@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Forestry, Faculty of"@en, "Forest Resources Management, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Schmidt, Quentin"@en ; dcterms:issued "2017-02-01T14:25:05"@en, "2016-04"@en ; dcterms:description "Understanding historic fire regimes in the dry forests of southern British Columbia has been the cause of contentious debate, with implications that will continue to influence the approach to wildfire management in the area. Making use of lighting-strike and human-caused ignition data from 1998 through 2012 for the Kamloops, Lillooet and Merritt fire zones, this study analyzes records on both spatial and temporal scales and draws connections between ignitions and the distribution of climatic zones and fuel types on the landscape. Using fire weather data for the Kamloops zone, individual fire events were then assessed for their potential behaviour in the absence of fire suppression. For the 2365 ignitions included in this study, 58% were attributed to human causes, which accounted for 76% of the total area burned. Fire numbers were disproportionately high in lower elevation ecosystems, but had larger impacts in upper elevation forests. The most telling result is that 92% of all fires did not make it over four hectares in size, either as the result of aggressive suppression or weather conditions at the time of ignition. This absence of large-scale events provides no natural fuel mitigation across the landscape, and will allow stands to become more densely structured and host much more severe wildfires."@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/59241?expand=metadata"@en ; skos:note " Analyzing Fire Ignition Data in the Kamloops, Lillooet and Merritt fire zones; with implications toward the effects of fire suppression on the landscape. Quentin Schmidt A technical report produced within the University of British Columbia’s Bachelor of Science in Forestry, in Forest Resources Management FRST 497 Dr. Peter L Marshall Primary Supervisor: Lori Daniels FRST497 Graduate Essay Quentin Schmidt II Abstract Understanding historic fire regimes in the dry forests of southern British Columbia has been the cause of contentious debate, with implications that will continue to influence the approach to wildfire management in the area. Making use of lighting-strike and human-caused ignition data from 1998 through 2012 for the Kamloops, Lillooet and Merritt fire zones, this study analyzes records on both spatial and temporal scales and draws connections between ignitions and the distribution of climatic zones and fuel types on the landscape. Using fire weather data for the Kamloops zone, individual fire events were then assessed for their potential behaviour in the absence of fire suppression. For the 2365 ignitions included in this study, 58% were attributed to human causes, which accounted for 76% of the total area burned. Fire numbers were disproportionately high in lower elevation ecosystems, but had larger impacts in upper elevation forests. The most telling result is that 92% of all fires did not make it over four hectares in size, either as the result of aggressive suppression or weather conditions at the time of ignition. This absence of large-scale events provides no natural fuel mitigation across the landscape, and will allow stands to become more densely structured and host much more severe wildfires. Key Words  Fire regime  Dry forests  BEC zone  Ignition  Fuel type  Suppression  Fire weather  Fire behaviour FRST497 Graduate Essay Quentin Schmidt III Table of Contents Abstract.......................................................................................................................................II Key Words ..................................................................................................................................II Table of Contents .......................................................................................................................III List of Figures ........................................................................................................................... IV List of Tables ............................................................................................................................. V Introduction .................................................................................................................................1 Background .................................................................................................................................2 Methods ......................................................................................................................................3 BC Wildfire Service Ignition Data.......................................................................................................... 3 Ignition Locations by BEC Zone and Fuel Type ...................................................................................... 4 Linking Fire Weather Indices with Ignition Data ................................................................................... 5 Results ........................................................................................................................................6 Monthly and Annual Trends................................................................................................................. 6 BEC Zone Summaries and Ignitions Locations....................................................................................... 7 Fuel Type and Ignition Analysis .......................................................................................................... 11 Fire Behaviour Analysis ...................................................................................................................... 12 Discussion ................................................................................................................................15 Summary ..................................................................................................................................17 Acknowledgements ...................................................................................................................17 References ...............................................................................................................................18 Additional Information ...............................................................................................................19 BC Wildfire Service ............................................................................................................................ 19 Natural Resources Canada ................................................................................................................. 19 Appendices ...............................................................................................................................20 Appendix A ........................................................................................................................................ 20 Appendix B ........................................................................................................................................ 21 Appendix C ........................................................................................................................................ 22 Appendix D ........................................................................................................................................ 24 FRST497 Graduate Essay Quentin Schmidt IV List of Figures Figure 1: Geographic positioning of the Kamloops, Merritt and Lillooet fire zones (clockwise from top right). Figure 2: Annual summary of every ignition of the 1998-2012 fire seasons in the Kamloops, Lillooet and Merritt fire zones. Figure 3: Monthly summary of every ignition throughout the 1998-2012 fire seasons in the Kamloops, Lillooet and Merritt fire zones, with lightning starts in blue and human starts in orange Figure 4: BEC zone map of the study area, showing the BG, IDF and PP zones. The white areas did not fall into any of the three BEC zones that this analysis focused on. Figure 5: Ignition map of all the individual fires from 1998-2012 across the Kamloops, Lillooet and Merritt fire zones. Split by the cause of ignition, lightning-strike fires are shown in brown and human-caused fires in green. Overlaid on top of the previous BEC zone map, showing the makeup of BG, IDF and PP zones. Figure 6: Breakdown by size category (0-1.0ha, 1.0-4.0ha, 4.0-39ha, 40-200ha, >200ha) of the final size of every ignition in the Kamloops, Lillooet and Merritt fire zones from 1998-2012. Figure 7: Summary of the number of ignitions (left) and the total area burned (right, in hectares) by fuel type, and further stratified by the cause of ignition. Using the same 2365 ignitions as the analysis above. Figure 8: Monthly distribution of 424 ignitions that occurred in C7, C3 and C4 fuels in the Kamloops fire zone between May and September from 1998-2012, in which each had a greater than 50% chance of being sustained based on fire weather data on the day of ignition. Categorized by the fire type (surface fires in green, intermittent crown fires in yellow, crown fires in red) that was predicted to have occurred without any attempts of fire suppression. FRST497 Graduate Essay Quentin Schmidt V List of Tables Table 1: Summary of the 4 major FBP fuel types found within the Kamloops/Lillooet/Merritt fire zones. Information sourced from the Canadian Wildland Fire Information System, provided by Natural Resources Canada (2014). Table 2: A select set of the information tracked for each individual fire event. Table 3: Summary statistics of the percentage of the total study area that fell into each individual BEC zone. Table 4: Summary of ignition data for each fire zone, categorized by the number of individual fires and the resultant area burned per BEC zone. Final percentages in the rightmost column display the proportion of ignitions and burned area that fell into each BEC zone. Table 5: Summary of the proportion of ignitions in each BEC zone that were caused by either humans or lightning. FRST497 Graduate Essay Quentin Schmidt 1 Introduction Wildfire is an integral component of ecosystems throughout British Columbia, as a natural and anthropogenic occurrence that brings change to landscapes all across the province. Over the previous decade (2004-2013), British Columbia averaged 1847 unique fire events per year that directly affected 115,464 hectares of land, with 39% of ignitions caused by humans and the remaining 61% by lightning (BC Wildfire Service, 2015). Throughout British Columbia, the frequency and severity of fires change geographically due to the diversity of ecosystems, with the dry forests of the southern interior commonly at the center of fire research; historically hosting a landscape likely to have been shaped by wildfire. Allen et al. (2002) presents that the natural physical structure of these hot and dry ecosystems was driven by frequent low-severity fires, but that post-settlement fire suppression efforts have caused a fundamental shift and a growing problem. The dry forests that were once well-spaced with grassland understories have been increasing in tree density; now subject to fuel loads that are conducive to higher severity fires (Covington & Moore, 1994). In British Columbia, fire suppression tactics were extremely aggressive throughout the late 1900’s, with 92% of wildfires being suppressed before they reached 4 hectares in size (UBC Forestry, 2012). With a new understanding of the fuel buildup this had caused in many ecosystems, the province was prompted to issue a new fire management strategy in 2010; including more proactive fuel management efforts and the re-introduction of fire to the landscape (BC Wildland Fire Management Strategy, 2010). Parts of these strategic plans assume a level of knowledge about the historic fire regimes on the landscapes. Klenner et al. (2008) examines fire regimes in the southern interior and the effects that fire suppression may have had on these landscapes. They argue that a mixed-severity disturbance regime is much more applicable than any assumption of a frequent, low-severity fire regime. Though credible, results of this analysis often focused on the total area burned by wildfires, which downplays the importance of the 92% of ignitions that were containable; almost exclusively looking at the effects of the largest and most severe fires, which represent only 8% of ignitions that exceeded suppression capability. If we are to accurately understand the effects that wildfires have historically had on a landscape, it is imperative that we employ an inclusive viewpoint of all unique fire events. This will allow us to further assess the impacts that British Columbia’s previous fire management tactics have had. This analysis looks to build on the ideas of Klenner et al. (2008), first taking an FRST497 Graduate Essay Quentin Schmidt 2 in-depth look at the spatial distribution of human and lightning-caused ignitions across a portion of British Columbia’s southern interior. In the absence of any suppression efforts, fire weather conditions can model the likelihood of these ignitions establishing into an active surface, intermittent crowning or crowning wildfire, providing a look at their potential effects. These results help paint a picture of fire regimes in this area. Background The study area for this analysis used three of the seven regional fire zones that are managed by the Kamloops Fire Center: Kamloops, Lillooet and Merritt. Combined, these three management areas cover 35,586km2 of the landscape, and are individually comparable in size, differing at most by just under 2000km2. These three encompass a historically hot and dry geographic area in the Southern interior of British Columbia, influenced largely by the rain-shadow of the Coastal mountain range to the west. See Figure 1 beside for their mapped location. As a result of this regional climate these fire zones have been very busy over the previous two decades, averaging a total of 158 active wildfires between May and September each year, and having burned over 86,000ha of forested land over this time frame, despite extremely aggressive suppression tactics. For this analysis, forest types at lower elevations were focused on; characterized by drier climate regimes with sparser canopy coverage than the wetter, colder and higher elevation forests (Klenner et. al, 2008). These forest types fall into the Bunchgrass (BG), Interior Douglas-fir (IDF) and Ponderosa Pine (PP) Biogeoclimatic (BEC) zones. As defined by Meidinger and Pojar (1991), the BG zone occurs in the lowest elevation areas of the dry, hot valleys in the interior of the province, dominated by wheatgrass and sage. The BG transitions into the PP zone which is the driest forested zone, often supporting savannah-like stands of Ponderosa Pine (Pinus ponderosa) and extending up to 900m elevation. Increasing up to 1400m in elevation, the IDF zone is the second warmest forested zone, more commonly comprised of denser stands of Douglas-fir (Pseudotsuga menziesii) intermixed with Ponderosa and Lodgepole Pine (Pinus contorta). Within these different BEC zones, there are a number of different fuel types across the landscape, defined at a national level through the Canadian Forest Fire Behaviour Prediction Figure 1: Geographic positioning of the Kamloops, Merritt and Lillooet fire zones (clockwise from top right). FRST497 Graduate Essay Quentin Schmidt 3 System (FBP). The structural organization of biomass within different forest stands defines these unique fuel types, while crudely named by their common species and seral stages. The dominant fuel types in our region of interest are the C3, C4 and C7 coniferous forests, as well as the O1-b open grasslands. Further details of each are shown below in Table 1, with characteristic photographs of each displayed in Appendix 1. Table 1: Summary of the 4 major FBP fuel types found within the Kamloops/Lillooet/Merritt fire zones. Information sourced from the Canadian Wildland Fire Information System, provided by Natural Resources Canada (2014). Fuel Type Name Description C3 Mature Jack/Lodgepole Pine  Fully stocked (1000-2000 stems/ha) pine stands, at complete crown closure.  Limited in ladder fuels and dead surface fuels, with a sparse understory. C4 Immature Jack/Lodgepole Pine  Densely stocked (10,000-30,000 stems/ha) pine stands.  Heavy loading of dead standing stems and surface fuels of downed wood. C7 Ponderosa Pine-Douglas-Fir  Open stands of uneven aged timber, often in clumped distributions.  Lower canopy closure than C3 or C4, with a consistent shrub and grass coverage between trees. O1-b Standing Grassland  Continuous grass cover, no consistent groups of trees present.  More specifically defined by the percentage of cured material. Methods BC Wildfire Service Ignition Data The ignition data used for this analysis was obtained through the BC Wildfire Service, as archived by the Kamloops Fire Center. Each data point received represents a unique fire event in which a BC Wildfire Service Initial Attack (IA) crew was actively dispatched to, and tasked with suppressing. The individual fires are assessed a unique Fire Number, as well as descriptive information to capture the most important aspects of the ignition: the date, the cause, the location (by geographic description, fire zone, BEC zone, fuel type and as a precise GPS point), and the final size of the fire. For example, a sample of the information logged from an individual fire is as follows: FRST497 Graduate Essay Quentin Schmidt 4 Table 3: A select set of the information tracked for each individual fire event. Ignition Date Fire Cause Fire Label Fire Zone Current Size Description BEC Zone Fuel Type 30-Jul-03 Person 2003-K20272 Kamloops 26,345 McClure #2 IDF C3 Due to a number of inconsistencies and irrelevant information in the data, portions of the data were omitted from the original set that was obtained from the Wildfire Service. Any reported fires that had a “Fire Type” of “Nuisance”, “Smoke Chase”, “Unknown” or “Duplicate” were left out, as well as any that had a “Fire Size” equal to 0. The ignitions that were analyzed in this report cover the 1998 through 2012 fire seasons, and stricter monthly filters were employed for some of the final analyses, which will be mentioned later in the report. Only fires that burned in C3, C4, C7 or 01-b stands were looked at, and all geographic analyses were limited to the IDF, PP and BG BEC zones. All of the applicable ignitions were first summarized annually and monthly, to assess any temporal trends in the data. Ignition Locations by BEC Zone and Fuel Type The first stage of this analysis sought to understand the relationship between the geographic positioning of ignitions and the distribution of BEC zones throughout our fire zones. By looking at the proportion of each specific BEC zone as a function of our total study area, we are better equipped to understand any spatial links between BEC zones and the total number of ignitions in each, the total area burned, and the cause of these fires. ArcMap 10.3.1 was used for this analysis, and every ignition point regardless of fuel type was included. After completing this, all of the ignitions were then analyzed separately from our BEC zones. All of the individual fire events were categorized by their final reported size, which carries meaningful (though generalized) information about the behaviour of the fire and the potential for suppression success. The size categories chosen organize our fire starts between 0-1.0, 1.0-4.0, 4.0-39, 40-200, and greater than 200 hectares, which represents different assumed thresholds. Fires below 1 hectare are not required to be mapped by the Wildfire Service, and those up to 4 hectares in size are generally accepted to be well within containment potential of an IA crew. Very extreme fire behaviour becomes exhibited when fires spread more than 40 hectares, and fires greater than 200 hectares historically became separately databased nationally under the Large Fire Database. Across Canada, all sizes of fires are now collected FRST497 Graduate Essay Quentin Schmidt 5 under the Canadian National Fire Database, though the largest category is often the only one mapped (Natural Resources Canada, 2016). Linking Fire Weather Indices with Ignition Data The final stage of this analysis links actual fire events with fire weather indices (FWI) on the day of ignition, and uses pre-established algorithms to determine potential fire behaviour for that day. In Canada, FWI values are calculated through the combination of six components which are then applied to unique fuel types. A summary of these components is provided by the Canadian Wildland Fire Information System (2016), with a flow chart presented in Appendix B. Three of these six are fuel moisture codes (Fine Fuel Moisture Code [FFMC], Duff Moisture Code [DMC] and Drought Code [DC]), which represent the effect of temperature, relative humidity, wind and rain on the moisture content of different sized fuels. The next two are fire behaviour indices (Initial Spread Index [ISI] and Buildup Index [BUI]), which pairs these moisture codes with predicted wind speeds to produce a final rating for potential fire intensity, our Fire Weather Index (FWI). Each of the roughly 260 weather stations throughout BC are constantly monitoring weather conditions to produce unique FWI data for localized areas of the province. Because of these differences, this final analysis was only completed for the Kamloops fire zone, chosen over Lillooet and Merritt as it perennially had the highest number of ignitions. Fire weather data was collected from the East Barriere weather station which is operated through the BC Wildfire Service, a fully automated station which records daily temperature, relative humidity, wind speed and precipitation. This station is located at 51o15’11’’N, 119o52’54’’W, placed at an elevation of 631 meters (BC Wildfire Service, 2015). This weather data was used to compile fire behaviour predictions in order to calculate the probability of an ignition sustaining itself given the weather on that date. After a value for this probability had been obtained for every day of the study, the data was then filtered to capture the days during our study period in which this probability was greater than 50%. All of the fire events in C3, C4 and C7 fuels from the beginning of May to the end of September from 1998-2012 were then compiled, and filtered to only look at those which occurred on a day that satisfied this 50% threshold for a sustained ignition. Using predictive fire behaviour algorithms based on the FBP system in Canada, we are then able to assess the fire type (surface, intermittent crowning or crowning) that was predicted FRST497 Graduate Essay Quentin Schmidt 6 to have occurred without any suppression efforts. This system uses the Buildup Index (BUI) and Initial Spread Index (ISI) reported for that day, while applying constants for foliar moisture content and crown base height to predict the type of fire, equilibrium rate of spread and the fire intensity class. Example FBP tables for the C3, C4 and C7 fuels have been attached in Appendix C. Fire intensity classes are used to predict physical fire behaviour and rank the intensity of active fires, from Rank 1 (smouldering ground/creeping surface fire) to Rank 6 (blow up or conflagration). Example photographs of the six ranks are displayed in Appendix D. Results Monthly and Annual Trends Figure 2 below shows the number of total ignitions for each fire season between 1998 and 2012 across our three zones of interest. The two busiest seasons came in 2009 and 1998, with 307 and 237 ignitions respectfully, while 2005, 2010 and 2011 all had a period-low of 90 ignitions. Figure 3 shows the same breadth of ignitions as categorized by the month in which the fire began, further split by whether the fire was caused by lightning (blue) or a human (orange). The busiest months for wildfires were July and August, hosting 51% of all ignitions. Human-caused starts occurred in every month of the season, while not a single lightning start fell in March or October. As conditions heat up and the fine fuels dry out, we see many more successful lightning ignitions, responsible for over 59% of all fires in June, July and August. Figure 2: Annual summary of every ignition of the 1998-2012 fire seasons in the Kamloops, Lillooet and Merritt fire zones. 0501001502002503003501998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012Number of IgnitionsFire SeasonFRST497 Graduate Essay Quentin Schmidt 7 Figure 3: Monthly summary of every ignition throughout the 1998-2012 fire seasons in the Kamloops, Lillooet and Merritt fire zones, with lightning starts in blue and human starts in orange. BEC Zone Summaries and Ignitions Locations To gain a better understanding of the physical landscape in our study area, a map of the current BEC zones of interest was produced in ArcMap, to find the proportions of each across these three fire zones. As mentioned above, this study was limited to the BG, IDF and PP zones. Figure 4 below shows a map of this information, and Table 3 shows the percentage of the study area that falls into each BEC category. The majority of the applicable area (83.2%) is represented by the IDF zone, with narrow bands of the BG (10.3%) and PP (6.5%) occurring within the lower elevation valleys. 0100200300400500600700Number of IgnitionsFRST497 Graduate Essay Quentin Schmidt 8 Figure 4: BEC zone map of the study area, showing the BG, IDF and PP zones. The white areas did not fall into any of the three BEC zones that this analysis focused on. Table 3: Summary statistics of the percentage of the total study area that fell into each individual BEC zone. % of Study Area BG IDF PP 10.3% 83.2% 6.5% After looking at the BEC zone makeup of our three fire zones, an ignition map was produced of every reported fire for the 1998 through 2012 fire seasons, totalling 2365 fires. The map is shown below in Figure 5, with ignitions split between their official cause: whether the result of a lightning strike (brown dots) or a direct human-cause (green dots). Lightning ignitions are confirmed by the BC Wildfire Service through the use of real-time strike data collected by the Canadian Lightning Detection Network (BC Wildfire Service, 2015). As visually shown in the map, ignitions appear to be more concentrated in the valleys of each fire zone, as the furthest stretches of our IDF zones are much less clustered. Human-caused fires quite often appear to FRST497 Graduate Essay Quentin Schmidt 9 be in close proximity to one another, often seeming to form more continuous and linear distributions along our low-elevation areas of the BG and PP zones. Neither of these relationships were quantitatively assessed. Figure 5: Ignition map of all the individual fires from 1998-2012 across the Kamloops, Lillooet and Merritt fire zones. Split by the cause of ignition, lightning-strike fires are shown in brown and human-caused fires in green. Overlaid on top of the previous BEC zone map, showing the makeup of BG, IDF and PP zones. Summary information of these ignitions is shown in Table 4 below, organized by the BEC zones in which each fire was located. Each fire zone was looked at individually, while the rightmost column contains the cumulative numbers for the three zones. Comparing this ignition information to the BEC zone statistics above, we see that there is a much higher proportion of ignitions in the BG zone compared to its total size, as it only makes up 10.3% of our area but contained 16.3% of fires. This same trend is even more apparent for the PP zone, covering 6.5% of the study area while housing 22.3% of ignitions. The lower half of Table 4 also presents the total area burned from the fires that ignited within each BEC zone, with results that can be conceptually connected to the dominant forest types that are found in each. Comparing 7176.7 hectares burned to 385 ignitions, fires in the BG FRST497 Graduate Essay Quentin Schmidt 10 zone had the lowest average fire size at 18.6 hectares. Moving up in elevation, the PP and IDF zones have much higher values for this, at 34.2 and 41.9 hectares respectively. Table 4: Summary of ignition data for each fire zone, categorized by the number of individual fires and the resultant area burned per BEC zone. Final percentages in the rightmost column display the proportion of ignitions and burned area that fell into each BEC zone. Kamloops Lillooet Merritt All Zones Number of Ignitions BG IDF PP 261 612 224 39 208 150 85 632 154 385 (16.3%) 1452 (61.4%) 528 (22.3%) Total # of Ignitions 1097 397 871 2365 Area Burned (ha) BG IDF PP 6611.5 27,311.1 8254.9 171.2 29,757.5 9052.4 394.0 3694.4 754.3 7176.7 (8.3%) 60763.0 (70.7%) 18061.6 (21.0) Total Area Burned (ha) 42,177.5 38,981.1 4842.7 86,001.3 On top of looking at the spatial attributes of each fire, it is also important to look at summary statistics of the cause of each of these ignitions. For the 2365 ignitions analyzed across all three BEC zones, 57.9% of them were attributed to human actions; the remainder the result of lightning strikes. Narrowing our view to individual BEC zones, the percentage of fires that were human-caused in the lower elevation and more sparsely forested BG and PP zones was nearly double that of the IDF zone. Summary numbers are presented in Table 5 below. Table 5: Summary of the proportion of ignitions in each BEC zone that were caused by either humans or lightning. Ignition %; Human-Caused Ignition %; Lightning-Caused BEC Zone BG IDF PP 90.0% 45.8% 82.4% 10.0% 54.2% 17.6% FRST497 Graduate Essay Quentin Schmidt 11 For another look at individual fire sizes, fires were grouped into the following hectare categories: 0-1.0, 1.0-4.0, 4.0-39, 40-200, and greater than 200. The results are displayed in Figure 6 below. Fires under 1.0 hectare make up 82% of all total ignitions, with an additional 10% only making it to a maximum of 4.0 hectares in size. The remaining 8% of fires represent those that exceeded the abilities of an IA crew to suppress, with only 2% of fires breaking the 40 hectare threshold. Figure 6: Breakdown by size category (0-1.0ha, 1.0-4.0ha, 4.0-39ha, 40-200ha, >200ha) of the final size of every ignition in the Kamloops, Lillooet and Merritt fire zones from 1998-2012. Fuel Type and Ignition Analysis After assessing all of the ignitions individually by BEC zone, fires were then analyzed for the fuel type in which they burned. This was again done by stratifying all of the ignitions into either lightning-strike or human-caused fires, and summarizing them by the number of ignitions and the total area burned. Shown in Figure 7 below, there were the fewest number of fires in C4 fuels (244), followed closely by C3 (307), but more than doubled by both C7 (725) and 01-b (1089). Lightning was the primary cause of ignition in all of our coniferous fuels at 59.6%, 66.8% and 52.0% respectively for C3, C4 and C7, while humans caused 75% of ignitions in grassland fuels. For the cumulative final sizes of these fires, it is important to consider that the dataset used for this analysis reports the fuel type in which the wildfire first ignited, but does not provide a comment of which fuel type the fire proceeded to move into, if different from the point of ignition. This can strongly influence the analysis, especially when considering our largest fires. 82%10%6%1%1%FIRE SIZE IN HECTARES0-1.01.0-4.04.0-3940-200 >200FRST497 Graduate Essay Quentin Schmidt 12 The cause of ignition presents a very telling story in this summary, as the vast majority of the area burned (65,090 out of the 86,001 hectares) resulted from an anthropogenic fire. Virtually all of the cumulative C3 fuel that burned was via human causes, with the exact opposite scenario in the C4. Area burned in C7 fuels was the most evenly spread out at 35.5% human and 64.5% lightning, while 88.5% of burned grasslands were ignited by humans. Figure 7: Summary of the number of ignitions (left) and the total area burned (right, in hectares) by fuel type, and further stratified by the cause of ignition. Using the same 2365 ignitions as the analysis above. Fire Behaviour Analysis Continuing to look at the 1998-2012 fire seasons, this section of the analysis only looked at the Kamloops fire zone, and restricted the ignitions to May through September. This narrowed our view to a total of 519 ignitions throughout our three coniferous fuel types. Of the 519 unique fire starts, 424 of them happened on days in which the fire weather offered a probability greater than 50% of an ignition turning into a sustained fire. Each of these potentially sustained ignitions is graphically represented in Figure 8 below. In each of the three fuel types, 82% of the total reported ignitions fell on days in which the fire weather satisfied our 50% threshold. In our C7, C3 and C4 fuels respectively, this was 224 out of 275 fires, 123 out of 150 fires and 77 out of 94 fires. From this, we can conclude that for this time period, 18% of the wildfires that were actively and aggressively suppressed had less than a 50% chance of becoming an active and spreading fire, with fire indices that would have significantly hampered any significant fire behaviour. Also shown in Figure 8, the predicted fire type in the case of a positive start changes with each of the three fuel types, based on that specific day’s fire weather. Within all three fuel types, the proportion of fires predicted to have intermittent and continuous crowning peaks in 020040060080010001200C3 C4 C7 O1-bNumber of IgnitionsFuel TypePersonLightning0500010000150002000025000300003500040000C3 C4 C7 O1-bTotal Area Burned (ha)Fuel TypePersonLightningFRST497 Graduate Essay Quentin Schmidt 13 July and August, which are also the two months in which we see the highest number of ignitions (as per Figure 3 above). As we see with our more open C7 fuel type, dominated by sparse Douglas-fir and Ponderosa Pine, not a single ignition was predicted to have evolved into a continuous crown fire, with the majority of fires predicted to have remained on the surface. Though our C3 and C4 fuel types saw a lower number of sustained ignitions throughout the fire season, the potential severity for an ignition is much higher. In C3 fuels, 61 out of 123 ignitions (50%) were predicted to continuously crown, likewise with 40 out of 77 ignitions (52%) in C4 fuels. Only 9% and 3% were predicted to remain as surface fires in C3 and C4 fuels respectively. FRST497 Graduate Essay Quentin Schmidt 14 Figure 8: Monthly distribution of 424 ignitions that occurred in C7, C3 and C4 fuels in the Kamloops fire zone between May and September from 1998-2012, in which each had a greater than 50% chance of being sustained based on fire weather data on the day of ignition. Categorized by the fire type (surface fires in green, intermittent crown fires in yellow, crown fires in red) that was predicted to have occurred without any attempts of fire suppression. 020406080100C7 FuelsTotal = 224SurfaceIntermittent020406080100Number of Sustained IgnitionsC3 FuelsTotal = 123SurfaceIntermittentCrown020406080100MonthC4 FuelsTotal = 77SurfaceIntermittentCrownFRST497 Graduate Essay Quentin Schmidt 15 Discussion Through analyzing these ignitions at a temporal and spatial scale, certain relationships can be drawn between the location of fires and their resultant effects. Patterns become apparent based on climatic attributes of the individual BEC zones, connecting this with properties of the different fuel types in our study area. Analyzing the average fire sizes in each of our BEC zones, our results demonstrate common differences in fire behaviour among characteristic fuels, with a potential anthropogenic influence that is tough to quantify. It is important to note that the sample size of fires in each of these zones is extremely different, and that the location of the fire (by BEC zone and fuel type) is archived by where the fire first began. This may introduce bias in the results in the case of our largest fire events, which may have spanned a number of zones or fuel types as they spread across the landscape. The BG zone hosted the smallest average fires, followed by the PP and then the IDF. With little to no timber establishment in the BG, the comparatively low fuel buildup of grassland areas presents a structural condition which would not be as likely to promote the development of an uncontrollable, severe fire. With this in mind, it is important to consider that these low elevation areas may have a higher density of transportation corridors and urban areas. This would likely allow for much easier access for suppression crews to action the fire, and would require an immediate response if located in the Wildland Urban Interface (WUI). As fuel size and fuel continuity increase in the PP and IDF zones, this buildup appears to go hand in hand with the largest fire events, despite hosting a climate that is not as hot or dry as the BG valleys. As aggressive suppression tactics or weather conditions held 92% of fires in this study under 4 hectares in size, it becomes clear that there are very few fire events occurring that are controlling the buildup of fuels at the landscape level. It is important to note that fuel types are crudely named after certain tree species, but that they are really defined by their vertical and horizontal distribution of fuels. With a lack of low-severity fire events in these dry ecosystems, fuel types are able to shift along the spectrum: 01-b fuels infilling with young timber and traditional C7 fuels hosting vast regeneration and re-structuring as C3/4. This fire regime and resulting change in historic Ponderosa Pine landscapes is well documented in draft by R.W. Gray et al (2003), while Wong and Iverson (2004) demonstrate that the physical characteristics of dry IDF ecosystems have been forced well beyond their range of natural variability. With this change in fuels across the landscape, my results show the major effect this will have on fire severity. As our fuels become more densely structured (shifting downward in Figure 8), the onset of intermittent crowning and crowning fires under the same weather conditions FRST497 Graduate Essay Quentin Schmidt 16 becomes very apparent. This will blatantly affect our ability to suppress active fires, nearly removing the potential for low-severity surface fires as we experience our hottest and driest weather. With the aforementioned 92% in mind, and recalling that 18% of our ignitions occurred under very moderate fire weather conditions, it suggests that natural disturbance regimes in our study area have been dominated by frequent, low-severity events – but are now risking a major change as the result of anthropogenic influences. While it is important to note the effect that fire suppression has had on the landscape, we also need to consider the destructive impact that humans have caused. With 76% of the cumulative hectares burned coming from human-caused fires, society’s influence on fire regimes has to be understood from both sides of the coin. Public education and proactive commitments from our communities need to ensure that we are limiting this issue, if we have any aspirations of returning our landscapes to their “natural” ways. While working to produce as comprehensive an analysis as possible, I faced a number of limitations which motivates a list of further ideas for research. Though this study looked at the proporition and distribution of BEC zones, it would provide a more thorough analysis to extend this same view to mapped fuel types across the landscape; though the potential change in fuel types in the absence of disturbance may make this difficult or invalid. As well, the Fire Behaviour Analysis would hold more merit if it was extended to all three fire zones, as well as to the ignitions that occurred in 01-b fuels. Results from the remaining two zones would likely be very similar due to regional climate regimes, and by the nature of its fuel structure the 01-b fuel type can only host surface fires. Finally, a more comprehensive look at the locations of human and lightning ignitions would be useful in assessing their distribution on the landscape, as well as the validity of any patterns that appear to occur. This could also be paired to look at newly updated maps of WUI-defined zones and the number of ignitions that occur here, and apply similar analyses to new areas deemed for Modified Response tactics as per the new BC Wildfire Management Strategy. FRST497 Graduate Essay Quentin Schmidt 17 Summary Fully understanding the historic fire regimes in the dry forests of BC has proven to be extremely difficult, researched and contested by a number of academics. My research sheds a light on some cumulative statistics of fire locations, types and sizes in the Kamloops, Lillooet and Merritt fire zones, and works to draw relationships between these fires and physical characteristics of the landscapes they occur in. The effects of anthropogenic actions are also considered; believed to have influenced fire regimes immensely in this study area, both by a down-playing from suppression and an increase in destruction from direct causation. These effects need to be considered alongside historic fire records when making decisions regarding wildfire management. Acknowledgements I would like to thank the following individuals and groups for their invaluable help in completing this project. Their contributions include, but are not limited to:  Lori Daniels, PhD and professor at UBC, for her constant assistance in formulating a topic, accessing data and writing the essay.  The BC Wildfire Service, for providing the ignition data needed to complete this project.  Greg Greene, PhD student at UBC, for compiling all necessary GIS files and greatly helping in all mapping analyses.  Peter Cherniwchan, fellow undergraduate student at UBC, for working with Fire Weather data in the region and sharing his results for this analysis. FRST497 Graduate Essay Quentin Schmidt 18 References Allen, D.C. et al. (2002). Ecological Restoration of Southwestern Ponderosa Pine Ecosystems: A Broad Perspective. Ecological Applications (12), No. 5. pp. 1418-1433. DOI: 10.2307/3099981 Covington, W. & Moore, M. (1994). Southwestern Ponderosa Forest Structure: Changes Since Euro-American Settlement. Journal of Forestry (92), No. 1. pp. 39-47. Retrieved from: http://www.ingentaconnect.com/content/saf/jof/1994/00000092/00000001/art00014?token=004d1c2d662bf6b0efb0437a63736a6f5e47462136663f44532e796f642f4642d9776bd05ac72 Gray, R.W., Riccius, E., Wong, C., Gayton, D. (2003). Comparison of current and historic stand structure in two IDFdm2 sites in the Rocky Mountain Trench. Unpublished draft report. Retrieved from: http://www.trench-er.com/public/library/files/lewis-isidore-fire-history.pdf Klenner, W. et al. (2008). Dry forests in the Southern Interior of British Columbia: Historic disturbances and implications for restoration and management. Forest Ecology and Management (256). pp. 1711-1722. Retrieved from: http://www.sciencedirect.com/science/article/pii/S0378112708002260 Meidinger, D. & Pojar, J. (1991). Ecosystems of British Columbia. British Columbia Ministry of Forests. pp. 330. UBC Forestry. (2012). Wildfire Management – Challenges in a changing environment. Branchlines (23). No. 4. Retrieved from: http://www.forestry.ubc.ca/files/2011/11/BL-23.4.1.pdf Wong, C. & Iverson, K. (2004). Range of natural variability: Applying the concept to forest management in central British Columbia. BC Journal of Ecosystems and Management (4), No. 1. Retrieved from: http://jem-online.org/index.php/jem/article/view/258 FRST497 Graduate Essay Quentin Schmidt 19 Additional Information BC Wildfire Service BC Wildland Fire Management Strategy. (2010). Retrieved from: http://bcwildfire.ca/Prevention/PrescribedFire/docs/BCWFMS.pdf (January 19th, 2016) BC Wildfire Service. (2015). Fire Averages. Retrieved from: http://bcwildfire.ca/History/average.htm (January 19th, 2016) BC Wildfire Service. (2015). Fire Rank. Retrieved from: http://bcwildfire.ca/fightingwildfire/firerank.htm (February 5th, 2016) BC Wildfire Service. (2015). Weather Stations. Retrieved from: http://bcwildfire.ca/weather/stations.htm (January 19th, 2016) Natural Resources Canada Canadian Wildland Fire Information System. (2016). Canadian Forest Fire Weather Index (FWI) System. Retrieved from: http://cwfis.cfs.nrcan.gc.ca/background/summary/fwi (February 5th, 2016) Canadian Wildland Fire Information System. (2016) Canadian National Fire Database. Retrieved from: http://cwfis.cfs.nrcan.gc.ca/ha/nfdb (January 19th, 2016). Canadian Wildland Fire Information System. (2016). FBP Fuel Type Descriptions. Retrieved from: http://cwfis.cfs.nrcan.gc.ca/background/fueltypes/c1 (February 5th, 2016). FRST497 Graduate Essay Quentin Schmidt 20 Appendices Appendix A Visual depiction of the 4 major FBP fuel types found in our study area. Obtained from the Canadian Wildland Fire Information System, at: http://cwfis.cfs.nrcan.gc.ca/background/fueltypes/c1 C3: Mature Jack/Lodgepole Pine C4: Immature Jack/Lodgepole Pine O1-b: Standing Grassland C7: Ponderosa Pine – Douglas-fir FRST497 Graduate Essay Quentin Schmidt 21 Appendix B Basic guide to the weather components that affect each individual component of the Fire Weather Index (FWI) system. Obtained from the Canadian Wildland Fire Information System, at: http://cwfis.cfs.nrcan.gc.ca/background/summary/fwi FRST497 Graduate Essay Quentin Schmidt 22 Appendix C Sample Field Guide cards from the Canadian Forest Fire Behaviour Prediction (FBP) System, for the C3, C4 and C7 fuel types. Different shading colours represent the intensity classes, different font colours represent the type of fire, and individual numbers represent the equilibrium rate of spread (in meters per minute). Obtained from the FBP “Red Book”, at: https://www.for.gov.bc.ca/hfd/pubs/docs/frh/Frh012.pdf FRST497 Graduate Essay Quentin Schmidt 23 FRST497 Graduate Essay Quentin Schmidt 24 Appendix D Sample photographs of the six fire intensity ranks. Obtained from the BC Wildfire Service at: http://bcwildfire.ca/fightingwildfire/firerank.htm Rank 1: Smouldering/creeping ground/surface Rank 2: Low vigour surface Rank 3: Moderately vigorous surface Rank 4: Highly vigorous surface, torching FRST497 Graduate Essay Quentin Schmidt 25 Rank 5: Extreme surface, actively crowning Rank 6: Blow up, extreme fire behaviour "@en ; edm:hasType "Graduating Project"@en ; edm:isShownAt "10.14288/1.0314379"@en ; dcterms:language "eng"@en ; ns0:peerReviewStatus "Unreviewed"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@* ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* ; ns0:scholarLevel "Undergraduate"@en ; dcterms:isPartOf "University of British Columbia. FRST 497"@en ; dcterms:subject "Fire weather"@en, "Fire regime"@en, "Dry forests"@en, "BEC zone"@en, "Ignition"@en, "Fuel type"@en, "Suppression"@en, "Fire behaviour"@en ; dcterms:title "Analyzing Fire Ignition Data in the Kamloops, Lillooet and Merritt fire zones : with implications toward the effects of fire suppression on the landscape."@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/59241"@en .