To accommodate how fire behaves among multiple units, the algorithm Finney developed divides the landscape into a series of parallel strips oriented perpendicularly to the main fire spread direction. For each strip, beginning with the upwind strip that was ignited, fire growth and minimum travel routes are computed. The procedure identifies treatment units within the strip that have the best sizes and shapes for efficiently retarding fire growth.

Living with the learning
When hoping to modify the behavior of a potentially voracious fire, the good solutions, according to Finney, seem to be “greedy” ones. Greedy solutions, he explains, are chosen from only locally available information—but information that considers how fire moves across the landscape. Places where fire moves easily are distinguished from places where fire spread is difficult. “These pathways are found by simulating fire movement across the landscape and account for the complex spatial patterns of the fuels, topography, and wind direction that are not local at all,” Finney explains. After identifying the places where fuels, topography, and wind direction allow fire to move easily, solutions can be made that consider only these local pathways. The advantages of these solutions are twofold: faster computation times in the model because fire growth does not have to be simulated far downwind from the strip that was first ignited and, more importantly for fire management applications, the solutions place a treatment unit on a locally major pathway of fire movement, which increases the likelihood that a well-placed treatment unit will be near a random ignition source on the landscape. Rather than reworking and modifying hundreds of thousands of acres on a landscape-wide level, at great cost, labor, and time, Finney has determined that applying annual treatments strategically can produce the desired effects in modifying fire’s behavior.  Fuel treatments can be designed to decrease burn probability by considering both the treatment used at the stand level and at the landscape level.

To achieve a pattern across a landscape that will inhibit fire’s behavior, land managers must make the annual rate of treatment or maintenance high enough to outpace the rate that vegetation will regrow, adding new fuels. Little is known about how long treatments last, but a few studies suggest that the benefits are limited to 10 to 15 years. Using the best treatment options (e.g., thinning, prescribed burning), fuel treatment arrangements that are optimal in disrupting the growth of large fires require that at least 1% to 2% of the landscape be treated each year.  Even if spot fires jump into units that have been treated, an extensive landscape pattern of treated units would interrupt any new fires. Simulations showed randomly arranged units with the same treatments applied as used in an optimal pattern required twice as many units to reduce fire growth.

Timeframes for treatment longevity mean that land managers must make a substantial effort for roughly two decades to realize fuel treatment benefits that can temper fire’s behavior. Since higher treatment rates seemed to accelerate the production of benefits, higher rates might be desirable in the first decade followed by later decreases. Variation in treatment unit sizes had the least impact on modifying large fires compared to treatment pattern and rate of treatment. In the absence of spot fires, large and small units were found to produce similar reductions in fire sizes, spread rates, and burn probabilities. Small units, though, may not effectively block fire through large corridors where fire moves easily.

As with rapid recovery of fuels after treatment, limitations on the placement of treatment areas reduce the effectiveness of a plan to inhibit large fires. Simulations showed restrictions on placing treatment areas (imposed for a variety of reasons, including concern for wildlife habitat, proximity to streams or rivers, road access, budget limitations, or ownership), despite the need for treatment at these locations, decreased the effectiveness of an optimal treatment by 50%. Treatment restrictions amounting to more than about 40% of a landscape would diminish any advantage an optimal solution would achieve over purely random treatment placement. If land managers want to reduce large fires, collaboration with all concerned parties would be necessary to permit treatment at locations necessary to achieve landscape-level changes that would prevent fire from raging out of control. Under healthier conditions, fire could return to its habitat.

Management Implications
• When hoping to modify the behavior of a large fire, the best solutions use locally available information on where fire movement is made easy by the alignment of fuels, topography, and wind direction. Treatment units, as a result, are placed on a locally major pathway for fire.

• Land management activities that exclude areas from treatment can completely remove the benefit of an optimal treatment strategy. Restricting treatment in necessary areas reduces the effectiveness of optimal treatment patterns by 50%. If land managers intend to achieve reductions in large fires, collaboration with all concerned parties would be necessary to achieve landscape-level effects.

Further Information: Publications and Web Resources

Finney MA. 2001. Design of regular landscape fuel treatment patterns for modifying fire growth and behavior. For Sci. 47(2):219-228.

Finney MA. 2003. Calculation of fire spread rates across random landscapes. Intl J Wildl Fire. 12(2):167-174.

Finney MA. 2007. A computational method for optimizing fuel treatment locations. Intl J Wildl Fire 16:702-711.

Finney MA., Seli RC, McHugh CW, Ager AA, Bahro B, and Agee JK. 2007.  Simulation of long-term landscape-level fuel treatment effects on large wildfires. Intl J Wildl Fire. 16:712-727.

Scientist Profile
Mark A. Finney is a Research Forester with the USDA Forest Service, Rocky Mountain Research Station, in the Fire Behavior Project at the Fire Sciences Laboratory in Missoula, Montana. His current research interests are fire behavior and fire behavior modeling. Before taking his current position, he spent 7 years as a research scientist and consultant, and 2 years as a fire ecologist with Sequoia National Park. He received his Ph.D. in wildland fire science from the University of California at Berkeley in 1991 studying prescribed fire and effects in the coast redwood forests. He earned his M.S. in fire ecology at the University of Washington in 1986 working on fire history and effects in North Cascades National Park. He has a B.S. degree in forestry from Colorado State University (1984).
Mark Finney can be reached at: USDA Forest Service, Missoula Fire Sciences Laboratory, PO Box 8089 Missoula, MT 59808 Phone: 406-329-4832 E-mail:

Alan A. Ager, USDA Forest Service, Pacific Northwest Research Station
Bernhard Bahro, USDA Forest Service, Pacific Southwest Region
Charles W. McHugh and Robert C. Seli, USDA Forest Service, Rocky Mountain Research Station, Missoula Fire Sciences Laboratory
James K. Agee, University of Washington, College of Forest Resources