How do we reduce the number of firefighter fatalities?

House in the Eiler Fire. Photo by Bill Gabbert.
David Shepard’s house survived the Eiler Fire, 40 air miles east of Redding, California. Photo by Bill Gabbert, August 6, 2014.

Our piece about trends in wildland firefighter fatalities generated discussion of what the data meant and the fact that there was a great deal of variation from year to year. I wrote a comment below the article that grew larger than I originally expected. Here it is:


With an average of 17 fatalities over the last 25 years the annual numbers will never be smooth or without spikes. If there were more than 30,000 deaths each year, like with motor vehicles and firearms, there would be less relative variation from year to year and it would be much easier to see a trend. The wildfire environment is dynamic and volatile, but human factors may be what most influences the number of fatalities, and that is difficult to measure or predict.

We have seen some interesting discussion, on this article and others, about how to reduce the fatality rate. A large percentage of the fatalities on wildfires are caused by medical issues or accidents in vehicles and helicopters. For example in 2014 there were 10 deaths on fires, but none involved burnovers. But having said that, off the top of my head, here are a few areas that need to be emphasized in order to reduce the number of burnover fatalities:

  • Realize that firefighter safety is far more important than protecting structures or vegetation. It’s hard to step back and watch homes burn, but it’s far more painful to watch a funeral.
  • Increase the use of simulation tools such as sand tables and computers to train leaders. Try to make it as realistic as possible, but don’t keep throwing problems at the trainee until they fail. Point out mistakes, but the simulation director needs to avoid getting on a power trip. This occasionally was a problem when we used a simulator with a bank of overhead projectors and a rear-projection screen, a system that was extremely flexible.
  • Find a way to make crew resource management more effective so that crew members feel empowered. If they see something, SAY something.
  • The first things every firefighter should consider before committing to a fire suppression effort are escape routes and safety zones. After that, anchor, flank, and keep one foot in the black. Then, escapes routes and safety zones, again and again.
  • Utilize existing technology that will enable Division Supervisors, Operations Section Chiefs, and Safety Officers to know in real time, 1) where the fire is, and 2) where the firefighters are. The Holy Grail of Firefighter Safety. When you think about it, it’s crazy that we sometimes send firefighters into a dangerous environment without knowing these two very basic things. Last month Tom Harbour told me that he was very concerned that, for example, someone in Washington would be accessing the data from thousands of miles away and order that a firefighter move 20 feet to the left. That can be managed. Making the information available to supervisors on the ground can save lives.

What are your recommendations?

Typos, let us know HERE, and specify which article. Please read the commenting rules before you post a comment.

Author: Bill Gabbert

After working full time in wildland fire for 33 years, he continues to learn, and strives to be a Student of Fire.

14 thoughts on “How do we reduce the number of firefighter fatalities?”

  1. Bill

    Concerning entrapment’s and how to reduce them, we first need to understand why they occur. Once we really understand why entrapment’s occur then we in the fire world can look for solutions. Do we currently really understand why entrapment’s occur? It appears to me that the answer is we “kinda” do. Agencies investigate serious accidents where there are injuries or fatalities but entrapment’s are not always investigated, for a wide variety of reasons. Perhaps we in the fire world need to do some kind of review of all entrapment’s no matter if there were injuries or not, and that review might simply consist of a simple questionnaire.

    One of the problems with obtaining information on an entrapment is there is sometimes a stigma attached to the folks that “popped the shelters” or had to “run for their lives” or “lay in the rockslide”. There was one agency that used to pull folks off the fireline for even opening a fire shelter and it never mattered whether you used it or not, and then they brought in a review team of several persons to “investigate”. We need to learn lessons from these incidents not point fingers at folks, and the current trend seems to be “learning lessons” which is great.

    Perhaps to learn more about entrapment’s we could ask the folks that were entrapped what happened (or witnesses). This could be anonymous/voluntary with a list of pertinent multiple choice questions in a web based survey. The right questions need to be asked such as:

    – Did you have a paper map?
    – Did you have a GPS unit?
    – Were there problems with using either GPS or paper maps?
    – Did you understand where the fire was located?
    – Did you understand where you were located on the map?
    – Did you receive a fire weather forecast?
    – Did you understand the weather forecast?
    – Did you receive a fire behavior forecast?
    – Did you understand the fire behavior forecast?
    – Was the fire behavior and fire location being discussed on the radio?
    – How could things have been done better?
    …and so on and so forth. At the end you leave space for comments.

    A central data base should then be created so that we can really analyze trends and why things happen like they do.

    There are some great web based customer survey tools out there that are cheap to use, and questionnaires can be quickly created.

  2. Bill

    I endorse your ideas. Concerning reducing entrapments; there are many good ideas out there but it is not readily apparent what initiatives wildland fire agencies are engaging in, or even considering. What really needs to be done is prioritize the steps that can be taken to reduce entrapments and focus on those. It’s my opinion that the following initiatives need to be promoted.

    – Establish US National Grid (USNG) as the “language of location” for wildland fire operations. Latitude/longitude is not a good option and has several weaknesses. USNG is easy to use and its a flexible/adaptable system. Require fire crews to report locations, back to communications or dispatch, using USNG at a minimum of two times a day. Require Divisions to report fire behavior and fire movement using USNG at a minimum of two times a day (or as needed for safe operations). If we know where fire crews are located and where the fire is located (in real time) that is a huge step in keeping all folks on the fire informed and will help in reducing entrapments. Everyone on the fire including drivers, situations, planning, field observers, fire crews, Division Supervisors etc. need to have this information available. Currently if one needs an update on fire behavior or crew locations (real time) you have to radio the appropriate Division Supervisor, which is a system designed for failure (Yarnell Hill comes to mind).

    – Establish protocols for reporting fire behavior and fire location (USNG). Generally there is very little talk over two-way radios during wildland fire operations concerning fire behavior. That’s because there are no standards or requirements for reporting fire behavior over the radio (or cell phone). We don’t train our fire folks on how to do this. We need standards and requirements and crew leaders need to practice this in simulations.

    – Use of simulators. There are sand box tables and some really great high tech. tools out there. Crew/team leaders need to practice and train on simulators, and go through refresher training say every three years. I am sure some agencies are doing this but it needs to become universal.

    – A paper map with a USNG grid laid over it should be mandatory and every crew leader needs one in hand before they hit the fireline. Not everyone carries a GPS or smart phone and those systems can fail for a whole variety of reasons.

    – Move the fire behavior analyst (FBAN) back to command and general staff. The FBAN position was demoted a decade or more ago to working under the Situations Unit Leader, which makes no sense.

    Thanks for starting this discussion.


  3. I see too many chiefs that still eat as if they were 18 years old. Find a way for the chiefs and the 50 year-old volunteers to maintain the physical fitness they had when they were young and the medical deaths will decline.

  4. My recommendation? You said it already Bill, ” Realize that firefighter safety is far more important than protecting structures or vegetation. It’s hard to step back and watch homes burn, but it’s far more painful to watch a funeral.”

  5. Place firefighter safety as number one priority…

    Ahead of homes, trees, you name it…


  6. Pretty much agree with all you say Bill. In answer to your question, In addition, I agree with other commenters that part of the answer lies in engaging fewer fires and/or engaging some fires less aggressively. I’d also say focus (extensively) on entrapment avoidance. Entrapment survival training is a necessary evil, but our focus should be on entrapment avoidance.

  7. I find it interesting that you wrote this article after the previous quote from Mr.Finney. We definitely need more fire on the landscape and we need to be smarter in determining which fire we choose to engage. There are obvious areas where rapid IA is the answer (Most of S Cal, front range CO), but the vast majority of fires should be managed less aggressively.
    This is not a new concept, but why are will still talking about it? For one thing, aggressive IA is a much easier decision to make. It has less immediate risk to the line officer and it is ingrained in firefighters. What do we tell all firefighters? If you don’t know what else to do, anchor and flank.
    Deciding to manage a fire has many hurdles. The Forest Service requires regional approval to manage a fire for resource benefit. Also, in the near future the Forest Service is also planning on discontinuing the practice of counting wildfire acres towards fuels targets. Outside of just the Forest Service, an entire (well-funded) industry has grown around suppressing fires. There is little to no incentives, especially short term, for line officers to choose this route. The biggest hurdle, though, in my mind is the firefighters on the ground. Its very easy to talk about the prudent decision to make in the winter at training, but when you are there and a small fire is becoming a big fire, its a different ball game. To overcome this we need fire managers to pre-plan. In the off-seasons we need to set up maps and other decision tools to help the initial responders. We need to talk with our cooperators that may not look on fire the same as us. We also need to remind people that a point protection or modified suppression fire is not “Let-burn.” Most of all though, we need to get the message out. The most successful ad campaign in the history of our nation has lead to a fear of all fire. We need a new “Smokey.” A spokesperson that can increase the acceptance of fire.
    One last thought, At an AAR for R1 fire mangers this last fall, a hotshot superintendent had a great quote. “It’s too bad that it takes a year like this (2015), for us to fight fire the way we should.” A simple concept, fight fire on our terms not the fire’s.


      If Alaska has had very little fire control since the inception of the out by 10 AM policy, why does it still have extreme fire behavior? Simply put, if this was allowed to rage through much of California, many civilians would die. The driving forces of change in the California landscape are the 40 million residents; a return to pre-european / asian american settlement is impossible.

      I disagree with the idea that many of these deaths are due to overly aggressive tactics. Don’t confuse stupid tactics with aggressive fire attack. Every day that a fire is allowed to burn in California exposes countless civilians and firefighters to danger; at some point Murphy will speak.

      Put the fire out while it is small, and the problem goes away.

      1. I agree with your first paragraph but I disagree with your statement ” Put the fire out while it is small, and the problem goes away” I think you are just transferring a problem to a later date.

      2. It has “extreme” fire behavior because that’s the way black spruce burns, with a fire regine more like northern Rockies lodgepole than California mixed conifer. Black spruce fires may be more active now due to warmer and drier winters and earlier fire seasons due to climate change, but large fires are not new to Alaska, as Steve Pyne notes in Between Two Fires, especially the Hotline chapter.

        The return to “natural” conditions does not mean a return to static1890 stand structures, or some other presettlement era. It does mean that under those historic conditions, patterns established by older fires limited the spread and behavior of new ones, especially in mixed conifer. Fire suppression has eroded that mosaic, and that’s an important reason why fires have become larger, more intense, and expensive.

        Therefore, prescribed and natural fires can reestablish that mosaic, if done on a sufficient scale. That’s the characteristic of the presettlement condition that needs to be recreated; we can’t do much about weather or topography, but we can do something about fuels. If we don’t make time for large-scale fuels management, to give firefighters a safer work environment, it gives the lie to the oft-repeated statement that firefighter safety is our highest priority, by sending them out into increasingly dangerous fuels.

        Finally, putting out the fire when it is small sounds good, but clearly isn’t working, simply because 100% IA success is never going to happen, and the fires that escape feed off of all those untreated fuels. Unless as much effort is put into fire restoration as fire suppression, we’re just chasing our tail.

        1. “Finally, putting out the fire when it is small sounds good, but clearly isn’t working, simply because 100% IA success is never going to happen, and the fires that escape feed off of all those untreated fuels. Unless as much effort is put into fire restoration as fire suppression, we’re just chasing our tail.”

          It works well in California. Close to 95% of the fires are contained at 10 acres or less, and have been for years. That won’t change – unless the land management agencies persist with WFU policies. Today, fires that escape are due to low fuel moistures (regardless of the model), high winds or unstable atmospheres coupled with low humidities.

          ” Fire suppression has eroded that mosaic, and that’s an important reason why fires have become larger, more intense, and expensive.”

          What fire suppression existed prior to the Big Burn or the Peshtigo fires? How are fires today, after decades of fire suppression, more intense than the Big Burn? If prior to fire suppression, there was a fuel mosaic that prevented large fires from happening, how did the Big Burn occur?

          What proof can you offer that the fire mosaic will truly create a docile fire environment? Close to 40 million people live in California. If several 100k acre fires are allowed to burn to recreate the natural fire environment simultaneously, and an off shore wind event strikes California, what will be the result? Why risk human life for a theory?

          1. It would be useful if you stopped mixing fire regimes; the Big Burn of the northern rockies and the Peshtigo up near the Lake states occurred in different vegetation types than occur in California, and were also fueled by logging slash, as any book on those events makes clear. There is ample research on the fire history of the mixed conifer type in California that the large stand replacing fires that are now occuring are unnatural; see literature by Biswell, Van Wagtendonk, or Scott Stephens. Also the results of allowing fires to burn in Yosemite, which is seeing a restoration of the historic mosaic which in turn affects fire spread. No one is suggesting that fires be allowed to burn in WUI, but many do suggest that fuels work plus suppression in that area is much more effective than reliance on just suppression, which works until it doesn’t. Is 95% IA success really that great when the other 5% are the King, Valley, Rim, Rough, and such fires plowing their way through untreated fuels?

          2. Interesting. I didn’t read where any of those authors could guaranty that successive fires would be docile when burning through the mosaic of fuel. Where is that discussed? Simply put all of the fires cited explode when confronted with low relative humidity and high winds or an unstable atmosphere. You can have whatever fuel mosaic you want, but if the forces align, the fire will rip. That’s fact and not a theory. Now if the WFU is located next to populated areas, as most national forests are in California, there is the potential to have those forces kill people. That’s fact, too.

            The King and the Valley were both pushed by strong winds during their major runs. I know, I was a Div Sup on the ground. The other two I wasn’t on, but I have my suspicions.

            Every ecosystem has its magic conditions that yield explosive fire growth; the longer that a fire is allowed to burn, the better chance those forces have to come into alignment, and hurt people.

  8. I think it is easier said than done… But there has to be more fires allowed to let burn… Prescribed fires or letting natural fires burn. Well managed fuels provide FFs with a safety net, more than any fire shelter can ever provide. I think the attitude – that generally speaking fire provides a host of benefits to the ecosystems it passes through and its plant and animal inhabitants – needs to be heard by the public and has to be re-emphasized to those who manage fires.
    On a different note, I think that your upward trend in fatalities seems spot on… It would be interesting to have a numbers person dissect that more thoroughly. For example, those fatality events – are there more fatality events today than there were in the past? How are these fatality events explained by climate or other things such as proximity to structures? I thank you for bringing this subject up… I think it is worth thinking about and crunching numbers.


Comments are closed.