Don McKenzie, Pacific Wildland Fire Sciences Lab; Faith Ann Heinsch, Rocky Mountain Research Station; and Warren E. Heilman, Northern Research Station.
An archived version of this topic paper is available.
Fire has been a part of most of Earth's ecosystems for millions of years (1). Fire interacts with climate and vegetation (fuel) in predictable ways that are documented in lake-sediment records (2,3,4,5), fire-scarred trees (6,7,8), forest age-class distributions (9), and contemporary instrumental records (10,11). Understanding climate-fire-vegetation interactions is essential for addressing nation-wide issues associated with climate change, particularly
- effects on regional circulation and other atmospheric patterns that affect fire weather (12).
- effects of changing fire regimes on the carbon cycle, forest structure, and species composition (13).
- complications from land-use change, invasive species, and an increasing wildland-urban interface (14).
Large fires (>300 acres) account for more than 95 percent of the area burned by wildfires in the United States in a given year. These fires are frequently associated with specific mesoscale (~5-1000 km) and broad scale (>1000 km) atmospheric circulation, temperature, and moisture patterns in different regions of the U.S. (15,16,17). The short-term weather conditions that are conducive to severe fires (i.e. fire weather) are manifestations of these patterns. For example, in maritime forests of the Pacific Northwest, hot dry conditions combined with easterly winds are necessary to lower fuel moisture to the point of supporting widespread fire (18). Similar conditions, associated with blocking ridges, account for much of the area burned in the Canadian boreal forest (19,20). In a very different ecosystem, southern California chaparral, seasonal weather in the form of Santa Ana winds accounts for most if not all of the large severe fires. Issues for land management generally arise at these regional scales, but projecting how they will change in response to larger-scale patterns is more challenging than projecting large-scale patterns themselves into the future.
Interactions between climate, fire, and vegetation, and their consequences, vary regionally and among ecosystem types within regions. For example, most forest ecosystems across the northwestern United States and Alaska have abundant fuels, and total fire extent across this region varies with fuel condition -- regional drought years produce low fuel moisture and high flammability (11). In contrast, in arid mountains and rangelands of the Southwest, alternating wet and dry years, associated with cycles of the El Niño Southern Oscillation (ENSO) and changing fuel abundance and connectivity, account for much of the interannual variability in area burned (6,11,21). In the American Southeast, where fire plays a major role in ecosystems, fuels are abundant, and total fire extent varies with fuel condition, but changes in fire extent are also tied to the ENSO cycle, with La Niña years bringing reduced precipitation and increased fire (22). In drier forests US-wide, and often in moist southeastern forests, large fires can be the result of a "perfect storm" of weather conditions, fuel availability, and stand conditions. For example, portions of the Tripod Complex Fire in north central Washington that burned with highest intensity had been attacked recently by mountain pine beetles, leaving abundant dead fuels in the canopy that were vulnerable to high-intensity crown fire (see images below) (23).
Climate warming associated with elevated greenhouse-gas concentrations may create an atmospheric and fuel environment that is more conducive to large severe fires. The consensus view of the Intergovernmental Panel on Climate Change is that projected higher summer temperatures will likely increase the annual window of high fire risk by 10-30%, leading to a potential increase in area burned in Canada of 74-118% by the year 2100 (24,25,26). General circulation model studies suggest that fire occurrence or area burned could increase across North America under a doubled CO2 environment because of increases in lightning activity, the frequency of surface pressure and associated circulation patterns conducive to surface drying, and fire-weather conditions in general that are conducive to severe wildfires (15,27,28,29).
Predictions past about 2040 are largely speculative given the current rate of increase in fossil fuel emissions (30), but a warmer climate will certainly amplify the effects of drought and is expected to increase the number of days in a year with flammable fuels, thereby extending fire seasons and area burned in ecoregions where fire extent is linked to fuel conditions (11). Most forests in the western United States fall into this category. In arid ecosystems, increased drought theoretically could reduce fuels to the point at which annual fire extent actually decreases (31). Statistical models linking area burned to water-balance deficit suggest that both these patterns may obtain (30), undermining coarse-scale predictions of huge increases in area burned in a warmer climate. Total forest area may also decrease with conversion of some forests to grasslands from the effects of more frequent fire (32). Further confounding predictions for arid ecosystems could be the continued spread of invasive plants such as buffelgrass and cheatgrass (33,34), which will increase fuel connectivity drastically. Overall, more fire is expected in western forests and rangelands for the foreseeable future, because of the preponderance of ecosystem types in which drought is strongly correlated with area burned. In the East, complex spatial patterns of land use and active prescribed fire programs make broad-scale predictions of area burned even more difficult than for the West; this requires ongoing research.
Future changes in fire frequency and severity are much harder to predict. Global and regional climate changes associated with elevated greenhouse gas concentrations could alter large weather patterns, thereby affecting fire-weather conducive to extreme fire behavior. The inherent complexity of analysis at "landscape scales" requires understanding both this atmospheric component and finer-scale controls on the ground. Even in a so-called "megafire" such as the Tripod Complex, fine-scale controls on fire spread, such as topographic barriers, can substantially change the spatial pattern of fire severity, both for individual fires and for fire regimes in the aggregate. At present we are able to suggest that at broad scales we can expect more large severe fires (increasing area burned at high severity), but are much less able to predict changes in fire severity within individual fires. Fire severity provides a negative feedback to fire frequency, particularly for area-based metrics of fire frequency such as the natural fire rotation or the fire cycle. It is therefore difficult to predict global changes in fire frequency in a warming climate, although estimates for specific fire regimes may be more tractable (35).
Options for Management
In some western dry forests, particularly those affected by 20th-century fire exclusion, thinning and surface fuel treatment (including prescribed burning) can reduce fire severity and fire hazard (36), although maintenance treatments may be required every 20 to 40 years. Strategic placement of treatments can greatly increase the effective area treated (37). In unmanaged forests, especially in areas in which fire suppression is difficult, expensive, or counterproductive to resource objectives, managers can take advantage of the self-limiting nature of wildfire. Fire spread rates and severity are reduced when a fire reaches a recently burned area (38).
Fuel treatments will be challenging to implement at spatial scales large enough to have much impact, especially if wildfire increases greatly in the future, but can enhance resilience on specific landscapes with high resource, economic, or political values (e.g., the wildland-urban interface). In the Southeast, undergrowth may grow even faster in warmer temperatures. Management practices may need to respond to an increase in available fuels, while anticipating a shortening of the prescribed burning "season", particularly in Florida (40), and increased scrutiny of prescribed fire for unacceptable degradation of air quality, which will be exacerbated by climate change (41).
Some general guidelines for adaptation (40,41,42)
- Increase landscape diversity -- increase large-scale resilience, size of management units, and connectivity.
- Maintain biological diversity -- experiment with species and genotype mixes, and identify species, populations, and communities that are sensitive to increased fire and develop conservation plans for them.
- Plan for post-disturbance management -- treat fire and other ecological disturbances as normal processes and incorporate fire management into planning.
- Maintain and improve the resilience of watersheds and aquatic ecosystems by implementing practices that protect, maintain, and restore watershed processes and services.
- Implement early detection and rapid response -- monitor post-fire conditions, and eliminate or control exotic species early on.
- Manage for realistic outcomes -- identify key thresholds and prioritize projects with a high probability of success; abandon hopeless causes; consider even alternatives that might be undesirable in an unchanging climate.
- Incorporate climate change into restoration -- avoid trying to replicate historical conditions, but continue to learn lessons from historical variation.
- Develop regulations and policies that take climate change into account -- raise awareness with stakeholders, and work with local stakeholders from the onset of projects.
- Anticipate big surprises -- expect mega droughts, larger fires, species extirpations, loss of resilience and system collapses, and incorporate these events in planning.
For information on ongoing Tripod Fire research, contact Susan Prichard.
1 acre = 0.405 hectares
Degree Fahrenheit (F) to Celsius (F-32) x 0.56 = C