Carbon (C) emissions from wildfires are a key terrestrial–atmosphere interaction that influences global atmospheric composition and climate. Positive feedbacks between climate warming and boreal wildfires are predicted based on top-down controls of fire weather and climate, but C emissions from boreal fires may also depend on bottom-up controls of fuel availability related to edaphic controls and overstory tree composition. Here we synthesized data from 417 field sites spanning six ecoregions in the northwestern North American boreal forest and assessed the network of interactions among potential bottom-up and top-down drivers of C emissions. Our results indicate that C emissions are more strongly driven by fuel availability than by fire weather, highlighting the importance of fine-scale drainage conditions, overstory tree species composition and fuel accumulation rates for predicting total C emissions. By implication, climate change-induced modification of fuels needs to be considered for accurately predicting future C emissions from boreal wildfires. Carbon emissions from fires are generally modelled and predicted from fire weather and climate. Fuel availability drives carbon emissions more strongly than fire weather in boreal forests, highlighting the importance of ecological dynamics for fire–climate feedbacks.

In recent years, time-critical processing or real-time processing and analytics of bid data have received a significant amount of attentions. There are many areas/domains where real-time processing of data and making timely decision can saves thousands of human lives, minimizing the risks of human lives and resources, enhance the quality of human lives, enhance the chance of profitability, efficient resources management etc. This paper have presented such type of real-time big data analytic applications and a classification of those applications. In addition it presents the time requirements of each type of these applications along with its significant benefits. Also, a general overview of big data to describe a background knowledge on this scope.

The Canadian Forest Fire Weather Index (FWI) System is the mostly widely used fire danger rating system in the world. We have developed a global database of daily FWI System calculations, beginning in 1980, called the Global Fire WEather Database (GFWED) gridded to a spatial resolution of 0.5° latitude by 2/3° longitude. Input weather data were obtained from the NASA Modern Era Retrospective-Analysis for Research and Applications (MERRA), and two different estimates of daily precipitation from rain gauges over land. FWI System Drought Code calculations from the gridded data sets were compared to calculations from individual weather station data for a representative set of 48 stations in North, Central and South America, Europe, Russia, Southeast Asia and Australia. Agreement between gridded calculations and the station-based calculations tended to be most different at low latitudes for strictly MERRA-based calculations. Strong biases could be seen in either direction: MERRA DC over the Mato Grosso in Brazil reached unrealistically high values exceeding DC = 1500 during the dry season but was too low over Southeast Asia during the dry season. These biases are consistent with those previously identified in MERRA's precipitation, and they reinforce the need to consider alternative sources of precipitation data. GFWED can be used for analyzing historical relationships between fire weather and fire activity at continental and global scales, in identifying large-scale atmosphere–ocean controls on fire weather, and calibration of FWI-based fire prediction models.

The objective of this paper is to examine the sensitivity of fuel moisture to changes in temperature and precipitation and explore the implications under a future climate. We use the Canadian Forest Fire Weather Index System components to represent the moisture content of fine surface fuels (Fine Fuel Moisture Code, FFMC), upper forest floor (duff) layers (Duff Moisture Code, DMC) and deep organic soils (Drought Code, DC). We obtained weather data from 12 stations across Canada for the fire season during the 1971–2000 period and with these data we created a set of modified weather streams from the original data by varying the daily temperatures by 0 to +5 °C in increments of 1 °C and the daily precipitation from −40 to 40 % in increments of 10 %. The fuel moistures were calculated for all the temperature and precipitation combinations. When temperature increases we find that for every degree of warming, precipitation has to increase by more than 15 % for FFMC, about 10 % for DMC and about 5 % for DC to compensate for the drying caused by warmer temperatures. Also, we find in terms of the number of days equal to or above an FFMC of 91, a critical value for fire spread, that no increase in precipitation amount alone could compensate for a temperature increase of 1 °C. Results from three General Circulation Models (GCMs) and three emission scenarios suggest that this sensitivity to temperature increases will result in a future with drier fuels and a higher frequency of extreme fire weather days.

Landscape fires show large variability in the amount of biomass or fuel consumed per unit area burned. Fuel consumption (FC) depends on the biomass available to burn and the fraction of the biomass that is actually combusted, and can be combined with estimates of area burned to assess emissions. While burned area can be detected from space and estimates are becoming more reliable due to improved algorithms and sensors, FC is usually modeled or taken selectively from the literature. We compiled the peer-reviewed literature on FC for various biomes and fuel categories to understand FC and its variability better, and to provide a database that can be used to constrain biogeochemical models with fire modules. We compiled in total 77 studies covering 11 biomes including savanna (15 studies, average FC of 4.6 t DM (dry matter) ha−1 with a standard deviation of 2.2), tropical forest (n = 19, FC = 126 ± 77), temperate forest (n = 12, FC = 58 ± 72), boreal forest (n = 16, FC = 35 ± 24), pasture (n = 4, FC = 28 ± 9.3), shifting cultivation (n = 2, FC = 23, with a range of 4.0–43), crop residue (n = 4, FC = 6.5 ± 9.0), chaparral (n = 3, FC = 27 ± 19), tropical peatland (n = 4, FC = 314 ± 196), boreal peatland (n = 2, FC = 42 [42–43]), and tundra (n = 1, FC = 40). Within biomes the regional variability in the number of measurements was sometimes large, with e.g. only three measurement locations in boreal Russia and 35 sites in North America. Substantial regional differences in FC were found within the defined biomes: for example, FC of temperate pine forests in the USA was 37% lower than Australian forests dominated by eucalypt trees. Besides showing the differences between biomes, FC estimates were also grouped into different fuel classes. Our results highlight the large variability in FC, not only between biomes but also within biomes and fuel classes. This implies that substantial uncertainties are associated with using biome-averaged values to represent FC for whole biomes. Comparing the compiled FC values with co-located Global Fire Emissions Database version 3 (GFED3) FC indicates that modeling studies that aim to represent variability in FC also within biomes, still require improvements as they have difficulty in representing the dynamics governing FC.

The top-down hypothesis for long-term condition of Rocky Mountain trembling aspen (Populus tremuloides) holds that these stands developed under conditions of frequent anthropogenic fires and low herbivory by ungulates due to intense predation by carnivores and humans. To test predictions of the hypothesis, we repeated 156 historic photographs (taken in years 1874 to 1949) showing detailed aspen stand conditions in seven areas along the east slope of the Rocky Mountains in Alberta, Cana­ da. We quantified historic and current aspen stand conditions (e.g., stem spacing by height class and species, barking, and browse class) for 195 stands visible in photographs. Large aspen in historic photographs have no bark-scarring from brows­ ing by elk (Cervus elaphus), an indication of low elk use before 1870. Aspen stand ages in historic photographs appear to increase from the period 1874 to 1894 to the period 1895 to 1914, likely due to declining fire frequency before organized fire suppression programs began. Aspen stands in all areas responded to a known period of low herbivore density (1880 to 1930) by consistently showing a multi-aged structure with abundant saplings (1 to 4 m high). Current repeat photographs in Banff and Jasper national park areas with high elk density (>4 elk/km2 since 1940) show heavy browsing and few aspen saplings in stands that were historically multi-sized and lightly browsed. Five other areas with low or moderate elk density (<4 elk/km2) maintained multi-aged aspen stand conditions similar to those visible in historic photographs. All areas had increased conifer cover and older trees due to reduced fire frequency. These results support the top-down model for long-term Rocky Moun­ tain montane ecosystem development. Those conditions changed, though, with reduced burning by First Nation cultures by the 1890s, and were accelerated in Banff and Jasper national parks where, in addition to fire, hunting and predators were also controlled. Maintenance of aspen will likely require restoration of long-term low elk density and wary behavior patterns, fol­ lowed by prescribed burning. keywords: Alberta, Cervus elaphus, elk, fire history, herbivory, Populus tremuloides, repeat photography, Rocky Mountains, trembling aspen. Citation: White, C.A., and M.C. Feller. 2004. Repeat photography of montane trembling aspen in the Canadian Rocky Moun­ tains. Pages 2–22 in R.T. Engstrom, K.E.M. Galley, and W.J. de Groot (eds.). Proceedings of the 22nd Tall Timbers Fire Ecol­ ogy Conference: Fire in Temperate, Boreal, and Montane Ecosystems. Tall Timbers Fire Ecology Research Station, Tallahas­ see, FL. INTRODUCTION provide useful information on changing ecological Historic photographs and current retakes from the conditions over time (Houston 1982, Kay 1990, Kay et same camera station, often called repeat photographs, al. 1999). Aspen twigs and bark are valuable forage for can be used to evaluate long-term (>50 years) landscape cervids such as elk, moose (Alces alces), white-tailed change over large regions. The technique was applied as deer (Odocoileus virginianus), and mule deer (O. early as 1888 by Finsterwalder to map glacier movehemionus) (Nelson and Leege 1982, DeByle and ment in the eastern Alps (Hattersley-Smith 1966). It has Winokur 1985). Twig-browsing and bark-stripping, also been used to evaluate long-term effects of wildlife, visible in historic and current photographs, can profire, and climate on landscapes and vegetation in varivide a long-term record of herbivore abundance (Kay ous situations from rangelands in the western United 1990). Further, aspen stands are generally found on the States (Gruell 1980, Rogers et al. 1984, Hart and Layvalley bottoms within the montane ecoregion (Achuff cock 1996, Meagher and Houston 1998) to the African and Corns 1982, Houston 1982, Suzuki et al. 1999), Serengeti (Sinclair 1995). where historically fires were frequent (Houston 1973, Repeat photography studies of trembling aspen Tande 1979, Arno 1980, White 1985a). Evidence of forests in the Rocky Mountains of North America can fire, such as recently burned trees and logs, and young

Federal land management agencies in the U.S. have responded to recent severe wildfire seasons with plans to greatly expand fuel treatment programs. These plans are often accompanied by ecological justifications to assuage environmental objections to fuel treatment activities (e.g., tree removal, smoke production). However, the chain of hypotheses that support ecological justifications for fuel treatments has received scant empirical assessment. With the aid of published repeat photography, we present an objective evaluation of changes in 20th-century wildland fire potential and substantiate the relationship of these changes to historic fire frequency. Unlabeled photo pairs depicting historic versus recent vegetation conditions at seven diverse locations in the western U.S. were evaluated by 32 wildland fire professionals. Their ratings demonstrated a large and significant increase in perceived crown fire potential (d [the standardized mean difference between ratings for recent and historic conditions] = 0.83, 95% CI = 0.34 ≤ d ≤ 1.23) and a moderate and significant increase in fire severity potential (d = 0.66, 95% CI = 0.25 ≤ d ≤ 1.04), but no change in spread rate potential (d = –0.29, 95% CI = –0.79 ≤ d ≤ 0.13). Perceived changes in crown fire potential and potential fire severity are both related to the historic fire regime of forested photo locations (P = 0.012 and 0.015, respec­ tively), with the greatest amount of change perceived where fire was historically most frequent. These results support the use of the historic fire regime–current condition class concept to justify and prioritize fuel treatments that reduce the potential for crown fires on forested lands. keywords: fire exclusion, fire hazard, fire regimes, fuel treatments, repeat photography, vegetation change, western U.S. Citation: Martinson, E.J., and P.N. Omi. 2004. Relating historic fire regimes to 20th-century fire potential may augment ecological justifications for expanded fuel treatment programs. Pages 36–42 in R.T. Engstrom, K.E.M. Galley, and W.J. de Groot (eds.). Proceedings of the 22nd Tall Timbers Fire Ecology Conference: Fire in Tem­ perate, Boreal, and Montane Ecosystems. Tall Timbers Research Station, Tallahassee, FL. INTRODUCTION Writings (e.g., Cooper 1960) and photographs (e.g., The severity of recent wildfire seasons (e.g., 1988, Veblen and Lorenz 1991) from the period of Euro­ 1994, 1996, 2000, 2002) in the western U.S. support American settlement anecdotally support suppositions assertions of increased fire potential as a result of of increased fuel accumulation in some ecosystems more hazardous fuel profiles created by fire exclusion since the advent of organized fire suppression in the during the 20th century. This chain of hypotheses is early 20 th century. Gruell et al. (1982) used hazard ratoften invoked to provide ecological justification for ings by a fuels specialist in their repeat photography expanded fuel treatment programs on U.S. federal study to infer a general hazard increase since 1909. We lands (e.g., USDA Forest Service 2000). Fire history took a similar but broader approach to quantify and studies provide widespread graphic evidence that fire differentiate changes in fire potential represented in exclusion during the 20th century has dramatically repeat photographs that have been published for a reduced fire frequency in many ecosystems (Swetnam variety of ecosystems in the western U.S. It is gener­ et al. 1999). Since fire is an agent of mortality (Ryan ally presumed that fire exclusion has had the greatest et al. 1988) and a process of accelerated decomposiimpact on ecosystems that historically experienced a tion (Ottmar et al. 1993), reduced fire frequency regime of frequent fires (Covington and Moore 1994). would be expected to result in greater fire hazard due We tested the hypothesis that the degree of perceivable to increased surface fuel loads, more abundant ladder difference in fire potential represented in repeat pho­ fuels, and denser canopy fuels (Agee 1996). tographs is related to the historic fire frequency at

During the next few decades, a considerable portion of the productive boreal forest in Canada will be harvested and there is an excellent opportunity to use forest management activities (e.g., harvesting, regeneration, stand tending) to alter the forest fuels for fire management purposes. This process, known as fire-smart forest management, has the potential to reduce the number and size of wildfires and the risk associated with the use of prescribed fire. We describe a landscape-level fire-smart technique in which strategically located fuel treatments, primarily species conversion, are incorporated into a long-term for­ est management planning model. Using a mechanistic-based fire simulation model, a comparative analysis of projected land­ scapes in central Alberta showed that fuel treatments could have a considerable impact on fire size. These findings have impor­ tant implications for sustainable forest management in crown fire-dominated boreal forest ecosystems now and in the future. keywords: Alberta, boreal forest, climate change adaptation, fire-smart forest management, forest management planning, fuels management, timber supply modeling. Citation: Hirsch, K., V. Kafka, and B. Todd. 2004. Using forest management techniques to alter forest fuels and reduce wild­ fire size: an exploratory analysis. Pages 175–184 in R.T. Engstrom, K.E.M. Galley, and W.J. de Groot (eds.). Proceedings of the 22nd Tall Timbers Fire Ecology Conference: Fire in Temperate, Boreal, and Montane Ecosystems. Tall Timbers Research Station, Tallahassee, FL. INTRODUCTION ther supported by simulation modeling results for Fire is an important natural disturbance in boreal Ontario that showed a small percentage of wildfires forest ecosystems and has significant economic, (2%–4%) are likely to continue to escape initial attack social, and ecological effects. During the last two and have the potential to become large despite increas­ es in suppression spending, due to diminishing decades, there has been an average of about 8,500 fires marginal returns on suppression investments per year in Canada, and the annual area burned has (McAlpine and Hirsch 1999). To reduce the area ranged between 0.3 million and 7.5 million ha (Canaburned below current levels and reintroduce fire where dian Council of Forest Ministers 1997). Most (97%) of it is ecologically desirable, it will be necessary to the area burned is caused by a small proportion (3%) implement a new, proactive approach to fire manageof all reported wildfires (Weber and Stocks 1998). ment that emphasizes standand landscape-level fuels Although forest managers recognize the ecological management in conjunction with fire suppression. benefits of fire, wildfires are suppressed in heavily Fuels management is the planned manipulation of inhabited and industrial forest areas. forest vegetation to decrease the intensity and rate of Traditionally, Canadian fire management agencies spread of a wildfire to improve suppression effectivehave focused on the prevention and suppression of ness and reduce fire impacts. Pyne et al. (1996) iden­ wildfires in an attempt to protect life, property, and tify three types of fuels management: reduction, con­ natural resources. This has been effective in some version, and isolation. These activities have generally regions, but it is neither economically possible nor been associated with the protection of relatively small, ecologically desirable to eliminate fire in most forest high-value areas, such as homes in the wildland–urban ecosystems (Weber and Stocks 1998). This is exeminterface, but they may also have application at the plified by recent fire seasons (e.g., 2002, 2001, and landscape scale (e.g., Weatherspoon and Skinner 1996, 1998 in Alberta, 1996 and 1995 in Quebec, 1995 in Agee et al. 2000). For example, Finney (2001) conOntario) during which large areas burned despite ducted a theoretical analysis of the shape and pattern of unprecedented fire suppression expenditures. The confuel breaks on a landscape to minimize fire spread. In a cept of a limit to fire suppression effectiveness is furmore applied approach, Sessions et al. (1999) discuss

The fire management goal in Banff National Park is to maintain or restore, where possible, historical fire regimes. Fire cycles are an important component of a fire regime, and historical fire cycles provide a reference to guide the use of prescribed fire. Weather, climate, vegetation, and ignition are strongly influenced by the extremely rugged topography in the park, resulting in fire cycles that vary spatially. By analyzing a forest stand-origin database, we found that four variables (valley orientation, elevation, aspect, and proximity to the Continental Divide) explained 64% and 70% of the variation of stand-age patterns (i.e., fire cycles) in subalpine and montane ecoregions, respectively. Based on this information, historical fire cycles in Banff National Park were mapped in 50-year fire cycle classes. For each fire cycle class, the areas burned by wildfire and prescribed fires were tabulated and subtracted from the theoretical mean fire activity to determine the fire deficit (or surplus) within each of the park’s land management units. These data can help managers to prioritize areas for burning and provide a method to monitor the prolonged effects of prescribed and wildfires through time. keywords: Alberta, Banff National Park, fire cycles, fire management, prescribed fire, stand-age patterns, topography. Citation: Rogeau, M.-P., I. Pengelly, and M.-J. Fortin. 2004. Using topography to model and monitor fire cycles in Banff Nation­ al Park. Pages 55–69 in R.T. Engstrom, K.E.M. Galley, and W.J. de Groot (eds.). Proceedings of the 22nd Tall Timbers Fire Ecol­ ogy Conference: Fire in Temperate, Boreal, and Montane Ecosystems. Tall Timbers Research Station, Tallahassee, FL. INTRODUCTION frequency, size, and pattern. In mountainous terrain, After many decades of fire suppression, managers of the fire cycle varies over time and space. As described below, features of topography affect the spatial expresCanadian national parks are using planned prescribed sion of fire cycles directly or indirectly in a number of burns to manage forest fuels, maintain wildlife habi­ tats, and preserve fire-adapted vegetation. Studies ways. from the Southern Canadian Rockies and Northern Elevation Rockies of the United States have shown that fire freElevation has often been identified as a controlling quency varies spatially in these mountainous environfactor of landscape patterns (Barrows 1951, Hawkes ments (Tande 1979, Hawkes 1980, White 1985, Bar1980, Kushla 1996). Depending on the type of ecosys­ rett et al. 1991, Tymstra 1991). It has therefore been tems under study, elevation can affect fire behavior in crucial for managers of Banff National Park (BNP) to different ways. Kushla (1996) found that in the Oreunderstand the park-wide fire cycles in order to restore gon Coast Range, lower elevations favor longer fire fire in a way similar to historical patterns. intervals due to wetter fuels, whereas in the Canadian Fire Cycles and Topographic Features Rocky Mountain forests, lower elevations favor short­ er fire intervals (Tande 1979) because of their drier A fire regime is the type, intensity, severity, frequenconditions. Older stands are also more commonly cy, size, and pattern of fires which characterize an found at higher elevations, even though lightning area, while the fire cycle is the number of years occurrence is greater than at lower elevations (Barrequired to burn over an area equal to the entire area of rows 1951, Hawkes 1980). At higher elevations, fires interest (National Research Council of Canada 1987). are limited by a combination of fuel discontinuity creThe fire cycle integrates three fire regime components: ated by treeline, patchy fuels due to poor growing

Multi-scale planning was used in the Upper Arkansas River subbasin and Box Creek watershed to prioritize and plan forest health and National Fire Plan restoration on San Isabel National Forest and Bureau of Land Management (BLM) lands. Restoration alternatives were designed to restore fire-adapted ecosystems, improve forest health, improve native species habi­ tats, and reduce unwanted wildland fire and other hazards to human communities. The Upper Arkansas River subbasin and Box Creek watershed are located in the mountains of central Colorado. National Forest and BLM plans and national policies and budgets identified the need to prioritize and plan to achieve effective multi-resource and fire-adapted ecosystem restora­ tion. A consistent and science-based approach was used for mapping and analysis of fire regime condition class, historical regime departure, vegetation, and other resource and social values. Findings from this analysis were used to determine the amount of area to restore, develop the management prescriptions, map operationally restorable outcomes, and conduct effects analysis. The results from the Box Creek watershed restoration project demonstrate a cost-effective and science-based attempt to provide consistent and repeatable risk data for assessment of conditions and development of alternatives. In addition, the interdisciplinary team demonstrated how to identify the full “decision space” available for restoration if an integrated approach to project prioritization, purpose and need, and proposed action formulation is implemented versus accepting the tra­ ditional “mitigation spin.” keywords: Colorado, condition class, fire regimes, forest health, fuel management, natural regime, planning, wildland–urban interface, wildlife habitats. Citation: McNicoll, C.H., and W.J. Hann. 2004. Multi-scale planning and implementation to restore fire-adapted ecosystems and reduce risk to the wildland–urban interface. Pages 294–316 in R.T. Engstrom, K.E.M. Galley, and W.J. de Groot (eds.). Proceedings of the 22nd Tall Timbers Fire Ecology Conference: Fire in Temperate, Boreal, and Montane Ecosystems. Tall Tim­ bers Research Station, Tallahassee, FL. INTRODUCTION by a fire-resistant tree species with a fairly open strucThe Box Creek watershed was identified as a priorture to a more dense structure that would facilitate a ity for restoration in an assessment of all watersheds in crown fire that might result in more severe fire effects the Upper Arkansas River subbasin of central Colto the vegetation versus a previously cooler surface orado (McNicoll et al. 1999). This assessment used a fire. Funding for the planning and restoration imple­ systematic rating system combined with an interview mentation came from Forest Health and the National approach for key publics to develop a suite of risk Fire Plan. The framework for the planning process rankings for each watershed. These data were used to came from Forest Health and the National Fire Plan, rank watersheds for restoration based on departure which was often referred to as the Cohesive Strategy from the historical fire regime condition class 3, (USDA 2001); the analysis of options was developed uncharacteristic levels of insects and disease, uncharfrom Hann and Bunnell (2001). acteristic wildlife habitat conditions, conflicts between During the Upper Arkansas assessment and Box user groups, and high wildfire risk to the Creek watershed planning, we hypothesized that an wildland–urban interface. Uncharacteristic (vegetation integrated approach was needed to resolve many of the or structure) was defined as a vegetation–fuel condiconflicts currently faced by agency leadership (USDA tion, disturbance behavior or effects determined to not 2001, USGAO 2002). Fire regime condition class, occur within the natural or historical regime, similar to departure from the historical regime, and other sci­ the definition from Hann (this volume). For example, ence-based risk measures were used as part of this an uncharacteristic condition of a vegetation commuintegrated approach to prioritize project areas, develop nity could be a change from a community dominated the project purpose and need, design the proposed

Forest floor data are important for many forest resource management applications. In terms of fire and forest carbon dynamics, these data are critical for modeling direct carbon emissions from wildfire in Canadian forests because forest floor organic material is usually the greatest emissions source. However, there are very few data available to initialize wildfire emission models. Six data sets representing 41 534 forest stands across Canada were combined to provide summary statistics and to analyze factors controlling forest floor fuel loads and depths. The impacts of dominant tree species, ecozone, drainage-class, and age-class data on forest floor fuel loads and depth were examined using ANOVA and regression. All four parameters were significant factors affecting forest floor fuel load and depth, but only tree species and ecozone were substantially influential. Although forest floor depths summarized in this study are similar to those of previous studies, forest floor fuel loads are higher. Average forest floor fuel loads and depths are summarized by species and ecozone and can be used to initialize dynamic stand-level forest models.

The boreal biome is characterised by extensive wildfires that frequently burn into the thick organic soils found in many forests and wetlands. Previous studies investigating surface fuel consumption generally have not accounted for variation in the properties of organic soils or how this affects the severity of fuel consumption. We experimentally altered soil moisture profiles of peat monoliths collected from several vegetation types common in boreal bogs and used laboratory burn tests to examine the effects of depth-dependent variation in bulk density and moisture on depth of fuel consumption. Depth of burning ranged from 1 to 17 cm, comparable with observations following natural wildfires. Individually, fuel bulk density and moisture were unreliable predictors of depth of burning. However, they demonstrated a cumulative influence on the thermodynamics of downward combustion propagation. By modifying Van Wagner's surface fuel consumption model to account for stratigraphic changes in fuel conditions, we were able to accurately predict the maximum depth of fuel consumption for most of the laboratory burn tests. This modified model for predicting the depth of surface fuel consumption in boreal ecosystems may provide a useful framework for informing wildland fire management activities and guiding future development of operational fire behaviour and carbon emission models.

New estimates of greenhouse gas emissions from Canadian forest fires were calculated based on a revised model for fuel consumption, using both the fire fuel load and the Drought Code of the Canadian Forest Fire Weather Index System. This model was applied to future climate scenarios of 2×CO2 and 3×CO2 environments using the Canadian Global Climate Model. Total forest floor fuel consumption for six boreal ecozones was estimated at 60, 80, and 117 Tg dry biomass for the 1×CO2, 2×CO2, and 3×CO2 scenarios, respectively. These ecozones cover the boreal and taiga regions and account for about 86% of the total fire consumption for Canada. Almost all of the increase in fuel consumption for future climates is caused by an increase in the area burned. The effect of more severe fuel consumption density (kilograms of fuel consumed per square metre) is relatively small, ranging from 0% to 18%, depending on the ecozone. The emissions of greenhouse gases from all Canadian fires are estimated to increase from about 162 Tg·year–1 of CO2 equivalent in the 1×CO2 scenario to 313 Tg·year–1 of CO2 equivalent in the 3×CO2 scenario, including contributions from CO2, CH4, and N2O.