Online Supplemental Material
Supplemental Material: Fire regime characteristics
Table S1. Observed fire regime characteristics for each ecosystem and the studied landscape. The five ecosystem types, referred to as Biogeoclimatic (BEC) Zones, are the Ponderosa Pine (PP) zone, Interior Douglas-Fir (IDF) zone, Montane Spruce (MS) zone, Interior Cedar Hemlock (ICH) zone and the Engelmann Spruce Subalpine Fir (ESSF) zone (Meidinger and Pojar 1991).
Ecosystem / Fire Interval Range(yrs) / Mean fire Interval
(yrs) / Minimum Fire SizeIV
(ha) / Mean Fire SizeIII
(ha) / Maximum Fire SizeIV
(ha)
PP / 1 – 250I, II / 50VI, 250II / 0.1 / 315 / 500
IDF / 1-250I, II / 50 VI, 250II / 0.1 / 633 / 5000
ICH / 80 - 240I / 150 II / 0.1 / 419 / 25000
MS / 35 – 150I / 150 II / 0.1 / 1069 / 5000
ESSF / 110 - 230I / 150 II / 0.1 / 462 / 10000
All EcosystemsV / 1 - 250 / 150 II / 0.1 / 650 / 6528
I: Wong et al. (2004)
II: Ministry of Forests and B.C. Environment (1995)
III: Lindgren (2001)
IV: Parminter (1992)
V: average fire regime for landscape
VI: Stand maintaining fire interval
References for Table S1
Lindgren P (2001) Fire regime of the southern Okanagan: 80 years of data, from 1919 to 1999. Draft Report prepared for Riverside Forest Products Ltd
Meidinger D, Pojar J (1991) Ecosystems of British Columbia. B.C. Ministry of Forests Special Report SRS06
Ministry of Forests, BC Environment (1995) Biodiversity guidebook: Forest practices code of British Columbia. ISBN 0–7726–2619–7, Province of British Columbia
Parminter J (1992) Old-growth forests: problem analysis. Research Branch, Ministry of Forests. Government of British Columbia
Wong C, Sandmann H, Dorner B (2004) Historical variability of natural disturbances in British Columbia: a literature review. FORREX Series 12
Supplemental Material:
Methods: Parameterization of Prometheus
Fuel Type Classification and Input
Fuel types, as defined by Canadian Forest Fire Behavior Prediction system, were identified using the British Columbia Ministry of Forests and Range (MOFR) (2006a) fuel plot studies for the forest region in which the study area is located. The fuel plot studies classified the regional forest types into the fuel types that have been identified and parameterized within the FBP system. The identified fuel type-forest type relationships where then used to reclassify the study area from forest types into the FBP system fuel types using ArcGIS 9.0. For a detailed description of the FBP fuel types see Hirsch (1996). Eleven FBP system fuel types were identified (see Table S2). Figure S1 summarizes the proportion of the TFL 49 landscape classified in each fuel type. A fuel type raster coverage, with a grid resolution of 50 x 50 m (0.25 ha), was created and imported into Prometheus as an ASCII raster grid. This raster grid was the vegetation/ fuel type input for the Prometheus model. The fuel type raster contains three fuel types that are considered non-flammable (non-fuel, water and unknown) which were included to provide features that could act as potential fire breaks.
Table S2. Fuel type classes identified on TFLL 49 landscape and use in fire modelling
Fuel Type / Fuel Type CharacteristicsC-2 / spruce (Picea sp.), subalpine fir (Abies lasiocarpa) and western red cedar (Thuja plicata) dominated stands
C-3 / mature lodgepole pine (Pinus contorta var. latifolia) dominated stands;
C-4 / immature lodgepole pine dominated stands
C-7 / ponderosa pine (Pinus ponderosa), Douglas-fir (Pseudotsuga menziesii var. glauca) and western larch (Larix occidentalis) dominated stands
D-1 / pure deciduous stands: trembling aspen (Populus tremuloides), paper birch (Betula papyrifera), black cottonwood (Populus balsamifera ssp. Tichocarpa)
M-1/M-2a / deciduous dominated mixed stands (≥ 50 % deciduous component)
M-1/M-2b / coniferous dominated mixed stands (≥50 % coniferous component)
O-1 / grasslands (< 10% tree cover)
Water / lakes and water reservoirs
Non-fuel / roads, rock outcrops and wetlands
Unknown / forest inventory data missing
Fig. S1. Proportion of TFL 49 landscape occupied by each FBP fuel type
Topographic Inputs
A digital elevation model (DEM) was generated from contour data to create raster coverages for elevation, slope and aspect. Three raster coverages, for elevation, slope, and aspect, with grid resolutions of 50 x 50 m (0.25 ha), were created from the DEM and then imported into Prometheus as ASCII raster grids. These raster grids were the topographic inputs for the Prometheus model.
Fire Ignition Input
Both lightning and human-caused ignition points were used. Ignition points represented points where fires were initiated but did not necessarily represent successful fire events. The ability of a fire to spread beyond the initiation point was governed by weather, fuel moisture and fuel type at any given time within Prometheus. Thirty-two human ignition points were identified on the TFL 49 landscape based on known recreation sites (Ministry of Forests and Range 2006b) (Fig. 1). Human ignition points represent fires started by matches, cigarettes and campfires; the three most common ignition mechanisms for human-caused fires (Tanskanen et al. 2005). The selection of lightning-caused ignitions was based on four components: (1) FWI, (2) the daily lightning-ignition frequency distribution, (3) the monthly lightning-ignition distribution, and (4) a landscape-level lightning-ignition risk rating system based on elevation and vegetation (Nitschke and Innes 2008b). Apart from climate, these two factors are the most important for determining lightning ignitions (Días-Avalos et al. 2001; Wierzchowski et al. 2002). Elevation and forest composition were used in conjunction with lightning occurrence data calculated for the area by Wierzchowski et al. (2002) and Environment Canada (2009). From these sources it was estimated that 10 lightning fire ignitions/100 km2/year occur within the region. The TFL 49 landscape is approximately 1450 km2, implying 145 lightning-ignitions per year for the study area. The numbers of fires calculated for each combination of vegetation and elevation were then used as targets to randomly selected ignition points from each combination using ArcGIS 9.0. The results of the random selection are illustrated in Fig. 1. Combining both lightning- and human-caused ignitions, a total of 178 ignition points were generated for the TFL 49 landscape (146 lighting fires and 32 human fires).
Ignition dates were chosen based on dates that had high to extreme fire weather severity ratings, calculated using the Fire Weather Index (FWI) of the Canadian Forest Fire Danger Rating System (Van Wagner 1987). For human-caused fires ignition was assumed to occur during times of high public use; such as weekends and holidays. The assignment of lightning ignition dates was based on the FWI ratings. Days rated as high to extreme (daily severity rating >3 [Stocks et al. 1998]) were randomly assigned to selected ignition points. The time of day selected for fire ignition was based on a frequency distribution for lightning ignitions in the Pacific Northwest observed by Rorig and Ferguson (1999) (see Fig. S2). Fig. S2 presents the hourly and monthly distribution of fire ignitions used in this study. The monthly lightning-ignition distribution that was used was based on the frequency distribution identified by Wierzchowski et al. (2002) for the central Cordillera region of British Columbia, but was adjusted to take into account days with the highest severity ratings from the FWI.
Fig. S2. Hourly distribution of lightning-caused ignition inputs (a) and monthly distribution of both lightning- and human-caused fire ignition inputs (b) generated for modelling fire simulations in Prometheus
Climate Inputs
Climate inputs used were minimum temperature, maximum temperature, relative humidity, wind speed, wind direction and precipitation. Climate data from two local weather stations were used as the base climate to develop historic and future climate scenarios which incorporate daily climate variation. The Fintry station was chosen because it is the most geographically central station for Block 1, while the Salmon Arm station was used for Block 2 because of its proximity. Eleven years of climate data were available for the Fintry weather station and 26 years of data for the Salmon Arm weather station. Seven weather scenarios from each station were randomly generated to provide a representation of below-, average, and above-average conditions, while maintaining high variability between scenarios (see Nitschke and Innes, 2008). The Prometheus model uses environmental lapse rates to interpolate weather variables recorded at a local climate station to each cell across the study area based on changes in slope, aspect, and elevation (see Tymstra et al. 2010). The methodology allows for the variability in observed climate conditions to be incorporated into the study. Incorporation of longer-term climate trends such as those driven by the Pacific Decadal Oscillation (PDO) or El Niňo Southern Oscillation (ENSO) were not possible due to the short-term nature of the region’s climate record; however, Meyn et al. (2010) found that fire variability and area burned in the ecosystems of the study region are driven by summer drought and not the PDO and ENSO climate oscillations.
Climate Change Inputs
A direct adjustment approach was used to integrate climate change predictions from global circulation models (GCM) into the historical climate scenarios (Wilks 1999; Nitschke and Innes 2008a). The global circulation models used were the Canadian Global Circulation Model II (CGCM2), and the Hadley Centre Global Circulation Model III (HadCM3). For a description of the Canadian models, see Flato et al. (2000) and for the Hadley model, see Johns et al. (2003). Climate change outputs were obtained from the Pacific Climate Impacts Consortium (2009). The IPCC’s SRES emission scenario, A2x (Nakicenovic et al. 2000), was selected because of its regional representation, a scenario justified by the observed differential change in regional emissions since 2000 between both developed and developing countries (Raupach et al., 2007). The CGCM2 and HADCM3 models were selected because they both ranked highly in their ability to model historic temperature patterns with the latter model be the best model for representing historic precipitation patterns in western Canada (Bonsal et al. 2003). Caution on the use of GCM predictions over British Columbia has been stated by Wotton et al. (2010); however, the GCM scenarios are being used to provide an envelope of potential climate not exact predictions with the degree of change the important factor considered in this study. The GCM predictions were applied on a month-by-month basis to create a daily time series of weather that represents the local daily variation along with monthly variation of the GCM predictions for each climate scenario. The weather scenarios for each weather station were imported into Prometheus as ASCII text files to provide the climate inputs for the model.
References for Methods Section
Bonsal BR, Prowse TD, Pietroniro A (2003) An assessment of global climate model-simulated climate for the western cordillera of Canada (1961–90). Hydrological Processes 17: 3703–3716
Días-Avalos C, Peterson DL, Alvarado E, Fergusan SA, Besag JE (2001) Space-time modelling of lightning-caused ignitions in the Blue Mountains, Oregon. Canadian Journal of Forest Research 31: 1579–1593
Environment Canada (2009) Lightning Hotspots in Canada.Government of Canada. http://www.msc.ec.gc.ca/education/lightning. Accessed 19 February 2009
Flato GM, Boer GJ, Lee WG, McFarlane NA, Ramsden D, Reader MC, Reader AJ (2000) The Canadian centre for climate modelling and analysis global coupled model and its climate. Climate Dynamics 16: 451–467
Hirsch KG (1996) Canadian Forest Fire Behavior Prediction (FBP) System: users’s guide. Natural Resources Canada Special Report 7. Government of Canada
Johns TC, Gregory JM, Ingram WJ, Johnson CE, Jones A, Lowe JA, Mitchell JFB, Roberts DL, Sexton DMH, Stevenson DS, Tett SFB, Woodage MJ (2003) Anthropogenic climate change for 1860 to 2100 simulated with the HadCM3 model under updated emission scenarios. Climate Dynamics 20: 583–612
Meyn A, Taylor SW, Flannigan MD, Thonicke K, Cramer W (2010). Relationship between fore, climate oscillations, and drought in British Columbia, Canada, 1920-2000. Global Change Biology 16: 977–989.
Ministry of Forests and Range (2006a) Kamloops Forest Region Photo Fuel Plot Study. Government of British Columbia
Ministry of Forests and Range (2006b) Ministry of Forests Recreation Sites. Government of British Columbia
Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J, Gaffin S, Gregory K, Grübler A, Yong Jung T, Kram T, Lebre La Rovere E, Michaelis L, Mori S, Morita T, Pepper W, Pitcher H, Price L, Riahi K, Roehrl A, Rogner HH, Sankovski A, Schlesinger M, Shukla P, Smith S, Swart R, van Rooijen S, Victor N, Dadi Z (2000) Emission Scenarios. A Special Report of Working Group II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge
Nitschke CR, Innes JL (2008) Climate change and fire potential in south-central British Columbia, Canada. Global Change Biology 14: 841–855
Pacific Climate Impacts Consortium (2009) Pacific Climate Impact Scenarios. http://www.pacificclimate.org/tools/. Accessed 5 Jan, 2009
Raupach MR, Marland G, Ciasis P, Le Quéré C, Canadell JG, Klepper G, Field CB (2007) Global and regional drivers of accelerating CO2 emissions. Proceedings of the National Academy of Science 104: 10288–10293
Rorig ML, Ferguson SA (1999) Characteristics of lightning and wildland fire ignition in the Pacific Northwest. Journal of Applied Meteorology 38: 1565–1575
Stocks BJ, Fosberg MA, Lynham TJ, Mearns L, Wotton BM, Yang Q, Lin J-Z, Lawrence K, Hartley GR, Mason JA, McKenney DW (1998) Climate change and forest fire potential in Russian and Canadian Boreal forests. Climate change 38: 1–13
Tanskanen H, Venäläinen A, Puttonen P, Granström A (2005) Impact of stand structure on surface fire ignition potential in Picea abies and Pinus sylvestris forests in southern Finland. Canadian Journal of Forest Research 35: 410–420
Van Wagner CE (1987) Development and Structure of the Canadian Forest Fire Weather Index System. Canadian Forest Service Forestry Technical Report 35
Wierzchowski J, Heathcott M, Flannigan MD (2002) Lightning and lightning fire, central cordillera, Canada. International Journal of Wildland Fire 11: 41–51.
Wilks DS (1999) Multisite downscaling of daily precipitation with a stochastic weather generator. Climate Research 11: 125–136.
Wotton BM, Flannigan MD, Nock CA (2010) Forest fire occurrence and climate change in Canada. International Journal of Wildland Fire 19: 253–271.
Supplemental Material:
Results: Change in Fire Regime