Deconstructing Rim Fire Retardant Use


The Forest Service has posted its 2013 aerial fire retardant statistics. Over 12 million gallons dropped, a 50% increase over last decade’s average annual amount (page 220 of the Forest Service’s aerial retardant FEIS summarizes use for the 2000-2010 decade).

As usual, Region 5 (California) led the nation, accounting for over 7 million gallons. And the Stanislaus NF led Region 5 with over 2 million gallons dumped. Ahh, the good old Rim Fire, of course.

So what did retardant accomplish at the Rim Fire?

Obviously, retardant didn’t achieve the initial attack objective of keeping the fire below 300 acres in size. But not for lack of effort. As one informed commentator pointed out over at “over 32,000 gallons of retardant were applied between detection (approx 1540) and dark on the first day, 65,000+ the second day and (largely due to visibility issues) 23,000 the third day.” Within 20 minutes of first detection the incident commander had decided that the steep, dry, remote canyon was too dangerous to insert boots on the ground (“CALFIRE helitack pilots refused to insert crews at the origin for lack of safety zones”), without which retardant is largely ineffective. With that sensible decision to put firefighter safety first, it’s a head scratcher why the retardant air attack continued.

But, continue it did. Not just during the next three days, but as “the fire grew, literally, exponentially for 6 days. That’s doubling in size every day. 55,000 acre runs.” Did retardant use slow that doubling, stop those runs? Of course not. The prevailing winds blew the fire across the Stanislaus, into Yosemite, until the fire ran out of fuel on barren granite hard surface of the Sierra mountains.

Question for readers. How would the Rim Fire’s outcome been different had retardant not been used?


  1. The Rim Fire might have burned as much without the retardants, but there are plenty of fires where retardants were successful. I have seen some first hand and had the red-stained fire shirts to prove it. Gallons applied is one measure. Does anyone have stats on the effectiveness of retardants?

    • Steve,

      “Does anyone have stats on the effectiveness of retardants?” Excellent question.

      The word “stats” implies empirical measures of effectiveness. What should those measures be? In other words, when fighting wildland fire, what are we trying to accomplish?

      The Forest Service measures firefighting effectiveness by the rate of initial attack success, defined quantitatively as the proportion of ignitions kept below 300 acres. Simple correlation analysis of Forest Service data shows that retardant use is only poorly correlated with initial attack success rate, and negatively at that. The more retardant used on a national forest, the lower the initial attack success rate. That negative correlation holds true at the national and within-region scales.

      The Forest Service does not dispute (see pages 460-462) this weak, negative correlation, but points out, accurately, that correlation alone is not the same as a randomized controlled experiment. The FS has not done that experiment; in fact, it refuses to do so, preferring to rely upon laboratory studies and anecdotes. As I’ve written before, if the FS’s justification for using retardant were the basis for medical practice, we would still be bleeding patients because, after all, some patients get better when you bleed them.

      • Andy, In my reading of pages 460-462, the USFS doesn’t say they do not dispute “this weak, negative correlation,” but instead say the correlation is not yet clearly or properly demonstrated:

        “To truly understand the influence of fire retardant on initial attack success and ultimate fire size, one would have to compare outcomes under similar fire behavior and fire retardant application circumstances. Simply put, what is the comparative fire size when fire retardant is used or not used under the same or similar conditions? Only a comparison of fire retardant use and non-use under similar circumstances can allow any valid conclusions to be drawn as to effectiveness of fire retardant. However, as noted, it is extremely difficult to accomplish this kind of controlled experimentation given the high degree of variability among wildfire incidents and the limited circumstances where fire can be allowed to burn unchecked for purely experimental purposes.”

        They go on to say that better analysis is required:

        “In the absence of controlled scientific experiments that could be used to support a valid statistical analysis, we must look to other means to evaluate the effectiveness of fire retardant. Laboratory experiments show that, under controlled conditions, fire retardant lowers fire intensity and slows the rate of spread in certain fuel types and fire behavior conditions. In addition, data show that long-term retardant applications show a reduction in fire spread and intensity of about 39 to 45 percent, when compared to water, when the fuels are still wet from application. When the water has evaporated off, the fuels treated with retardant still show a reduction in spread and intensity of 0.53 to 0.57, or a reduction of greater than 50 percent compared to untreated fuels. Data for water-treated fuels show no reduction in spread or intensity (0 reduction factor) after they are dried (Appendix O). From these experiments, it is reasonable to conclude that similar results are obtained under field conditions, where fire retardant can be effectively applied.”

        And present anecdotal evidence from fire managers:

        “Wildland firefighters who have experienced hundreds of fires under similar fire behavior conditions and observed fire behavior with and without the use of fire retardant have attested to the conclusion that the use of fire retardant reduces fire intensity and rate of spread, allowing earlier control and containment, and reducing overall fire size (Appendix O). In the absence of an ability to conduct controlled, scientific experimentation, the observations of these experienced firefighters provides a reasonable basis to conclude that fire retardant is an effective firefighting tool, which can be used to reduce the size of wildland fires and resulting damage to natural resources and human improvements.”

        See Appendix O, “Fire Professionals Comments on Retardant Effectiveness Summary”

        • Steve, the FS does not dispute that it has failed to establish any statistically significant correlation between retardant use and initial attack success or average fire size. None exists. The FS agrees that none exists. The burden of proof is on those who promote retardant to demonstrate its effectiveness.

          To meet that burden, the FS relies upon anecdotal evidence. Let’s take a closer look at how those anecdotes were gathered. The FS sent a questionnaire to a bunch of firefighters, 50 of whom responded, asking for their stories of successful retardant use (“Following are some questions the team has crafted to obtain responses from fire professionals to provide additional citations relative to the successful use of fire retardant on wild land fires” (emphasis added)). The FS did not ask for unsuccessful stories or for a balanced appraisal — it requested only success stories. Then the FS high-graded the responses for its EIS, eliminating those that reported failures (these folks didn’t read the directions carefully).

          Here, for example, is a firefighter’s response that did not make it into the FS’s EIS:

          “Overall, the vast majority of retardant drops I’ve seen (and often personally requested) were either ineffective or turned out (after the fact) to have not been needed (rain, wind change, etc.). However I believe this is because the vast majority of retardant drops I’ve seen and/or requested were in locations where in hindsight I should have known the probability of success was quite low.
          In my opinion, the federal wildland fire organization could substantially reduce the amount of retardant used if we would change how we teach and mentor aerial firefighters. In my experience ATGS training and mentoring is biased towards the aggressive use of retardant. I believe this is because expert ATGSs may have the most experience personally witnessing the consequences of what happens when retardant is not available in time or misses the target. These consequences become tangible as the loss of homes (not just structures), escaped fires that could have been caught, and on occasion injured firefighters. In contrast, the consequences of “wasted” retardant is effectively inconsequential and certainly much less memorable. In my opinion ATGS training would be enhanced with more integration and reliance on fire behavior information and particularly fire behavior predictions. This integration would give our aerial firefighters a better ability to predict and know when retardant would be ineffective.”

    • Sharon, as far as I know an after action review is not standard practice; only when something goes wrong, e.g., a fatality. Thankfully, there were no fatalities associated with fighting the Rim fire.

      As to your question, I don’t know what “using retardant correctly” means. If a cancer drug doesn’t make a difference when administered according to label directions, can its use ever be “correct?”

      • I think that this is pretty much the same as other disciplines. Practitioners know that retardants work from their own experience. If there is no “science” it’s because none has been funded. That does not make practitioner experience invalid.

        As I said once giving a talk in DC, if you want to know how safe this neighborhood is, would you ask a person who lives there or consult a computer model?

  2. Of course, wildfire suppression activities are, above all, a political tool to show the public “they are doing their best”, with big, shiny and new fire engines, giant 4-engine jets and heavy-lift helicopters. Yes, it is MUCH easier to throw money at a problem’s symptoms but, much more difficult to deal with the real problems in our forests (and our land management Agencies).

  3. NAMP, NATOPS I’m familiar with Naval Aviation Maintenance Program and Naval Aviation Training and Operation Procedures Standards. I would think Forest Service and Fire Fighters would have similar guides for data collection, reduction and application of lessons learned.
    I get frustrated when local fires are attributed to “red tailed” hawks. Hawks are blamed because a carcass was found. The implication is hawks are an act of God and nothing can be done about hawks. Do hawks prefer to start fires along particular electrical lines? Did the hawk really cause the fire? What do the logs at power generating station say about intermittent faults? What do the after action reports say regarding delay due to mapping inconsistencies? Weather anomalies? If the electrical generator or transmission has fault logs, go they indicate where the fault happened?
    The quoted report (Rim Fire EIS) seems to concentrate on environmental issues. USC 40. Maybe I missed it, but I didn’t read the question on how effective retardant, or even water drops are. I didn’t see any comment regarding after incident requirements to assess goodness of tactics or identification of contributing factors or suggestions of how to do it better next time.
    All of us know more than any of us. Are after action reports in the public domain? Or are they privileged with limited distribution?

  4. Question for readers. How would the Rim Fire’s outcome been different had retardant not been used?

    Fire outcome might not have been much different. Not using it would probably have made a big difference in the outcome (futures) of the Incident Commander, Fire Staff Officer, Forest Supervisor and so on and so on….gotta try something because you can’t throw your hands up and say “this fire is going to burn all the way into the goat rocks, so let’s do nothing”

    • JZ hits the nail on the head: “gotta try something because you can’t throw your hands up and say ‘this fire is going to burn all the way into the goat rocks, so let’s do nothing.’”

      A corollary question, why isn’t it okay to do nothing? When it comes to fighting disease, when “nothing” is known to work, we “do nothing” much of the time. But when it comes to fighting fire, we do everything even when nothing is known to work. Are homes more important than lives? Or is the “science” of firefighting less rigorous than the medical sciences?

      • Andy, I think it’s remarkably similar. People want treatment when they are hurting and doctors feel pressured to prescribe meds. Check out the “pressure to prescribe” article here.
        And if you go to mental health, well there’s drugs, psychotherapy, other counseling, none of which work every time. (same pdf, piece on evidence-based mental health..

        “Despite these undoubted advances, however, a con­siderable gap remains between research and practice. For example, important variations exist in the treatment of depression, in the use of electroconvulsive therapy, and in the use of stimulant medication for attention deficit hyperactivity disorder. In mental health nursing the recent increase in the amount of published research has rarely been reflected by
        changes in practice. In clinical psychology it has been asserted that “in clinical practice empirically sup­ported methods are routinely ignored in favour of
        intuition and clinical experience.” Moreover, the pub­lic perception of mental health services has not kept up with advances in research and practice. Others have
        argued that mental health policy has usually been influenced more by political values than evidence.”

        Some practitioners in the medical biz tell me that since studies are funded by companies making drugs, non-drug treatments don’t get the same kind of study. So of course these can’t be shown to be effective (empirically based), because there’s no study. So they rely on their own experience.. days and years of observations and talking with their peers, training, etc.

        Another similarity to fire is that Science” takes bucks and what gets studies is a function of who has them.

  5. I am very sure that a lot of drops were put into the area within this aerial photo.,-120.1197033,7197m/data=!3m1!1e3?hl=en

    The fire was ignited near Lumsden Bridge, near the right edge of the picture. The houses near the edge of the canyon are a part of the Pine Mountain Lake housing area. THIS is the main reason for retardant drops, and the fight took at least three days before they were successful at fully protecting every home. Without tankers, I’d guess that hundreds of homes would have burned. Maybe 1000’s of structures, too.

  6. No one has mentioned the Station Fire, where tankers weren’t used early in the fire. There were a lot of people pointing at that place and time as where the Station Fire was lost.

    Of course, most of us knows that retardant, alone, does not put fires out. At times, the combination of pink goo and boots on the ground is what is needed to “hook” the fire. Without retardant, crews are often blocked from cutting line, for safety reasons. Clearly, the Rim Fire wasn’t going to be held to a few hundred acres in the bottom of a canyon with EXTREME terrain, even with retardant. In the area of Pine Mountain Lake, I’m VERY sure that a wide line of pink allowed firefighters to safely engage the fire front, instead of the indirect tactics they had to use up the canyon.

    For some, I’m sure that the costs of using retardant were far, far outweighed by the political shitstorm that would occur had they not used it. For those who don’t know the details of fire retardant, water, alone, does not penetrate fuels very well, without a “wetting agent”. A similar substance is added to the tanks of fire engines, to keep the water from merely running off bone dry fuels. How come Andy doesn’t complain about the “wet water” that is added to thousands of fire engines, nationwide? Basically, the pink stuff is a weak mixture of fertilizer-like materials, with pink added to be seen from the air. Since the mixture penetrates the fuels, it stays put when the winter precipitation comes.

    So, it comes down to what is worse? Incinerated landscapes or pink goo?

Leave a Reply

Your email address will not be published. Required fields are marked *