Fire Management Today Volume 62 ¥ No

Fire Management today Volume 62 ¥ No. 3 ¥ Summer 2002

THIRTYMILEHIRTYMILE FIREIRE

United States Department of Agriculture Forest Service CORRECTION: BEAR RIVER FIRE

The cover of the Summer 2001 issue of Fire Management Today (volume 61, number 3) shows a silhouetted firefighter identified as a member of the Payson Hotshots. A reader notified Fire Management Today that the firefighter in the photo was actually a member of his crew, the Flagstaff Hotshots.

The photo shows the 1994 Bear River Fire (not “Bear Creek Fire,” as the caption states) on Idaho’s Boise National Forest. The caption states that the fire had a low suppression cost, which is true but misleading. The fire was overrun by the Rabbit Creek Fire shortly after containment. Suppression costs were low in part because mopup costs were minimal.

Thanks for the correction go to Paul F. Musser, Superintendent, Flagstaff Interagency Hotshot Crew, Flagstaff, AZ.

Fire Management Today is published by the Forest Service of the U.S. Department of Agriculture, Washington, DC. The Secretary of Agriculture has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department.

Fire Management Today is for sale by the Superintendent of Documents, U.S. Government Printing Office, at: Internet: bookstore.gpo.gov Phone: 202-512-1800 Fax: 202-512-2250 Mail: Stop SSOP, Washington, DC 20402-0001

Fire Management Today is available on the World Wide Web at .

Ann M. Veneman, Secretary April J. Baily U.S. Department of Agriculture General Manager

Dale Bosworth, Chief Robert H. “Hutch” Brown, Ph.D. Forest Service Managing Editor

Jerry Williams, Director Madelyn Dillon Fire and Aviation Management Editor

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, sex, religion, age, disability, political beliefs, sexual orientation, or marital or family status. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD).

To file a complaint of discrimination, write USDA, Director, Office of Civil Rights, Room 326-W, Whitten Building, 1400 Independence Avenue, SW, Washington, DC 20250-9410 or call (202) 720-5964 (voice and TDD). USDA is an equal opportunity provider and employer.

Disclaimer: The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are responsible for the technical accuracy of the material presented in Fire Management Today.

2 Fire Management Today Fire today Management Volume 62 ¥ No. 3 ¥ Summer 2002 CONTENTS On the Cover: Revisiting the Loss of Our Own ...... 4 Dale Bosworth Lessons From Thirtymile: Transition Fires and Fire Orders...... 6 Jerry Williams Causal Factors in the Thirtymile Fire Accident ...... 9 Thirtymile Fire Accident Review Board Recommended Changes in Wildland Fire Canyon of the Chewuch Management ...... 13 River on the Okanogan Thirtymile Fire Accident Review Board National Forest near Thirtymile Fire Prevention Action Plan...... 14 Winthrop, WA. On July 10, Thirtymile Fire Accident Review Board 2001, the Thirtymile Fire entrapped 16 people in the Initial Hazard Abatement Plan...... 19 canyon. Firefighters de- USDA Forest Service ployed their fire shelters on Crosswalk Between OSHA Violations and Accident the road and scree slope Prevention Plan ...... 22 above the widening of the USDA Forest Service river visible at bottom Thirtymile Fire: Fire Behavior and Management center; four perished in the Response ...... 23 subsequent burnover. Photo: Hutch Brown Ben Croft, USDA Forest Service, Missoula Technogy Rush to Judgment on the Thirtymile Fire ...... 31 Development Center, Hutch Brown Missoula, MT, 2001. Injuries, Illnesses, and Fatalities Among Wildland Firefighters ...... 36 The FIRE 21 symbol (shown below and on the Richard J. Mangan cover) stands for the safe and effective use of wildland fire, now and throughout the 21st Keeping Haines Real—or Really Changing Haines? ..... 41 century. Its shape represents the fire triangle (oxygen, heat, and fuel). The three outer red Brian E. Potter, Dan Borsum, and Don Haines triangles represent the basic functions of wildland fire organizations (planning, operations, and Burning Under Old-Growth Ponderosa Pines on aviation management), and the three critical aspects of wildland fire management (prevention, Lava Soils ...... 47 suppression, and prescription). The black interior Peter Z. Fulé, Greg Verkamp, Amy E.M. Waltz, represents land affected by fire; the emerging green points symbolize the growth, restoration, and W. Wallace Covington and sustainability associated with fire-adapted ecosystems. The flame represents fire itself as an ever-present force in nature. For more information on FIRE 21 and the science, research, and SHORT FEATURES innovative thinking behind it, contact Mike Apicello, National Interagency Fire Center, 208-387-5460. Websites on Fire...... 8 National Fire Plan at Work ...... 12 Guidelines for Contributors...... 50 Photo Contest Announcement ...... 51

Firefighter and public safety is our first priority.

Volume 62 • No. 3 • Summer 2002 3 REVISITING THE LOSS OF OUR OWN

Dale Bosworth

efore we get to the primary vened to ensure the adequacy of the able, such as the recent terrorist reason why we are all here,* I investigation team’s findings. After attacks. Bwould like to share a personal careful review and analysis, and thought or two with you. Two weeks with their input included, the board Let us do our brave wildland ago, the United States suffered a has recommended the report and I firefighters a great service by terrible attack upon its home- have accepted it. remembering the things they have land.** As you all know, the pro- accomplished. This summer, these jected loss of life from those terror- I wish to thank Jim Furnish and his courageous men and women have ist acts is inconceivable. While we dedicated team for their diligence succeeded in saving thousands of mourn the loss of lives on Septem- and professionalism in completing homes. They successfully fought ber 11, we also grieve the tragic loss this most difficult investigation thousands of lightning- and hu- of four of our young firefighters assignment.*** Jim has assembled man-caused fires, controlling them here on the Okanogan–Wenatchee a team of highly experienced at sizes smaller than this room. National Forest on July 10. members with expertise in many Overall, they have accomplished areas, including wildland fire and many millions of hours of danger- Sense of Loss weather behavior, suppression ous, back-breaking work accident The tremendous sense of loss I tactics, safety, fire crew skills, free. personally feel for these firefighters’ training, and equipment, just to families compelled me to be here— name a few. The investigation Never Again in your community—to receive the report reflects the fact that five But sometimes, no matter how investigation team’s final report. members of the investigation team, much safety is stressed, accidents The loss of any firefighter is tragic; including Jim, have more than 30 do occur. And they can occur with and now, today, we must, for good years of agency experience, and that tragic results, such as serious injury reasons, revisit the loss of four of several have prior experience in or death. I promise you that the our own. serious accident investigations, recommendations of this investiga- including fire fatalities. tion team will not be forgotten. I I am here today to personally am charging my leadership team receive the findings and recommen- Jim, I am especially appreciative of with ensuring—and I am accepting dations presented to me by the your bringing in the expertise of a responsibility for the Forest Service Thirtymile Fire Accident Investiga- widely respected chief investigator to do everything within its power to tion Team. Before that could from the private sector, one who ensure—that what happened on the happen, a formal Thirtymile Fire has some 20 years of investigative Thirtymile Fire does not happen Accident Review Board was con- experience, including experience in again. fatality accident investigation, safety management, and human Dale Bosworth is the Chief of the USDA Although the investigation team’s Forest Service, Washington, DC. error analysis. job might be done, the board of review’s job will continue. The * On July 10, 2001, four firefighters died in a burnover I would like to remind everyone on the Thirtymile Fire on the Okanogan National Forest board is now developing an action near Winthrop, WA. The USDA Forest Service once again that firefighting is a plan that addresses the findings and immediately formed a team to investigate the accident. dangerous vocation—whether you This article is based on opening remarks by Chief recommendations of the report, Bosworth at a press conference in Yakima, WA, on are engaged in fighting wildland with the intent of improving September 26, 2001, to announce release of the fires or are called upon to respond Thirtymile Fire Investigation Report. wildland firefighter safety for all to the worst urban tragedy imagin- agencies.**** ** On September 11, 2001, terrorists flew two hijacked airliners into the World Trade Center in New York City, collapsing its twin towers. They flew a third airliner into *** Jim Furnish, the Deputy Chief for the National **** The action plan is completed and being the Pentagon outside Washington, DC; a fourth hijacked Forest System (retired), led the Thirtymile Fire Accident implemented. It is reprinted on pages 14–18. The airliner crashed in the Pennsylvania countryside. In all, Investigation Team. He was present with Chief Bosworth board’s findings and recommendations are reprinted many thousands of people were killed and injured. in Yakima to release the investigation report. on pages 9–13.

4 Fire Management Today Let me close by saying this: I the Thirtymile Fire and the loss of families, friends, and colleagues of continue to search for words to other fire personnel across the those who have suffered—those express how deeply sorry I feel for Nation this summer. Let us keep who have perished—and, finally, the loss of the four firefighters on our thoughts and prayers with the with those who have survived. ■

Firefighter gear. Photo: Josh Shroyer, Missouri Department of Conservation, Clinton, MO, 1998.

Volume 62 • No. 3 • Summer 2002 5 LESSONS FROM THIRTYMILE: TRANSITION FIRES AND FIRE ORDERS*

Jerry Williams

n November 2001, I was flying over Helena, MT, when I looked With few exceptions, transition fires Iout the window and saw Mann are our tragedy fires, the fires where most Gulch. As a wildland firefighter, you of our entrapments and deployments occur. cannot see Mann Gulch without being deeply moved.** More than most places, Mann Gulch speaks to small to large (see sidebar). They National Fire Management Analysis our past and our future. The story became tragedy fires as they grew System for deriving the most of Mann Gulch reminds us of how from something we thought was efficient level of resources. We also far we have come and what chal- under control into something that have flexible options for managing lenges we still face in improving suddenly overwhelmed us. Some- large fires through our Wildland fireline safety. times, such as at Mann Gulch, Fire Situation Analysis. transition fires involve blowup Mann Gulch reminds us of other conditions; at other times, such as However, we remain vulnerable in fires in other places as well, fires at Thirtymile, they do not. In most managing the fires in between. with names like Dude, South cases, we did not recognize the full Transition fires are relatively few, Canyon, Island Fork, and now gravity of the situation until it was but their small number belies their Thirtymile.*** Like Mann Gulch, too late and there was nowhere to significance. Probably because they these were all tragedy fires—fires escape. occur so infrequently, we are where lives were lost. They invite us sometimes unprepared for their to see a connection to Mann Gulch. For most of the fires we manage, potential consequences. We get into Especially after the Thirtymile Fire, our policy is sound. We have strong trouble when we continue offensive they call on us to act on that strategies for initial attack through tactics in a situation that puts us on connection. our preparedness planning—our the defensive. People sometimes get Transition to Tragedy These fires were all transition incidents—fires in transition from WHAT ARE TRANSITION FIRES?

Jerry Williams is the Director of Fire and On average, the USDA Forest Service suppresses about 10,000 wildfires Aviation Management, USDA Forest Service, per year on the national forests and grasslands. Of these, about 92 to Washington Office, Washington, DC. 93 percent are controlled quickly, at little cost, with relatively little effort, and with scant loss or damage. * Based on remarks made by the author on November 6, 2001, at a meeting of wildland fire safety managers in Missoula, MT. At the other end of the spectrum, about 2 to 3 percent of our fires are

** On August 5, 1949, 13 firefighters lost their lives on large almost from the outset. Factors such as volatile fuels, drought, the Mann Gulch Fire on the Helena National Forest, MT and wind combine to produce a fire that quickly leads to a dispatch (see Richard C. Rothermal and Hutch Brown, “A Race That Couldn’t Be Won,” Fire Management Today 60(2): call for a type 1 incident command team. Though rare, such fires 8–10). The accident shocked the wildland fire account for nearly 70 percent of our total suppression expenditures. community, precipitating safety reforms that laid the basis for today’s methods of wildland firefighting. The remaining 5 percent of our fires—about 500 per year—are be- *** On June 26, 1990, six firefighters were killed on the Dude Fire on the Tonto National Forest, AZ; on July 6, tween rapid, successful initial attack and quickly recognized, almost 1994, 14 firefighters died on the South Canyon Fire near immediate large-fire mobilizations. They are fires in transition from Glenwood Springs, CO; and on April 6, 1999, two firefighters perished on the Island Fork Fire near small to large. Cranston, KY.

6 Fire Management Today It is time to develop and adopt that look during every fire season. a more formal approach to transition fires, The shine comes off a little as the season wears on, but it is always one that makes us less vulnerable. there. It projects our “can-do” strength of character. But it is a strength that can defeat us; if hurt or even killed. With few able. Many of the items in the untempered by sound judgment, exceptions, our tragedy fires—the Thirtymile Fire Prevention Action “can-do” can devolve into “make- fires where most of our entrap- Plan are specifically designed to do,” which can lead to tragedy. The ments and deployments occur—are improve our approach to managing role of management supervisors transition fires. transition fires (see the action plan and crew leaders is to instill the items beginning on page 14). At the sense that the biggest part of our Special Challenges fall 2001 meeting of the National job is doing the job right. Doing the Transition fires raise special ques- Wildfire Coordinating Group, the job right by following the rules is tions: interagency leadership committed what keeps “can-do” from slipping itself to such an approach. Support toward tragedy. •What is our strategy for dealing is needed from the entire inter- with a fire specifically while it is agency wildland fire community. Our rules of engagement are the in transition? Ten Standard Firefighting Orders. •What predictive models do we use “Can-Do” Attitude Though simple, they are sometimes to see a transition coming, and Beginning with Mann Gulch, overlooked or ignored. Our “can- how do we use them? transition fires have shocked the do” orientation can impel us to put •Operationally, how do we adjust wildland fire community and operations ahead of safety. The Fire to mitigate the risks of a rapidly precipitated safety reforms. A key Orders remind us that there is a expanding incident? reform was the introduction of the right way to do things before we • In terms of management over- Ten Standard Firefighting Orders.* engage, while on the fireline, and in sight, supervisory control, and The Fire Orders and other safety our after-action reviews. The Fire crew leadership, what do we do measures complement our funda- Orders are not an obstacle to differently on transition fires? mental beliefs, behaviors, and getting the job done; instead, they How do we obtain the astute values as wildland firefighters. They are the way to get the job done situational assessments and fluid are a vital counterweight to a right. operational control required remarkable strength in our culture under rapidly changing condi- that, if left unattenuated, can Discipline Needed tions? jeopardize our safety on the fireline: We especially need to follow the our “can-do” attitude. Fire Orders at times when we might I believe that we do not have a be tempted not to: coherent strategy for what are Our “can-do” attitude is written arguably the most important fires across the face of every firefighter •When the challenge is great we face. We find ourselves doing the on every incident. The attack on the because the consequence greatly best we can with that we have, World Trade Center on September matters; hoping that it will suffice. Usually, it 11, 2001, produced some remark- •When we are tired, fooled by a does suffice; faced with a transition able examples. One photograph deceptive fire, preoccupied, or fire, experienced leaders will disen- shows people in a stairwell stream- complacent; or gage and regroup. But as long as ing downstairs past a New York City •When managers and supervisors good judgment during transition fireman on his way up. His expres- show that they are human by fires is more a matter of chance sion is typically “can-do”: bright- missing or forgetting something than design, another tragedy seems eyed, eager, and ready for a firefight. that they should have noticed or inevitable sooner or later. remembered. That look, that attitude, and that It is time to develop and adopt a spirit define the firefighter. I see The Fire Orders have less value in a more formal approach to transition controlled environment than in an * For more on the Ten Standard Fire Orders, see Karl fires, one that makes us less vulner- Brauneis, “Fire Orders: Do You Know Their Original environment that is uncertain: Intent?”, Fire Management Today 62(2): 27–29.

Volume 62 • No. 3 • Summer 2002 7 The Ten Standard Fire Orders are not an obstacle to getting the job done; instead, they are the way to get the job done right. when equipment fails, when the In our business, especially on The Thirtymile Fire, like other weather changes, when we get tired transition fires, the unexpected can tragedy fires, reminds us that we and our thoughts drift, or when happen without warning. That’s are accountable at all times for our somebody lets us down. Such why: performance as fire managers, crew things happen in the world of supervisors, and firefighters. The firefighters, though infrequently. • Safety comes first on every fire, Ten Standard Firefighting Orders When they do happen, especially every time. are the measure against which our during a transition fire, their • The Ten Standard Firefighting performance as professionals consequences can be fatal. When Orders are firm. We don’t break should be judged. It is time we used the unexpected happens, the Fire them, and we don’t bend them. the Fire Orders not only as guide- Orders assume their real value. And •We don’t just do the job; we do it lines for safe firefighting, but also because the unexpected does right. as criteria for measuring our happen, we must be disciplined in firefighting performance. ■ observing the Fire Orders day in, day out.

WEBSITES ON FIRE* Wildland Fire Training Interagency Aviation Training If wildland fire training is on your to-do list, then The Interagency Aviation Training Education, this Website is the place for you. The site provides Qualification, and Currency System was developed access to workshops, meetings, and training courses under the direction of the Aviation Management by geographic area. Users can access a multiagency Council to establish aviation training standards for training schedule and then complete a training natural resource agency personnel. Personnel nomination form electronically. Links to seven meeting these standards are qualified to perform a different national training centers are available, variety of aviation-related tasks. The system is including the National Interagency Fire Center’s training based. The primary criterion for qualifica- National Fire and Aviation Training Group/Course tion is completion of the training modules with a Development, Office of Aircraft Services, Southwest passing score. The aviation training subject matter is Fire Use Training Academy, National Advanced designed to be progressive and to build on past Resource Technology Center, National Fire Academy, training experience. Training is offered online, in the National Interagency Rx Fire Training Center, and classroom, through aviation conferences and educa- Technical Fire Management. The site offers a field tion, and through a combined airplane/helicopter manager’s course guide containing the National safety course. Wildfire Coordinating Group’s course descriptions, prerequisites, and target audience, along with news Found at and other information of interest to the wildland fire community.

Found at

* Occasionally, Fire Management Today briefly describes Websites brought to our attention by the wildland fire community. Readers should not construe the description of these sites as in any way exhaustive or as an official endorsement by the USDA Forest Service. To have a Website described, contact the managing editor, Hutch Brown, at USDA Forest Service, Office of Communication, Mail Stop 1111, 1400 Independence Avenue, SW, Washington, DC 20250-1111, 202-205-1028 (tel.), 202-205-0885 (fax), hutchbrown/[email protected] (e-mail).

8 Fire Management Today CAUSAL FACTORS IN THE THIRTYMILE FIRE ACCIDENT

Thirtymile Fire Accident Review Board*

Editor’s note: On July 10, 2001, four firefighters died in a burnover on the Thirtymile Fire on the Okanogan National Forest near Winthrop, WA. The USDA Forest Service formed a team to investigate the accident and a board to review the team’s findings and determine causal factors for the accident. In its Management Evaluation Report, part of the Thirtymile Fire Investigation Report of September 26, 2001, as amended on October 16, 2001, the board established 14 significant causal factors and 5 influencing factors. The section of the report describing these factors is reprinted here, lightly edited.

causal factor is any behavior or omission that starts or sustains All 10 Standard Fire Orders Aan accident occurrence. For and 10 of the 18 Watch Out Situations were this investigation, the causal factors violated or disregarded during the incident. were classified as either significant or influencing. They were identified from the four categories of findings by the Thirtymile Fire Accident associated subject category and Fatigue Investigation Team (see the sidebar incident phase(s). Category: Management on page 10). Phases: Preparedness, initial Inadequate safety consideration attack, transition, entrapment, The causal factors on the Thirty- Category: Management deployment mile incident are interrelated, and Phases: Preparedness, initial Work/rest cycles for incident and it is difficult to point to one causal attack, transition, entrapment, fire program management person- factor or one finding as the most deployment nel, both at the forest and district important. Additionally, several The safety considerations were not levels, were disregarded, resulting causal factors were identified from appropriate to respond to the in mental fatigue. This significantly more than one phase of the inci- current, potential, and subsequent degraded the vigilance and decision- dent. The five phases of the incident fire conditions on this incident. All making ability of those involved. were preparedness, initial attack, 10 Standard Firefighting Orders transition, entrapment, and fire and 10 of the 18 Watch Out Situa- Command and control shelter deployment (see sidebar). tions were violated or disregarded Category: Management during the incident. Phases: Preparedness, initial Significant Causal attack, transition, entrapment, Factors Lack of situational awareness/ deployment inaccurate assessment Failure to maintain clear command The causal factors determined to be Category: Management and control resulted in poor risk significant are listed below, in relative order, together with the Phases: Preparedness, initial management and inhibited decisive attack, transition, entrapment, actions, which contributed to the deployment entrapment and deployment of fire * Board members were Tom L. Thompson, Regional Forester, Rocky Mountain Region, USDA Forest Service, At critical points throughout the shelters. Denver, CO (chairperson); Mark Boche, Director of Fire incident, the lack of situational and Aviation Management, Eastern Region, Forest Service, Milwaukee, WI; Paul Broyles, National Fire awareness by key incident, district, Strategy, tactics, and transition Operations Specialist, USDI National Park Service, Boise, ID; Ron Hooper, Business Operations Specialist, and forest personnel led to inaccu- Category: Management Washington Office, Forest Service, Washington, DC; Jim rate assessments of fuels, fire Phases: Initial attack, transition Golden, Supervisor, Coconino National Forest, Forest Service, Flagstaff, AZ; Paul Orozco, Fire Operations behavior, and fire potential. The suppression strategy failed to Specialist, Santa Fe National Forest, Santa Fe, NM; and Marc Rounsaville, Emergency Operations Specialist, Southern Region, Forest Service, Atlanta, GA.

Volume 62 • No. 3 • Summer 2002 9 adequately consider objectives, not a corresponding change in crown fire growing from a few acres fuels, fire behavior, and fire poten- strategy or tactics. to several thousand acres on the day tial, as well as the capability, avail- of the accident: ability, and condition of the sup- Fire behavior pression resources. This led to the Category: Environment •Valley bottom and slope fuels selection of tactics that could not Phases: Preparedness, entrapment, were dense, with abundant ladder succeed. As the fire complexity deployment fuels. changed significantly and initial A variety of environment factors • The moisture content of the fuels attack was unsuccessful, there was supported the development of a was at historically low levels.

DEFINITIONS

Subject Categories Transition. Based on the Fireline Handbook As directed in the USDA Forest Service’s Accident (1998), transition can be defined as the next level Investigation Guide (2001), the Thirtymile Fire of management that is expected and required Accident Investigation Team placed its findings in when it becomes apparent that the assigned four subject categories: resources will not meet containment objectives within the expected timeframes and/or the fire • Environment of the location of the incident; escalates to another level of complexity. • Equipment involved in or contributing to the incident; Entrapment. The Glossary of Wildland Fire • People involved in or contributing to the incident; Terminology (1996) defines entrapment as: “A and situation where personnel are unexpectedly caught •Management issues or principles associated with in a fire-behavior-related, life-threatening position the incident. where planned escape routes or safety zones are absent, inadequate, or compromised. An entrap- Incident Phases ment may or may not include deployment of a fire shelter for its intended purpose.” The Thirtymile Fire Review Board identified five phases of the incident. Fire shelter deployment. The Glossary of Wild- land Fire Terminology (1996) defines fire shelter Preparedness. The Wildland and Prescribed Fire deployment as: “The removing of a fire shelter Management Policy Guide (1998) defines prepared- from its case and the using of it as protection ness as: “Activities that lead to a safe, efficient, and against fire.” cost-effective fire management program in support of land and resource management objectives References through appropriate planning and coordination.” Accident Investigation Guide. 2001. Missoula, MT: USDA Examples include activities in preparation for the Forest Service, Missoula Technology and Development fire season, such as annual refresher training, work Center. capacity testing, review of plans and guides, and fire Fireline Handbook. 1998. National Wildfire Coordinating equipment and personnel readiness checks. Group, Incident Operations Standards Working Team. PMS 410-1; NFES 0065. Boise, ID: National Interagency Fire Center. Initial Attack. The Wildland and Prescribed Fire Glossary of Wildland Fire Terminology. 1996. National Management Policy Guide (1998) defines initial Wildfire Coordinating Group, Incident Operations Stan- dards Working Team. NFES 1832. Boise, ID: National attack as: “An aggressive suppression action consis- Interagency Fire Center. tent with firefighter and public safety and Wildland and Prescribed Fire Management Policy: Implemen- values to be protected.” Initial attack is made by the tation Procedures and Reference Guide. 1998. Boise, ID: first resources to arrive on a fire. It is designed to National Interagency Fire Center. protect lives and property and prevent further extension of the fire.

10 Fire Management Today • The combination of extremely low Site selection for the deployment of Phase: Deployment relative humidity, high tempera- the shelters above the road contrib- The improper use of personal ture, and atmospheric instability uted to the four fatalities. The rocky protective equipment contributed created weather conditions con- nature of the deployment site made to injuries. Three people occupied ducive to the rapid movement, it difficult to seal out the super- one shelter. This exceeded the growth, and intensity of the fire at heated air. The large size and the design capacity (although providing the times of entrapment and arrangement of the rocks made it shelter protection of the two deployment. difficult to fully deploy the shelters. civilians was appropriate and justified by the emergency). One Failure in road closure and area Personal protective equipment crewmember and the two civilians evacuation Category: Equipment/people did not have gloves; other Category: Management Phase: Initial attack The entrapment of two civilians was due to the failure to close the road and to subsequently evacuate the upper valley in a timely fashion.

Management intervention Category: People Phase: Transition There were missed opportunities for intervention by management personnel on this incident. Leader- ship’s failure to respond to concerns and observations by key individuals exacerbated circumstances that led to the entrapment.

Lack of escape routes and safety zones Category: People Phase: Entrapment Given the rapidly increasing fire intensity and changing fire situation, adequate consideration was not given to identifying escape routes and safety zones.

Failure to prepare for deployment Category: People Phase: Deployment Leadership of the entrapped firefighters failed to utilize available time and resources to coordinate and prepare crewmembers and civilians for fire shelter deployment.

Deployment site selection Category: Equipment/people Phase: Deployment Tree torching at night. On the Thirtymile Fire, nighttime torching contributed to initial fire spread. Photo: Paul S. Fieldhouse, USDA Forest Service, Missoula Smokejumper Base, Missoula, MT.

Volume 62 • No. 3 • Summer 2002 11 crewmembers did not wear their Influencing Factors Ineffective water operations gloves. Some of the line gear that The causal factors determined to be Category: Equipment/people was left close to the shelters ignited, influencing are listed below. Phase: Preparedness, initial attack and there was burning vegetation Water operations, both aerial and close to and under the shelters. Overextension of fire personnel ground-based, were ineffective or Category: Management delayed during the initial suppres- Sudden upcanyon extreme fire Phase: Preparedness sion actions. behavior Unit fire personnel were overex- Category: Environment tended. Although weather and fuel Helicopter delay Phase: Deployment conditions were near historic highs Category: Management The dense forest and the strong and there was significant fire Phase: Preparedness, initial attack fire-induced winds on the southern activity on the forest, additional fire Assignment of a helicopter to the canyon wall contributed to intense program management personnel incident was delayed. This may have spotting, causing the fire on the and additional initial- and extended- reduced the effectiveness of sup- canyon floor to intensify suddenly attack resources were not readily pression actions. The lack of a clear and surge over the deployment available. process and determination of area. responsibilities to deal with Endan- Development of crew cohesion gered Species Act issues contrib- Heat from fire Category: Management uted in part to this delay, as did Category: Environment Phase: Preparedness, deployment dispatch actions and confusion Phase: Deployment A number of issues limited the associated with availability. The fatalities were caused by development of crew cohesion for inhalation of superheated air and the Northwest Regular #6 crew, Organizational relationships exposure to high levels of radiant including collateral duties of Category: Management and convective heat. The presence command, fatigue, incident com- Phases: Initial attack, transition, of burnable fuels around and under plexity, lack of previous opportunity entrapment, deployment the chosen deployment sites also to work together, and management Unclear organizational relation- contributed to the fatalities and effectiveness. ships among forest, district, and injuries. The higher temperatures incident personnel reduced man- of the rock scree slope made condi- agement effectiveness on the tions worse for deployment than incident. ■ conditions on the road.

NATIONAL FIRE PLAN AT WORK*

Watershed Lassen Creek is a tributary of the Upper Pit River on Restoration the Modoc National Forest in northern California. The in California National Fire Plan is funding the Upper Pit River Watershed Restoration Project, one of more than a Streams are ecological ribbons dozen large watershed projects across the Nation. The that tie together the land Lassen Creek component of the project is designed to through watersheds. Water- restore historical ecosystem structure, function, sheds are key to healthy, diversity, and dynamics along Lassen Creek. Goals resilient ecosystems. Fire include the conservation of basic soils, the restoration damage to watersheds can impair ecological health of historical waterflows, and the prevention of invasion across entire landscapes. That’s why Congress by nonnative species. Specific projects include the authorized funding for watershed restoration planting of native grasses, forbs, shrubs, and trees to through the National Fire Plan. restore flourishing biological communities. ■

* Occasionally, Fire Management Today tells a success story or describes an exemplary project under the National Fire Plan. Readers can find many more such accounts on the Website for the National Fire Plan at .

12 Fire Management Today RECOMMENDED CHANGES IN WILDLAND FIRE MANAGEMENT

Thirtymile Fire Accident Review Board*

Editor’s note: On July 10, 2001, four firefighters died in a burnover on the Thirtymile Fire on the Okanogan National Forest near Winthrop, WA. The USDA Forest Service formed a team to investigate the accident and a board to review the team’s findings and make appropriate recommendations. In its Management Evaluation Report, part of the Thirtymile Fire Investigation Report of September 26, 2001, as amended on October 16, 2001, the board made 10 recommendations. They are reprinted here, lightly edited.

he following recommended changes to avoid incidents Develop and fully implement T similar to what happened on a fatigue countermeasures program. the Thirtymile Fire are proposed for agency consideration and action. As appropriate, it is recommended that •Develop and strengthen the they be responded to with the Fatigue annual preparation of firefighters involvement, understanding, and •Develop and fully implement a for personal responsibilities and support of the interagency wildland fatigue countermeasures pro- leadership. fire community (National Wildfire gram. Coordinating Group). These recommended changes should be consid- Incident Operations Safety Management ered in the context of this larger • Strengthen command-and- and Accountability interagency perspective. control performance by agency •Improve fire program safety administrators, fire program management by adopting and Situational Awareness, managers, and type 3–5 incident aggressively implementing Assessment, and Tran- commanders. proven components of a compre- sition • Strengthen operating procedures hensive safety program. • Ensure that fire program manag- to ensure accurate determination ers, fireline supervisors, and of complexity, with proper align- Equipment firefighters have situational ment of resources to match the • Continue improvements in awareness, assessment, and incident complexity and potential, personal protective equipment to decisionmaking abilities neces- and requisite command and provide for firefighter safety. sary to successfully and safely control. transition command from initial Endangered Species attack to extended attack on Leadership Act Protocols incidents. • Critically review the fire manage- • Clarify the relationship between • Ensure that fire program manag- ment leadership program on a the Endangered Species Act (ESA) ers and incident commanders national basis to ensure that all and fire suppression actions to have situational awareness, individuals in leadership posi- establish a coherent process that assessment, and decisionmaking tions, at all levels of organiza- accounts for ESA requirements abilities necessary to react to tions, have the skills and capabili- with respect to the full range of significant changes in fire danger ties to unquestionably lead in a fire suppression activities. ■ thresholds. responsible way.

* For a list of board members, see the footnote on page 9.

Volume 62 • No. 3 • Summer 2002 13 THIRTYMILE FIRE PREVENTION ACTION PLAN

Thirtymile Fire Accident Review Board*

Editor’s note: On July 10, 2001, four firefighters died in a burnover on the Thirtymile Fire on the Okanogan National Forest near Winthrop, WA. The USDA Forest Service formed a team to investigate the accident and a board to review the team’s findings and formulate an action plan for preventing similar future accidents. The action plan, released on October 22 and updated on December 3, 2001, is reprinted here, with condensed preface and lightly edited action items. For timelines and progress updates, see the Fire and Aviation Management Website at .

Preface he Thirtymile Fire Prevention Ensure that there is adequate direction Action Plan was developed from in fire management plans to trigger fire danger T the recommendations made by awareness with escalating fire potential. the Thirtymile Fire Accident Review Board (see page 13). The plan seeks to prevent future accidents of the officers of their responsibility to be A-3 The Director of F&AM will type that occurred on the Thirty- engaged and involved in the devel- ensure that there is adequate mile Fire. Although all firefighters involved in the Thirtymile Fire were opment of these policies and direction in fire management plans procedures, nor does it relieve to trigger fire danger awareness Forest Service employees, the agency representatives to the with escalating fire potential. More corrective actions in this plan will affect wildland fire management NWCG of their responsibility to specifically: keep senior management of the and policy in all jurisdictions and should be addressed by the National agencies informed and engaged. • A-3-a Notification from forest to districts (with regional assistance) Wildfire Coordinating Group of key increases in fire danger (NWCG). The Accident Investiga- Action Items tion Team and the Accident Review A-1 The Director of F&AM will thresholds that typically affect safety and control. At regional, Board were both made up of inter- initiate changes in the Federal forest, and district levels, identify agency representatives. Wildland Fire Management Policy to include recommendations for fire departures from historical weather patterns that will signifi- There are 31 action items in this suppression and for firefighter cantly influence increased fire plan; 26 items address issues that safety, in preparation for and in the affect all agencies involved in transition phase between initial- potential or fire behavior thresholds for safety and control in the wildland fire management. Many attack and extended-attack fires. operational area. action items are assigned to the Forest Service’s Director of Fire and A-2 The Director of F&AM will • A-3-b District fire program managers will review fire danger Aviation Management (F&AM), who work with the NWCG to review indices and convey their meaning is the agency’s representative to the current policy on preparation and NWCG. The NWCG, through its direction for fire management plan and significance to crews and incident management teams (e.g., interagency working teams, will development, and amend the use pocket cards to ensure address and reconcile the action direction to ensure that these plans items. specifically address staffing modifi- availability to firefighters). cations (firefighters, crew, and • A-3-c Identify the thresholds at which large fires typically occur. The Director of F&AM will bring supervisory span of control); these items to the NWCG. This does management coordination; and These thresholds indicate fire danger levels that significantly not relieve senior agency line allocation of resources as fire danger and occurrence escalate. compromise safety and control. When thresholds are approached, * For a list of board members, see the footnote on page XX. fire program managers will

14 Fire Management Today request additional supervisory factor in all ground and aviation management skills necessary for and suppression support. incidents, including entrapments the successful transition from and accidents. This should include a initial attack through a type 1 A-4 The Director of F&AM will 72-hour work/rest history with incident. initiate changes to the Federal quantitative analysis. Wildland Fire Management Policy A-10 The Director of F&AM will to ensure that there are defined A-8 The Director of F&AM will issue evaluate current training courses indicators for the need to transition direction that: for type 3–5 incident commanders from initial attack to extended to seek opportunities to strengthen attack. An example that might be • A-8-a Requires forest fire man- competencies. considered is establishing a fire- agement organizations, including specific perimeter limit trigger the agency administrators, fire A-11 The Director of F&AM will point. program managers, and incident develop a standardized briefing commanders, to meet annually to format for type 3–5 incidents. A-5 The Directors of Human review the responsibilities, Resources and F&AM should review expectations, and authorities of A-12 The Director of F&AM will policy, procedures, and perfor- the type 3–5 incident command- ensure that local fire mobilization mance expectations to reduce ers in fire suppression operations plans provide direction that re- firefighter fatigue, working with and incident operations protocols. quires dispatch centers to: NWCG to coordinate the effort. Review should include the follow- Consider: ing: • A-12-a Implement a standard 1.Provide for the safety and protocol ensuring that dispatch • A-5-a Requiring agency adminis- welfare of all personnel and centers inform all resources of trators to periodically review civilians. the name of the assigned incident time-and-attendance records for 2.Develop and implement viable commander and all other perti- compliance with work/rest strategies and tactics for the nent information. guidelines (including local agency incident. • A-12-b Announce all changes in administrators, fire program 3.Monitor the effectiveness of the incident command leadership to managers, fire support personnel, planned strategy and tactics. all assigned and incoming re- and firefighters). 4.Disengage suppression activities sources for initial- and extended- • A-5-b Working to build electronic immediately if strategies and attack incidents and relay that review options in the existing tactics cannot be implemented information to the duty officer automated payroll systems. safely. and forest fire management staff. • A-5-c Setting a daily duty hour 5.Maintain command and control limitation. of the incident.* A-13 The Director of F&AM will • A-5-d Reviewing S–200 and S– • A-8-b Establish procedures to initiate a proposed revision of the 300 courses to ensure that there ensure that arriving resources NWCG Fireline Handbook (PMS is adequate emphasis on the have positive contact with the 410-1) and all other related docu- incident commander’s responsi- incident commander, operations ments that would change incident bility in managing fatigue. section chief, or other appropriate classifications from the current incident management personnel. structure—initial attack, extended A-6 The Director of Human Re- This must occur prior to com- attack, type 2, and type 1—to sources will evaluate existing mencing work on the fire. another structure—type 5, type 4, training in fatigue awareness and type 3, type 2, and type 1. Initial other associated management A-9 The Director of F&AM will attack and extended attack will be training, and make training avail- review all existing fire training reserved as terms describing the able to all employees. courses to ensure that the course stages of an incident, not the content adequately teaches the command level required. A-7 The Directors of Human Resources and F&AM will develop *A policy letter on January 11, 2002, from Forest Service A-14 The Director of F&AM will Chief Dale Bosworth, designed to implement action plan protocols for the accident investiga- items, added a sixth bullet: “Use local rules and specific assess the need for a complexity tion process to evaluate fatigue as a criteria to determine when a fire has moved beyond analysis for type 3–5 incidents that initial attack.”

Volume 62 • No. 3 • Summer 2002 15 would assist fire program managers Evaluate existing training in fatigue awareness in determining the appropriate level and other associated management training | of management. Factors such as historical levels of fire danger, fuels, and make training available to all employees. fire history, fire potential, and historical fires in the vicinity A-20 The Director of F&AM will investigation. Additionally, the should be considered. initiate actions to widely distribute Director will work with the NWCG the findings, recommendations, and to ensure that the results of the A-15 The Director of F&AM will causal factors from this accident accident investigation are ensure that fire management plans require a single dedicated incident commander for all type 3 to type 5 incidents. Incident command responsibilities should not be diluted with collateral duties. The exception would be as a trainer or as an evaluator of an assigned trainee.

A-16 The Director of F&AM will initiate development of a standard operations guide for type 3, type 4, and type 5 incidents.

A-17 The Director of F&AM will initiate the adoption of an interagency “Standards for Fire Opera- tions” handbook modeled upon the Bureau of Land Management’s “red book.” The interagency handbook should be developed to ensure that each agency’s unique standards are maintained.

A-18 The Directors of F&AM and Human Resources will work with the Forest Service Line Officers Team to develop core fire management competencies for agency administrators with fire program responsibilities. They will also seek inclusion of these competencies in the position descriptions and in selection criteria for agency administrators.

A-19 The Director of F&AM will adopt and implement newly developed “Interagency Fire Program Management Qualifications” for the key fire program management Smokejumpers in training. The Thirtymile Fire Prevention Action Plan modifies manage- positions. ment of the type 3–5 incidents often handled by smokejumpers. Photo: Paul S. Fieldhouse, USDA Forest Service, Missoula Smokejumper Base, Missoula, MT.

16 Fire Management Today Disengage suppression activities immediately San Dimas Technology and if strategies and tactics cannot be Development Center to ensure its release for the 2002 training implemented safely. season. incorporated into the leadership A-21 The Director of F&AM will A-22 The Forest Service Chief, training curriculum. review the fire simulation program regional foresters, forest supervi- currently being developed by the sors, and district rangers will personally communicate their expectation of leadership in fire management. This should be completed prior to the 2002 fire season and in conjunction with leadership team meetings and annual fire schools.

A-23 The Directors of Human Resources and F&AM should work with the Forest Service Line Officers Team (Tom Thomp- son, chair) to assess the roles, responsibilities, and methodolo- gies needed to develop future fire management leaders. This would include fitness assessments to deal with performance expectations and issues surrounding fire leadership.

A-24 The Director of F&AM shall utilize the findings from the Thirtymile Fire Accident Investi- gation Report in preparedness training. This training will be available to all firefighting personnel. It will be mandatory for all new fire employees.

A-25 The Director of F&AM will work with the NWCG in the development of the leadership curriculum to ensure that there is adequate emphasis on prepar- ing fire personnel to effectively exercise personal responsibilities and leadership. The curriculum must specifically address how group dynamics influence situational awareness, communi- Torching in a coniferous forest. Torching on the Thirtymile Fire helped spread the fire into cations, group judgments, drier fuels and steeper terrain, starting the transition to a crown fire. The Thirtymile Fire decisionmaking, and responsibil- Prevention Action Plan calls for increased situational awareness on transition fires. Photo: ity for individual personal Paul S. Fieldhouse, USDA Forest Service, Missoula Smokejumper Base, Missoula, MT. actions.

Volume 62 • No. 3 • Summer 2002 17 Ensure that fire management plans require a single dedicated incident commander for all incidents, type 3–5.

A-26 The Directors of Human management personnel, and line preparation and training, and Resources and F&AM should officers to perform safely at all deployment site command and develop and implement a compre- times. control. hensive safety and health program utilizing all of the tools available. A-27 The Director of F&AM will A-30 The Director of Engineering This should include risk manage- adopt and implement the formal will work with the Missoula Tech- ment, system safety analysis, risk management process endorsed nology Development Center to compliance, inspection, oversight, by NWCG. review and appropriately incorpo- human factors, and behavior rate into the program of work the modification. In addition, use a A-28 The Directors of F&AM and recommendations found in the behavior-based safety program for Engineering will review the current Equipment Appendix of the fire management that: fire shelter development program to Thirtymile Fire Investigation determine the possibility of acceler- Report. 1.Focuses on active agency admin- ating the development of an im- istrator involvement; proved fire shelter. The current A-31 The Directors of Wildlife, Fish 2.Encourages monitoring and timeline provides for completion in and Rare Plants and F&AM will intervention; 2003. review and clarify issues of 3.Promotes individual safe behavior firefighter safety, fire potential, and on the fireline; A-29 The Directors of F&AM and use of tactical resources during 4.Rewards safe behavior; and Engineering will revise fire shelter developing fire emergencies with 5.Reinforces the agency’s commit- training to emphasize entrapment regard to threatened and endan- ment to safety through the use of avoidance. Incorporate information gered species. Such review will incentives, recognition, and from this accident in revising the include coordination and disciplinary procedures. training. This training will include clarification of existing protocols the proper use of personal protec- with the U.S. Department of Com- Additionally, develop procedures and tive equipment (including gloves merce, National Marine Fisheries protocols to ensure accountability at and shrouds), the importance of Service, and U.S. Department of the all levels of the organization. These crew cohesion in deployment, Interior, U.S. Fish and Wildlife procedures and protocols should alternative deployment site Service. ■ enable fireline supervisors, fire selection, shelter deployment

18 Fire Management Today INITIAL HAZARD ABATEMENT PLAN

Editor’s note: On July 10, 2001, four firefighters died in a burnover on the Thirtymile Fire on the Okanogan National Forest near Winthrop, WA. The Occupational Safety and Health Administration investigated the accident and cited the USDA Forest Service for five violations of the Code of Federal Regulations, part 1960— Elements for Federal Employee Occupational Safety and Health Programs. On February 7, 2002, as required by law, the Forest Service released an Initial Hazard Abatement Plan. The plan is reprinted here, lightly edited; the plan’s appendix A is reprinted on page 22.

he Occupational Safety and Health Administration (OSHA) The Chief’s letter of January 11, 2002, T notified the Forest Service of requires that every fireline-qualified individual the findings of its investigation of will receive training on entrapment recognition the Thirtymile Fire on the Okanogan–Wenatchee National and deployment protocols. Forest in July 2001. OSHA cited the Forest Service for two willful and three serious violations. OSHA Hazards Hazard 1-B. An incident com- Identified mander for all stages of the In response to the violations, the Hazard 1-A. Work/rest cycles Thirtymile Fire was not clearly Forest Service is immediately developed by the Forest Service assigned. Incident command was implementing all of the actions were not followed. This resulted in not formally passed between various described below. In addition, the a lack of situational awareness and leaders. Forest Service is amending the impaired judgment in responding Accident Prevention Plan (dated to critical fire situations. Abatement Actions: December 14, 2001, and updated January 14, 2002) by letter to all Abatement Action: Washington •Document all actions on appro- Forest Service field units; and Office Human Resources staff will priate incident action records and consulting with member agencies coordinate with the Partnership include the formal assignment of of the National Wildfire Coordinat- Council to clarify or modify work/ incident command responsibili- ing Group to address the hazards rest guidelines in order to ensure ties in the D–311 Initial Attack identified by OSHA across the adequate levels of rest. Dispatcher curriculum, to be wildland firefighting community. released in 2003.

OSHA Violation 1— OSHA VIOLATION CATEGORIES* Serious 29 CFR 1960.8(a): The agency did The Occupational Safety and Health Administration (OSHA) gives the not furnish employees employment following definitions for the two violation categories cited in its report and a place of employment which on the Thirtymile Fire: was free from recognized hazards that were causing or likely to cause • Willful violation: A violation that the employer intentionally and death or serious physical harm in knowingly commits. The employer either knows that what he or she that employees were exposed to the is doing constitutes a violation, or is aware that a hazardous condi- hazards of burns, smoke inhalation exists and has made no reasonable effort to eliminate it. tion, and death from fire-related • Serious violation: A violation where there is substantial probability causes. that death or serious physical harm could result and that the employer knew, or should have known, of the hazard.

* Based on the Small Business Handbook posted by the U.S. Department of Labor on the World Wide Web at .

Volume 62 • No. 3 • Summer 2002 19 • In addition, the completion of Management and the Director of official in charge of an establish- items in the Forest Service Engineering to revise fire shelter ment, and supervisory employee, or Accident Prevention Plan will training to emphasize entrapment other appropriate management abate this hazard. Specifically, avoidance and incorporate official, measures that employee’s these action plan items require: information from this accident performance in meeting require- – Item A-12-a: Implement a into revision of the training. This ments of the agency occupational standard protocol whereby training will include the proper safety and health program. dispatch centers inform all use, care, and storage of personal resources of the name of the protective equipment (including OSHA Hazard Identified assigned incident commander gloves and shrouds); the impor- Hazard 2-A. Evaluations of super- and all other pertinent informa- tance of crew cohesion in deploy- visory and management officials at tion. ment; alternative deployment site the Okanogan and Wenatchee – Item A-12-b: Announce all selection; shelter deployment National Forests, above the level of changes in incident command preparation and training; and crew boss, did not have perfor- leadership to all assigned and deployment site command and mance elements relating to their incoming resources for initial control. support of, or meeting the require- and extended attack incidents • The Chief’s letter of January 11, ments in, the occupational safety and relay that information to 2002, requires that every fireline- and health program. the duty officer and wildland qualified individual will receive fire management staff. training on entrapment recogni- Abatement Action: Washington • The Forest Service Chief’s letter tion and deployment protocols. Office Human Resources staff will of January 11, 2002, requires that This training should be con- coordinate with the Partnership the National Mobilization Guide ducted in conjunction with Council to ensure that performance will include direction to dispatch refresher training and/or annual elements for line officers, fire centers that will ensure that all fire schools. The principles managers, and incident command- resources know the name of the outlined in the entrapment ers evaluate performance in the assigned incident commander and avoidance attachment will be area of safety and health. These announce all changes in incident incorporated into the next itera- safety and health elements, at a command. Geographic Area tion of wildland fire shelter minimum, will include perfor- Mobilization Guides, Zone Mobili- training. mance measures based on the Ten zation Guides, and Local Mobili- • The Chief’s letter of January 11, Standard Firefighting Orders and zation Guides should include this 2002, also requires that the Fire the Eighteen Situations That Shout direction as they are revised for and Aviation Ground Safety Watch Out. the 2002 fire season. Manager will issue guidelines, prior to the 2002 western fire OSHA Violation 3— Hazard 1-C. Fire shelter deploy- season, for crew actions in the Serious ment procedures had not been event of entrapment and in 29 CFR 1960.55(a): The agency did developed for firefighters whose preparation for deployment. not provide occupational safety and escape routes were compromised. These guidelines will include health training for supervisory All firefighters must begin prepar- specific actions necessary for employees. ing for deployment of fire shelters entrapment avoidance, safety when they are surrounded by fire, zone characteristics and selec- OSHA Hazard Identified even if they believe they are in a tion, crew deployment training, Hazard 3-A. A member of North- safety zone. and emergency deployment west Regular Fire Crew #6 who had supervision when surrounded by not completed his task book and Abatement Actions: fire. required courses was assigned as a squad boss. • Completion of Accident Preven- OSHA Violation 2— tion Plan item A-29 will abate this Serious Abatement Action: All Fire Qualifi- hazard. Specifically, the item 29 CFR 1960.11: Each agency did cation Review Committees will requires the Forest Service not ensure that any performance annually review qualifications, as Director of Fire and Aviation evaluation of any management required by Forest Service Hand-

20 Fire Management Today book 5109.17. Results of this to raise awareness and penalize 5 fires on their unit(s), both annual review will be shared with infractions in order to emphasize onsite and through after-action affected employees. Forest fire the importance of the orders and activities. These reviews will management officers will maintain situations. include evaluation and documen- a complete and current record of all • The Forest Service Accident tation of fire management opera- qualified personnel dispatched by Prevention Plan extensively tions, with the Ten Standard the forest. addresses the serious failure to Firefighting Orders and the follow the Fire Orders on the Eighteen Situations That Shout OSHA Violation 4— Thirtymile Fire. Twelve of the Watch Out as the standard for Willful action items (Accident Report measuring safety and health 29 CFR 1960.8(a): The agency did Items A-1, A-2, A-3, A-5, A-6, A-7, performance. not furnish employees employment A-8, A-11, A-12, A-15, A-17, and A- • In addition, fire managers and and a place of employment which 25) require, taken together, 28 line officers will review a suffi- was free from recognized hazards separate actions, including cient number of type 3, 4, and 5 that were causing or likely to cause changes in transition operations, fires to document trends and death or serious physical harm in changes in briefing format, identify safety and health hazards that employees were exposed to the changes in communication and/or violations of the Fire hazards of burns, smoke inhala- direction, and continuing empha- Orders and Watch Out Situations. tion, and death from fire-related sis on the leadership training • The Fire and Aviation Ground causes. curriculum. Safety Manager will complete and implement the Safe Practices OSHA Hazard Identified OSHA Violation 5— Spreadsheet program to improve Hazard 4-A. All Ten Standard Willful and standardize documentation of Firefighting Orders in the National 29 CFR 1960.8(a): The agency did violations and good safety and Wildfire Coordinating Group’s not furnish employees employment health performance in order to Fireline Handbook 410-1 were and a place of employment which identify hazards and trends. violated. Supervisors at the was free from recognized hazards Wenatchee National Forest and at that were causing or likely to cause Hazard 5-B: After-action reports the Thirtymile Fire did not ensure death or serious physical harm in prepared for out-of-forest fire- that the Fire Orders were followed. that employees were exposed to the fighting crews on type 3, 4, and hazards of burns, smoke inhala- 5 fires did not identify safety and Abatement Actions: tion, and death from fire-related health hazards. causes. • Regional foresters, forest supervi- Abatement Action: After-action sors, and district rangers will OSHA Hazards reports, individual overhead perfor- emphasize their commitment to mance evaluations, and crew Identified evaluations will be completed on the Ten Standard Firefighting Hazard 5-A. Management failed to Orders through annual meetings every fire assignment. These conduct inspections of firefighting evaluations will use the Ten Stan- with incident commanders and operations, including onsite, fire management personnel. dard Firefighting Orders and the frontline evaluations of type 3, 4, Eighteen Situations That Shout Documentation of these meetings and 5 fires to ensure that estab- will be required. Additionally, any Watch Out as the standard for lished firefighting practices were measuring performance. Corrective corrective actions related to the enforced. Fire Orders will be documented. action, even if satisfactorily completed, will be documented in the • The Forest Service is also devel- Abatement Actions: oping standard disciplinary and reports and evaluations. Local units will keep these records and share accountability standards for • Fire managers and line officers violations of the Ten Standard them with the evaluated resource will periodically review a repre- and the resource’s home unit. ■ Firefighting Orders and Eighteen sentative number of type 3, 4, and Situations That Shout Watch Out

Volume 62 • No. 3 • Summer 2002 21 CROSSWALK BETWEEN OSHA VIOLATIONS AND ACCIDENT PREVENTION PLAN* The following table shows how safety violations identified by the Occupational Safety and Health Administration (OSHA) during its investigation of the Thirtymile Fire accident correspond to action items called for under the USDA Forest Service’s Thirtymile Fire Accident Prevention Plan.

OSHA Investigation Accident Prevention Plan Violation Hazard Description Item Description 29 CFR 1960.8(a) 1-A Work/rest cycle A-1 Fatigue policy A-5-a–d Manage fatigue A-6 Fatigue awareness A-7 Fatigue evaluation 1-B Command unclear A-8-a Type 3–5 incident command responsibilities A-8-b Documented contact A-9 Transition training A-10 Strengthen competencies A-11 Briefing format A-12-a Command broadcasted A-12-b Broadcast command changes A-25 Leadership curriculum 1-C Shelter deployment procedures A-29 Entrapment avoidance training A-30 Equipment recommendations A-17 Fire operations handbook A-20 Share findings 29 CFR 1960.11 2-A Performance elements for safety A-18 Administrator competencies and health A-23 Fire management leadership A-26 Behavior-based safety and health program 29 CFR 1960.55(a) 3-A Uncompleted task book -- Standard operating procedure 29 CFR 1960.8(a) 4-A Ten Standard Firefighting Orders A-17 Fire operations handbook 1. Safety first -- Standard operating procedure 2. Overextended A-3 Fire danger awareness 3. Spot forecast -- Standard operating procedure 4. Unclear/conflicting instructions A-8-a Type 3–5 incident command responsibilities A-8-b Documented contact A-12-a Command broadcasted A-12-b Broadcast command changes A-15 Dedicated incident commander A-25 Leadership curriculum 5. No current fire information A-1 Transition from initial to extended attack A-2 Fire danger/staffing A-3 Fire danger awareness A-8 -a Type 3–5 incident command responsibilities A-11 Briefing format 6. Communications A-8 Type 3–5 incident command responsibilities A-12 Command broadcasted A-15 Dedicated incident commander A-17 Fire operations handbook 7. Safety zones and escape routes -- Standard operating procedure 8. Lookouts -- Standard operating procedure 9. Leadership fragmented A-8-a Type 3–5 incident command responsibilities A-8-b Documented contact A-12-a Command broadcasted A-12-b Broadcast command changes A-15 Dedicated incident commander A-25 Leadership curriculum 10. Fatigue A-5-a,b,c Manage fatigue A-6 Fatigue awareness A-7 Fatigue evaluation A-8-a Type 3–5 incident command responsibilities A-11 Briefing format A-25 Leadership curriculum 29 CFR 1960.8(a) 5-A No inspections/oversight A-1 Transition from initial to extended attack A-2 Fire danger/staffing A-3 Fire danger awareness A-8-a Type 3–5 incident command responsibilities A-11 Briefing format 5-B No safety/health hazards -- identified in after-action reports

* Based on appendix A in the USDA Forest Service’s Initial Hazard Abatement Plan, released on February 7, 2002, in response to OSHA’s Thirtymile Fire investigation report.

22 Fire Management Today THIRTYMILE FIRE: FIRE BEHAVIOR AND MANAGEMENT RESPONSE*

Hutch Brown

t’s licking; it’s rolling; it’s alive; it’s screaming at us!” That’s Topography, fuels, and weather conditions “Iwhat the Thirtymile Fire looked in Chewuch Canyon all pointed to the possibility and sounded like to one firefighter of extreme fire behavior. just before it hit with a wall of heat and flame (Rutman 2001). On a blazing July afternoon, the fire had entrapped 16 people in a steep three-person initial-attack crew made by those responsible for safely canyon formed by the Chewuch arrived after a plane first detected managing the fire. What does the River on the Okanogan National the fire in Chewuch Canyon. The Thirtymile Fire teach? What lessons Forest in north-central Washing- incident commander (IC) reported can be learned from the fire’s ton. Four firefighters perished. that 3 to 8 acres (1.2–3.2 ha) were behavior and the management burning on the canyon floor just response? Sizeup Disparities north of the Chewuch River, with two spot fires visible across the Earlier, things had looked quite Environmental different. The 21 members of the stream. Conditions Northwest Regular (NWR) #6 Type On the morning of July 10, condi- One spot fire, the IC believed, was 2 Fire Crew arrived on the fire just tions in Chewuch Canyon were after 9 a.m. on the morning of July already burning dangerously close dangerous. Topography, fuels, and to the canyon’s heavily forested 10, 2001. They were eager to sup- weather conditions all pointed to press the spot fires in Chewuch south slope. Unless controlled, he the possibility of extreme fire warned, “it will grow, hit the slope, Canyon and perhaps move on to the behavior. Escape from the canyon and get larger.” Libby South Fire, a much larger would be difficult, and there were incident with a smoke column no safety zones inside the canyon. Asked by the district duty officer visible 50 miles (80 km) to the and Okanogan National Forest south. In briefing the crew boss Topography. The Thirtymile Fire trainee, the fire management officer Dispatch whether the fire could ignited about 20 miles (32 km) wait until morning, the IC replied for the Methow Valley Ranger north of Winthrop, WA, on the that it needed “to be taken care of District, Okanogan National Forest, banks of the Chewuch River, a said the Thirtymile Fire was “basi- tonight, because if it hits that slope, beautiful stream in a V-shaped it is going to the ridgetop.” That cally a mopup show,” a view that canyon carved by glaciers (fig. 1). A assessment ultimately proved would be repeated by the media. single unimproved road follows the correct. On the next afternoon, the river upcanyon along its north fire not only went to the ridgetop, The earliest sizeup report had been bank. The road dead-ends at but also swept upcanyon. Within more ominous. Late on July 9, a Thirtymile Campground, a trailhead hours, four firefighters lay dead. for wilderness travel. Hutch Brown is the managing editor of Fire Following the accident, the USDA Management Today, USDA Forest Service, The canyon runs from northeast to Washington Office, Washington, DC. Forest Service immediately as- southwest, the same orientation as sembled an investigation team. The the prevailing southwest winds. The * This article closely follows the Thirtymile Fire Thirtymile Fire Investigation canyon floor narrows from about Investigation Report: Accident Investigation Factual Report, released on September 26 Report and Management Evaluation Report (Winthrop, 1,500 feet (460 m) at the point of WA: USDA Forest Service; September 26, 2001, as and amended on October 16, 2001, amended October 16, 2001), particularly the Fire ignition to about 900 feet (270 m) Behavior Appendix, pp. 54–65. Unless otherwise suggests that serious mistakes were referenced, all information in this article is from the at the point where firefighters investigation report and followup investigation.

Volume 62 • No. 3 • Summer 2002 23 deployed their fire shelters (Roeder sidebar), although some areas areas. Fuels and vegetation became 2001). The elevation at the point of resembled fuel models 8 (timber, more continuous near the ridge- ignition is about 3,440 feet (1,040 low fuel load) and 12 (slash, moder- line. Vegetation on both sides of the m). The canyon slopes are 70 ate fuel load). canyon was broken by occasional degrees or more and about 3,000 rockslides and scree slopes. feet (910 m) high. Forest cover in Chewuch Canyon is normally heavier on the cooler, Weather. The winter of 2000–01 Fuels. The canyon floor supports a shadier south (north-facing) slope was the area’s second driest in 30 forest of spruce, Douglas-fir, lodge- than on the drier north (south- years. Precipitation from April pole pine, and (on streambanks) facing) slope. Before the fire, the through June remained well below cottonwood. At the point of igni- south slope was densely covered by normal. The Palmer Drought Index tion, the floor was strewn with Douglas-fir and lodgepole pine. of July 14, 2001, showed severe to decomposing woody debris, much Branches hung down to within a extreme drought for the region. By of it highly flammable. Ladder fuels few feet of surface fuels; fuel models early July, weather and fuel mois- were plentiful, including shrubs, 8 and 10 applied. On the more open ture conditions in Chewuch Canyon understory conifers, and drooping north slope, ponderosa pine and resembled normal conditions for boughs from overstory spruce. Douglas-fir prevailed, with grasses early August, the peak of fire Overall, fuel model 10 applied (see and brush dominating in many season.

FUEL MODEL 10

Fuel models, based on factors such as type and amount of vegetation on a site, help fire managers assess fire hazards. On much of the Thirtymile Fire, fuel model 10 prevailed, described by Anderson (1982) as follows:

The fires burn in the surface and ground fuels with greater fire intensity than the other timber litter models. Dead/down fuels include greater quantities of 3-inch (7.6-cm) or larger limb wood resulting from over- maturity or natural events that create a large load of dead material on the forest floor. Crowning out, spotting, and torching of individual trees are more frequent in this fuel situation, leading to potential fire control difficulties. Any forest type may be considered if heavy down material is present; examples are insect- or disease- Figure 1—Chewuch Canyon after the Thirtymile Fire on July 10, 2001. Scree slopes are ridden stands, windthrown visible on both sides; heavier vegetation is on the south slope (right). The fire burned stands, overmature situations upcanyon from lower right to upper left, crossing the “S” in the road at center and with deadfall, and aged light entrapping the firefighters. Photo: Ben Croft, USDA Forest Service, Missoula Technology thinning or partial-cut slash. Development Center, Missoula, MT, 2001.

24 Fire Management Today On July 10, relative humidity Escape from the canyon would be difficult, and dropped to 8 percent and tempera- there were no safety zones inside the canyon. tures reached a high of 94 ºF (34 ºC). Fuel moistures approached historical lows: measures atmospheric dryness and of the blaze—a streamside cooking • 10-hour fuels = 3 percent stability on a scale of 2 to 6; un- fire abandoned by picnickers. Forest • 100-hour fuels = 5 percent stable atmospheric conditions officials believed that if the fire was • 1,000-hour fuels = 10 percent (Haines Index 5 or 6) are often allowed to burn under the severe associated with large fire growth.* environmental conditions in Large-diameter fuels were so dry Chewuch Canyon, it could spread that many burned almost com- Overall, conditions in Chewuch across thousands of acres, threaten- pletely in the fire. Live foliar Canyon were unseasonably ripe for ing lives and crossing onto non- moisture probably dipped below 100 extreme fire behavior. The Federal lands. Another major fire in percent, making conifers more firefighters did not obtain the fire the region might overtax fire susceptible to torching. weather forecast issued at 8:30 a.m. suppression resources already on July 10. Their only weather stretched thin. In accordance with The closest fire weather stations information came from a spot the Okanogan National Forest’s fire classified the fire danger on July 10 weather report for the Libby South management action plan, forest as “very high.” The station at First Fire issued a day earlier. officials decided to suppress the fire Butte Lookout, about 12 miles (19 as quickly as possible. km) to the south, registered the Initial Attack highest energy release component On July 9, just as on the previous 5 Recognizing the opportunity to (ERC) for that day since record- days, the relative humidity in apply water in the riparian zone, keeping began in 1970. ERC is a Chewuch Canyon did not reach its the first IC requested two engines, a measure of the potential available usual nighttime peak. In the Mark 3 pump, hoses, and a hand energy per unit area within the relatively dry nighttime conditions, crew of at least 10 people. A little flaming front at the head of a the Thirtymile Fire spread at the after 1 a.m. on July 10, the Entiat moving fire. Since April 14, the rate of about 86 feet per hour (26 Interagency Hotshot Crew relieved ERC at First Butte had approached m/h) through the large-diameter the first crew on the fire. The new record levels almost every day. fuels on the north bank of the IC decided that lining spot fires Chewuch River. Flame lengths would most effectively control the The prevailing winds on July 10 reached about 2.3 feet (0.7 m). fire. remained light—2 to 4 miles per hour (0.9–1.8 m/s). However, as Profuse deadfall and ladder fuels The hotshots went to work on the temperatures rose during the day carried the surface fire into some north bank. At 2:15 a.m., they and the flames grew in intensity, treetops. The torching trees spewed crossed the river to tackle the fires the fire sucked air into its base, embers, igniting at least six spot on the south bank. The IC now saw creating its own winds. During the fires and sending the fire across the that the job was bigger than at first afternoon, a helicopter pilot ob- river. The fire moved generally thought. He ordered a reconnais- served drafts of 20 to 30 miles per upcanyon from the point of igni- sance aircraft, another crew, and hour (9–13 m/s), and the entrapped tion, with spot fires scattered two Mark 3 pumps with kits and firefighters had trouble deploying toward the base of the south slope. 1,500 feet (457 m) of hose. their fire shelters in the fire- induced winds. The first crew arrived on the scene Working through the night, the at about 11 p.m. on July 9. The hotshots dug containment line and Atmospheric conditions were firefighters soon found the source surrounded spot fires. By 6 a.m., conducive to the formation of fire the fire had lost intensity; by columns. At Spokane, WA, the mid- midmorning, only eight scattered * For a good description of Haines Index application in level Haines Index on July 10 was 6, the West, see John Werth and Paul Werth, “Haines Index hotspots remained. At about 9 a.m., Climatology for the Western United States,” Fire and the high-level Haines Index Management Notes 58(3): 8–18. See also the article by the forest and district fire manage- reached 4. The Haines Index Brian Potter and others beginning on page 41 in this ment officers (FMOs) arrived. The issue.

Volume 62 • No. 3 • Summer 2002 25 Large-diameter fuels were so dry that many ing runs at about 1.25 miles per burned almost completely. hour (2 km/h). Within 15 minutes, the fire doubled in size to about 100 acres (40 ha). A huge convection column formed, with pervasive forest FMO estimated that about Abandoning the pumps, the crew spotting as far away as 0.4 miles 3 acres (1.2 ha) of active fire were started constructing fireline ahead (0.6 km). “We were losing it,” a left, spread over a 5-acre (2-ha) of the fire, trying to pinch the fire’s firefighter remarked (Solomon and area. He judged fire behavior to be head with a line from the river to a Welch 2001a), “but it was one of the “very benign.” scree slope (Solomon and Welch most fantastic things I’ve ever 2001a). However, the many roots in seen.” The hotshots badly needed sleep, so the soil hampered progress. the NRW #6 crew took over at about The crowning runs signaled a new 9:30 a.m. According to one arriving At 12:08 p.m., the crew boss trainee level of fire activity. At the point of firefighter, only a few flames were requested a helicopter with bucket origin, spruce dominates the site, visible (Hamilton and Earls 2001). and longline. Twenty minutes later, with alder in the understory and The firefighters were confident that he ordered additional crews. At aspen nearby. The site is wetter they could easily finish the job. about 2 p.m., the Entiat Hotshots than surrounding areas of pine and Although several crew members returned to reinforce the NRW #6 Douglas-fir, so the fire was initially had little or no experience, all had crew; by 2:38 p.m., a type 3 helicop- limited to torching and spotting. In the requisite training. Their leaders ter was making water drops. Still, fact, most fuels in the area of had many decades of fireline the fire kept crossing containment ignition remained unburned (fig. experience among them, although lines. 2). But as the day warmed and the the FMOs had no recent experience fire reached drier sites, its intensity with initial and extended attack. The crew boss trainee decided that grew and the rate of spread in- the fire could not be controlled and creased. Indrafts and spotting from Transition moved the firefighters back across the blaze on the south slope further The NRW #6 crew boss passed IC the Chewuch River to the north intensified fire activity on the responsibility to the crew boss bank. The crew boss now assumed canyon floor. From about 3 p.m. to trainee, standing by in case the the IC role. Sometime after 2 p.m., 4 p.m., as the firefighters lunched trainee needed help. At about the crew boss removed the sole and rested, conditions further 11 a.m., the trainee led the crew lookout from a patch of scree on deteriorated. By 4:20 p.m., the fire across the river to the south bank. the south slope for fear that her on the canyon floor had begun a Using two pumps, the crew applied escape route might be cut off. An sustained run through tree crowns. water to the remaining spot fires air attack plane diverted from the while digging fireline. However, the Libby South Fire now acted as Entrapment firefighters had trouble keeping the lookout, but smoke limited its By 3:30 p.m., two engines had pumps running. Several hoses burst effectiveness. arrived on the fire. Both were and four pulaskis broke. ordered by Air Attack at 2:27 p.m. By 2:50 p.m., the fire had gained a Without checking in with the IC, By noon, conditions were deterio- foothold on the south slope. The they drove past the NRW #6 crew to rating. Since midmorning, the fire, now entirely unconstrained, attack spot fires upcanyon. relative humidity had been rapidly spread at a rate of about 710 feet falling and the temperature rising. per hour (220 m/h). By 3:20 p.m., it At about 4 p.m., the IC made the Now, surface fires were becoming covered about 50 acres (20 ha) on fateful decision to send a squad more active, with flame lengths the valley floor. from the NRW #6 crew to support approaching 8 feet (2.4 m). More the engines ahead of the main fire. trees were torching, and spot fires At 3:20 p.m., the fire at the base of By 4:30 p.m., two squads were were crossing containment lines. the south slope reached critical fighting spot fires along the road mass. Driven by convective winds upcanyon from the lunch spot, At about noon, the crew boss through dense fuels on 100-degree although the engines had with- trainee decided to change tactics. slopes, the fire began uphill crown- drawn.

26 Fire Management Today “No one associated with this fire civilians who had been camping at gave it the respect it was due.” the end of the road (see the sidebar on page 28). –Thirtymile Fire Investigation Report By 5 p.m., the fire was scouring the canyon on both sides of the The fire was spreading upcanyon river and threatened to cross the Chewuch River, from the base of the generally parallel to the road along road behind the NRW #6 crew. north slope to the top of the south the Chewuch River. However, just ridge. Two convection columns upcanyon from the lunch spot, the Seeing the danger, Air Attack and formed, one from the south slope road makes a 90-degree turn the crew leaders ordered everyone fire and another from the fire on towards the river (fig. 3). At this out. But for the IC and 13 other the canyon floor. The hillside fire point, the fire spread was perpen- firefighters, it was too late. At 4:34 drew strong upcanyon winds that dicular to the road. The area was p.m., the fire cut off their only probably drove the fire on the densely forested, with a heavy escape. Led by the IC, the fire- canyon floor, flattening its convec- understory of shrubs and young fighters retreated about 1 mile tion column. “It was rolling like a conifers. At about 4:30 p.m., fed by (1.6 km) upriver to a site selected little tornado on its side,” observed the ladder fuels, the fire made a for survivability. Shortly after the Air Attack pilot (Solomon and torching run that swept across the 5 p.m., the crew was joined by two Welch 2001a). With its flattened convection column aimed straight at the deployment zone, the fire swept upcanyon at a rate of more than 1.6 miles per hour (2.5 km/h). Deployment The IC had chosen a good deployment zone (fig. 4). A bend in the river widened the stream, forming a sandbar against the north bank. Across the river was an extensive, lightly vegetated rock slump behind a narrow band of conifers along the stream. Adjacent to the road was a scree slope and lightly vegetated hillside. Just to the north and east of the deployment site, within about 20 feet (6 m), was a timbered area (fuel model 10) that extended upcanyon for several miles.

The rock slump across the river hid the approaching fire on the canyon floor from view. However, the firefighters could see the plume from the fire on the south slope. The plume might have led the IC to believe that the fire’s convection column would be vertical, carrying Figure 2—Area of early fire activity. Initially, the Thirtymile Fire was a series of slow- heat and flame upslope across the moving, occasionally torching spot fires, leaving most fuels near the point of origin river, well away from the deploy- unburned (lower left). As it spread upcanyon, the fire reached drier fuels and transitioned ment zone. “I think he thought we to a crown fire, severely burning most stands (upper right). Photo: Ben Croft, USDA Forest Service, Missoula Technology Development Center, Missoula, MT, 2001. were completely safe where we

Volume 62 • No. 3 • Summer 2002 27 were,” one firefighter later told Just before 5:24 p.m., the situation feet (30 m) above the road. In a reporters (Roeder 2001). Although abruptly changed. The sky darkened tight group, they deployed in rocks the firefighters waited in the and embers began to pelt the that were 1 to 3 feet (0.3–0.9 m) in deployment zone for about half an firefighters, who draped themselves size, with duff and rotting wood hour before the fire arrived, they with their fire shelters for protec- lodged in the crevices. Four died did nothing to prepare the site, tion. The fire came into view, from asphyxia due to inhaling such as removing flammable howling and roaring as it sped superheated gases, probably be- materials. Some spent the time through the trees. Then it hit with a cause they could not get a good seal taking photos or making journal plume of hot gases. Caught by for their fire shelters. Two survived entries. surprise, the firefighters hurriedly the fire’s initial blast, then fled deployed their fire shelters. Some downhill—one into the river, the Nobody considered moving to the mistakenly deployed with their other into a van on the road. Both sandbar, possibly the safest deploy- heads toward the fire; others lived, one with severe burns. ment site. Most formed a loose dropped their packs next to their group on the road near the IC, but shelters or even took them inside— The deployment zone was well some joined a squad leader on the an extreme fire hazard, because chosen; it was probably the only scree slope, partly for a better view packs and other gear can ignite, survivable site in the canyon. The of the approaching fire (Roeder compromising the effectiveness of scarcity of fuels on the hillside 2001). One firefighter on the scree fire shelters. above and on the rock slump across slope wore no gloves. A second the river limited the amount and squad leader climbed through the Despite the mistakes, all eight duration of heat on the road. Also, scree, assessing its viability as a firefighters who deployed on the the narrow band of conifers on the deployment site. Nobody offered the road survived. The two civilians south side of the river might have civilians extra fire shelters or other took shelter with one of them; all partially shielded firefighters on the personal protective equipment. three survived with minor injuries. road from the initial blast of heat However, six firefighters were from the fire. However, the deploy- caught on the scree slope up to 100 ment zone was no safety zone. By

CIVILIANS ENTRAPPED AT THIRTYMILE

In a twist of fate, a civilian couple spend the night. On the way, they The civilians put on heavy was saved by a firefighter en- saw burned areas but no fire- clothes to protect themselves trapped on the Thirtymile Fire. fighters. from falling embers, but they did How did it happen? not have personal protective Between 3 p.m. and 4 p.m., the equipment. When the fire arrived The only road into Chewuch superintendent of the Entiat and the firefighters deployed Canyon ends at Thirtymile Interagency Hotshot Crew, together their fire shelters, the couple Campground, a trailhead for with his next-in-command, drove to sought shelter with one of the wilderness travel. On July 10, the campground to make sure firefighters. It saved their lives. between 10 a.m. and 11 a.m., the nobody was there. He noticed three district fire management officer cars parked at the trailhead but did Investigators concluded that the ordered the road closed due to not see the couple or their vehicle. civilians were entrapped because danger from the Thirtymile Fire. the road was not closed in time However, no action was taken for Just before 5 p.m., the couple and because evacuation of the the next 4 to 5 hours. At 3:17 noticed a smoke column down- upper canyon failed. If fire- p.m., an engine finally posted a canyon and decided to get out. fighters had not themselves sign at the road’s entrance. Driving down the road, they met become entrapped, the civilians the 14 entrapped firefighters at the probably would not have sur- Before the sign was posted, at deployment zone near a sandbar on vived. about 1 p.m., a couple drove up the river. The firefighters explained the road to Thirtymile Camp- that they were all entrapped, ground, where they planned to offering advice and encouragement.

28 Fire Management Today the time the firefighters realized a chain leading to tragedy. Fire into a high-intensity crown fire. that, it was too late for some of managers and ICs must provide the Transition fires are complicated by them. leadership necessary to break the the need for constant reassessment chain. and readjustment to a rapidly Preventable Tragedy changing situation. Common Forest Service investigators found Deceptive Situation denominators of fire behavior on that the entrapment and subse- In congressional testimony, Forest tragedy fires—fires that cost lives— quent fatalities were preventable. Service Chief Dale Bosworth offered include small size and deceptively “The lessons to be learned as a a compelling reason for what went innocent appearance before a rapid result of the fatalities on the wrong (Bosworth 2001). “The transition to extreme fire behavior Thirtymile Fire in July 2001 are people on this fire were dedicated (NWCG 1996). Most tragedy fires mostly about what was not done people,” he said. “They intended to are transition fires (see the article that should have been done,” do the right things, but they were by Jerry Williams beginning on concluded the Thirtymile Fire deceived by the fire, and the situa- page 6). Accident Review Board. Indepen- tion changed on them quickly.” dent investigations reached similar The Thirtymile Fire Prevention conclusions (Roeder 2001; Solomon The Thirtymile Fire was a transition Action Plan is designed, among and Welch 2001b). fire—a fire that rapidly changed other things, to improve situational from a low-intensity surface fire awareness on transition fires (see Investigators found 14 significant causal factors for the accident (see the list beginning on page 9). Among the most important were the violation of all of the Ten Standard Firefighting Orders and the failure to observe 10 of the Eighteen Situations That Shout Watch Out. In addition, a lack of situational awareness, exacerbated by fatigue, led to poor strategic and tactical decisions.

After the fire, some of the survivors commented to reporters on their own mistakes (Solomon and Welch 2001b). “We even made a joke of it,” said the assistant foreman on one of the engines. “‘Where’s our escape routes? Where’s the safety zones?’ Why didn’t we say anything?” “I failed to do a lot of things after lunch,” admitted one of the entrapped squad leaders on the NRW #6 crew. “I put my trust in other folks.”

That trust had seemed well placed. Good leaders were on the fire, Figure 3—Entrapment site, facing north across the river. The main fire was burning in people deeply respected for their dense fuels (center, in the crook formed by the river), spreading upcanyon to the right. knowledge, experience, and ability. Firefighters moved ahead of the main fire to fight spot fires along the “S” in the road However, fatal incidents often (right and far right). At 4:34 p.m., the fire made a crowning run across the “S,” entrapping 14 firefighters. Photo: Ben Croft, USDA Forest Service, Missoula Technology Development involve a series of events that form Center, Missoula, MT, 2001.

Volume 62 • No. 3 • Summer 2002 29 the plan beginning on page 14). By fully understanding what the fire is References implementing the action plan, the doing at all times—and always Anderson, H.E. 1982. Aids to determining wildland fire community can putting safety first. fuel models for estimating fire behavior. strengthen its capacity for safely Gen. Tech. Rep. INT–122. Ogden, UT: USDA Forest Service, Intermountain managing transition fires. Acknowledgments Research Station. The author wishes to thank every- Bosworth, D. 2001. Statement before the Could Thirtymile happen again? Subcommittee on Public Lands and one who contributed to the Forests, Committee on Energy and Investigators on the South Canyon Thirtymile Fire Investigation Natural Resources, United States Senate, Fire concluded that “a similar Report, especially those who wrote concerning the Thirtymile Fire Accident alignment of environmental factors the Fire Behavior Appendix, for the Investigation. November 14. Washington, and extreme fire behavior is not DC. outstanding work on which this Butler, B.W.; Bartlette, R.A.; Bradshaw, L.S.; uncommon and will happen again” article is based. The author also Cohen, J.D.; Andrews, P.L.; Putnam, T.; (Butler and others 2001). The same thanks Ben Croft of the Forest Mangan, R.J.; Brown, H. 2001. The South holds true for the Thirtymile Fire. Canyon Fire revisited: Lessons in fire Service’s Missoula Technology behavior. Fire Management Today. 61(1): Firefighters will face other transi- Development Center in Missoula, 14–20. tion fires under similar circum- MT, for furnishing photos for this Hamilton, J. 2001. Blaze stays with stances. Their safety will depend on firefighter who saved two campers. article. Yakima Herald-Republic (Yakima, WA). July 17. Hamilton, J.; Earls, S. 2001. Caught by fire in the canyon. Yakima Herald-Republic (Yakima, WA). September 1. NWCG (National Wildfire Coordinating Group). 1996. Common denominators of fire behavior on tragedy and near-miss wildland fires. Safety and Health Working Team. PMS 407; NFES 2225. Boise, ID: National Interagency Fire Center. Roeder, T. 2001. The investigation: Com- manders didn’t follow safety guidelines. Yakima Herald-Republic (Yakima, WA). September 1. Rutman, M. 2001. The sights, sounds and smells of a fire gone out of control [from a firefighter’s journal]. Yakima Herald- Republic (Yakima, WA). September 1. Solomon, C.; Welch, C. 2001a. Firefighters who were there tell the story of the deadly Thirty Mile blaze: “It’s snowing fire.” Seattle Times (Seattle, WA). July 29. Solomon, C.; Welch, C. 2001b. Rules broken to fatal effect in July 10 wildfire. Seattle Times (Seattle, WA). September 23. ■

Figure 4—Deployment zone, facing downriver. Eight firefighters deployed their fire shelters on the road above the sandbar (lower center), one together with two civilians; all survived. Six firefighters deployed in the scree above the road (center right); four died. Photo: Ben Croft, USDA Forest Service, Missoula Technology Development Center, Missoula, MT, 2001.

30 Fire Management Today RUSH TO JUDGMENT ON THE THIRTYMILE FIRE

Hutch Brown

ale Bosworth, Chief of the USDA Forest Service, was Stories based on half-truths Dconcerned. “There has been a circulate in the media, distorting public whole lot of speculation about why perceptions and undermining the investigation. this tragedy occurred,” he told a gathering in Coeur d’Alene, ID (Bosworth 2001). “Some people get one piece of the puzzle, then loudly •A helicopter was delayed from over justifiably triggers questions proclaim that they have the whole dropping water due to the Endan- about weather and wind. picture. This speaking out of turn gered Species Act. without all the facts does a disser- For weeks after the Thirtymile Fire, vice to our fallen firefighters, to Each explanation contained enough media stories focused on a “weak their families, and to the truth.” plausibility to feed speculation. disturbance” in the upper atmo- Each, however, represented a sphere over the fire and a possible Postfire Speculation misleading rush to judgment, and inversion break in the steep, V- Bosworth was speaking about the none proved to be a significant shaped canyon where the fire was Thirtymile Fire on July 10, 2001, causal factor.* burning (Garber and others 2001; when four firefighters perished in a Paulson 2001a; Solomon and Welch remote canyon on the Okanogan Wind Event? 2001a; Solomon and others 2001). National Forest near Winthrop, WA. Weather-induced winds can change According to one story (Garber and He put his finger on a problem that fire behavior in seconds. In 1910, a others 2001), a cold front nosing has long plagued the interagency windstorm powerful enough to snap into the area clashed with the wildland fire community: prema- mature trees whipped smoldering warm, dry air over the fire. Within ture public speculation about fires in the Northern Rockies into seconds, tremendous winds swept tragedy fires—incidents that cost the Big Blowup. In 1937, a passing through the canyon and caused a lives. Professional investigators weather front in Wyoming changed blowup, entrapping the firefighters. often take months to prepare a the wind direction by 90 degrees, It was, the reporters announced, detailed report. Meanwhile, stories entrapping dozens of firefighters on “the equivalent of a meteorological based on half-truths circulate in the the Blackwater Fire. In 1994, sud- sneak attack,” impossible for media, distorting public perceptions den winds blew a surface fire on firefighters to predict or detect in and undermining the investigation. Colorado’s Storm King Mountain advance. into a raging inferno that took 14 The Thirtymile Fire was no excep- lives.** In view of these and other Observations taken by firefighters tion. Long before the investigation tragedy fires, news of every burn- in the canyon do not support media team released its findings, one speculation about weather-induced explanation after another surfaced winds. From 1:30 a.m. to 7:00 p.m. * See Thirtymile Fire Investigation Report: Accident on the day of the entrapment, in the media: Investigation Factual Report and Management Evaluation Report (Winthrop, WA: USDA Forest Service; windspeeds generally ranged from September 26, 2001, as amended October 16, 2001). calm to 2 to 4 miles per hour (0.9– •A wind event caused the tragedy; Unless otherwise referenced, all information in this • Firefighters were too young and article is from the investigation report. 1.8 m/s). The fire itself generated gusts and strong winds after it inexperienced; ** For more on the Big Blowup, see Stephen J. Pyne, “A • Fire shelters failed; and Story To Tell,” Fire Management Today 60(4): 6–8; and began upslope runs at about 3:20 Hutch Brown, “The 1910 Fires: A New Book by Stephen p.m., but these winds were not J. Pyne,” Fire Management Today 61(4): 45–47. On the Blackwater Fire, see Karl Brauneis, “1937 Blackwater weather related. Fire Investigation: Boost for Smokejumpers?”, Fire Hutch Brown is the managing editor of Fire Management Today 62(2): 24–26. On the South Canyon Fire, see Bret W. Butler and others, “The South Canyon Moreover, the investigation team Management Today, USDA Forest Service, Fire Revisited: Lessons Learned,” Fire Management Washington Office, Washington, DC. Today 61(1): 14–20. found no reason to suppose that an

Volume 62 • No. 3 • Summer 2002 31 Observations taken by firefighters and experienced; the Northwest in the canyon do not support speculation Regular (NRW) #6 Fire Crew boss alone had more than two decades of about weather-induced winds. firefighting experience. Investiga- tors concluded that causal factors other than youth and inexperience inversion break (see sidebar) caused were too young, inexperienced, and contributed to the fire tragedy, a blowup. Fire behavior on the poorly trained to safely fight fires especially management’s underesti- Thirtymile Fire was entirely consis- (Holt 2001; Milstein 2001; Rivera mation of the fire danger. tent with unseasonably dry condi- and Solomon 2001; Stover and tions, unstable atmospheric condi- Roeder 2001). Investigators found two areas where tions, and high fire danger in deficient training and experience explosive fuels. Although the Firefighters must be adults (at least might have played a role in what firefighters did not obtain a local 18 years old), pass a physical happened on the Thirtymile Fire: weather forecast, conditions in the examination and rigorous exercise canyon were known or knowable, test, and take about 40 hours of • According to some NRW #6 crew and extreme fire behavior was training in fire behavior and members, the crew boss trainee foreseeable. firefighting techniques. After was “fast-tracked” for the crew meeting the basic requirements, boss position. He might not have Youth and Inexperi- hundreds or even thousands of had sufficient experience to fully ence? firefighters with little or no experi- understand the risk inherent in some of the tactics employed “Leadership flows from experience ence safely fight fires each year. under the dangerous conditions and training, and the Forest Service Moreover, the fourth victim on the prevalent on the Thirtymile Fire. appears to have a bias against both,” Thirtymile Fire was a 30-year-old • Several crew members did not intoned an editorial after the with 12 years of firefighting experi- properly deploy their fire shelters, Thirtymile Fire (Seattle Times ence, a well-trained squad leader perhaps because the fire caught 2001). Two of the four firefighters revered by the young firefighters he them by surprise. Some deployed who died on the fire had no experi- led. with their heads toward the fire, ence, and a third had spent only and others took their packs inside 1 year fighting fires; all three were According to Forest Service investi- their shelters or dropped them less than 22 years old. Critics asked gators, the firefighters on the nearby (a fire hazard). whether the Forest Service was Thirtymile Fire were generally well putting people on firelines who qualified. Many were highly skilled In addition, an investigation by the Occupational Safety and Health Administration (OSHA) found that INVERSION BREAKS supervisory personnel on the fire above the level of crew boss lacked Nighttime inversions are common in the deep, steep valleys and sufficient OSHA-related training, canyons of the West. At night, the ground cools more quickly than the and that one of the squad bosses on air. The ground cools adjacent air, which sinks downslope and collects the fire could not document in the valley bottom. A warm layer of air above traps cooler air below completion of every task required on the valley floor. Smoke from smoldering nighttime fires rises only for his position. (See OSHA viola- to the point where its temperature equals the warmer temperatures tions 2 and 3 in Forest Service’s above. One sign of an inversion on a fire is a flat layer of smoke trapped Initial Hazard Abatement Plan, in a valley. reprinted beginning on page 19.)

However, as daytime temperatures rise and fire activity grows, air on Fire Shelter Failure? the valley floor can become warm enough to break through the inver- Four firefighters perished inside sion layer. Warm air from below rapidly rises, and cool air from above their fire shelters, raising questions rushes in to replace it. Within seconds of an inversion break, winds can about the technology (CBS 2001; rise and fire behavior can become extreme. Garber and others 2001; Paulson

32 Fire Management Today Firefighters are trained to use fire shelters for the specific protections they provide. For example, firefighters learn that they should choose deployment sites that will give them a good seal against the ground, well away from natural fuels and flammable equipment.

Since they were first introduced in 1974, fire shelters have saved more than 250 lives and have prevented hundreds of serious injuries and illnesses from burns and smoke inhalation. On the Thirtymile Fire, fire shelters saved 12 lives; even Firefighters learning how to deploy a fire shelter. Every firefighter takes about 40 hours of improperly deployed fire shelters basic training in wildland fire behavior and firefighting techniques before ever stepping onto a fireline. Photo: USDA Forest Service. offered good protection on suitable sites. For example, one fire shelter 2001b; Slivka 2001; Wiley 2001). The main cause of burnover fatali- held three people—a firefighter and Critics charged that the Forest ties is damage to lungs and airways two civilians—even though, like Service had ignored fire shelter from flames and hot gases. Fire every fire shelter, it was designed development or prevented better shelters protect firefighters by for only one. All three survived with fire shelter designs from being trapping breathable air and by only minor injuries. adopted. However, a process for reflecting 95 percent of a fire’s improving fire shelter design was radiant heat; they are not designed On the Thirtymile Fire, the en- already underway (see sidebar). to withstand direct flames. trapped firefighters had enough time—about half an hour—to

FIRE SHELTER DEVELOPMENT

Any product can be tested to look ments in Canada’s Northwest The tests had to be repeatable in a good or bad. Firefighter safety Territories in summer 1999. laboratory so that each shelter demands a standard, reliable, could be evaluated under the repeatable, and impartial perfor- The next step was to use the data to same conditions to ensure mance test for fire shelters. Until develop appropriate performance impartiality. recently, there has been no such tests for the fire shelter. In Decem- test. The USDA Forest Service’s ber 1999, the University of Alberta’s All test protocols were available Missoula Technology and Devel- Combustion and Environment for testing fire shelter designs by opment Center (MTDC) in Group, Department of Mechanical June 2001. Full-scale testing Missoula, MT, set out to get one. Engineering, contracted to develop began in September 2001, with a battery of tests: the current fire shelter used as a The first step was to describe the performance baseline. Successful fire environment for the fire • Full-scale radiant and convective designs were presented to shelter. Working with the Forest thermal tests, decisionmakers, who adopted a Service’s Fire Sciences Laboratory • Strength and durability tests, new design in June 2002. Field in Missoula, MT, and the Univer- • Flammability tests, delivery of a new fire shelter is sity of Alberta in Edmonton, • Protocols for volume measure- scheduled for spring 2003. Alberta, Canada, MTDC gathered ments, and the necessary data during the •Toxicity tests. International Crown Fire Experi-

Volume 62 • No. 3 • Summer 2002 33 select and prepare a good deploy- Investigation findings do not bear Dispatch notified them that the ment site before the burnover. The out the story. At 2:15 a.m. on July helicopter had arrived and was four firefighters who did not survive 10, the incident commander poised for action, but the leaders deployed their fire shelters away ordered two Mark 3 pumps and a still did not request deployment. from the main group, among rocks helicopter for reconnaissance. The However, the crew was having from 1 to 3 feet (0.3–0.9 m) thick, pumps were critical; the fire was on trouble running the pumps, and the with duff and other flammable flat terrain in a riparian zone, so leaders at last decided to change material embedded in the crevices. pumps with hoses could deliver tactics. At 12:08 p.m., they asked for They died from asphyxia due to about 10 times more water, with far the helicopter. inhaling superheated gases, prob- more regularity and precision, than ably because they could not get a a type 3 helicopter with a longline. Before launching, dispatch decided good seal against the ground. In- Indeed, the heli-copter’s limited to obtain approval from the district vestigators concluded that deploy- capabilities, had they been brought ranger for fear of violating the ment site selection, not fire shelter to bear much earlier on the Thirty- Endangered Species Act. The fear failure, was a significant causal mile Fire, would likely have made was misplaced; a 1995 memoran- factor in all four fatalities. little difference (see sidebar), except dum from the USDI U.S. Fish and possibly in providing better infor- Wildlife Service made clear that Helicopter Delay? mation on fire activity. firefighter safety takes precedence During a congressional hearing on over the protection of endangered the National Fire Plan in late July The helicopter pilot had already species or their habitat.* 2001, members of the House spent long hours on another fire Subcommittee on Forests and nearby; to accommodate his need The district ranger gave his ap- Forest Health grilled the Forest for an off-duty rotation, the helicop- proval at 2:00 p.m., a delay of about Service Chief about an obscure ter was ordered for 10 a.m. on July 2 hours. By 2:38 p.m., the helicop- event on the Thirtymile Fire: a 10. By then, initial attack using ter was making water drops on the delay in helicopter arrival. They pumps had succeeded in reducing fire. Despite the water drops, the seemed to believe that it held the the fire to a few scattered hotspots fire gained a foothold at the base of key to the Thirtymile Fire tragedy. with benign fire behavior. the canyon slope and began making upslope runs. The helicopter played Media outlets immediately picked At about 11 a.m., the NRW #6 crew no significant role in ensuing up the story (Dlouhy 2001; Price began attacking the fire’s remains. events, including the actual entrap- 2001; Solomon 2001; Soraghan The leaders decided to continue ment and fatalities. 2001; Zahn and La Jeunesse 2001). using pumps; they did not immedi- “For nine hours, the trapped ately ask for the helicopter. At 11:52 * For a discussion of the negligible danger to salmonids from helicopter dipping in lakes, see Justin Jimenez and firefighters begged for a helicopter a.m., Okanogan National Forest Timothy A. Burton, “Are We Scooping More Than Just to drop water from the nearby Water?”, Fire Management Today 61(1): 34–36. Chewuch River,” one reporter declared (Price 2001). Citing HELICOPTER CAPABILITIES unnamed sources, reports suggested that lives were lost to save A type 3 helicopter with longline supports firefighters on the ground by fish. The helicopter was delayed, so cooling down hotspots and extinguishing small spot fires. However, it the story went, because endangered cannot by itself stop even a small wildland fire. A 5-acre (2-ha) fire— salmonids were in the river and the about the size of the Thirtymile Fire in its initial stages—covers about Endangered Species Act therefore 220,000 square feet (20,000 m2), an area about equal to a city block of prohibited helicopter dipping. 10 to 15 houses, plus streets and improvements. A 25-acre (10-ha) fire Statements by firefighters only covers more than 1 million square feet (93,000 m2), about equal to a fanned the flames. “I think if we’d neighborhood of more than 100 homes, including streets and improve- had the water when we asked for it, ments. A type 3 helicopter with a longline can drop 75 gallons (284 L) none of this would have happened,” (the equivalent of two large bathtubs) every 5 to 10 minutes—not the NRW #6 crew boss told report- nearly enough to have played a decisive role at any time during the ers (Solomon 2001; Soraghan Thirtymile Fire. 2001).

34 Fire Management Today The investigation concluded that •Openly discuss known facts about Garber, A.; Solomon, C.; Welch, C. 2001. the helicopter was indeed delayed, the fire; Dangerous mix fueled a blaze considered minor into the State’s deadliest wildfire partly due to confusion about legal •Deplore the accident and offer since 1974. Seattle Times (Seattle, WA). requirements for endangered condolences to the victims and July 15. species protection. According to the their families; Hamilton, J.; Earls, S. 2001. Caught by fire in the canyon. Yakima Herald-Republic investigation report, the delay “may •Describe the investigation, (Yakima, WA). September 2. have reduced the effectiveness of offering detailed information on Holt, G. 2001. Even today’s training can fall suppression actions” before the fire its purpose, nature, and scope; short on the fire lines. Seattle Post- Intelligencer (Seattle, WA). July 13. established itself at the base of the and Milstein, M. 2001. Crew left unready to fight slope. However, because its overall •Decline to comment on any wildfire. The Oregonian (Portland, WA). effect was minimal, investigators circumstances under investiga- September 27. classified the helicopter delay as an tion. Paulson, T. 2001a. Shift in weather can make fires explode. Seattle Post- influencing factor and not a signifi- Intelligencer (Seattle, WA). July 12. cant causal factor in the Thirtymile Declining to comment might seem Paulson, T. 2001b. Lifesaving foil tents are Fire tragedy. evasive. However, media speculation last resort if overcome by flames. Seattle Post-Intelligencer (Seattle, WA). July 12. following the Thirtymile Fire, partly Price, J.H. 2001. Lawmakers question why Failed Speculation based on comments made by firefighters died; target environmental firefighters and other “official” rules, indecisiveness. Washington Times Overall, the rush to judgment on (Washington, DC). August 3. the Thirtymile Fire ended in failure. sources, produced a failed rush to Rivera, R.; Solomon, C. 2001. Raging None of the speculative reasons for judgment. The disservice done to wildfire leaves clues about deaths in its the tragedy—a wind event, the public gives fire managers a path. Seattle Times (Seattle, WA). July 13. sound reason for declining to Roeder, T. 2001. The investigation: Com- firefighter inexperience, fire shelter manders didn’t follow safety guidelines. failure, or a helicopter delay— comment on circumstances under Yakima Herald-Republic (Yakima, WA). proved to be a significant causal investigation following any future September 1. incident. Seattle Times. 2001. The fire last time: factor in the entrapment or subse- Forest Service mismanagement. Seattle quent fatalities. The investigation Times (Seattle, WA). September 30. team found 14 totally different Acknowledgment Slivka, J. Flawed fire shelters equal “tinfoil coffins.” The Arizona Republic (Phoenix, causal factors, mostly related to The author wishes to thank Willette AZ). August 21. management decisions on the fire Squire, a media assistant for the Solomon, C. 2001. Why Thirty Mile Fire (see the list beginning on page 9). Forest Service, Office of Communi- raged without water. Seattle Times (Seattle, WA). August 1. Independent investigations largely cation, Washington Office, Wash- Solomon, C.; Welch, C. 2001a. Firefighters corroborated the team’s findings ington, DC, for compiling the who were there tell the story of the (Roeder 2001; Solomon and Welch media clippings used for this deadly Thirty Mile blaze: “It’s snowing 2001b). article. The Office of Communica- fire.” Seattle Times (Seattle, WA). July 29. Solomon, C.; Welch, C. 2001b. Rules broken tion clipping service is available for to fatal effect in July 10 wildfire. Seattle Speculation about tragedy fires is use by Forest Service employees on Times (Seattle, WA). September 23. driven by the public’s need—and the World Wide Web at . wind: New details available on Winthrop and fatalities occur on a fire. De- blaze. Seattle Times (Seattle, WA). July tailed investigation reports meet References 17. Soraghan, M. 2001. Water delay blamed in the need but take time to prepare. Bosworth, D. 2001. A long-term fire fire deaths. Denver Post (Denver, CO). The timelag creates opportunities strategy for ecosystem health. Address to August 1. for premature speculation by the Western Governors’ Association, Stover, E.; Roeder, T. 2001. Fire report is Plenary Session on New Partnerships for reporters eager for a scoop and by flawed, victim’s dad asserts. Yakima Preventing and Managing Wildfires and Herald-Republic (Yakima, WA). July 18. others with axes to grind. Restoring Fire-Prone Ecosystems. Coeur Wiley, J.K. 2001. Shelters didn’t save 4 d’Alene, ID: August 13. firefighters. Washington Post (Washing- After a tragedy fire, fire managers CBS. 2001. Building a better fire shelter. ton, DC). July 12. CBS News (Pasadena, CA). August 14. Zahn, P.; La Jeunesse, W. 2001. In focus: Did should prepare for the inevitable Dlouhy, J.A. 2001. Lack of water on fire line a Federal policy to protect rare fish in postfire speculation. The best at issue. Seattle Post-Intelligencer Washington cost four firemen their lives? response is to: (Seattle, WA). August 1. Fox News: The Edge With Paula Zahn. August 1. ■

Volume 62 • No. 3 • Summer 2002 35 INJURIES, ILLNESSES, AND FATALITIES AMONG WILDLAND FIREFIGHTERS*

Richard J. Mangan

ildland firefighting activities take place in a high-risk Firefighters often come to the fireline W environment. The fire- with the same physical conditions and problems fighters involved in these activities as the general population. are often at risk, both in the short term and the long term, of illnesses, injuries, and sometimes even death. In the United States, 133 individu- levels: Federal, State, rural, volun- Firefighter Workforce als died in activities associated with teer, and contractor. The individuals who participate in wildland fire suppression from 1990 wildland firefighting operations are to 1998. Australia has also experi- Work Environment as varied as the fuel and terrain enced numerous fire-related fatali- A wide variety of environmental types that they fight fire in: females ties during the same period, and conditions exist in the world of and males of all racial backgrounds, other firefighters around the world wildland fire suppression: From the at least 18 years old but often into have died in Greece, Mongolia, Arctic tundra to the Florida ever- their 60s and 70s. They might Russia, and South Africa. glades; from the eucalypt forest of weigh less than 100 pounds (45 kg) Australia to the chaparral of south- or more than 250 pounds (113 kg) This article discusses factors that ern California; and from the ponde- and range from less than 5 feet (1.5 are critical to both firefighters and rosa pine forests of Montana to the m) tall to more than 6.5 feet (2.0 m) fire managers in ensuring a safe and pine barrens of New York and New tall. productive workforce. First, it Jersey, the extent of ecosystems that discusses such items as the work experience fires is truly worldwide. Firefighters are truly a cross- environment, the firefighter section of the population that they workforce, physical fitness, nutri- Numerous factors compound the serve. Although some fire agencies tion, work/rest cycles, lifestyle already stressful work of suppress- have physical fitness requirements, choices, and job requirements. ing fires: elevations that range from firefighters often come to the fire Next, it reviews firefighter illnesses, sea level to more than 6,500 feet environment with the same physi- injuries, and fatalities, with the (2,000 m); steep, uneven ground; cal conditions as the general popu- purpose of identifying mitigation high ambient air temperatures that lation: allergies to smoke and dust; measures for reducing and/or often exceed 95 ºF (35 ºC); and trick knees; weight and fitness eliminating the risks from the fire above-average levels of smoke and problems; and other preexisting environment. The mitigation dust. All these conditions have the conditions that might surface on measures suggested are applicable potential to affect the on-the- the fireline. to firefighters at all organizational ground performance of the wildland firefighter; they can ultimately Firefighting Job Dick Mangan is the program leader result in illness, injury, or even (retired) for Fire and Aviation Management, Besides environmental and human Missoula Technology and Development death. These factors, especially for factors, another critical factor that Center, USDA Forest Service, Missoula, MT; individuals not acclimated to them, contributes to the illnesses, inju- he is currently President of Blackbull can have a cumulative effect on a Wildfire Services, Missoula, MT. ries, and deaths suffered by wildland firefighter’s ability to resist these firefighters is the actual job itself. * The author presented a version of this article at the exposures and risks. Long hours of arduous work under symposium “Fire Conference 2000: the First National Congress on Fire Ecology, Prevention and Manage- difficult physical conditions, ment,” November 2000, San Diego, CA.

36 Fire Management Today controlled by the environment (such as heat, humidity, and elevation) are interesting to contem- plate, but are beyond the scope of our ability to affect in the context of wildland firefighting.

There are, however, a number of items that individual firefighters, whether volunteer or full-timer, can affect through their own actions and attitudes. Although physical height is a genetically inherited factor, an individual has a range of options regarding lean body weight, physical fitness level, and muscular endurance. These factors are a direct result of the firefighter’s choices regarding nutrition and Environmental smoke is a constant health hazard for wildland firefighters, whether on wildfires or prescribed burns (as here, on Nevada’s Toiyabe National Forest). Photo: USDA exercise regimes—that is, the Forest Service, 1994. firefighter’s motivation to prepare for the job at hand. coupled with reduced sleep and Montana’s Human Performance Whereas these factors are generally dietary changes, plus working Laboratory in Missoula, MT, have considered long term, other factors closely with a new group of indi- shown that aerobic fitness is the tend to be affected more by short- viduals in a less-than-hygienic primary limiting factor in the term actions. For example, acclima- setting, with the potential for firefighter’s ability to sustain hard tization with respect to both heat exposure to previously unseen work throughout the long and elevation can be changed infections during a period of workshifts. within a relatively short time. As reduced immunity: All these are temperatures rise during the early prime conditions for illness and/or Like athletes, serious firefighters stages of a fire season, firefighters injury to the firefighter, especially realize that physical activity and should begin moderate levels of on multiday fire assignments. training are a year-round require- outside activity to prepare them- ment if they are to successfully selves for the inevitable fires that Fighting wildland fires has unique meet the demands of the job. This will require extended physical physical fitness requirements, requirement is often difficult to activity. Similarly, higher levels of unlike most other jobs in the meet, especially in a workforce that hydration and nutrient supple- civilian workforce. Both lower and has many other demands on its ments will be necessary during upper body strength are needed to available time. prolonged periods of strenuous complete the necessary tasks, and activity during periods of high heat endurance is essential to work for Individual Factors loads, both from the ambient air the extended periods of time A number of factors affect the and from the fires. required to control the unwanted ability of an individual to perform fires. In addition, there is always the wildland fire suppression activities Firefighter Illnesses unexpected action of responding to in a safe and efficient manner Infectious Disease. The illnesses a flareup on the control line, or, (Davis 1999). Some are beyond the that firefighters are subject to are even worse, the need to make a individual’s ability to influence, but not that different from those rapid retreat when a fire threatens many are well within the indivi- suffered by other large groups of the firefighter’s personal safety, dual’s total control. Factors that are individuals thrown together in a especially after long hours on the inherited (such as physical height close environment—such as sailors line. Studies at the University of and weight) and those that are at sea, or teachers and students in a

Volume 62 • No. 3 • Summer 2002 37 Long hours of arduous work under difficult physical conditions, coupled with reduced sleep and dietary changes, expose firefighters to increased risk of illness and injuries. classroom—for extended periods of and developed mitigation measures time. Infection and disease in any for on-the-ground fire operations to one individual can cause illness in reduce exposure to smoke (Sharkey other individuals who have not had 1997a). previous exposure and the opportunity to develop an immune re- The long fire season in northern sponse. In addition to bringing a Idaho and western Montana in 1994 large group of individuals together, offered another opportunity to look wildland fires also complicate at the incidence of illness among matters by requiring long hours of firefighters on large incidents hard work, coupled with a change managed by fire overhead teams. An in diet and sleep patterns. These informal review of medical records factors, in addition to the exposure conducted by Mark Vore from the Smokejumpers in training. Like athletes, to smoke and dust, result in a Idaho Panhandle National Forests serious firefighters realize that physical variety of illnesses among showed that nearly 40 percent of activity and training are a year-round firefighters, especially as the the visits to the incident medical requirement if they are to successfully meet the demands of the job. Photo: Paul S. duration of a fire assignment units were documented as respira- Fieldhouse, USDA Forest Service, Missoula progresses beyond the first week. tory problems. These findings are Smokejumper Base, Missoula, MT. consistent with the problems that Environmental Smoke. The short- surfaced in 1987–88 and could stress. Project “Aquarius” noted term and long-term exposure to surface again, given the mountain- that two-thirds of the firefighter’s high levels of environmental smoke ous terrain and inversion potential from wildland fires was most on many large wildfires and pre- heat load was generated internally, with only one-third coming from apparent during the 1987 and 1988 scribed burns in the Western United the radiant heat of the fire. The fire seasons. In those years, smoke States. inversions plagued not only the study recommended that the design of protective clothing should be to immediate fire area, but also the Heat Stress. Another illness issue “let heat out, not keep heat out.” incident base camps and surround- that appears to be on an upward ing communities for days on end. trend on wildland firefighting Additionally, it recommended that wildland firefighters consider the For firefighters spending multiple operations is the incidence of heat need to consume as much as 1 liter 21-day assignments under these stress injuries. Under conditions of conditions, the incidence of upper- both high ambient air temperatures (2 pints) of fluids per hour under high temperatures and heavy respiratory-tract infections was and high radiant heat flux, the workload conditions. high; infections lasted for periods as firefighter can easily become long as 3 to 4 months after the dehydrated and subject to heat The logistics of supporting this firefighting operations were over. As stress if positive preventative level of fluid replacement during a a result, the Health Hazards of measures are not implemented as a Smoke project sponsored by the normal way of doing business on a 12-hour operational period can be challenging, but it is certainly National Wildfire Coordinating daily basis. essential to prevent heat stress Group was undertaken at the Missoula Technology and Develop- An Australian study (Budd and illness. Dehydration and heat stress illness can be the result of a pro- ment Center (MTDC). The 6-year others 1996) on work productivity gressive deterioration that occurs project culminated in 1997 with a among bush firefighters indicated conference in Missoula, MT, that that personal protective clothing over several days of reduced fluid intake. They can also be com- summarized the research findings was a key factor in reducing heat

38 Fire Management Today pounded by other factors, such as firefighter illness in the bushfire and down cut slopes; clearing other illnesses or medications being environment. obstacles and using full muscle taken by the individual. control when swinging handtools; Firefighter Injuries using proper lifting techniques for Countermeasures. Fire managers In difficult terrain, under condi- heavy objects; and paying full and crew leaders should take tions of long hours and arduous attention to driving techniques on positive actions to avoid working work, injuries are one of the major winding, steep, and/or unsurfaced firefighters to the point of exhaus- perils that wildland firefighters are roads. tion or exposing them to excessive subject to. Although no docu- levels of smoke. Additional actions mented records exist showing Although accidents are not well that can help reduce firefighter trends of firefighter injuries, on- enough documented to show their illness include: the-ground observations by experi- rate of occurrence during enced personnel show several major firefighting operations, experienced • Reducing both physical and areas where injuries occur: personnel are well aware of the emotional stress; risks. Better documentation will • Enhancing rest and recuperation •Vehicle accidents; more clearly define the problems periods, with a target of a 2-to-1 •Tool use; and lead to mitigation practices that work/rest cycle (16 hours of work, • Slips, trips, and falls; and will ultimately reduce the risk. The followed by 8 hours of rest); and •Muscle strains. MTDC publication Fitness and • Providing adequate energy and Work Capacity (Sharkey 1997b) nutrients to meet the special By inference, several of these injury documents many of the condition- requirements of the arduous fire areas can be related back to the ing techniques that can reduce job. causal factors of fitness levels and firefighter fatigue by increasing fatigue. work stamina. Firefighters have an individual responsibility to ensure their own Fatigue. As firefighters become Fitness Levels. A number of recent ability to perform the job by getting more fatigued from the long hours studies have documented the and staying in good physical condi- and arduous work, they become less relationship between fitness levels tion; making correct nutritional attentive to the small things that and injury rates. In the U.S. Army, a choices to sustain them on prevent injuries under different study of 861 female and male multihour and multiday fire assign- circumstances: using care in trainees indicated that the fittest ments; and making healthy lifestyle walking on steep slopes, over logs, soldiers (measured by their choices (such as not smoking), which will help them remain on the job during periods of reduced immunity to illnesses.

Dr. Steve Wood from Abbott Labora- tories has identified “immune- friendly nutrients” that enhance the function of the human immune system (Wood 1999). They include vitamins C and E, which both stimulate and enhance immune response; Beta carotene, which stimulates natural killer cells; vitamin B6, which promotes white- cell proliferation; selenium, which promotes antibacterial activity; and zinc, which promotes wound healing. All these nutrients can be helpful in reducing the risk of Firefighters relaxing. Firefighters should manage work/rest cycles to avoid needless fatigue. Photo: USDA Forest Service.

Volume 62 • No. 3 • Summer 2002 39 pushups, situps, and 2-mile [3.2- In the progression of events, it can • During multiday firefighting km] runs) experienced the lowest be surmised that fatalities on operations, fire managers and injury rates. Another study showed wildland firefighting operations are, crew leaders should ensure that that the most fit individuals, as in many cases, the logical extension fluid and nutritional needs are indicated by running speed, experi- of early failures to address issues of met and that work/rest cycles are enced the fewest injuries in sports illness and injuries that manifest managed to prevent unnecessary training. Finally, a 1999 Australian themselves throughout the fire fatigue among both firefighters Army study of recruits indicated a season. It is imperative that we and fire managers. negative relationship between break the chain if we are to ulti- • Fire agencies should develop, fitness and injuries. The implica- mately reduce firefighter fatalities. maintain, and monitor an illness tions of these studies for firefighters and injury data base, preferably at are obvious, especially for such a Toward Safety and the national level, to identify physically demanding activity. Health health and safety trends in the The safety and health of the wild- wildland fire community. Firefighter Fatalities land firefighting workforce is The first half of the 1990s saw two critically important to the References major wildfire fatality events that firefighters and their families, the Budd, G.; Brotherhood, J.; others. 1996. riveted the attention of the Nation: Safe and productive bushfire fighting fire management organization, and with handtools. Canberra, Australian the Dude Fire in 1990 killed 6 the community it serves. There are Capital Territory, Australia: Australian firefighters, and 14 firefighters died numerous opportunities, both Government Publishing Service. on the South Canyon Fire in 1994. short-term and long-term, to Davis, P.O. 1999. Individual factors related to health, safety and performance. In: Although these tragic events were improve the health and safety of the Sharkey, B., ed. Wildland Firefighter horrific reminders of the risks wildland workforce for all Health and Safety: Recommendations of inherent in wildland fire suppres- firefighters: the April 1999 Conference. Wildland Firefighter Health and Safety Conference; sion activities, they were only a few 27–29 April 1999; Missoula, MT. 9951– of the deaths that occurred from • First and foremost, individual 2841–MTDC. Missoula, MT: Missoula 1990 to 1998. In that period, 133 firefighters must take positive and Technology and Development Center: 61– firefighters and others involved in 64. affirmative actions to ensure their Mangan, R.J. 1999. Wildland fire fatalities in wildland firefighting operations own health and safety. This the United States: 1990–1998. 9951– died from a variety of causes. The includes maintaining an appropri- 2808–MTDC. Missoula, MT: USDA Forest MTDC report Wildland Fire Fatali- ate height/weight ratio, partici- Service, Missoula Technology Develop- ment Center. ties in the United States: 1990 to pating in an exercise program, Sharkey, B., ed. 1997a. Health hazards of 1998 (Mangan 1999) documents the and minimizing high-risk activi- smoke: Recommendations of the April causes, including aircraft accidents ties that threaten good health. 1997 Consensus Conference. 9751–2836– MTDC. Missoula, MT: USDA Forest (30 deaths), heart attacks (28 • Fire agencies have a major Service, Missoula Technology Develop- fatalities), and vehicle accidents (25 obligation to provide an environment Center. deaths). ment that fosters a safe and Sharkey, B., ed. 1997b. Fitness and work capacity. Second edition. PMS 304–2, healthy workforce. This can NFES 1596. Missoula, MT: USDA Forest Numerous opportunities exist to include health-screening pro- Service, Missoula Technology Develop- reduce firefighter fatalities away grams; exercise facilities; and, in ment Center. from the immediate fire ground some cases, work capacity testing. Wood, S. 1999. Wildland firefighting and the immune response. In: Sharkey, B., ed. through many of the same actions • Fire agencies should provide Wildland Firefighter Health and Safety: that will reduce illness and injuries. specialized training in high-risk Recommendations of the April 1999 Prevention of heart attacks offers activities, such as emergency Conference. Wildland Firefighter Health and Safety Conference; 27–29 April 1999; the best opportunity to reduce the vehicle operation; and create a Missoula, MT. 9951–2841–MTDC. number of deaths. However, pre- culture that does not condone or Missoula, MT: Missoula Techology and venting heart attacks will require a tolerate unsafe work practices, Development Center: 39–52. ■ major lifestyle change for many even during a fire emergency. firefighters.

40 Fire Management Today KEEPING HAINES REAL—OR REALLY CHANGING HAINES?

Brian E. Potter, Dan Borsum, and Don Haines

ost incident command teams can handle low- to moderate- The Haines Index is a quick and easy tool M intensity fires with few for summarizing two atmospheric parameters. unanticipated problems. However, high-intensity situations, especially the plume-dominated fires that the air overlying the fires, the of building high pressure, followed often develop when winds are low environmental lapse rate (tempera- by a trough of low pressure. High- and erratic behavior is unexpected, ture difference within a vertical pressure areas are characterized by can create dangerous situations layer of that air), and negative subsiding air that is warmed even for well-trained, experienced vertical wind shear (Haines 1988). through adiabatic compression fire crews (Rothermel 1991). He noted that researchers could use (when expanding air cools and Plume-dominated fires have a measurements of these atmospheric contracting air warms), which strong convection column that features to construct a severity occurs because atmospheric pres- towers above the fire rather than index. However, no features could sure decreases with height, whereas leaning over before the wind. They be identified from the usual surface the rising air cools and sinking air differ from wind-driven fires in that weather measurements; therefore, warms. The air is prevented from the winds are lower and primarily mathematical descriptors of the reaching the surface because of the fire induced. Some authors (Byram features had to be constructed cooler, underlying mixing layer. 1954) have called plume-dominated using above-surface observations. Temperatures within the mixing fires “blowup fires,” but that name layer decrease with altitude, but the is now commonly used for any The HI is an indicator of the poten- air just above the mixing layer, sudden increase in fire activity. tial for extreme fire behavior based having been warmed through the on two of the three features Haines adiabatic process, is often much This article updates the uses of the (1988) described—the dryness and warmer than the air at the top of fire severity index called the Haines the stability of the atmosphere. HI the mixing layer. Therefore, a Index (HI). We discuss the original uses measures of the dryness of air temperature inversion usually intended use of HI, its current above the fire to calculate the separates the mixing layer from the operational use, some ways that likelihood that an unstable lapse compressed, warm air above (fig. 1). users have modified it, and different rate will help that air reach the aspects of HI that researchers are ground. Haines did not include a This situation is ideal for plume- examining to improve its predictive wind shear term in HI because of dominated fire development. A value. disagreement among researchers breakdown of the upper level ridge, over the meteorological importance with a cooling of the air aloft and Plume-Dominated Fires of various wind profiles (see the the subsequent instability, often Haines (1988) suggested that the sidebar on page 42). produces the potential for intense growth of plume-dominated fires plume-dominated fires, especially in depends on the moisture content of Ideally, atmospheric features should the Northwestern United States be measured in the region just (Gibson 1996). above the mixing layer, where air is Brian Potter is a research meteorologist at mixed by convection. One of the the USDA Forest Service, North Central Structure of HI Research Station, East Lansing, MI; Dan synoptic patterns (atmospheric, HI ranks the moisture and stability Borsum is a meteorologist with the weather, or other conditions that of the lower atmosphere by assign- National Weather Service, Billings, MT; and exist simultaneously over a broad Don Haines is a retired meteorologist, ing to each term a value from 1 to formerly with the North Central Research area) that fire weather meteorolo- 3, as follows: Station, East Lansing, MI. gists look for is an upper level ridge

Volume 62 • No. 3 • Summer 2002 41 WIND SHEAR AND The simplicity inherent in the Haines Index limits its THE HAINES INDEX use as an indicator of broad fire potential.

The Haines Index is an indicator of the potential for extreme 1 = Stable air/Moist air Using HI fire behavior based on two of 2 = Moderately unstable air/Moder- Use HI when you have substantial the three features Haines (1988) ately dry air available fuels, when you have an described—the dryness and the 3 = Unstable air/Dry air ongoing fire or are confident of stability of the atmosphere. The ignition, and when you might Haines Index is used to calcu- When the values of the two terms expect a plume to build above the late the likelihood that atmo- are added, HI can range from a fire. A high HI value will indicate spheric instability will bring dry minimum value of 2 to a maximum the likelihood of rapid fire growth air to the ground, contributing value of 6. For example, a moist and and erratic, extreme fire behavior. A to fire activity. stable atmosphere above the fire has low HI will mean that the smoke an HI value of 2; dry and unstable column should not extend to a Another way to bring dry air to air has an HI value of 6 (table 1). significant height and that there is a fire is through wind-induced a low possibility of rapid fire growth mixing by means of negative HI can be calculated over one of and erratic, extreme fire behavior. vertical wind shear, the third three layers between 950 and 500 feature described by Haines millibar, depending on the surface Could low moisture and instability (1988). However, a wind shear elevation. The layer used should be in the overlying air contribute to a term was not included in the high enough above the surface, wind-driven fire? Possibly, but a Haines Index because of dis- usually just above the mixing layer, mechanism to bring the overlying agreement among researchers to avoid the major diurnal variabil- air down to the fire is not obvious. over the meteorological impor- ity of surface temperature extremes The downdrafts that, on a plume- tance of various wind profiles. and surface-based inversions. dominated fire, would carry the air For example, Byram (1954) Although this lessens the diurnal down are conveyed away from the listed six types of wind profiles effects, there is no way to totally fire by the wind. Fires are complex, that could be “potential trouble negate their influence. For step-by- turbulent structures; like the makers,” too many to incorpo- step procedures about how to proverbial snowflake, no two are rate into a predictive model. calculate HI, see Haines (1988) and exactly alike. Further observations, Another problem with includ- Werth and Ochoa (1990). research, and computer modeling ing wind profiles is the duration of wind events, which can vary from downburst outflows (including those generated by Figure 1— fire), to gust fronts, to noctur- Temperature profile of an idealized nal drainage, to synoptic or atmosphere that geographically preferred areas could produce an of low-level jets. explosive, plume- dominated fire. Although not the Recently, computer simulations only situation of fire spread have produced where this might occur, it is an interesting results when wind occasion where shear is considered (Coen and downdrafts others 1998; Jenkins 2000). originating in the Perhaps this line of research region above the mixing layer can will provide a future, usable bring warmer and solution. drier air to an existing fire.

42 Fire Management Today Table 1—Numeric values of the Haines Index and descriptions of extreme fire risk. mid-1990s. Within the National Weather Service and the fire Haines Index Potential for rapid fire growth weather program, fire-weather- or extreme fire behavior forecasting responsibilities were spread to more individuals in each 2 or 3………….. very low office; some offices that had not had fire weather programs in the past 4………………. low began to assume forecasting responsibilities. The new forecasters 5………………. moderate readily looked to computer-modeling data output for guidance in 6………………. high forecasting HI. Simultaneously, traditional forecast techniques involved analyzing morning sounding data and then predicting are needed to answer this question dominated fire to develop. During changes and deriving an HI fore- (Jenkins 2000). For now, we know such times, an HI value of 5 or 6 cast. With access to gridded model that the components of HI can could alert firefighters to unex- data, forecasters began to put more make a major contribution during pected fire growth and erratic fire emphasis on model calculations to the growth of plume-dominated behavior. determine projected Haines In- fires. dexes. Figure 2 shows a sample of a HI is not a predictor of wind- model-generated HI forecast. Because there is no wind compo- controlled fire, ignition potential, nent in HI, it should not be used to fuel conditions, or the number of The model output calculated higher predict the behavior of wind-driven expected fire days. There are other HI frequencies of 5 and 6 than fires. HI can tell you little when methods for evaluating these expected by experienced fire strong horizontal winds cause fire- factors, and the wise firefighter will weather forecasters, numbers that induced winds to shear ahead of the use the corresponding tools. less experienced forecasters were fire. The resulting separation of the willing to accept because the fire and the winds generated lessens HI Misuse calculations appeared correct. This led to concerns among more the possibility of the fire feeding Significant changes in fire weather back on its own circulation. experienced forecasters that users forecasting and, specifically, in would think the forecasters were forecasting HI occurred during the Nevertheless, strong winds can abate during the course of a fire, allowing a vertical smoke column Figure 2—A sample of a to develop. This happened in May computer model output 1980 during the Mack Lake Fire for the midlevel Haines Index from the 00Z (24- near Mio, MI (Simard and others hour clock that is based 1983). Two hours after ignition, on midnight at the 0th strong wind gusts and lowered meridian) on June 11, 2001. relative humidity caused a prescribed burn to escape across firelines. The fire burned a total of 24,000 acres (9,700 ha), mostly by the evening of the first day.

During its major run, the Mack Lake Fire spread about 8 miles (13 km) at an average rate of 2 miles per hour (3 km/h). Later, winds slackened, allowing a plume-

Volume 62 • No. 3 • Summer 2002 43 Future studies will enable us to establish a strong relationship between the physical processes behind the Haines Index and how it is computed.

“crying wolf.” In response, some tions. However, changes to HI just above the Haines moisture forecasters began rolling factors calculations should be made in a level so that the values used by HI such as winds, fuel moistures, and systematic way across the Nation, do not depict the true nature of fire dangers into calculations that not office by office. the atmosphere (fig. 4). they still called HI. Thus, forecast- • The mixing layer of the atmo- ers-in-training have heard experi- Examination of computer model sphere extends through the enced fire weather forecasters soundings can help forecasters tell Haines layer, yielding abnormally comment, “It’s not really a 6 Haines where calculated HI values do not high instability values (fig. 5). day.” fully reveal the true atmospheric This is of particular concern in potential, for example when: areas using the midlevel HI due Whereas that comment might to their elevation, but where deep sometimes be valid, the guidelines •A strong inversion is present mixed layers may develop during for making it are at best below the lower level used for the summer. nonstandardized and are often Haines calculations and the fire nonexistant. The problem illus- will not likely interact with the Even in these situations, a fore- trates the need to emphasize proper Haines layer (fig. 3). This is caster would be limited to arbi- HI use and calculation on the part common along the Pacific Coast trarily lowering or raising the HI. of both users and forecasters. It is of the United States. For example, a forecaster would not important to understand that HI is • The moisture content of the know whether to drop an HI value used on a national basis and that atmosphere dramatically changes of 6 to 5 or to 4. The only truly there is no systematic method for modifying HI values, even though forecasters might develop local methods for doing so. HI is a quick and easy tool for simultaneously summarizing two atmospheric parameters, moisture and stability. Any effort to use HI as a broad fire potential index is constrained by its inherent simplicity.

HI is similar to another meteorological parameter, the Lifted Index, which is used to assess potential fire instability. Like HI, the Lifted Index does not take all levels of the atmosphere into account in assessing the convective potential. Forecasters now use multiple parameters to assess the overall potential indicated in the atmosphere. Methods for modifying HI through more detailed analysis are appropriate, especially where Figure 3—Upper air sounding from Rapid City, SD, at 00Z (24-hour clock that is based on locations are near the dividing line midnight at the 0th meridian) on July 8, 2000. The low-level inversion would have isolated the surface from the upper air and inhibited the Haines Index layer—700 to 500 milli- between high and middle eleva- bar—from reaching the ground.

44 Fire Management Today appropriate practice would be to provide the calculated HI while making note of any atmospheric conditions that might limit or enhance it. Improving HI The uses and interpretation of HI are gradually changing, for two primary reasons:

1.There is no “finish line” in science—each question that is answered raises several new questions; and 2.Operational use of HI is pushing it beyond its original purpose and ability.

The climatology used in Haines Figure 4—Upper air sounding from Rapid City, SD, at 00Z (24-hour clock that is based on (1988) relied on two station loca- midnight at the 0th meridian) on July 16, 2000. The moisture at 700 millibar was not tions and a single year. At the time, representative of a high-elevation Haines Index layer—700 to 500 millibar. such a climatology required many hours of tabulation and calculation. Today, with greater computer power and data that are more accessible, it is possible to produce a climatology for multiple stations and years. Werth and Werth (1998) did this for 20 stations in the Western United States, covering 5 years at each station. Their analysis showed that the high frequency of HI 5 and 6 values noted in the 1990s was not a product of the computer models— in some areas it is real. Some stations in the high-elevation regions of the West differ significantly from the climatology used in Haines (1988).

With funding from the National Wildfire Coordinating Group and the National Fire Plan, meteorologists at Jackson State University in Jackson, MS, and at the USDA Forest Service’s North Central Research Station in East Lansing, MI, are creating an HI climatology Figure 5—Upper air sounding from Great Falls, MT, at 00Z (24-hour clock that is based on for the United States, including midnight at the 0th meridian) on July 13, 2000. The 700 millibar level is deep within the Hawaii, Alaska, and Puerto Rico, for mixing layer, possibly causing a misleadingly high Haines Index for this high-elevation site. 1961 to 1990. When complete, the

Volume 62 • No. 3 • Summer 2002 45 Meteorologists are creating a Haines Index climatology for the United States, including Hawaii, Alaska, and Puerto Rico, for 1961 to 1990. climatology will allow fire weather Instead, we must rely on data from References forecasters across the United States a few fires where all of the relevant Byram, G.M. 1954. Atmospheric conditions to see how HI behaves in their measurements were made, and on related to blowup fires. Sta. Pap. 35. region during the year and how it computer simulations of fire– Asheville, NC: USDA Forest Service, Southeastern Forest Experiment Station. compares to other areas of the atmosphere interactions. Using Coen, J.L.; Clark, T.L.; Hall, W.D. 1998. Nation. The climatology will also computer models similar to those Effect of atmospheric winds and shear on provide a starting point for re- used for regular weather forecasts fire behavior experiments using a coupled atmosphere–fire model. In: Second searchers as they try to determine and for studying thunderstorms, Conference on Fire and Forest Meteorol- whether or how they should adjust researchers at Los Alamos National ogy; 1998 January 11–16; Phoenix, AZ: their methods of calculating HI. Laboratory, the National Center for 81–86. Atmospheric Research, the Univer- Gibson, C.V. 1996. Break-down of the western U.S. upper-level ridge: A critical The inspiration for HI and the basis sity of Utah, and the North Central fire weather pattern. In: Twenty-Second for its reliance on stability and Research Station, among others, Conference on Agricultural and Forest moisture came from the observa- are trying to refine our understand- Meteorology; 1996 28 January–2 February; Atlanta, GA: 261–262. tions and experiences of numerous ing of fire weather. Using these Haines, D.A. 1988. A lower-atmosphere firefighters, researchers, and models, researchers can specify the severity index for wildland fires. National forecasters over many years (Haines stability, wind, and moisture. In Weather Digest. 13: 23–27. Jenkins, M.A. 2000. The Haines Index: Uses, 1988). They enabled Haines to turn, the computer simulation misuses and needs. In: Fire Conference create an index that clearly reflects provides information on how the 2000; 2000 November 27–December 1; two important variables in fire atmosphere interacts with an San Diego, CA. Audiotape 440–23. weather. Because HI is based on intense fire. Available at . formal and informal observations, it Rothermel, R.C. 1991. Predicting behavior is not explained in terms of physical Future studies will enable us to and size of crown fires in the Northern processes, air motion, or the way establish a strong relationship Rocky Mountains. Res. Pap. INT–438. Ogden, UT: USDA Forest Service, that the elevated layer of air used to between the physical processes Intermountain Research Station. calculate HI affects a fire on the behind HI and how it is computed. Simard, A.J.; Haines, D.A.; Blank, R.W.; ground. Understanding these Whatever the changes, HI will Frost, J.S. 1983. The Mack Lake Fire. Gen. Tech. Rep. NC–83. St. Paul, MN: processes would perhaps allow retain its original focus and charac- USDA Forest Service, North Central improvements in HI generally, or at ter—measuring the ability of the Forest Experiment Station. least specifically at locations where atmosphere to turn a low- or Werth, P.; Ochoa, R. 1990. The Haines Index it is thought to break down. moderate-intensity fire into an and Idaho wildfire growth. Fire Manage- ment Notes. 51(4): 9–13. explosive, dangerous, “blowup” fire Werth, J.; Werth, P. 1998. Haines Index Piecing Together the under low-wind conditions. This climatology for the Western United piece of information from HI will fit States. Fire Management Notes. 58(3): 8– Puzzle 17. ■ We cannot ignite fires under together with other puzzle pieces, controlled atmospheric conditions allowing us to calculate the fire risk to make the needed observations. for a specific situation.

46 Fire Management Today BURNING UNDER OLD-GROWTH PONDEROSA PINES ON LAVA SOILS

Peter Z. Fulé, Greg Verkamp, Amy E.M. Waltz, and W. Wallace Covington

very year, fire managers successfully restore surface fires to Prescribed burning on volcanic soils in the Eponderosa pine forests that have Grand Canyon’s Parashant National Monument not burned for a century or more. produced unanticipated results. Nevertheless, the power of fire can have unanticipated effects on stressed trees. In northern Arizona, of an alarming amount of old- observed in 2001 at the Old Folks an episode of unusual old-growth growth ponderosa pines. In October Home site (figs. 1 and 2). Younger tree mortality, after two prescribed 1999, 5 acres (2 ha) at the Old Folks ponderosa pine trees at both burn burns on shallow volcanic soils, Home site were burned and in sites were not killed. provided managers with a learning March 2000, 25 acres (10 ha) at opportunity and a lesson in caution. experimental block (EB) 1 were The unusual mortality seemed burned. The sites were on two edges confined to the lava flow region. High Tree Mortality of a large, shallow-soiled lava flow, High mortality was not observed on Ponderosa pine forests at the base which historically had high-fre- the adjacent thinned 650 acres (260 of Mt. Trumbull in the Grand quency, low-intensity surface fire ha) or unthinned 396 acres (160 Canyon–Parashant National Monu- regimes but had remained largely ha), which were burned between ment have been the focus of exten- fire free for 130 years. The area had 1996 and 2001. However, Swezy and sive research in ecological restora- experienced severe drought in 1977, Agee (1991) found high old-growth tion since 1995. Sponsored by the 1989, 1996, and 2000. ponderosa pine mortality following USDI Bureau of Land Management, a fire at Crater Lake, OR, in another Arizona Game and Fish Depart- Before conducting the burns, all ponderosa forest on shallow lava ment, and Northern Arizona Uni- old-growth trees at both sites were soils. They suggested that the cause versity, the research gives managers protected against heat damage to was heat injury to fine roots (fig. 3). information about how to restore the cambium layer by raking Other hypotheses include reduced natural surface fire regimes and accumulated forest floor litter away tree vigor, drought stress, or lack of open forest conditions using pro- from the tree trunks; research has wind to disperse heat during the ductive native vegetation. Since shown that removal of deep forest Old Folks Home burn. 1995, 650 acres (260 ha) have been floors prevents tree girdling by fire treated with thinning, fuel reduc- (Sackett and others 1996). The 1999 A census of trees greater than 5.9 tion, and prescribed fire. An addi- burn site had no tree thinning inches (15 cm) in diameter was tional 396 acres (160 ha) have been before burning, whereas the 2000 conducted on the site of the Old treated with fire alone. burn site was previously thinned to Folks Home burn on June 5 and 6, help restore the conditions that 2001 (fig. 4). Of 247 ponderosa pine Most treatments in the forest have existed before the frequent fire trees, 91 (37 percent) were dead. had good initial results (Fulé and regimes were disrupted (Covington Five of the dead trees had great others 2001). However, two pre- and others 1997). decay, suggesting that they probably scribed burns resulted in the death died before the prescribed burn. Not Shortly after the 2000 fire, postburn counting these trees, the mortality Peter Fulé is an assistant professor, Greg measurements at EB1 indicated rate was 35 percent. The pattern of Verkamp is a research technician, Amy minimal fire effects; however, field mortality varied with tree size. Only Waltz is a research specialist, and Wallace observation in May 2001 showed 19 percent of the trees that were Covington is a regents’ professor at the Ecological Restoration Institute and School that many old trees were dying. less than 20 inches (50 cm) died, of Forestry, Northern Arizona University, Mortality of the large trees was also compared with a 67-percent mortal- Flagstaff, AZ.

Volume 62 • No. 3 • Summer 2002 47 ity rate for trees that were greater than 20 inches (50 cm) in diameter. A similar comparison will be done at EB1 in a few years, when the fire- caused mortality is fully evident. Future Trends Conservation of the old-growth trees that are scattered across the landscape is an important project goal. The Old Folks Home and EB1 sites make up less than 3 percent of the total area burned at Mt. Trumbull, so the death of old trees is not widespread. However, it might be necessary to stop burning on the lava flow pending further Figure 1—Mortality of mature trees on lava soils at the Old Folks Home burn site at the base of Mt. Trumbull, AZ. The burn occurred in October 1999. Photo: Greg Verkamp, information from both experimen- Ecological Restoration Institute and School of Forestry, Northern Arizona University tal sites. Researchers might repeat Flagstaff, AZ, 2001. pre- and postburn measurements on permanent monitoring plots and Figure 2—Large compare them with a nearby trees dying after control site; they might also study prescribed burning fine-root mortality and drought on lava soils in an unthinned stand. stress on old-growth trees in the Note the surviving lava area. Additionally, managers trees in back- should consider using only nonfire ground and at left. Photo: Greg treatments, such as thinning alone, Verkamp, for forests on lava soils. Ecological Restoration Institute and School of Forestry, Northern Arizona University, Flagstaff, AZ, 2001.

48 Fire Management Today Unfortunately, the true extent of tree mortality on the EB1 and Old Folks Home experimental sites remains unknown. Several years might pass before needles on the remaining large ponderosa pine trees, which appear to be alive and well, begin prematurely to brown and fall. References Covington, W.W.; Fulé, P.Z.; Moore, M.M.; Hart, S.C.; Kolb, T.E.; Mast, J.N.; Sackett, S.S.; Wagner, M.R. 1997. Restoration of ecosystem health in southwestern ponderosa pine forests. Journal of Forestry. 95(4): 23–29. Fulé, P.Z.; Waltz, A.E.M.; Covington, W.W.; Heinlein, T.A. 2001. Measuring forest Figure 3—Exposed tree root on shallow soils. A mix of dying and surviving trees is evident restoration effectiveness in hazardous in the background. Photo: Greg Verkamp, Ecological Restoration Institute and School of fuels reduction. Journal of Forestry. Forestry, Northern Arizona University, Flagstaff, AZ, 2001. 99(11): 24–29. Sackett, S.S.; Haase, S.M.; Harrington, M.G. 1996. Lessons learned from fire use for restoring southwestern ponderosa pine ecosystems. Gen. Tech. Rep. RMFRES– Number of trees 278. Fort Collins, CO: USDA Forest 70 Service, Rocky Mountain Forest and Range Experiment Station. Swezy, D.M.; Agee, J.K. 1991. Prescribed- 60 fire effects on fine-root and tree mortality Live trees Dead trees in old-growth ponderosa pine. Canadian 50 Journal of Forest Research. 21: 626–634. ■ 40

0 20 30 40 50 60 70 80 90 100 Diameter (cm)

Figure 4—Diameters of living and dead ponderosa pine trees at the Old Folks Home burn site. Most large trees were killed, whereas most small trees survived.

Volume 62 • No. 3 • Summer 2002 49 GUIDELINES FOR CONTRIBUTORS Editorial Policy Paper Copy. Type or word-process the manu- label all photos and illustrations (figure 1, 2, 3, Fire Management Today (FMT) is an interna- script on white paper (double-spaced) on one etc.; photograph A, B, C, etc.). At the end of the tional quarterly magazine for the wildland fire side. Include the complete name(s), title(s), manuscript, include clear, thorough figure and community. FMT welcomes unsolicited affiliation(s), and address(es) of the author(s), as photo captions labeled in the same way as the manuscripts from readers on any subject related well as telephone and fax numbers and e-mail corresponding material (figure 1, 2, 3; photo- to fire management. Because space is a information. If the same or a similar manuscript graph A, B, C; etc.). Captions should make photos consideration, long manuscripts might be is being submitted elsewhere, include that and illustrations understandable without reading abridged by the editor, subject to approval by the information also. Authors who are affiliated the text. For photos, indicate the name and author; FMT does print short pieces of interest to should submit a camera-ready logo for their affiliation of the photographer and the year the readers. agency, institution, or organization. photo was taken.

Submission Guidelines Style. Authors are responsible for using wildland Electronic Files. See special mailing instruc- Submit manuscripts to either the general fire terminology that conforms to the latest tions above. Please label all disks carefully with manager or the managing editor at: standards set by the National Wildfire Coordinat- name(s) of file(s) and system(s) used. If the ing Group under the National Interagency manuscript is word-processed, please submit a 3- USDA Forest Service Incident Management System. FMT uses the 1/2 inch, IBM-compatible disk together with the Attn: April J. Baily, F&AM Staff spelling, capitalization, hyphenation, and other paper copy (see above) as an electronic file in one Mail Stop 1107 styles recommended in the United States of these formats: WordPerfect 5.1 for DOS; 1400 Independence Avenue, SW Government Printing Office Style Manual, as WordPerfect 7.0 or earlier for Windows 95; Washington, DC 20250-1107 required by the U.S. Department of Agriculture. Microsoft Word 6.0 or earlier for Windows 95; tel. 202-205-0891, fax 202-205-1272 Authors should use the U.S. system of weight and Rich Text format; or ASCII. Digital photos may be e-mail: [email protected] measure, with equivalent values in the metric submitted but must be at least 300 dpi and system. Try to keep titles concise and descriptive; accompanied by a high-resolution (preferably USDA Forest Service subheadings and bulleted material are useful and laser) printout for editorial review and quality Attn: Hutch Brown, Office of Communication help readability. As a general rule of clear writing, control during the printing process. Do not Mail Stop 1111 use the active voice (e.g., write, “Fire managers embed illustrations (such as maps, charts, and 1400 Independence Avenue, SW know…” and not, “It is known…”). Provide graphs) in the electronic file for the manuscript. Washington, DC 20250-1111 spellouts for all abbreviations. Consult recent Instead, submit each illustration at 1,200 dpi in a tel. 202-205-1028, fax 202-205-0885 issues (on the World Wide Web at ) for such as EPS, TIFF, or JPEG, accompanied by a placement of the author’s name, title, agency high-resolution (preferably laser) printout. For Mailing Disks. Do not mail disks with electronic affiliation, and location, as well as for style of charts and graphs, include the data needed to files to the above addresses, because mail will be paragraph headings and references. reconstruct them. irradiated and the disks could be rendered inoperable. Send electronic files by e-mail or by Tables. Tables should be logical and understand- Release Authorization. Non-Federal Govern- courier service to: able without reading the text. Include tables at ment authors must sign a release to allow their the end of the manuscript. work to be in the public domain and on the USDA Forest Service World Wide Web. In addition, all photos and Attn: Hutch Brown, 2CEN Yates Photos and Illustrations. Figures, illustrations, illustrations require a written release by the 201 14th Street, SW overhead transparencies (originals are prefer- photographer or illustrator. The author, photo, Washington, DC 20024 able), and clear photographs (color slides or and illustration release forms are available from glossy color prints are preferable) are often General Manager April Baily. If you have questions about a submission, please essential to the understanding of articles. Clearly contact the managing editor, Hutch Brown.

CONTRIBUTORS WANTED We need your fire-related articles and photographs for Fire Management Today! Feature articles should be up to about 2,000 words in length. We also need short items of up to 200 words. Subjects of articles published in Fire Management Today include:

Aviation Firefighting experiences Communication Incident management Cooperation Information management (including systems) Ecosystem management Personnel Equipment/Technology Planning (including budgeting) Fire behavior Preparedness Fire ecology Prevention/Education Fire effects Safety Fire history Suppression Fire science Training Fire use (including prescribed fire) Weather Fuels management Wildland–urban interface

To help prepare your submission, see “Guidelines for Contributors” in this issue.

50 Fire Management Today PHOTO CONTEST ANNOUNCEMENT Fire Management Today invites you Rules Service (see sample statement to submit your best fire-related • The contest is open to everyone. below). Include your full name, photos to be judged in our annual You may submit an unlimited agency or institutional affiliation competition. Judging begins after number of entries from any place (if any), address, and telephone the first Friday in March of each or time; but for each photo, you number. year. must indicate only one competi- • Photos are be eliminated from tion category. To ensure fair competition if they have date Awards evaluation, we reserve the right stamps; show unsafe firefighting All contestants will receive a CD– to change the competition practices (unless that is their ROM with all photos not eliminated category for your photo. express purpose); or are of low from competition. Winning photos • Each photo must be an original technical quality (for example, will appear in a future issue of Fire color slide or print. We are not have soft focus or show camera Management Today. In addition, responsible for photos lost or movement). (Duplicates— winners in each category will damaged, and photos submitted including most overlays and other receive: will not be returned (so make a composites—have soft focus and duplicate before submission). will be eliminated.) • 1st place—Camera equipment Digital photos will not be • Photos are judged by a photogra- worth $300 and a 16- by 20-inch accepted because of difficulty phy professional whose decision is framed copy of your photo. reproducing them in print. final. • 2nd place—An 11- by 14-inch •You must own the rights to the framed copy of your photo. photo, and the photo must not • 3rd place—An 8- by 10-inch have been published prior to Postmark Deadline framed copy of your photo. submission. First Friday in March • For every photo you submit, you Categories must give a detailed caption Send submissions to: •Wildland fire (including, for example, name, USDA Forest Service • Prescribed fire location, and date of the fire; Fire Management Today Photo •Wildland-urban interface fire names of any people and/or their Contest • Aerial resources job descriptions; and descriptions Attn: Hutch Brown, Office of •Ground resources of any vegetation and/or wildlife). Communication •Miscellaneous (fire effects; fire •You must complete and sign a Mail Stop 1111 weather; fire-dependent commu- statement granting rights to use 1400 Independence Avenue, SW nities or species; etc.) your photo(s) to the USDA Forest Washington, DC 20250-1111

Sample Photo Release Statement (You may copy and use this statement. It must be signed.) Enclosed is/are ______(number) slide(s) for publication by the USDA Forest Service. For each slide submitted, the contest category is indicated and a detailed caption is enclosed. I have the authority to give permission to the Forest Service to publish the enclosed photograph(s) and am aware that, if used, it or they will be in the public domain and appear on the World Wide Web.

Signature Date

Volume 62 • No. 3 • Summer 2002 51 5614

subscription(s) to Fire Management Today for $ 13.00 each per year ($ 16.25 foreign).

4/95