Black Tiger Fire Case Study

Prepared by the

NFPA9 National Fire Protection Association

Sponsored by the National Wildland/Urban Interface Fire Protection Initiative

Members of the Initiative:

United States Department of Agriculture Forest Service

National Association of State Foresters

United States Department of Interior Bureau of Land Management

National Fire Protection Association

For more information or to order additional copies of this report, contact Fire Investigations Division, National Fire Protection Association P.O. Box 9101, Quincy, Massachusetts, 02269 ABSTRACT

A human-caused wildland fire starting on July 9, 1989 in a scenic part of the Rocky Mountains near Boulder, Colorado, swept through residential areas nestled among the trees. Within the first five to six hours after ignition, 44 homes and other structures were destroyed and many others were damaged. The fire was not completely extinguished until four days later, after burning almost 2,100 acres. Loss estimates of homes and natural resources amounted to $10 million, and the cost to control the fire was another $1 million. More than 500 fire fighters from local, state and federal fire agencies worked to eventually contain the fire and protect the numerous other homes built in the rustic surroundings. Some of the fire fighters’ own homes were threatened or destroyed by the fire. Only a few minor fire fighter injuries were reported and one resident was hospitalized from burns. The result of this wildfire, especially the loss of the homes, represents an increasingly common example of the risks of building homes in what is called the wildland/urban interface, the term for a border zone where structures-mainly residences-are built in wildland areas that by nature are subject to fires. This fire, which soon outran the fire defenses in difficult terrain, demonstrated the predictable effects of a combination of factors: lack of rainfall; prolonged heat spell; wind; sloping topography; buildup of forest fuels; construction factors affecting the susceptibility of the home to fire; use of combustible construction materials; poor site access for emergency vehicles; and lack of a home’s site maintenance for fire protection. These factors plus the number of homes that were quickly threatened compounded the problems for the fire fighters. The Black Tiger Fire was the worst wildland fire loss in Colorado history, but the conditions that led to it are still prevalent in many parts of Colorado as well as in other states. The trend of building combustible homes in the flammable wildlands continues. In many of these areas the potential for similar or worse disaster currently exists, needing only an ignition source and the unfortunate development of hot, dry, windy weather conditions that come with dangerous regularity every year.

For several years fire protection agencies have been attempting to warn affected home- owners nationwide of the risks of thesewildland areas, but most homeowners remain not fully aware of, or insufficiently concerned about, the problem. Many publications also offer guidance for homes in the wildland/urban interface (see the Appendix for a sampling). Proposed NFPA 299, Protection of Life and Property from Wildland Fire, will be a national standard that will present fundamental planning and design criteria for fire agencies, planners, architects, developers and government for the protection of life and property. It includes information on procedures and practices for safe development in areas that may be threatened by wildfire. To assure that it will be an acceptable document, it-as are all standards developed by the National Fire Protection Association-is being prepared by a committee of those who would be most affected: homeowners; interested individuals; architects; urban planners; and fire officials from local, state and federal agencies.

3 T

ACKNOWLEDGMENTS

This report has been prepared by the Na- This wildland fire is only one of many that tional Fire Protection Association (NFPA) at occur throughout the world each year. Under the request and in cooperation with the Boul- the sponsorship of the Initiative, the National der County Sheriff’s Department. The project Fire Protection Association will review, ana- was sponsored by the National Wildland/Urban lyze, and document additional wildland/urban Interface Fire Protection Initiative to further interface fires that cause destruction to homes the goals of the Initiative established in 1986. and structures. Those goals are to create general public The preparation of this report would awareness of the wildland interface prob- not have been possible without the able lems, to encourage the formation of partner- assistance of the following people: Sheriff ships among problem-solvers and interest Brad Leach and Sgt. Larry Stern of the Boul- groups, and to focus on the development of er County Sheriffs Office; Chief Ruth Rave- local solutions to wildland/urban interface el of the Sugarloaf Volunteer Fire Depart- fire problems. The Initiative is sponsored ment; Ron Zeleny and Dave Parker of the currently by the U.S. Department of Agricul- Colorado State Forest Service; Shelly Nolde ture Forest Service, U.S. Department of the and Bob Wilmot of the Arapaho/Roosevelt Interior, the National Association of National Forest; Brooke B. Smith, State Foresters, and the National Fire Jr., Fire Protection Engineer, As- Protection Association. Contact in- pen Engineering, Inc.; and nu- formation for each of these organi- merous other people from the or- zations is provided in the Appendix. ganizations listed above who con- As part of achieving the goals of tributed to this report. the Initiative, an analysis of the Black In addition, the on-site assis- Tiger Fire was undertaken. The tance and technical guidance pro- purpose of the analysis was to docu- vided to NFPA personnel by Al ment the fire, determine to the ex- Roberts, U.S.D.A. Forest Service tent possible the variables causing Regional Office in Denver, greatly the destruction, and make recom- enhanced our ability to prepare mendations on how to prevent simi- this report. Next, technical assis- lar occurrences. Thomas Klem, Di- tance and input to the fire growth rector of NFPA’s Fire Investigations and intensity analysis of the re- Division, served as project manager and tech- port were provided by Dick Rothermel, U.S.D.A. nical advisor. William Baden, Senior Fire Serv- Forest Service Intermountain Forest Range Ex- ice Specialist, NFPA, served as the technical periment Station in Missoula, Montana. Each of advisor for the project. Dr. John R. Hall, Jr., these individuals has made significant contri- Director, Fire Analysis and Research Division, butions to the technical accuracy of the report. NFPA, specifically contributed the section “Sta- The methodology used for the project and ap- tistical Analysis of Factors in Damaged Homes.” proach used in the analysis is described stan- The information from this report can be ing on page 37. used to assist planners, local officials, fire serv- Laurie Ruszcyk, as project secretary, served ice personnel and homeowners in Colorado numerous functions throughout the project and in other parts of the country in developing including coordination of technical reviews, firesafe homes and communities in the procurement of photos, and arrangement of wildland/urban interface, a term referring to the report for layout and printing. the geographical area where two diverse sys- Finally, Jerry Laughlin of Books On Fire tems-in this case, wildland and residential- served as technical editor and prepared the join and affect each other. 4 layout of the final document. OVERVIEW

e officials across the nation are reporting associated with living in the wildlands. dramatic increases in the loss of homes to The population density of most cities is ex- wildfires. The 1980s have seen some of the tremely high. For example, Denver has ap- most severe wildfires in this century. In 1985, proximately 4,200 residents per square mile; for example, 1400 homes and other structures Los Angeles and Miami both have some 6,400 were damaged or destroyed in fires reported to per square mile; and New York City squeezes the U.S. Forest Service. Every year since 1985, in about 23,300 people per square mile. In ad- more than 300 homes have been lost to such dition, many city and suburban residents must wildfires, but the tragic losses are not just to contend with soaring property and rental costs property-lives of homeowners and fire fight- coupled with reduced services. They say urban ers are also lost every year. In 1987, more fire A dream areas feature excessive taxes, too-restrictive fighters died fighting wildland fires than any regulations, and endless noise, crime and grime. other single type of fire. homesite As a result, people are moving from cities to The fire danger to homes in the wildland/ the less-crowded wildlands because they offer urban interface is affected by five major fac- can become a scenic environment with generally lower tors: property costs, more privacy, fewer regula- a nightmare tions, less noise and less crime. w People continue to move to the This new trend is made possible by the scenic wildlands and build homes there. for the availability of a combination of services previ- This trend will continue to be influenced ously obtainable only in the cities and suburbs. by the less desirable factors of city and subur- unprepared. Now good connecting highways allow people ban life, as compared to the desirable factors to retain their jobs in the cities but escape to live

Photo by NFPA in and enjoy the aesthetics of a home in the a long-defunct mining company. woods. Extending outward with the highways, generally good communications and other l Wildfires continue to ignite and utilities have provided important bridges be- threaten homes in the wildlands. tween the urban world of work and the rural To a city resident, a forest may represent world of new homes. Some other basic serv- only a beautiful environment, its quiet punctu- ices-education, for example-are also now ated by the soothing sounds of birds or rushing more readily accessible in ruralwildland areas. streams or leaves rustling in the wind. But to a Indeed, one of the basic but erroneous as- fire, the forest represents tremendous fuel. sumptions made by people relocating to Fire is a fact of life in a forest. Taking a wildland residential areas is that seemingly naturalist’s view, fire serves a “cleaning” func- most of the services they enjoyed in the cities tion and helps recycle nutrients. Frequent small and suburbs are equally available in the new fires-started naturally by lightning, for ex- areas. But the shock comes when it is time to ample-assure that fallen leaves and limbs are Fire is a pay for the relatively high level of services periodically removed and recycled before the desired, because the typical lower population fuel builds up to a quantity that, if ignited, density in the wildlands means there are fewer fact of life would threaten the forest itself. On a longer- people to share the costs of the services. Thus, in the forest. term period, as the forest ages and decays, fires per capita costs rise. inevitably burn out large areas that are subse- To keep costs lower, rural communities quently replaced by new forest stands. may decide to provide minimal services, or It removes Lightning is not a phenomenon control- choose not to expand the current level of lable by humans. Thousands of lightning dis- services. The result is often that, in an emer- litter and charges strike the earth each day with the gency, residents discover that critical services recycles electrical characteristics needed to start a fire. taken for granted in city life may not be avail- Although most will not find the necessary able in the new rural setting. If available, they combination of fuels and dryness for ignition, nutrients. may not be quickly available. weather conditions experienced every year do But fire is one type of emergency that is provide dangerous combinations of prolonged very likely to occur-and develop quickly- heat, wind and dryness. Then, any lightning when combustible homes are built in close strike in the forest, brush or grasslands could proximity to the combustible vegetation found produce a rapidly spreading fire. in forests and other Where nature fails to wildland areas. Unfortu- ignite a fire, humans are nately, development of 34 (thousands of acres) capable of igniting fuels- these rural areas has eas- 32 whether accidentally or ily outpaced wildfire 30 intentionally. Nation- hazard awareness on the wide, a great majority of part of homeowners and 28 GOING UP... wildfires are caused by governments. 26 Colorado Wildfires humans. And while some 0 BLACK TIGER FIRE 24 (State and Private Lands) /v areas of this country have In Colorado an estimated a greater fire problem than 22 three million of the state’s I others, no area is immune 22 million acres of for- 20 to dangerous wildfires. ested land are dotted with 18 c BLACK TIGER FIRE homes. Boulder County In the area of the Black 16 has 18,000 people living Tiger Fire, the U.S.D.A. in the rural mountains in 14 Forest Service has wild- 6,000 buildings. Homes 12 fire protection responsi- found in the area of the 10 bility on national forest Black Tiger Fire ranged lands. The county sheriff 8 from shacks for pan-time in Colorado is the official use to large, expensive 6 agent responsible for sup- homes on the most scenic 4 pressing wildfires on pri- sites. One residential vate and unincorporated 2 structure was a remodeled lands. No other state has a railroad caboose. One 74 76 78 80 82 84 86 88 similar organization. Lo- cal volunteer fire depan- was an old bunkhouse for 6 ments provide the first line of response but on these roads can force them to leave some in- transfer control to the sheriff on large fires. The dividual homes unprotected at critical times. state forester can also assume the control duty Another problem occurs when roads are too with concurrence of the sheriff. Radio commu- narrow to allow simultaneous evacuation of nications are controlled at the county level in threatened residents while large emergency Boulder County. vehicles move in the opposite direction. Wildfires occurring during hot, dry and l These wildfires continue to pres- windy conditions find fuels prepared for easier ent particular problems to affected fire burning and rapid fire spread. When moisture protection agencies. levels for some typical fuel types are reduced Established procedures for controlling from 25 to 20 percent, the rate of spread factor wildland fires-sacrificing some acres by pre- can double. A moisture reduction from 25 to 10 paring a perimeter firebreak and percent in these same fuel types “backfiring” to remove adjacent can result in a rate-of-spread fac- fuels and starve the main fire, for tor increasing seven times. example--conflict with the neces- Fire intensity is perhaps even sity for direct intervention to pro- more important. One fuel model- tect individual homes. At the same ing system considers fuel types time, committing available person- and converts flame front energy nel to an intensive battle to protect and rate of spread into a number individual homes may result in not expressed in Btu (heat) per foot being able to control the main wild- (of fireline front) per second of fire itself. burning. A rating of 500 repre- Fires threatening homes in the Wildland/Urban Interface sents the theoretical, or “rule of wildland/urban interface area thumb,” limit of control by any (defined in box at right) may gen- This describes the geographical areas where organized means. Beyond 1,000 erate initial emergency response formally urban structures--mainly resi- Btu/ft/sec, a fire can be expected from either a wildland fire agency dence-are built in close proximity to the to feature dangerous spotting, or a structural fire department, de- flammable fuels naturally found in wildland firewhirls, crowning and major pending on which one is closer. areas, including forests, prairies, hillsides runs with high rates of spread However, wildland fire fighters are and valleys. The results can be aesthetically and violent fire behavior. (These not always equipped or trained to desirable.. or disastrous. We can better numbers are useful for compari- fight fires in structures, and struc- achieve the desirable aspects by better under- son purposes but they remain tural fire fighters are not always standing thepotential danger and prevent- theoretical calculations based on equipped or trained to fight ing or preparing for them. best estimates usually made after wildland fires. Therefore, the first- the fire from computer models.) responding fire fighters may not Spotting has particularly diffi- be able to quickly handle the fire at hand. cult implications. Spotting occurs as wind-borne Funding and manpower limitations faced by burning embers are carried far ahead of the both types of agencies suggest that this prob- main fire front and land in receptive fuels. The lem will not be quickly eliminated. embers can fall on the roofs of homes or In addition, many wildland homes are built woodpiles and start new fires while along narrow roads and on cul-de-sacs that fire fighters are occupied else- present access difficulty for larger emergency where with the main vehicles. The Black Tiger Fire area is character- ized by many small private ownerships arising from old gold-mining claims (Black Tiger was the name of one of those early gold mines) overlapping on federal lands. Most of the exist- ing narrow and winding roads originaily served the mining claims in the late 1800s. The danger of fire crews becoming trapped and unable to turn Across the Black Tiger Fire area For every 100 feet of distance, the land rises an average of 23 feet To make matters worse, if the embers fall that builds up on the ground. Homeowners on untreated wood shingle or shake roofs, the could also manually thin out thick stands of new spot fires can develop even more rapidly. brush and trees to reduce the intensity of any Yet, building code regulations to prevent the fire approaching a structure. Total clearing of use of combustible roofing are not often in all vegetation from the ground is not preferred place or enforced in wildland areas. because this could lead to erosion problems. 0BLACK TIGER FIRE State law in Colorado re- However, homeowners could plant greenbelts, quires a county to contact the Colorado State which are aesthetically pleasing zones of low Forest Service for comments on developments flammability vegetation that help protect a home in wildlands, but some counties don’t bother. and also hold the soil. These irrigated, land- Colorado fire officials are frustrated that some scaped and regularly maintained areas act as a mountain subdivision covenants actually re- fuel break. quire shake-shingle roofing to maintain the 0 BLACK TIGER FIRE The principal vegetation perceived rustic charm of the community. across the Black Tiger Fire area was tall grass Boulder, however, became the first county A safety under open ponderosa pine. Pockets of dense in Colorado to mandate fire-retardant roof lodgepole pine and Douglas fir were found on shingles, effective May 4, 1989. The greatest zone of low shaded slopes and along riparian zones-ar- limitation of this requirement-as well as simi- eas of high-moisture vegetation around lakes, lar ones elsewhere-is that it covers only new fuel density streams or ponds that serve to reduce fire construction. All existing homes are unaffected, intensity. which means they may be quite vulnerable to all around the The fire danger in this part of Colorado was wildfire, Fire officials would like to see tougher made worse by the recent effects of tree-killing laws, but they will only be created and passed home offers insects such as budworms and beetles. The if homeowners and local officials better under- spruce budworm larvae feed on the needles of stand the problem. important spruce trees, killing them slowly while build- ing up a thick carpet of needles on the ground. l Lack of good vegetative manage- protection. In the past ten years the Black Tiger area had ment predisposes areas to wildfiies. been ravaged by mountain pine beetles, which Arrangement of natural fuels is an impor- left many of the pine trees dead. The Douglas tant factor affecting the type of fire that could fir beetles were also in those remaining trees, occur in an area. This includes not only the entering the bark and damaging the internal amount of forest litter and the density of the water distribution system of the trees. In the trees, but also the heights of adjacent fuels. previous winter, a deep freeze followed by a Grass fires do not easily ignite tree canopies. quick thaw weakened many trees and made But so-called ladder fuels can allow flames to them even more vulnerable to the insect dam- escalate from grass to bushes to lower tree age. Dead and downed fuels had been re- limbs. Clearing ladder fuels is a prime objec- moved on some areas but remained on others. tive of prescribed or controlled burning. This term describes the knowledgeable application W Unless specific preventive meas- of fire to a limited land area under controlled ures are taken by homeowners and local conditions by forest management experts to governments, homes will continue to be accomplish specific objectives. In addition to lost and people’s lives will continue to be removing litter, prescribed fires may also be in danger. intended to thin out competing vegetation or When the owners and builders place their to prepare an area for the natural growth of primary emphasis on aesthetic values and vegetation more resistant to ignition and fire economic considerations without regard to fire spread. protection, the potential for catastrophic loss However, when people move to the for- increases. In the eyes of fire protection offi- ests, land management officials often have cials, these homes are built to burn. Homes in more difficulty conducting prescribed burning the wildlands that are at special risk usually because of the opposition of residents to the share several dangerous traits: smoke and possible risk of fire escaping con- trol. Combustible vegetation Other fuel management programs seek to An approaching fire will ignite surrounding limit the quantity of highly flammable fuels. vegetation in a step-by-step attack on a home. On an individual basis, homeowners could re- A safety zone of low fuel density all around the move fallen limbs and other flammable debris 8 home offers important protection. So does removing fallen leaves and limbs from roofs, Home located on a slope boxing eaves and screening vents. Landscap- Sloping hillsides provide for scenic views ing with fire-retardant plantings can actually and are some of the most desirable sites for help protect homes by repelling fire or giving homes in the wildlands. Slopes, however, cre- it no place to burn. ate natural wind flows that increase the spread Untreated of a wildfire. Slopes with gulleys can create Combustible exterior walls chimney effects that further increase a fire’s Radiated heat from a nearby fire can ignite wood shingle rate of spread. Compared to level ground, a 30 wooden walls without actually touching them. percent slope will double fire spread rates Wooden decks are also vulnerable. Similarly roofs are a while often cutting fire control efforts in half, dangerous is the placement of combustible The survivability of the home is further re- material such as woodpiles under decks, too major factor duced when there are large glass windows or close to walls, or immediately downhill from a wooden deck or unprotected wooden sup- the home. for wildland port piers facing the downslope.

Untreated wood roofs Remote location This is a major risk factor for wildland Homes a greater distance away from fire homes. Wind-carried burning embers can ignite protection equipment will obviously take longer these roofs far ahead of the main flame front. A to reach. Homes with poor access for larger test has been developed to measure the rela- emergency vehicles are in greater danger, as tive ability of roof coverings to withstand igni- are those hidden by vegetation and with poor tion from different size brands. A Class A roof or no markings to help fire fighters locate the protects against larger home. Steep, narrow brands (weighing trunk roads also slow over 4 pounds) than a emergency response Class C roof (with- Black Tiger Fire Time Line: Sunday, July 9,1989 and evacuation for resi- standing brands dents. weighing only I/3 12:35 pm Approximate time of fire origin ounce). Untreated 12:35 - 12:40 First structure (mobile home) involved (approx.) Lack of water wood shingle and 12:40 Fire discovered by residents and extinguishment Piped water sys- wood shake roofing attempted tems with sufficient will not withstand 12:43 Fire alarm received by Sugarloaf Volunteer FD. pressure and hydrants Class C brands. This Estimated size: 40 x 10 feet for fire fighting are means that a brand 12:55 First fire fighters arrive on scene. Estimated size: usually a product of half the size of a 40 x 100 feet higher population kitchen match will 12:59 Crown fire densities and may not ignite a weathered 12:59 Fire fighters on scene request equipment assistance be available in wildland wood roof under the 1:00 Estimated time of steady state burning. areas. Ponds and cis- wind and weather 1:00 Fire fighters on scene request mutual aid departments terns can provide addi- conditions accompa- 1:04 First request aerial tanker support tional options for fire nying most conflagra- 1:05 Additional mutual aid departments requested departments, but they tions. Burning wood 1:30 First evacuations begun (approx.) may be dry in the worst roofing then gives off 1:00 -1:39 Fire spread 78 ft/min average conditions; fire tanker more burning brands. 1:59 Radio confirmation of second structure trucks are seldom New treatments can 1:39 - 2:10 Fire spread 25 ft/min average enough. In the Black successfully improve 2:10 - 2:30 Fire spread 57 ft/min average Tiger Fire area, homes the ignition resistance 2:30 Type 1 interagency fire management team ordered were usually served by of wood roofing, but mobilized individual wells with only if the treatment 2:30 - 3:30 Fire spread 55 ft/min average small pressure tanks allows the roofing to 3:30 - 6:00 Fire spread 45 ft/min average that stopped function- be certified as passing 6:00 Last home lost ing during the fire when fire tests for at least a 6:30 Fire area: 1,500 acres burned poles and fall- Class C rating. Overnight Bulldozers cut 120-ft line ing wires cut off the July 10 Greater state and federal fire fighting assistance arrived electricity. July 13 Final control announced

9

THE FIRE

Weather Conditions and Topography tional mile approaching the Sugarloaf commu- The conditions on Sunday, July 9 in that nity. Then the slope increases to about 34 part of Colorado had all the elements in place percent from the base to the top of Sugarloaf for a dangerous fire, lacking only an ignition Mountain, with an elevation of 8,917 feet above source. sea level. The slope over the total distance of Rain had not fallen for at least 30 days the fire averages 23 percent. during an extended period of high tempera- Weather tures. The dry conditions were long term; snow Fire Origin pack the previous winter was only 25-75 per- Report: The Black Tiger Fire began on private cent of normal On Sunday, July 9, the tem- property along Highway 119 in the Sunnyside perature was again near 100 degrees. Humid- Hot, Dry area about seven miles west of Boulder, Colo- ity in the previous week was reported to be in rado, around 12:35 p.m. on Sunday, July 9, the single digits, although at the nearest official Windy 1989. weather station three miles east of the fire the The county sheriff determined that the fire reading at 1:00 p.m. on July 9 was 24 percent. was accidentally set-probably by a carelessly Dry winds were blowing up the Black discarded cigarette, The fire was first reported Tiger Gulch with greater force than usual. Fire as a small grass fire, 40 feet by 10 feet, by an fighters on the scene estimated the upslope area resident who, with the aid of other resi- wind speeds in the dents, unsuccessfully early stages of the fire attempted to extin- to have been varying guish the fire. between 15 and 25 Sugarloaf Volun- miles per hour. At the teer Fire Department weather station men- personnel were avail- tioned above, observa- able at the station pre- tions at 1:00 p.m. indi- paring for a commu- cated a wind speed of nity picnic when the 8 miles per hour. first radio alarm from The distance from the county was re- the point of origin to ceived at 12:43. The the northwest terminus alarm notification also of the fire at Sugarloaf included the U.S. For- Mountain is 2.5 miles. est Service and the Along the first 3,000 Colorado State Forest feet northward from Service. This message the point of fire origin, indicated that there Black Tiger Gulch rises was a “possible Wailer 720 feet, for an aver- endangered at that lo- age slope of 24 per- cation.” cent. Parts of this area Arriving on the Sliced Coffee Cup slope as much as 35 scene at 12:55, Sugar- percent. loaf VFD personnel re- Black Tiger Gulch can be visualized by comparing it to a sugar bowl or At the top of Black ported, “The fire is coffee cup sliced down the middle. The result is a side that slopes upward and Tiger Gulch the slope spreading and looks outward toward the top and also curves around. Heatedair at the bottom becomes less steep fur- like it is covering about naturally funnels up the slope and accelerates over the ridge. ther northward for ap- 40 by 100 (feet).” proximately one addi- 11 By 12:59 fire per- sonnel on the scene had requested more tank- spread in different areas, the timed progress of ers, a portable pump and as many portable the fire was noted from aerial observation, as containers as possible for collecting tanker shown in Figure 2. Measurements were taken water in the field, adding that they had some of the distance of fire spread and this was structures that were exposed. A storage trailer divided by the minutes of travel. adjacent to where the fire started was the first Energy density is a number describing the structure to be lost to the fire. (Figure 1 shows energy released by a square foot of a particular the approximate distribution of structures in fuel. Ponderosa pine had an energy density of the affected area.) The fire burned through the slightly more than 1,000 Btu/ft2. Mixed conifer grass into the forest and spread up the steep fuels had an energy density near 2,000 Btu/ft2. hillside. Dry meadow areas had energy densities be- One minute later-at 1:00 p.m.-the first tween 100 and 200 Btu/ft2. requests began for mutual aid from nearby fire Calculation of fireline intensity for specific departments. This is also the time estimated for Why one areas of this fire demonstrates that two differ- a condition of steady-state burning used in fire ent types of fire can have the same resistance to spread calculations. At 1:04 the requests for home is lost control: assistance escalated to inquiries about needed - a fire spreading rapidly through fuels response from air tankers to drop fire retardant and another with low energy density, and chemicals, and helicopters for aerial recon- - a fire spreading more slowly but naissance and water drops. Aerial support was survives: through fuels with higher energy densities. not immediately available. The nearest air If either calculation exceeds a fireline in- tanker base was 16 air miles away, but the air many variable tensity 500 Btu/ft/sec, control is not consid- tanker stationed there was already working ered possible by normal organized means. another fire that threatened structures in south- factors. Figure 3 shows the different fireline inten- western Colorado. (As the number of homes sities at the Black Tiger Fire, The fire started threatened by the Black Tiger Fire grew, and made an initial rapid run up Black Tiger however, the air tanker was diverted to Black Gulch, shown as area A. The heat from the Tiger, a distance of 260 air burning brush and fallen miles.) tree debris served to ignite Additional fire depart- ladder fuels until this be- ments were requested at came a crown fire among 1:05. A nearby department the predominate ponderosa preparing to respond with pine. Radio confirmation of mutual aid asked for clarifi- the crown fire came at 1:10 cation at 1:09 about which when the fire commander type of apparatus was reported: “We need all the needed: “Would you ask help we can get as quick as Sugarloaf whether they you got it. We’re crowning want forest fire-brush fire out rapidly here.” At that equipment or structure time nothing in the terrain protection, or both? offered a major obstacle to The answer was clear: slow the fire and help pro- “Respond all the equipment tect the homes uphill in the you can." path of the fire. Crowning generally is a Fire Intensity and two-step process. Accumu- Growth lated ground material, in- Fireline intensity is the cluding downed limbs and best indicator of the fire’s trees, burned and created destructive force and resis- intense heat to further dry tance to control. The inten- out and ignite tree cano- sity is computed by multi- pies. Where the fire burned plying the rate of spread by into timber stands with tree the energy density for the canopies touching or inter- particular fuel type found mingled, crowning oc- in the area. curred. Crowning could not To determine rates of be sustained if the ground fires ran out of heavy ladder fuels. Where cano- that its effectiveness as a fire break would be pies were not touching, the crowns generally inadequate. Fire crews fell back to the next did not burn although they were usually road, Sugarloaf Road. Evacuation of home- scorched to the top. In general, however, owners began, affecting about 100 homes. crowning occurred in all map reference areas All homes along Waterline Road were ulti- except I, J, L, P, Sand R. (Areas D, E and Uwere mately lost. The first radio confirmation of a not as completely documented with timelines. destroyed home came at 1:59, the victim of fire Due to the generally lower intensities in these spotting ahead of the main conflagration Mean- areas it is probable that the areas were burned while, valuable time was gained to set up de- by either flanking, backing or originating spot fenses for the homes along portions of Lost fires.) Angel Road and Sugarloaf Road. The steep terrain and rate of spread upslope Fuel types and their proximity to structures made direct fire attack on the fire’s head from determined in large part where defenses might the point of origin impossible. No roads pro- be possible. In areas of highest fireline inten- vided access to the area immediately above the Fuel types sity the fire break would have to be so wide as point of origin. The fire simply outran the fire to be impractical with limited numbers of fire fighters who were on foot in difficult terrain, and their fighters. Figure 4 shows fire characteristics for putting them into a defensive mode from the fuels in areas that would be controllable by earliest minutes. Indirect attack on the flanks proximity to hand crews, controllable by bulldozers clear- was not an option due to the speed of the fire ing wider breaks, and areas not controllable by and structural threats ahead of the fire. structures any immediate means. Table 1 presents fire Areas B and F were of different fuel cover behavior information in a matrix form. type (mixed conifer), but they burned rapidly determined Reaching the eastern crest of the gulch, the and with similar intensity to area A. fire slowed further at area H, and the fireline As the fire burned into more ponderosa where de- intensity decreased. pine-but in different topography-at areas C Area L represents a riparian zone of either and G, it slowed somewhat. At a calculated fenses might no burning or reduced energy density, Fire fireline intensity usually above 500 Btu/ft/sec, extension to areas I and P beyond this riparian these areas were still impossible to control. be possible. zone occurred from spotting over area L. The next defensive position was deter- A wind shift at the top of the ridge sent the mined by the next available roadway crossing fire in a northwesterly direction, This helped the area that would provide some ready break lower the fireline intensity at area 0 and al- in the fuel and allow access by the fire equip- lowed fire fighters to protect some of the homes ment. Fire fighters began to race around the in the area. fire and up to Waterline Road (see Figure 1), More mutual-aid fire equipment arrived following a ridge, which also gave them access and was set up to fight the fire. Communi- to the first homes to be threatened by the fire. cations became more complex as additional Running along the ridge top, Waterline Road units began operations. A third major stand connects with Lost Angel Road. was attempted on the east flank of the fire Waterline Road, however, was too narrow along Sugarloaf Road; if the fire extended far to allow two fire trucks to pass side by side, and past the road, another hundred homes would it featured close-in trees and brush on both have been at risk in other residential areas. Fire sides Although smoke obscured the size of the fighters had a chance to hold here because the fire, fire fighters could hear its roaring and area was out of the main path of the high determined that the fire intensity approaching intensity fire up Black Tiger Gulch and on up to them was high. Sugarloaf Mountain. Some homes were saved These fire fighters made a quick decision and some were lost. The outcome was often that Waterline Road was not a defensible posi- influenced by a home’s location, whether tion, due to the size of the approaching fire and construction features aided or resisted the the danger of having fire crews and equipment spread of the fire, and whether human inter- caught and overrun by the fire before they vention was available at the critical moment. could maneuver to escape. One fire officer Homes farther from Sugarloaf Road and those was almost trapped near where Waterline Road at the end of long dead-end driveways were at meets Lost Angel Road; he had to drive to a greater disadvantage and many were de- safety through fire burning on both sides of the stroyed. road. Back in the center of the main fire front, The narrowness of Waterline Road meant 15 area J shows a decreased fireline intensity due to the lighter, dry meadow fuels found there. any other area, but again spotting from areas N However, the rate of spread was increased by and K allowed the fire further to spread into wind in the flashy grass fuels. ponderosa pine at area T (Sugarloaf Mountain) Moving again into ponderosa pine fuels at an uncontrollable fireline intensity. Areas W with high energy density and higher rates of and V showed similar intensities but were in spread, area M shows increased fireline inten- mixed conifer fuels. sity equal to the level of area A up Black Tiger By now some 40 fire departments had re- Gulch. These areas were not controllable. sponded, some from other counties. The rapid For the third time, fire crews had to pull growth of the fire and the numerous mutual- back and re-group, but the news at 2:18 was aid departments made coordination and com- that the ETA for the closest aerial tanker was munications especially difficult. Some depart- only another 10 minutes. Then at 2:21 one unit ments in the Boulder County Fire Association reported: “I’ve been advised by the [spotter] had trained together, but this fire was larger aircraft that we’ve got erratic fire behavior. than anyone had anticipated. One result was We’ve got a crown fire making a run onSugar- that many fire companies were operating inde- loaf Road at this point. If you could haveNed- pendently. erland [Fire Department] send After 6:00 no more homes up their gear we’re going to need were burned. The last two homes it for structural protection.” Narrow escape lost were obscured by vegeta- Area N was similar to areas B tion, and mutual-aid fire crews and F, with mixed conifer fuel The timing of this fire provided for the unfamiliar with the area did not cover but continuing high fire greatest possible number of available volun- see the homes until it was too late intensity beyond control by teer fire fighters on the scene in the shortest or feared that the excessive vege- available fire agencies. possible time. But a Sunday afternoon meant tation would compromise the In area K the fuel was again that morepeople were at home and atrisk from safety of fire personnel as it began ponderosa pine but the topogra- a fast-spreading fire. The evacuation load was to burn. By 6:30 the fire had cov- phy and wind shift toward the higher. One resident was burned enough to ered 1,500 acres. west slowed the rate of spread. require hospitalization when be delayed his At nightfall the fire was at the As in similar areas C and G, the escape too long. Several others experienced the base of Sugarloaf Mountain. Three intensity here was beyond pos- terror of a wildfirefrom unexpected close range. bulldozers cut one line 120 feet sible control. wide along the back side of Sug- Area R was more dry mead- "I barelygot my - - - out of there, “said one arloaf, and nighttime weather ow grasses with a high rate of resident quoted in the newspaper. "First I conditions contributed to a slow- spread but lower energy den- smelled smoke, and then I saw flames. When I ing of the fire’s spread. sity. left, there were flames on both sides of the road. Although no more homes Meanwhile, a Rocky Moun- I was choked with smoke, and once I bad to were lost after Sunday, the fire tain Interagency Type 1 fire wait for a wind to blow the smoke away be- continued to burn hot on Mon- management team was ordered cause I couldn’t see the road. I cried like a baby day, advancing very little. Greater mobilized by 2:30 along with when I got to the end of the road. " state and federal fire fighting as- seven 20-person fire fighting sistance on Monday helped ef- crews. An additional five crews forts to contain the fire, as did came later, and more bulldozers rainshowers on Tuesday. Final were brought in to cut an additional fire line. control was announced on Thursday, July 13. Help also came in the form of contract planes Figure 5 shows the location of saved and dropping fire retardant and helicopters drop- destroyed structures in relation to the areas of ping water. similar fireline intensity. Figure 6 shows the Another riparian zone was at area s, where location of saved and destroyed structures in a higher proportion of homes survived than relation to the timeline of fire spread.

17

23 ANALYSIS

This wildfire resulted in the loss of more predominates. When the fire reached the next homes and other structures than at any other fuel type it was mixed conifer, which contin- time in Colorado wildfire history. The signifi- ued to feed fire intensities impossible to con- cance of this fire is that similar conditions exist trol. Table 1 and Figure 4 describe areas of fuel elsewhere in Colorado and in other states. types and resulting fireline intensities. One of the most frequent comments of- fered by homeowners in explanation for not Fuel arrangement. Fuel arrangement preparing for a wildfire is that the fire depart- facilitated the conversion of a surface fire to a ment will “handle” the fire. Once ignited and crown fire, The lack of small, natural, or con- without immediate attack, however, a fire under trolled fires or vegetative management meas- the conditions found in Black Tiger Gulch on ures in the area resulted in a heavy forest litter this July day would have been unstoppable by buildup of dead trees, limbs and brush. These any available fire fighting forces. fuels, in conjunction with remaining low The major factors contributing to this fire branches of live trees, formed ladder fuels. The and the loss of structures are numerous: dead forest litter was also more susceptible to drying out from decreased atmospheric hu- Weather conditions. The hot, dry, Aerial midity than live fuels. If this surface litter had windy weather lowered fuel moistures and been replaced with vegetation of low energy prepared the area vegetation for easy ignition reconn- density, crowning would not have been sus- and rapid fire spread. These conditions occur tained, (Note: Extremely dry live fuels could seasonally in Colorado and elsewhere. Fire aissance sustain a crown fire despite removal of com- protection agencies can predict when these bustible litter. Also, complete removal of conditions present the greatest fire danger but indicated needles and other vegetation could contribute cannot otherwise control them. Homeowners to soil erosion.). who wait for an announcement of high fire burning roofs danger to prepare their properties have lost the Fire spotting. Crown fires in dried time for appropriate and thorough prepara- a quarter-mile fuels under windy conditions caused fire spot- tion. ting ahead of the main fire front and further ac- or more celerated fire spread, as well as expanded the Topography. The steep topography affected area. This exposed more homes to the of Black Tiger Gulch created a chimney effect ahead of the fire while reducing the time available for fire that funnelled the fire toward homes at the top fighters to respond to each new outbreak. The of the canyon. The direction of the prevailing main fire additional fires diluted the number of available winds, the direction of the slope, and the fire fighters for each particular burning area. normal rise of heated air combined to increase front. Fire conditions leading to spotting also re- strong upslope winds that pushed heat closer duced the effectiveness of firelines, roads and to unburned fuels ahead of the flames, pre- other fuel-free areas normally expected to help heating the fuels and accelerating the combus- slow or stop a spreading wildfire. Aerial obser- tion process. vation indicated that fire spotting occurred as far as one-third of a mile ahead of the main fire Fuel types and densities. Fuel types front. in the area contributed to the unmanageable intensity of the fire. Fuel types have been rated Construction features. Homes in the according to their relative wildfire hazard. path of the highest fire intensitywere often not Ponderosa pine, Douglas fir and mixed conifer savable regardless of total fire suppression are rated in the highest hazard category. The efforts. Given the quantity and type of fuels in fire stated in an area where ponderosa pine 24 the fire zone, the amount of radiant energy

generated could ignite combustible homes the studied structures, this property owner had across normally recommended firebreaks. In a cleared fuel break of 75 feet in front of the most cases the survival or destruction of a approaching fire. The roof was asphalt and home was affected not by a single factor but by was wetted down by the owner before he the combined effects of multiple factors. For a evacuated. Although radiant heat caused heavy more in-depth examination, refer to the fol- damage, the structure survived. lowing section, “Statistical Analysis of Factors Structures 17 and 28 were saved by fire in Damaged Homes. " department action along Lost Angel Road, Figure 7 shows the location of 28 structures where a stand was attempted after waterline selected for detailed study of multiple con- Road had to be abandoned. Number 17 had a struction features and loss factors. wood shingle roof but stucco exterior and no Table 2 shows the features of the 28 homes open deck. Although adjacent vegetation came selected for detailed comparison. The most to within 5 feet of the home, the fuel bed was consistent factor associated with structure loss a riparian zone and the approaching fire was of the studied homes in this fire was the dis- on the ground, not in the crowns. The resulting tance of the homes from adjacent combustible fireline intensity was a manageable 100 Btu/ft/ vegetation. In every case except one, the prox- sec and good access allowed fire fighters to imity of homes to wildland fuels was 20 feet or save the home with damage to the roof. Only less. One structure (map refer- one studied home faced fire- ence number 171 had the ad- line intensities lower than 100 Editorial vantage of an adjacent moist ri- Btu/ft/sec. Number 7 shows a parian zone, but combustible low fireline intensity of 60 with “Our foothills and canyons are a beautiful vegetation was allowed to grow a rate of spread of only 3 feet and often perilous place to live. Many people who to within 5 feet of the structure. per minute, but fire department choose to do so take prudent steps to protect their Table 2 also shows a high action was at a disadvantage property. Others, as officials working the Sugarloaf occurrence of wood exterior because the fire reached the fire pointed out, have not done so, making efforts construction and open wood area before full mobilization to save their homes futile. decks, porches or balconies, was possible and the long dead- "As we have stated here before, the City of which can provide a ready end driveway discouraged en- Boulder has devoted countless hours and resources source of easily ignitable fuel. try by emergency vehicles. to preparing for a flood that may come this year or Decks, porches and balconies Furthermore, vegetation clear- in a century from now. Asimilareffortbas not been can serve as an easy path for ance was only 5 feet from a made to anticipate and mitigate the results of the burning vegetation to reach combustible exterior on a slope fires that flare up in the bills every year. " structures, and they are vulner- of 20 percent. As a result, the able to wind-carried brands home was lost. -Boulder Daily Camera, from the main fire front. Such Number 27 showed the decks become even more haz- July 11, 1989 benefit of fire department and ardous when the areas under- owner action. Its location meant neath them are used to store that the fire did not reach it for combustibles such as firewood. several hours, giving time for fire fighters to The multiple-factor effect can be seen in better organize defenses and bring in bulldoz- several examples. Structures 5 and 6 in Table 2 ers to build firebreaks. A fireline was cleared had metal roofs, which provide an extra meas- 200 feet from this home and the owner stayed ure of protection compared to combustible to extinguish spot fires. The owner had also roofing. However, in both cases combustible prepared the site by removing trees and limbs vegetation was only 5 feet away when the fire killed by beetle damage, although some vege- front moved through. Structure 5 was adjacent tation remained as close as 10 feet. to dry meadow fuels that produced a relatively Fire department intervention was attempted low fireline intensity, but the home had an at number 26, where another bulldozer began open pier foundation that allowed rapid igni- to clear a fireline. However, this was an area of tion. the highest intensity of the fire (1500 Btu/ft/ On the other hand, map reference3 was on sec) and vegetation was as close as 5 feet from a 10 percent slope adjacent to mixed conifer the structure, which also had open wood decks fuels that produced some of the highest inten- on two levels. The bulldozer had to abandon sity flame of the entire incident. The rate of the line due to the high intensity and rapid spread was a rapid 45 feet per minute, and spread of the fire. The home was destroyed. flame length was at 11 feet. Yet, unique among 26 -continued on page 30

eatures and Loss Fact

Fire Type Flame Rate of Fireline N Length Spread Intensity (ft) (ft/min) (Btu/ft/sec) 1 heading surface 6-8 25 40 total -limited defensible space, open wood deck 2 heading crown 9-11 45 800 total -high intensity fire, limited def. space

3 heading crown 11 45 1500 saved -hvy. damage, large def. space, owner wet roof 4 flanking surface 4 80 140 total -rap. spread thru grassy slope, open wood porch heading surface 4 60 120 total -limited def. space, open pier foundation heading crown 8 30 600 total -high intensity fire, limited def. space, deck 7 flanking crown 2-4 3 60 total -steep slope, limited def. space, long driveway

8 heading crown 4-6 15 300 total -wood roof, limited def. space, gone in 15 min.

9 heading crown 4-6 15 300 total -wood roof, slope, little def. space, poor access 10 heading crown 11 45 1500 total-intense fire, 11 flanking surface 4 15 250 total-fire spotting, limited def. space,

12 heading surface 68 25 450 total-limited defensible space, 13 heading surface 68 25 450 total-owner action too late, limited space, balcony 14 heading crown 10 45 900 total-intense fire, 15 heading crown 8 45 800 total-3 bldgs, limited def. 16 flanking surface 68 25 450 total-garage only, more def. space saved home 17 heading surface 2 10 100 saved, wood roof damage-FD exting., access good 18 flanking surface 4 10 180 total-fire spotting to wood roof, open deck 19 flanking surface 4 10 180 total-spotting, open deck, long dead-end driveway 20 heading crown 10 3 540 total-spotting, steep slope, open deck, wood exterior

21 heading surface 10 70 140 total-wood foundation, limited def. space, grassy slope 22 heading crown 10 60 1000 total-intense fire, limited def. space, comb. constr.

23 heading crown 10 47 875 total-limited def. space, open wood decks, intense fire 24 heading crown 10 47 875 total-heavy surr. vegetation, intense fire, wood constr. 1 25 heading crown 10 47 875 total-intense fire, limited def. space, open wood deck ’

26 heading crown 11 45 1500 total-fire line unsucc., intense- fire, limited def. space 1 27 flanking surface 4 25 450 saved-successful owner & FD dozer line intervention 28 flanking surface 4 10 180 saved-garage damage only, FD used hot tub water

29 Map reference 28 denotes a home saved With radio communications strained, many with damage to the garage only. Lost Angel mutual-aid departments were left to operate Road helped slow the fire in this area, which independently after being assigned a general had an intensity of only 180. However, total area. Local coordination of ground and aerial loss would have resulted without fire depart- water tankers, evacuation, and awareness of ment intervention, due to the close proximity the fire’s overall progress were hampered. of combustible vegetation and the wood roof Mutual-aid departments were not as familiar and exterior. Combustible materials were also with the area as the home department. Another stored adjacent to the garage. Although fire recommendation from the local critique of the department personnel were available here, fire suppression activities was in the future for water tankers were not, so fire fighters extin- each fire district to prepare maps of their own guished the garage fire with water from the area to present to arriving mutual-aid depart- owner’s hot tub. ments So that mutual-aid per- r sonnel could be tracked better, w Command and con- it was suggested that a check- trol. The rapid growth of the Mountain Residents Must Take Care in sheet listing all responding fire required a multijuris- personnel be carried on each dictional fire attack that compli- "The recent fire on SugarloafMountain and apparatus for collection at the cated the command structure the continuing concern for the residents who lost staging area. and strained communications Some requested mutual-aid homes or who bad severe fire damage has both- capabilities. departments could provide ered me. After the fire made its first only limited assistance to this rapid run across a broad front “While I feel deeply for those who lost their fire area because of a concern where fire intensities were espe- homes, there does not seem to be any assignment that the fire would extend into cially devastating, fire fighting of blame on those who build in a wooded area their own areas. forces were still not available in and take no precautions to protect their invest- sufficient time and number to ment. w Emergency access. simultaneously protect all re- “Gov. Romer is proposing small business loans Roadways and restricted drive- maining homes and continue and other relief for tbose people who had no in- ways to homes limited re- the broad attempt to stop the surance or were underinsured. I believe this type sponse and operational choices wildfire. Mutual aid fire depart- of concern dilutes the responsibility homeowners of fire fighters. The mountain- ments can provide critical assis- must take to protect tbeir property.. ous terrain featured roads that tance when a major fire threat- twisted to follow the hillsides “People who choose to build in flood plains, at as a single jurisdiction. In this and ridges. Old mining activity or near the end of runways, or in wooded areas case several fire departments influenced the Iocation of many had trained together to handle must assume the consequences of the risks they roads, which in some cases a major fire, but the unprece- take. " followed the same paths used dented size of this fire caused by horse-drawn wagons. Al- -Dave Haughey unexpected problems. though paved, these roads JuIy 20, Fort Collins The radio command chan- typically were not widened suf- nel was occasionally over- ficiently for easy use by large whelmed due to the quantity of emergency vehicles. messages attempted. On sev- Fire command officials eraI occasions the communications center tried reported making some decisions to defend or to remind personnel to reserve the command sacrifice individual homes based on their ac- channel for critical communications and to use cessibility. If there was a danger of fire crews or other channels for local coordination. One of apparatus becoming trapped from rapid fire the recommendations from local critiques of blowup - or the structure was prone to espe- the fire suppression activities was to acquire a cially easy and rapid ignition, such as with repeater radio channel for better penetration wood shingle roofs-then z+ structure would and coverage in mountainous areas. more likely he abandoned,

30 Statistical Analysis of Factors and some of the factors do not divide sharply in Damaged Homes into “good” and "bad" categories, a statistical analysis has to be considered tentative, ex- ploratory and illustrative. Nevertheless, the Table 2 provided a two-part matrix over- results of this exercise are plausible and should view of home features and loss factors for the provide food for thought. Some factors (e.g., 28 houses that were badly damaged or de- dead-end distance, clearance from vegetation) stroyed and for which detailed information can be seen as potentially important because was available. It may be seen from the fireline they were rated as bad in nearly all cases. Other intensity column that these homes include some factors (e.g., wood roofs, slope) often were not where the fire was so severe that no combina- bad in homes when the fireline intensity made tion of home features and fire department A less severe severe or total damage a nearly foregone con- response could be expected to prevent de- clusion but often were bad in homes where the struction. The other homes, where the fire was fire will be fire intensity left the outcome more in doubt. less severe, are likely to have suffered losses Specific findings and comments on each factor that were more sensitive to their home features more likely to are as follows: and fire department response, because the lesser fire severity left the outcome more in need some Roof Material doubt. For more severe fires, 15 of 17 homes (88%) The deciding line for these two groups may flaws in the had good roofs, while for less severe fires, only be drawn at 500 Btu/feet/second, because, 6 of 11 homes (55%) had good roofs. This 33 according to Richard Rothermel of the U.S. home design percentage point spread was the largest spread Forest Service, fires of this severity are usually for any factors (although slope showed the beyond control by any manual means. Tables and/or mainte- same spread). 3 and 4 therefore divide the 28 homes into a Put simply, in a very severe fire area, homes group of 17 that encountered fireline intensi- nance to sup- didn’t need to have untreated wood shingle ties of at least 500 Btu/feet/second and 11 that roofs to succumb to fire. In a less severe fire encountered lesser intensities. port spread to area, it was much more likely that an untreated To further support an explanatory statisti- wood shingle roof was there to facilitate the cal analysis, the major factors also have been the home. entry of the fire into the house. The 33 percent- divided into "good" and “bad” groups. For roof age point spread is evidence that this was one materials and exterior construction, wood is of the most important factors in the damage to bad and anything else is good. For eaves, open those homes where the outcome was most is bad and closed or none is good. For decks, sensitive to home features. porches and balconies, having one is bad and Important as it was, there is reason to be- not having one is good. All of these reflect lieve that the potential hazard created by un- significant differences in the vulnerability of treated wood shingle roofs did not more fully exposed exterior features to ignition by an emerge in this fire, as it has in previous fires, established hostile fire. because ofthe happenstance ofprevailing wind The lead investigator detemined 200 feet is velocity. In the most serious wildfires of recent the maximum dead-end road access distance decades where untreated wood shingle roofs that could be called good under these severe were a major spread factor, wind velocity had burning conditions. A slope of 20% or more been much higher than the roughly 10 to 15 was categorized as bad, and a good rating on mile per hour rate in the Black Tiger fire. For clearance from vegetation required at least 30 example, the $50 million Anaheim, California, feet distance. fire in 1982 featured gusts up to 60 miles per With this scheme, it may be seen that the 17 hour. In such conditions, not only the flamma- homes facing the more severe fire averaged 3.1 bility of untreated wood shingles but also their good ratings out of seven possible, while the 11 suitability as a source of flying brands create a homes facing the less severe fire averaged 2.5 role as vector of extensive fire spread that can good ratings out of seven. This is consistent be even clearer and more unmistakable than with the hypothesis that a very severe fire will was the case in the Black Tiger fire. overpower even a "good" home, while a less NFPA maintains records of a distressing severe fire will be more likely to need some number of large-loss multiple structure fires flaws in the home design and/or maintenance where untreated wood shingle and shake roof- to support spread to the home. ing was a contributing factor. More recent ex- Because only 28 homes were in the study 31 ample include fires at Davis, California, March Table 3 Home Features Rated (Fireline Intensity At Least 500Btu/feet/second)

Map Degree Roof Eaves Deck/Porch Exterior Dead End Slope Clearance Ref. of Loss Material Open/Closed /Balcony Construction Distance % from Veg.

1 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Good (150’) Good (5%) Bad (10’)

2 Total Good (Asphalt) Good (Closed) Good (No) Bad (Wood) Bad (1100’) Good (15%) Bad (10’) 3 Saved Good (Asphalt) Good (Closed) Good (No) Good (Con/Wd) Bad (1000’) Good (10%) Good (75’) 6 Total Good (Metal) Good (None) Bad (Yes) Good (Metal) Bad (2000’) Good (0%) Bad (5'') 10 Total Bad (Wood) Bad (Open) Bad (Yes) Bad (Wood) Bad (4200’) Good (15%) Bad (10’) 12 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Bad (900’) Good (10%) Bad (10’)

13 Total Good (Asphalt) Good (Closed) Bad (Yes) Good (Metal) Bad (1200’) Good (10%) Bad (20’) 14 Total Good (Asphalt) Bad (Open) Bad (Yes) Bad (Wood) Good (100’) Good (10%) Bad (10’)

15 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Bad (600’) Bad (25%) Bad (15%) 16 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Bad (850’) Good (15%) Bad (20’) 20 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Good (200’) Bad (35%) Bad (20’) 21 Total Good (Asphalt) Bad (Open) Bad (Yes) Bad (Wood) Bad (900’) Good (10%) Bad (5’)

23 Total Bad (Wood) Good (Closed) Bad (Yes) Bad (Wood) Bad (600’) Good (15%) Bad (5’) 24 Total Good (Asphalt) Good (Closed) Good (No) Bad (Wood) Bad (300’) Good (0-5%) Bad (10')

25 Total Good (Asphalt) Good (Closed) Bad (Yes) Bad (Wood) Bad (300’) Good (0-5%) Bad (20’) 26 Total Good (Asphalt) Bad (Open) Bad (Yes) Bad (Wood) Good (200’) Good (15%) Bad (5’)

27 Saved Good (Asphalt) Bad (Open) Bad (Yes) Bad (Wd logs) Bad (900’) Good(l5%) Bad(l0’)

Table 4 Home Features Rated (Fireline intensity Less Than 500 Btu/feet/second)

Roof Eaves Deck/Porch Exterior Dead End Slope Clearance Material Open/Closed /Balcony Construction Distance % from Veg.

4 Total Good (Asphalt) Bad (Open) Bad (Yes) Bad (Wood) Good (150’) Good (15%) Bad (5’) 5 Total Good (Metal) Good (None) Bad (Yes) Bad (Wood) Bad (1200’) Good (5%) Bad (5’) 7 Total Good (Asphalt) Bad (Open) Good (No) Bad (Wood) Bad (1800’) Bad (20%) Bad (5’) 8 Total Bad (Wood) Good (Closed) Bad (Yes) Bad (Wood) Bad (2000’) Good (5%) Bad (5’) Good (Closed) Bad (Yes) Bad (Wood) Bad( 1200') Bad (20%) Bad (10’) Good (Closed) Bad (Yes) Bad (Wood) Bad( 1200') Good (10%) Bad(l5’)

Good (Closed) Good (No) Good (Stucco) Bad (250’) Good (15%) Bad (5’) Good (Closed) Bad (Yes) Bad (Wood) Bad (800’) Bad (20%) Bad (20’) Good (Closed) Bad (Yes) Bad (Wood) Bad (800’) Bad (20%) Bad (15’)

Good (None) Good (No) Bad (Wd logs) Bad (400’) Good (10%; Bad (5’) Good (Closed) Bad (Yes) Bad (Wood) Good (200’) Bad (20%) Bad (10')

32 14, 1988; Dallas, Texas, March 21, 1983; where fire department access was likely to be Anaheim, California, April 21, 1982; Tulsa, impaired. Only four of the 17 more-severe fires Oklahoma, April 5, 1980; Los Angeles, Califor- (24%) and two of the less-severe fires (18%) nia, November 16,1980. Each large loss fire in showed acceptable distances, and half of these the record has reported damages of $500,000 had distances right at the assumed upper limit or higher. Brush fires involving wood-shingle of 200 feet. roofing in San Bernardino, California, destroyed 290 homes in 1980 and destroyed 239 homes in Slope Santa Barbara, California, in 1977. With wood roofs, slope was the other fac- tor that clearly discriminated between more- Other Exterior Features severe and less-severe fire intensities. In more- The presence of decks, porches or balco- severe fire intensity areas, I5 of I7 homes nies, and the use of wood in exterior construc- (88%) had good slopes of 15% or less, while tion were problems for the overwhelming only six of 11 homes (55%) in less-severe fire majority of homes, regardless of fire severity, intensity areas had these lower slopes. A large so the differences between more-severe and percentage point difference (17 percentage less-severe fires were modest and not always points) also exists if slopes of 15% and higher in the expected direction. Open eaves were are regarded as bad. Therefore, it appears that rare in both cases. A statistical analysis alone more-severe fire intensity areas were able to would find little evidence to focus attention on overpower homes even when slopes were any one of these features, mild, while less-severe not because they were Your Chance to Become Involved in the Process fire intensity areas were not important in fire more likely to need, or at development, but be- least to benefit from, the NFPA 299, Protection of Life and Property from Wildland Fire, when cause they tend to vary steeper slopes to in- completed and when adopted by a local jurisdiction will be an together. Among the volved homes. less-severe fire intensity important foundation for a majority of the factors discussed above cases, for example, and for the recommendations listed on tbe following pages. Its stated Clearance from purpose is to ‘provide criteria for fire agencies, land use planners, seven of 11 had used Vegetation architects, developers, and local government for fire safe develop- exterior wood construc- Nothing could be ment in areas which may be threatened by wildfire. “Its continuing tion; a deck, porch, or said statistically about development mirrors the first recommendation from this section: balcony; and open or no this factor because only namely, that the best way to solve a complexproblem is to apply the eaves. Seven of the 17 one home in 18 had an combined efforts of representatives from all affected groups. Indeed, more-severe fire cases acceptable clearance the particulars of NFPA 299 are still being considered prior to final had the same combina- factor. It is worth noting issuance in and the NFPA standards-making system is de- tion. This fairly standard 1991, that the one home, signed to facilitate input from all the interested parties listed. construction approach which also was the only was sufficiently popular home to rate good on all To learn more about the current status of NFPA 299 or about how you and the alternatives to it four exterior features, could provide input to theproposed standard, contact the Public Fire were sufficiently varied was also one of only two Protection Division of NFPA at 1 Batteymarcb Park, P.O. Box 9101, that it proved too diffi- homes to be saved from Quincy, MA 02269-9101. cult to get a clear statis- serious or total destruc- tical reading on differ- tion in the face of a more- ences in importance of these three features. severe fire. In fact, the fireline intensity for that Their collective importance in fires of this type fire was the highest recorded; only three homes is not an issue, of course, and is apparent in the were faced with that severe a fire intensity. This descriptions of fire development in the indi- case - Figure 7, map reference 3 - can be vidual homes. counted as a partial success story, illustrating the potential impact of good fire protection Access Road Dead End Distance design and practices to help defend homes in Little emerged statistically on this factor an urban/wildland interface. because nearly all the homes were at a distance

33 RECOMMENDATIONS

Ttzgic losses of homes in the wildlands are preventable. However, combined effotts of the community organizations; fire services; federal, state and local governments; and individual home- owners are necessary to minimize kxses, The Black Tiger Fire has resulted in a greater local and statewide awareness of the problems associated with the wildland/urban interface Nevertheless, a continuing and expanded effort must be undataken to inform the public nationwide of the potential hazards involved in interface areas, to inform them of how they can assess the hazards in their area, and to assist them in alleviating the hazards. Ckezuly, this effort cannot be accomplished by just the indiuidualefforts of one of the listed gro”ps.

FIR Protection Agc?ncies d If wildland fire agencies may be called Fire protection agencies are the groups upon to fight structure fires within wildland with the greatest knowledge of and areas, they should arrange for the experience with the current necessary cross-training and equip- wildland/urban interface fire prob- rnent (including communication lem. Actions need to be taken by d “Fire protection agen- equipment) to deal effectively with these agencies to fwther prepare that type of fire fighting. themselves, as well as the people cies need to conduct an assess- they protect, from the identified fire / Conversely, if structural fire hazards associated with the ment of the risks present in their agencies maybe called upon to fight wildland/urban interface. The pub- jurisdictions and prepare a stra- wildland fires that threaten strut- lic and lawmakers also rely on the tures, they should arrange for the fire service of the community to in- tegic plan to attack those risks.” necessary cross-training and equip- form them of fire protection risk and menttodeal effectivelywithwildland mitigation strategies. fire fighting.

fl Fire protection agencies / As the fire fighters discov- need to conduct an assesstnent of the pa&w ered in the early minutes of the Black Tiger lar risks present in their jurisdictions and pre- Fire, conditions may be present that makeitvir- pare a strategic plan to reduce those risks, The bully impossible for available forces to quickly plan should answer these questions: Are there control a wildfire. Therefore, fire agencies measures currently within their jurisdiction to should concurrentlytmin the public to prevent prevent construction of easily ignitable homes fires and to react properly when a fire occurs. or that provide adequate access to themfo r fire suppression purposes? Which homes can be /A fire prevention and public educa- defended during a wildland fire without jeep- tion effort should combine direct-contact train- ardizing the safety of fire crews? Also, as part of ing with the distribution of additional informa- this assessmat, a baseline study to determine tional materials to those individuals who tray the level of squareness of the public concern- not initially respond to the offer of direct train- ing these issues should be accomplished. ing, Libraries are also willing to retain or dis- tribute provided materials. The details of this tiOnce the prospects for local wildland/ public education effoa will depend on the urban interface fire disaster is known, specific results of the local assessment of the wildland/ training should b-z conducted to prepare fire urban interface risks present. suppression agencies for the fires that can be expected, flFire agencies should determine what 34 interagency cooperative agreements are needed to improve multijurisdictional coordi- quired for structures built on steep slopes, nation in both fire prevention efforts and fire above canyons, combustible materials and suppression. Major fires will require this com- exposures. bined attack, and when the fire occurs it is too late to begin the preparations for different or- d When homeowners become aware of ganizations to work together effectively. the wildfire risk of their own areas, they should join forces with other interested individuals / Fire agency personnel should become and groups to urge lawmakers to respond with better aware of the political process that influ- legislative assistance to require appropriate ences items varying from agency funding to fire safety measures by all of those who live in fire safety and building code regulations. Agen- the affected areas. cies should seek out information from other organizations on how best to work with law- makers to prepare and justify proper regula- Community Planners and Officials tions to help prevent unsafe development and A community’s planning and building offi- lack of maintenance of fire safety features in cials are often the first individuals who com- the wildlands. municate local practices and standards to those who want to buy or build in the wildlands. Their understanding of the potential hazards of The Public/Homeowners building in these areas is therefore vitally im- The people who choose to live in the portant. scenic wildlands have the basic responsibility to take necessary precautions when / A wildland fire behavior po- facing predictable hazards. Informed tential map should be created for homeowners would be better pre- existing and planned structures. The pared for surviving a wildfire, but d “It is the responsibility features specified for the map would some homeowners in the area of the include several topography-related Black Tiger Fire admitted to a lack of of the individualproperty owner factors: elevation, slope percent, knowledge about the wildfire risks to provide a defensible space drainages, prevailing wind direction, where they lived. worst-case wind direction (toward around structures to help pro- structures), and broken topography ti Potential homeowners features. should determine the wildfire haz- tect them.” ard potential of the immediate area d The wildland fire behavior before buying or moving into any potential map should also include home. This information can be ob- fuel types. Zones of possible high- tained from the local fire depart- intensity fire must be identified and merit. NFPA 299 will provide guidelines for rat- communicated to property owners. Fuel modi- ing the wildfire potential of an area, fication-the removal, spacing or volume re- duction of fuel types to accomplish a reduction / Homeowners should contact federal, in fuel loading-is a primary mitigation meas- state and local fire and forestry agencies for ure. educational programs and materials to address the fire hazard in general. Information should / Areas with abnormal accumulations also be shared with children. Information and of forest litter should be identified, and a re- publications covering numerous wildland view made of past fire history in each area fuel home fire safety details are available free from bed. many sources, but until individual apathy is overcome the homeowner may not be moti- / The Authority Having Jurisdiction (de- vated to take proper precautions. (Here, the termined locally) should evaluate all existing fire services can function as fire protection re- or planned housing developments to deter- source centers for the public.) mine relative wildland fire protection ratings. In doing this, jurisdictions must review fire d It is the responsibility of the individual danger weather records to determine patterns property owner to provide a defensible space of rain, heat, humidity and fuel moisture. Then around structures to help protect them. Extra property owners must be advised of conditions measures to provide additional space are re- 35 and their responsibilities. Developers / Developers can provide a valuable This event has also focused on the need to service to new buyers, who may initially be have construction standards for homes in the distracted by other moving details, by creating wildlands. The published version of NFPA 299 appropriate fuel breaks or greenbelt areas. will provide important guidance in this area, but it becomes fully effective only when adopted by local lawmakers. Lawmakers In the absence of clear and meaningful Although the public determines acceptable regulations for the common good, the prac- levels of risk from fire in wildland areas, law- tices of uninformed developers may create makers react to the perceived needs of con- potential hazards. Fire protection features, or stituents and enact the regulations controlling their costs, may not be appreciated by unin- that level of risk. Therefore, it is generally up to formed buyers. However, decisions made at homeowners and fire protection agencies to the early stages of a development will affect a articulate and justify acceptable and unaccept- home’s fire safety for many years in the future. able levels of risk. When tragic losses occur, they usually focus attention on the risks, but /All developments should have more preventive actions are preferable. than one ingress-egress route and employ looped road networks. Roads should be wide tiLawmakers should take the initiative enough for simultaneous access for emergency to examine existing laws, regulations and stan- vehicles and the evacuation of residents. In dards from other jurisdictions that are available consideration ofthe lane, wheelbase of tankers for local use in mitigating fire hazards associ- and other emergency vehicles, roads ated with wildland fires. should be constructed with anade- quate curve radius. Homes along tiLawmakers are encouraged deadend roads and long driveways to adopt NFPA 299, when it becomes provide extra privacy for residents d“Decisions made at the available, as one part of the protec- but also provide the potential for tion provided for new construction fire apparatus to become trapped early stages of a development in the wildlands. by spreading fire. These roads and will affect a home’s firesafety driveways should not prevent easy ti Authorities should provide access by large emergency vehicles. for many years in the future.” other strong building regulations to restrict non-treated wood shingle ti Developers should recon- roofs and other practices known to sider their frequent use of combus- decrease the fire-safety of a structure tible exterior building materials, or in the wildlands. In the past, untreated at least offer options for more fire- wood shingle roofs have repeatedly safe materials for potential buyers who may been shown to be a major contributing factor in not yet understand the differences. the loss of structures to wildfires, yet some resi- dential subdivisions in Colorado and elsewhere flDevelopers should also consider the still encourage and even require wood shingle long-range implications of siting unprotected roofs for aesthetic reasons. homes on slopes or where water supplies for fire fighting are weak or non-existent.

36 APPENDIX Methodology of the Study of the Black Tiger Fire

Uoon after the Black Tiger Fire, represen- have had on the ignition/survival of the homes tatives from the National Fire Protection Asso- was discussed. Again, these technical observa- ciation, U.S. Forest Service, Boulder County tions were corroborated with physical evidence. Sheriffs Office, Colorado State Forest Service, The survey began at the base of the Black and Sugarloaf Volunteer Fire Department met Tiger Gulch where the origin of the fire was to discuss a study of the fire to be conducted The survey pinpointed by local authorities. Coupled with by the NFPA’s Fire Investigations Division. A observations of the available videotapes of the one-day meeting of the participants was held, team began to fire, interviews conducted with local fire au- resulting in the discussion of various ap- thorities and other eyewitnesses, and upon proaches to the analysis. Participants identi- understand examination of the physical evidence, the group fied the available data and how it might be col- began to piece together a scenario of how the lected and discussed. In general terms, the the full im- fire developed and spread. These determina- wildland/urban interface fire problem was tions were then coupled with the fire spread also discussed and how those factors may or pact of the scenario that was documented during the fire may not be applicable to the Black Tiger Fire. suppression effort by the incident command- (Participants included: U.S.Forest Service: effects of ers. These fire spread observations plotted the Bob Swinford, Michelle Nolde, Bob Willmot, “head” of the fire as it varied over time, Chad DeVore, Dick Chase, Dick Rothermel, Al fuels, tem- The survey team also examined the fuels in Roberts; Boulder County Sberiss' s Office: Ernst the immediate area of origin and found that Little, Al Staehle, Larry Stern; Colorado State perature, they were typical of many grass-type fires. Forest Service: Dave Farmer, Ron Zeleny; Sug- However, beyond several acres an extreme arloaf Fire Protection District: Ruth Ravenel; slope, and uphill slope began to affect the terrain and this NFPA: Tom Klem, Bill Baden, Jim Smalley, area was the beginning of forest fuels consist- Heather Hintz.) wind that ex- ing of ponderosa pine and mixed conifer. From In addition to the meeting, an on-site sur- this vantage point, the survey team observed vey of the fire area was arranged through the isted on the evidence of a “crowning” fire that began within Boulder Sheriff’s Office so that participants sight of the area of origin. The survey team could examine the factors that might have led day of the began to understand the full impact of the to the destruction of homes and how to miti- effects of fuels, temperature, slope, and the up- gate other severe fires from causing such ex- fire. slope wind that existed on the day of the fire. tensive damage. A thorough one-day survey of The survey team moved to Waterline Road the burn site including examination of those to view conditions of fire growth. The survey homes destroyed/survived was undertaken. team was able to view several of the homes A complete drive/walk through the fire destroyed in the initial run of the fire. The team area was conducted by the representatives. observed that the homes were located near the During the tour, the representatives spoke to rim of the extreme uphill slope and that com- several residents to supplement previous ob- bustible vegetation was in close proximity to servations and determinations. Researchers their foundations. The remains of the homes from the U.S. Forest Service and other wildland were examined to attempt to identify other fire experts relayed their observations/deter- variables that might also have resulted in the minations with other representatives. Details ignition of the homes. However, these other of the rapid development and spread of the fire suspected variables, i.e. combustibility of were provided and physical evidence shown construction, roof, etc., were more difficult to to corroborate determinations. Further, the identify from the remains because of their de- effects that such rapid growth and spread might 37 struction Other factors also began to emerge, such as would have contributed to the ignition of the what role fire department access to the homes homes. They were: 1) inaccessibility for fire might have played. The long, narrow-approach fighting during a severe wildland fire; 2) lack of driveways were assessed by the team as dan- water supply; 3) close proximity of combus- gerous for fire department use, and consider- tible vegetation to the home; 4) combustibility ing the intensity and speed of the fire, these of roof and structure; and 5) storage of com- homes likely had no fire department interven- bustible materials beneath wooden porches. tion. Based on the assessment of the similar area, Summary comments regarding the homes the survey team then decided to examine local at the rim of the canyon along Waterline Road records of the homes in the fire area to deter- were that the position of the homes, the inten- mine the construction and arrangement of the sity of the fire, the proximity of combustible homes. As a result, the team visited the City of vegetation, the lack of access for fire fighting, Boulder Assessor’s Office and determined that and the short interval of time between the information was available from the assessor’s awareness and the threat of the property were records that include photos of the homes from extremely important variables in determining the affected area. Records were examined and why these homes were destroyed. Variables, The 28 homes it was determined that they contained informa- such as combustibility of the home, that may tion regarding construction arrangements of have been important variables could not be studied were the homes and adjacent combustible vegeta- confirmed. Similar analytic methods were used tion as depicted by a photograph attached to by the survey team the remainder of the day as selected on each record. It was felt that the vital informa- they moved through the burn area. The signifi- tion regarding the construction of a home, cance of the variables, it was felt, varied from the basis of proximity to combustible vegetation, and other location to location. factors could be overlayed with the other vari- Several members of the survey team, in- the best avail- ables to gain insight into the behavior of the cluding Tom Klem, NFPA, and Al Roberts, U.S. home under assault. Forest Service, Denver, continued an in-depth able data and The team decided that a more thorough survey of the burn area for two additional days. analysis of the assessor’s records, coupled with This in-depth survey also included detailed the location an overview of the location of a home and its interviews with numerous occupants in an proximity to the intensity of the fire, might attempt to corroborate the technical determi- of the homes yield important data regarding the variables nations being made of the fire spread and that contributed to the loss or survival of the intervention by the homeowner, if any. relative to homes of the Black Tiger Fire. This survey team conducted an on-site As a result, NFPA contracted with Aspen analysis of every home that burned or survived four identified Engineering, Inc., to conduct this thorough in Black Tiger Gulch. The team began to hy- analysis of the assessor’s records. The Aspen pothesize that the homes that did not burn: fire intensity Engineering representative met with the repre- probably survived due to fire department in- sentative of the U.S. Forest Service Denver Of- tervention; had noncombustible vegetation areas. fice for input regarding the intensity of the fire around the homes or had combustible vegeta- and background information. The contractor tion cleared from the proximity of the homes; also visited the Boulder Tax Assessor’s Office were constructed of noncombustible con- to acquire the necessarydata for such a rigor- struction materials; and/or were more distant ous analysis. The tasks also included a thor- from the main intensity of the fire. It was felt ough survey of the burn site by the contractor that some of the homes that survived could and the U.S. Forest Service representative well have burned had it not been for fortunate Provided with the fire intensity map and circumstances. One such home observed had having been briefed on the factors that may hay stored under its wooden porch, yet it was have affected the burning or survival of the not ignited despite enough heat near the com- homes, the contractor began the in-depth bustible hay pile to scorch the paint on a analysis. The findings from this analysis are window frame. The survey team also decided contained in the Black Tiger Fire report and in to conduct a survey of unaffected home sites the contractor’s submittal to the NFPA. within the same general area with similar to- In addition to this task, a statistical analysis pography, home construction and other fac- of the data from affected homes was con- tors. This thorough survey of homes lead to a ducted by Dr. John Hall, NFPA Fire Analysis number of additional variables, if representa- and Research Division, to further verify the tive of the homes in the Black Tiger Gulch, that 38 implications of the study. NFPA State and Local Agencies - Wildfire Strikes Home, 2nd Edition. - California Department of Forestry and Fire - Wildfire Strikes Home video. Protection: Fire Safe Guides for Residen- - Protecting Your Home From Wildfire video tial Development in California. and booklet. - Colorado State Forest Service: Wildfire - NFPA 224, Homes and Camps in Forest Safety Guidelines for Subdivisions and Areas, 1985. Developments. - “The Los Angeles Conflagration of 1961: - Colorado State University: The Role of The Devil Wind and Wood Shingles,“NFPA Private Individuals and Firms, States, and Quarterly, January 1962 (reprints avail- the Federal Government in Protection of able). State and Private Lands from the Hazards - “Statement of the National Fire Protection of Wildfire, A. A. Dyer. Association on the subject of prohibiting - East Bay (California] Regional Park Dis- untreated wood shingles and shakes as a trict: Report of the Blue Ribbon Urban roof covering material,” February13, 1979 Interface Fire Prevention Committee. - Houston Conflagration Spreads Across - Fire Research Institute(FRI) (California): Rooftops, Fire Journal, January 1980. Urban/Wildland Bibliography. Contact - “The Sycamore Canyon Fire” Fire Com- FRI, 1129 Western Drive, Santa Cruz, CA mand, March 1978. 95060. - “Apartment Complex Fire: Wood Shingles - Florida Interagency Fire Group: Proceed- Ignited by Cigarette,” Fire Journal, May ings of the Florida Wildland/Urban Fire 1982. Protection Initiative. - Montana Department of State Lands: Fire U.S. Forest Service Safe Guidelines for Residential Develop- - Fire Safety Considerations for Develop- ment in Montana. ments in Forested Areas. - New Jersey Department of Environmental - A Guide for Homeowners, Lawmakers, Protection: Development in the Pine Bar- and Planners, Howard E. Moore, General rens, A Design for Disaster, Joseph R. Technical Report PSW-50, Pacific South- Hughes. west Forest & Range Experiment Station. - Northwest Interagency Fire Prevention - Wildfire Protection: A Guide for Home- Group (Portland, Oregon): Fire Safety Con- owners and Developers, Mark Hummel, siderations for Developments in Forested Toiyabe National Forest. Areas: A Guide for Homeowners and - A Study of the Impacts of Severe Wildland Buyers, Copies available from USDA For- Fires and Disaster Relief Programs on est Service, P.O. Box 3623, Portland OR Homeowners Residing Along the Wildland 97208. Urban Interface, P. D. Gardner and Linda -Ohio State University Press: New Deal in Pies El-Abd, Pacific Southwest Forest and the Suburbs; a history of the greenbelt Range Experiment Station. town program, Joseph L. Arnold. - Forest Wildlands and Their Neighbor: - Oregon State Department of Forestry: Interactions, Issues, Opportunities; Wil- Planning for Survival: How to Protect Your liam E. Shands. Copies available from State Home from Wildfire. and Private Forestry, P.O. Box 96090, -University of Washington Press: Land Use Washington, DC 20090-6090. and Forest Resources in a Changing Envi- - People and Fire at the Wildland/Urban ronment; the urban/forest interface, Gor- Interface: A Sourcebook, Robert Gale and don A. Bradley, editor. Hanna Cortner, editors. Proceedings of - Utah Department of Natural Resources: the wildland/urban fire interface work- Protect Your Summer Home. shop for social scientists. - Wisconsin Department of Natural Resources: Forest Fires Burn More Than Boise Interagency Fire Center Trees. - Wildland/Urban Interface Resource Materials, PMS-050. 39 Source List of Organizations

For additional information, contact your local fire department or forestry agency, or contact the following organizations:

National Fire Protection Association Public Fire Protection Division 1 Batterymarch Park P.O. Box 9101 Quincy, MA 02269-9101 617-770-3000

United States Department of Agriculture Forest Service Fire &Aviation Management P.O. Box 96090 Washington, DC 20090.6090 703-235-3220

National Association of State Foresters 444 N. Capitol Street, NW Washington, DC 20001 202-624-5415

Boise Interagency Fire Center Publications Management System 3905 Vista Avenue Boise, ID 83705 208-389-2512