april 2015

by Bradley Shoemaker

I took my basic wildland fire classes while at Virginia Tech while pursuing a bachelor’s degree in forestry. The course was taught on a different schedule than normal classes to allow for longer lab time where the class dug line and performed all the skills we were being taught and also studied the concepts we were learning like fuel moisture and topographic influences. Shortly after finishing my last day of fire class, 32 students and teachers were killed on campus in the largest school shooting in US history. I spent that summer in Glacier National Park as a backcountry ranger. That summer I had the opportunity to act as a radio repeater for the IA of a small fire outside the Belly River Ranger Station. A few years later I went on a T6 engine to Florida. While there, two dozer operators were killed on the Blue Ribbon fire down the street from where I was working. My engine helped cover IA responsibilities while Florida DOF held memorials and funerals in following days. It was that day, while mopping up a small IA that should have been a now dead man’s responsibility the importance of safety really sank in.

Today, as I continue to progress through the wildland fire world, my interest in wildland fire, particularly incident management has continued to grow. Following those interests I started working on a master’s degree in all-hazard emergency management. I focused my research on the S-course and task book qualification system with the question in mind “have we managed to train a better, safer firefighter?” The task book system was implemented in late 1993, and has had 20 years to produce and educate a new generation of wildland firefighters and leaders. It seems time to take a look and decide if the system has been a success or not.

My research focused on the last 70-years, because fatality data was generally available for the period. Over that span of time a large number of changes have occurred. The 10 standard firefighting orders and 18 watch-out situations were established in the late 1950s. The fire shelter became required gear on the fireline in 1977, in the 1990s LCES were established and soon thereafter the S-course and task book system was rolled out as a training and certification system.

A number of environmental conditions also changed over this time period including use of more engines, increased use of aircraft, an increase in forest fuel loading due to 100 years of suppression, and an increase of development in the Wildland Urban Interface. Each of these individually factors has had an effect on wildland firefighting safety and has contributed to the difficult conditions firefighters face. This increase in complexity, which has occurred gradually over time, has no correlation with data about burnover fatalities. If the two were related the data should show fatalities increasing with fire complexity, followed by a decrease after implementation of a training program or change in policies. However no such relationship has been observed.

Increased fuel loading and climate change (natural or human caused) are commonly considered to increase fire severity and intensity, but conflicting research exists about whether or not fire severity and intensity are actually increasing. Research studies sometimes conflict on the various ecological roles fire plays in different ecosystems. Fire interval and severity are very different between the areas studied, based on fuels, elevation, annual moisture, and how each forest species grows and interacts with fire. Studies in different forest types should be expected to have different results.

Recent research by Donato et al. (2013) on the effects of bark beetle mortality on fuel profiles and fire intensity in the Wyoming Yellowstone Ecosystem suggested that infestation areas “reduced fire potentials in post-outbreak stands, particularly for crown fire after the red stage (1-3 years old), but fuel loading of coarse fuels in silver stands (25-30 years old) may increase burn residence time and heat release,” not flame length or other factors we would consider such as erratic fire behavior.

Research by Hanson & Augustine (2013) notes that fire severity is not increasing in the Sierra Nevada Mountains of California, and Odion et. al. (2004) states that “fuel buildup in the absence of fire did not cause increased fire severity as hypothesized. Instead, fuel that is receptive to combustion may decrease in the long absence of fire in the closed forests of our study area, which will favor the fire regime that has maintained these forests.”

On the other side, detailed historical study shows that researchers were publishing papers stating that not all wildland fires should be suppressed aggressively as early as 1925 (Spearhawk, 1925 & Hornby, 1926). Similar studies continue today and include Walsh (2013) which states that “rising temperatures and drier woodlands are leading to a longer burning season and a significant increase in forest fires” while the US General Accounting Office (1999) states that “The most extensive and serious problem related to health of national forests in the interior West is the over-accumulation of vegetation, which has caused an increasing number of large, intense, uncontrollable and catastrophically destructive wildfires.”

Other recent research published by Moritz Et Al. (2014) states that in dry, low-elevation forests, which historically experienced frequent low-severity fires, which maintained open forest structures of fire-resistant trees such as ponderosa pine. Timber harvesting and decades of fire suppression have lengthened fire return intervals, increased the density of smaller trees and shifted regimes of mostly low-severity fires to include more high-severity, stand-replacing fires. This shift to higher severity fires in these environments is important to note and is clear in my own observations managing ponderosa pine forests in Montana.

Many other factors continue to increase complexity across the US. According to the US Fire Administration (2002) and the International Association of Wildfire (2013). Urban development in the wildland urban interface (WUI) is increasing at an alarming rate. In the western US; 38% of new home construction is adjacent to or intermixed with the WUI and only 14% of the available WUI lands have been developed.

Training continues to adapt to the conditions facing responders and it is the hope of all who manage training systems that their efforts would result in a safer environment for themselves and the people they work with. The S-course and task book system were put in place to help bring wildland fire into a new era. Prior to the S-course and task book system the wildland world operated off a mentor-mentee program which was relatively unofficial and lacked legal credibility. The intent of the S-course and task book system was to assure that trainees were successfully performing necessary skills to perform the position they were filling, to provide validation of performance and consequently make the system credible.

Even today, while agencies operate under this system, fatalities in the wildland fire work environment are unfortunately common from everyday hazards such as heart attack, vehicle collisions or rollovers and aircraft accidents. These never receive the same media or agency attention that an entrapment or burnover incident receives due to the nature of these incidents and their potential for high consequence. While a heart attack or vehicle collision may only have the potential to affect 1-4 firefighters, entrapment and burnover has potential to affect many more as shown on the Yarnell Hill fire in 2013 which killed 19 firefighters. Fatalities in this manner tend to happen with disastrous results thereby making an average of annual fatalities more accurate than comparing numbers of single years. Large fatality incidents are generally rooted in changes in fire behavior, topography or weather conditions. These root causes are what the S-course and task book system focus on training. Leadership and fire behavior are common themes throughout all courses. Successes of the S-course and task book system may thereby be measured by a reduction in incidents and fatalities associated with these root causes.

To really take a look at whether this system is working the data was grouped into 10-year periods starting at 1993, when the system was started, and moving in both directions while averaging annual burnover fatalities. Measuring safety by fatalities is a difficult move in itself. Since the number of unsafe acts greatly overshadows the number of times we are caught doing those acts, it is a difficult to measure success or failure. It is unfortunately the information we have to work with. The 50 years prior to implementation of the S-course and task book system had 10-year averages for fatalities which ranged from 3.5 to 4.8/ year and more commonly hovered around 4.4/year. Since implementation of the task book system fatalities have averaged 3.6/year from 1994-2003 and then 3.3/year from 2004-2013 in each ten year period. While this might look like a reduction, a calculation of statistical significance shows that the two number sets are too close to show a statistically significant difference. Further information from the USFS states that fire shelters are estimated to have saved 275 firefighters since their introduction (USFS, Data Unk.). If those numbers were added to these figures, fatality numbers would have skyrocketed in recent years. Through that simple analysis we have determined no significant change has happened and the question stands glaring: what are we doing wrong, and how can we do it better?

Decreasing fatalities and entrapments requires a number of changes across the wildland fire community. The system starts, continues, and ends with training and personnel development. Truly becoming students of fire requires lifelong commitment and learning. Research (Peak Performance Center, 2013) states that information retention of passive learning styles such as lecture teaching is only about 5%, reading is 10% and audio visual retention is 20%. Information retention of an active or participatory learning method such as group discussion is 50% and practice by doing is 75%. S-courses are typically taught in lecture and audio visual presentations. This means that S-courses are not capitalizing on the best teaching methods, resulting in minimal information retention of the students. Moving the S-courses to an active or participatory teaching method would result in greater retention of information, or utilizing multiple types of teaching will result in increased information retention.

Looking at these information retention statistics suggests that a change in style of teaching and also revisiting the task book system is in order. The NWCG has already started moving towards a change in the teaching style, increasing blended and online classes. I would argue this is for the best, and for an increase in the use of blended classes. I took the crew boss and engine boss classes as blended courses the first time they were offered in my region. Allowing students to take time on their schedule before class to read the information at their speed gives students more time to absorb and learn the material. Allowing this learning to take place prior to class also shortens the amount of time needed for classroom lectures which allows more time for active discussion and learning in the actual class. The classroom portion of the blended crew boss and engine boss still had a significant portion of time spent on lecture and PowerPoint presentations but there was time to review real fatality incidents in a group setting and take part in active learning where information retention is significantly higher.

Here the questions still stand, where are we headed in the future? How are we to approach the future to reduce the number of firefighter fatalities caused by burnover?

So what is next? No single change will fix the problem. The information is clear that under the current system there has been no reduction in average annual fire fatalities from burnovers. Next year, the numbers of firefighters that will probably be killed in burnovers are the same as they have been at the beginning of each season for the last 70 years. More than one change is needed to solve this problem; one part may be moving courses to a blended system and to increase the active learning portion of classroom time. Maybe the real solution is a better final evaluation system; combining the current task book system with more focus on the mentor-mentee relationship which also offers another avenue for more active learning. These changes will need to happen at a manager, agency and local level to effect real success. The only thing I know for sure is that no one change will reduce the number of firefighters giving their lives. Whatever the changes are they have already been paid for in blood by those who lost their lives on the hot, black ground we work on.


Key Points

Historical fatality data shows no statistically significant change in average annual fatalities across the past 70 years. Between 1944-1953 the data showed an average of 0.9 incidents per year with an average of 3.5 fatalities per year. This period is defined by the Mann Gulch incident which killed 13 firefighters including 12 smokejumpers. From 2004-2013 data showed an average of 1.0 incidents per year with an average of 3.3 fatalities per year. This period is defined by the Yarnell Hilll fire which killed 19 hotshots.

Significant study into common factors of fatalities has been conducted and continues to be accurate today. Changes in fire behavior, topography and weather continue to be significant factors in fatality incidents.

Fire shelters have been mandatory since 1977 and have saved an estimated 275 firefighters in the short time they have been used. Adding these numbers to fatalities would result in a near tripling over the last 36 years compared to pre-1977 fatalities.

“In the western US, 38% of new home construction is adjacent to or intermixed with the WUI” (USFA, 2002) and only 14% of the available WUI lands have been developed (IAWF, 2013).

The link between weather, topography and fuel type is trained on but many large fatality incidents occur on responder’s home units. Examples include Mann Gulch (1949), Thirtymile Fire (2001), Esperanza Fire (2006) and the Yarnell Hill Fire (2013) which all occurred in crew’s native fuels and topography.

It may be time to include letters of recommendation and an interview at the local or regional level for certain positions such as Single Resource Boss, Strike Team Leader or Division Supervisor, similar to the process in professional structural organizations.

The current S-courses are not taking advantage of learning techniques which result in the highest information retention. Wildland fire is a field based job, training and education should take place in a field environment as much as possible.

Local resources are much more likely to be involved in large fatality incidents (4 or more fatalities). In the last 70 years there have been 18 incidents with 4 or more fatalities. 17 of these incidents and 136 fatalities were local crews, 1 incident and 14 fatalities were crews operating out of their home region.

Burnover Incidents with greater than 4 fatalities

  • 2013 Yarnell Hill Fire- 19, Local Hotshot Crew
  • 2006 Esperanza Fire-5 Local Engine Crew
  • 2001 Thirtymile Fire- 4 Local Hand Crew
  • 1994 South Canyon Fire- 14 (9 Out of Region Hotshots, 3 Out of Region Smokejumpers, 2 Helitack)
  • 1990 Dude Fire- 6 Local Hand Crew
  • 1979 Spanish Ranch Fire- 4 Local, Various Agencies
  • 1977 Bass River Fire- 4 Local VFD
  • 1971 Romero Fire- 4 Local USFS employees
  • 1968 Canyon Inn Fire- 8 Local Fire Department Crew
  • 1966 Loop Fire- 12 Local Hotshot Crew
  • 1962 Timber Lodge Fire- 4 Local USFS Engine
  • 1959 Decker Fire- 6 Local (3 from local Hotshots and engine crew, District Ranger and 2 local Department)
  • 1956 Inaja Fire- 11 Local (3 Local FS and 8 Local Prison Crew)
  • 1955 Hacienda CA-6 Local (1 Local LAFD Captain and 5 local juvenile probation crew members)
  • 1953 Rattlesnake Fire- 15 Local (1 Local USFS Emp. 14 Local Residents)
  • 1950 Pilitas No 1 Fire- 4 Local (1 CalFire, 3military)
  • 1949 Mann Gulch Fire- 13 Local Smokejumpers
  • 1943 Barrett-Cottonwood Morena Fire- 11 Local (9 marines, 2 army, 1 civilian)


For a list of supporting references, see the online version of this article at wildfiremagazine.org.

About the author

Brad Shoemaker is an Area Forester for the Montana Department of Natural Resources and Conservation (DNRC). He recently completed a master’s degree in all-hazard emergency management and researched the S-course and task book qualification system (it has been 20 years since implementation), seeking to answer the question: Have we produced a more qualified/ knowledgeable firefighter, as measured by the average annual firefighter fatalities caused by burnover.