2005).

​

Numerous research studies have used GCMs to simulate the future climate; these models include three-dimensional representations of the atmosphere, ocean, cryosphere and land surface and the parameterizations of the associated physical processes. Future climate scenarios are built based on the effects of various concentrations of greenhouse gases and other pollutants within the atmosphere on our earth-atmosphere system. Transient simulations are available from GCMs which allow examination of the possible rates of change in the climate. The major areas of uncertainty in GCMs include clouds and their radiative effects, the hydrological balance over land surfaces and the heat flux at the ocean surface. Despite these uncertainties, GCMs provide the best means available to estimate the impact of changes in the future climate on the fire regime at larger scales. Most models predict the greatest warming at high latitudes in winter. Confidence is lower for estimates of precipitation fields but many models suggest an increased moisture deficit particularly in the center of continents during summer. In addition to temperature and precipitation, other weather variables such as atmospheric moisture (e.g., relative humidity), wind, cloudiness, etc. will be altered in a changing climate. The variability of extreme events may also be altered (Mearns et al., 1989; Solomon & Leemans 1997) which could have a significant impact on fire activity as many of the largest fires occur on a few critical days with extreme fire weather (Flannigan & Wotton 2001).

​

Many studies have addressed the impact of climate change on fire weather severity. Flannigan and Van Wagner (1991) compared seasonal fire severity rating values (Seasonal Severity Rating - a rating index to provide a measure of fire control difficulty and is a component of the Canadian Forest Fire Weather Index System) from a 2 × CO2 scenario (mid 21st century) versus the 1 × CO2 scenario (approx. present-day) across Canada. Their study used monthly anomalies of temperature and precipitation from three GCMs (Geophysical Fluid Dynamics Laboratory (GFDL), Goddard Institute for Space Studies (GISS), and Oregon State University (OSU)). The results suggest increases in the Seasonal Severity rating across all of Canada with an average increase of nearly 50%, translating roughly into a 50% increase of area burned. Stocks et al. (1998) used monthly data from four GCMs to examine climate change and forest fire potential in Russian and Canadian boreal forests. Forecast seasonal fire weather severity was similar for the four GCMs, indicating large increases in the spatial extent of extreme fire danger in both countries under a 2 × CO2 scenario. Stocks et al. (1998) also conducted a month-to-month analysis, which showed an earlier start to the fire season and significant increases in the area experiencing high to extreme fire danger in both Canada and Russia, particularly during June and July. Flannigan et al. (1998) used daily output from the Canadian GCM to model potential fire danger through the use of the Canadian Fire Weather Index (FWI)(the FWI is a dimensionless relative numerical rating of fire intensity used as a general index of fire danger) for both the 1 × CO2 and 2 × CO2 scenarios for North America and Europe. Most of these studies showed large regional variation in the response of fire weather severity to climate change, ranging from significant increases to regions of no change or decreases in fire weather severity (e.g., Bergeron & Flannigan 1995; Flannigan et al., 2000). The consequences of climate change on fire disturbance must be viewed in a spatially dependent context. Only a few studies have quantified the potential changes in area burned due to climate change. Flannigan et al. (2005) used historical relationships between weather/fire danger and area burned in tandem with two GCMs to estimate future area burned in Canada (Fig. 1). The results suggest an increase of 74 - 118% increase in area burned by the end of this century. Table 1 displays the current trend in fire for the five regions as well as what we anticipate for the near term (2010-2040), medium-term (2040-2070) and long term (2070-2100). Note however that 95% of the area burned occurs in the Boreal East and Boreal West regions.

​

Bergeron et al. (2004) discuss the implications of a changing fire regime on sustainable forest management in Canada. They find that estimates of future fire activity this century are less than the historical fire activity (pre-industrial) for many sites across the boreal forest and suggest that forest management could potentially be used to recreate the forest age structure of fire-dominated pre-industrial landscapes. There are other factors such as ignition agents, length of the fire season and fire management policies that will likely greatly influence the impact of climate change on fire activity. Ignition probabilities may increase in a warming world due to increased cloud-to-ground lightning discharges (Price & Rind 1994), although they did not account for changes in vegetation, which may greatly influence the lightning ignitions and area burned. Projected changes in fire weather due to climate change have been used to examine resultant changes in fuel moisture and consequently changes to fire occurrence rates in parts of the Canadian boreal forest. Both human and lightning-caused fires are strongly influenced by moisture in the fuels of the forest floor (Wotton 2007), though both ignite through different processes and should be considered separately in the analysis of future fire occurrence potential. Apart from the obvious need for human activity to provide an ignition source, human-caused fire occurrence is most strongly influenced by the receptivity of surface litter to ignition. Wotton et al. (2003) carried out a detailed study of future human-caused fire occurrence using daily projections of fire weather and fuel moisture from two GCMs ( CCC GCM2 and HAdCM2) which showed that while changes in human-caused fire occurrence to vary spatially across the province, overall an 18% increase is expected by the year 2020 and a 50% increase by the end of the 21st Century. While human-caused fires make up just over 50% of fires occurring in Canada, it is lightning-caused fires that tend to be the major contributor to the area burned in the boreal forests of Canada (Stocks et al., 2002). Lightning fire ignition is strongly influenced by moisture content in the organic layers of the forest floor. Wotton et al., (2005) examined lightning fire occurrence using fire weather and fuel moisture scenarios derived from the CCC CGCM2 and found a projected increase in lightning fire activity of 24% by 2040 and 80% by the end of the 21st Century. This analysis used the relatively conservative approach of using monthly anomalies to generate fire weather scenarios (following the method of Stocks et al. 1998), did not account for a lengthening fire season, and did not include changes in lightning activity; thus it was hypothesized that these projected increases in fire activity were quite conservative. The exposure of large areas of previously frozen and wet peatlands to fire is expected to increase significantly as a result of climate change. In addition to the increase in emissions and carbon loss associated with more peatland burning, there is also a growing concern that fires will burn more deeply in exposed organic material, greatly increasing fire suppression difficulties. Much more time and effort would be required to extinguish such fires, thus occupying resources that might otherwise be used to attack new fires. In addition, there is the very real possibility that holdover fires (fires that continue to smolder under snow cover during the winter and reappear the following spring) could become much more common with more peat fires.

​

Climate warming is expected to make fire seasons longer. Wotton and Flannigan (1993) estimated that the fire season length in Canada, on average, will increase by 22%, or 30 days in a 2 × CO2 world. Also, research has suggested that the persistence of blocking ridges in the upper atmosphere will increase in a 2 × CO2 climate (Lupo et al., 1997), which could have a significant impact on forest fires as these upper ridges are associated with dry and warm conditions at the surface and are conducive to the development of large forest fires (Skinner et al., 1999, 2001). Fire management policies and effectiveness will continue to change. Changes in prevention programs, initial attack capabilities and restricted access/fire restriction policies will influence potential fire activity in this century. These other factors are all confounding effects that may dampen or amplify the impact of a changing climate on the fire regime.