Above: Map, showing the approximate location of the planned prescribed fire at Devils Tower (300 acres in white) and the 56 acres (in red) that escaped beyond the planned perimeter. Image from Google Earth. Perimeters by Wildfire Today.
Yesterday after Wildfire Today made inquiries about reports that may have been completed regarding the escaped prescribed fire on May 8, 2013 at Devils Tower National Monument in Wyoming, the National Park Service released the official review of the incident.
The NPS ignited the 300-acre unit on May 7, 2013. Several spot fires occurred outside the planned perimeter that were contained the first day. But during mopup at 12:50 p.m. on May 8 fire became established again at one of the spot fire locations on the southwest side of the project. At that time the cooperating U.S. Forest Service resources present the day before had been released. Some of the remaining firefighters were concentrating on the previous day’s spot fires at another location, but most of the firefighters were attending an After Action Review of a non-injury tipover of a Utility Task Vehicle (UTV) that occurred the day before. The 56 acres burning outside the burn unit were mostly stopped at the Monument boundary, but a few acres crossed over onto private land leased by Wyoming state Senator Ogden Driskill, but no structures were damaged.
Devils Tower Rx fire, May 7, 2013. NPS photo.
In the interest of full disclosure, Devils Tower was one of the seven National Parks for which I was the Area Fire Management Officer, from 1998 until 2003.
The review pointed out several times that one of the primary issues related to the escape was that in computing the fine dead fuel moisture, it was assumed that the fire would be shaded by the smoke column. However, some of the area was not shaded, and for two hours each day on May 7 and 8, at those locations the fine dead fuel moisture dropped below the 4 to 10 percent allowed in the prescription, down to 3 percent.
Other than how the weather affected the fuel moisture, the review barely mentioned the weather conditions and the forecast. Two spot weather forecasts were issued before the escape — one at 7:57 a.m. MDT on May 7 and another at 9:29 a.m. on May 8, the day of the escape. For May 8, both forecasts predicted fairly strong northeast winds, of 7 to 15 mph and 8 to 14 mph.
The Remote Automatic Weather Station at Devils Tower is very close to the location of the prescribed fire, in a low-lying area partially sheltered by trees from winds from all directions (see map above). Northwest, north, and northeast winds are additionally partially blocked by higher ground and the Devils Tower itself. Below are the weather observations from the weather station between 17:23 on May 7 through 17:23 on May 8. They show mild sustained wind speeds, with gusts around mid-day to late afternoon on May 8 of 13 to 22 mph. If the weather station was in a more exposed location the recorded speeds would have been higher.
A report issued by the U.S. Forest Service for a prescribed fire that escaped on the Dakota Prairie National Grasslands in northern South Dakota last year cited weather as being one of the primary factors in losing control of the fire.
The Pasture 3B prescribed fire was planned at 210 acres, but strong winds on April 3, 2013 caused by the predicted passage of a cold front pushed the fire across a mowed fire line into tall grass and ultimately burned 10,679 acres, (3,519 acres federal and 7,160 acres private). The wildfire, named Pautre Fire, was stopped at 11 p.m. that night.
An article in The Rapid City Journal has more information about the impacts of the fire on the ranchers. Privately owned grazing, hay stacks, and miles of fencing were damaged or destroyed during the ranchers’ calving season. Here is an excerpt:
“Laurie Casper, 36, said the fire destroyed 95 percent of her family’s farmland, which is more than 1,000 acres.
‘We lost all of our calving pasture, we lost our summer grazing, we lost our fall grazing, we lost 100 percent of our alfalfa— which we cut for hay bales in order to feed the cattle this oncoming winter— all that’s completely gone,’ she said. ‘And there’s just just miles and miles of fences that are completely gone.’ “
The prescription in the project’s burn plan for the maximum wind speed at eye level was 15 mph and the maximum wind speed at the 20-foot level was 20 mph. One of the spot weather forecasts for the morning of the prescribed fire predicted the passage of the cold front, with winds shifting from the south at 5 to 11 mph in the morning, to northwest at 25 mph with gusts to 30 mph in the afternoon. The actual weather that day was very similar to the forecast.
About a month later, another prescribed fire not too far away, at Devils Tower National Memorial in western Wyoming, also escaped, due partially to strong winds. That report still has not been posted on the Wildland Fire Lessons Learned Center website. We were told today that was due to a glitch, and it will appear there very soon, possibly as early as today, February 4. (UPDATE, February 5, 2014: the National Park Service arranged to have the Devils Tower report posted late in the day on February 4. We wrote about it here.)
Pautre Fire. US Forest Service photo.
The report said firefighters concluded they would not do much different next time (page 12):
Firefighters did an excellent job of planning, organizing and executing this prescribed fire and adhering to the prescribed fire plan. Following the control of the escaped fire some firefighters had difficulty thinking of anything they would do differently next time. While it is true that we work in a dangerous environment with unexpected changes in weather, we strive to be a learning culture and continuously improve our ability to make decisions that evaluate risk and get work done on the ground.
Some of the issues listed by the Facilitated Learning Analysis team included:
Improved weather forecasts are needed.
Consider additional research on methods to predict effects of drought on fire behavior in grass fuel models.
The nearest remote automated weather station (RAWS) is more than 90 miles away.
The project was conducted at the critical edge of the prescription.
Consider gaming out worst case scenario “what ifs” during the planning process, and discuss with participants during the on-site briefing.
There were problems with radio communications [note from Bill: I don’t remember EVER seeing a report like this that did not cite radio communications as being an issue].
The commendations section included this:
The personnel involved in all levels of the Pasture 3B prescribed fire were motivated, worked well as a team, were adequately trained, and appropriately briefed. They had a keen awareness that this was the first burn of the year, and took numerous precautions to ensure successful completion of the prescribed fire.
We did not see any mention in the report of damaged fences, hay, or pastures.
The U.S. Forest Service has released an aviation safety report titled “FY 2013 Aviation Safety Summary” which theoretically analyses, or at least lists, accident trends. Their presumed safety goal, although we could not find in the report any goals or objectives, is to reduce accidents. We were astounded to read on page 4 a statement that was repeated in various ways on pages 8, 18, and 33:
The Forest Service did not have any accountable accidents again in FY 2013; this was the third year in a row without an accident.
That statement was backed up by these two charts, and others in the report:
At least four accidents since 2008 with a total of nine fatalities do not show up in these stats:
2008, September 1: crash of a Neptune air tanker at Reno with three fatalities;
2012, June 3: crash of a Neptune air tanker in Utah with two fatalities;
2012, June 3, crash of a Minden air tanker at Minden, Nevada (one landing gear did not lower), irreparable damage, no fatalites;
2012, July 1, crash of a MAFFS air tanker in South Dakota, four fatalities.
There may have been other accidents between 2004 and 2007 that also were not listed.
We checked with the USFS about the discrepancy and spokesperson Jennifer Jones told us that the accidents “were not included in the document because it was a U.S. Forest Service aviation safety report and the airtankers were under the operational control of other agencies when the accidents occurred, so they are not considered reportable accidents for the U.S. Forest Service.”
It turns out that if an air tanker under contract to the USFS is flying on a fire for another agency and crashes, the USFS will not include that accident in the report. However, the MAFFS air tanker crashed while making a drop on the White Draw fire on the Black Hills National Forest.
C-130 MAFFS air tanker crash, July 1, 2012. US Air Force photo
A statement in the report absolves the USFS from responsibility for accidents involving military aircraft:
Military aircraft remain under the operational control of the military even while supporting USFS operations.
Fire suppression management is under civilian control.
And later on page 29:
Second, due to the need for swift reaction to live fires, the practical supervision of executing a MAFFS mission, by default, is under civilian control.
We could not find the word “MAFFS” anywhere in the 33-page FY 2013 Aviation Safety Summary document, or any reference to the nine fatalities we listed above.
In 2012 MAFFS air tankers dropped 2.45 million gallons of retardant on fires, frequently under USFS operational control.
The 2012 landing gear failure on the Minden air tanker was not listed, the USFS said, because “the National Transportation Safety Board determined that it did not meet the definition of an accident”. But part of the definition of an “accident” in this summary report (page 3) is one “in which the aircraft receives substantial damage.”
A reasonable person would think that an aviation summary document that compiled accident statistics would at least mention that aircraft on long term exclusive use contracts to the USFS crashed and killed nine crewmembers, even if they were on temporary loan to another organization for an hour or a few days. The agency selected these aircraft and the contractors, and the fact that there were four major accidents involving their chosen aircraft and contractors deserves mention, at least to honor their service. The nine fatalities and four crashes in a five-year period is a very disturbing trend that should not be ignored. And even more so when you also consider the 2010 accident that does show up in the stats. That one may be the June 26, 2010 accident in which Neptune’s T-44 went off the end of the runway at Rocky Mountain Regional Airport (Jeffco) near Denver due to a hydraulic system failure.
If the USFS analyzed the crash trends involving their contractors, including those occurring on non-USFS fires, they might find, for example, they should reconsider the specs in the contracts, the crash history of contractors, the suitability of aircraft designed for maritime patrol in the 1950s that are then used for flying in and out of canyons under air frame stresses the engineers did not consider, and the age of the aircraft. If what you are doing is not working, and these crashes and fatalities indicate it is not, then you need to do something different. The next-generation air tanker concept is a step in the right direction, but using jet airliners to fly into canyons is a concept that needs to be proven.
At a minimum, future reports should have a separate section to list the mishaps and accidents that involve their contracted aircraft even if they are on a non-USFS fire. And, accidents that involve MAFFS air tankers working under an agreement with the USFS, and accidents that result in major damage, should be listed as reportable accidents, regardless of specific jargon used by the NTSB.
It should not make any difference, for statistical, reporting, and accident prevention purposes, if the cause of an accident is mechanical, weather, or pilot error — they all should be recorded and reported. If the objective is learning lessons and preventing future accidents. they must be tracked and remembered. Splitting hairs and using imaginative criteria for leaving out certain accidents can turn the entire accident reporting program into a farce.
Yarnell Hill Fire at 7:30 p.m. MST, June 29, 2013, approximately 21 hours before the 19 fatalities. Photo by ATGS Rory Collins, Oregon Department of Forestry.
During the two investigations by teams of people working for the Arizona Division of Occupational Safety and Health (ADOSH) trying to document what happened and why during the Yarnell Hill Fire that killed 19 firefighters on June 30, the U.S. Forest Service provided so little information that it was described as “useless” by one of the teams.
The ADOSH “Inspection Narrative” said the USFS provided redacted copies of documents produced by members of the Blue Ridge Interagency Hotshot Crew, which was working nearby during the accident. However, the documents were redacted to the point where they were “useless in ADOSH’s investigation”. The removed portions included not only the names of the Blue Ridge personnel, but all names, as well as information the USFS stated “was of a sensitive nature”.
…we were given access to all information and personnel that we requested with the exception of the employees of the USDA Forest Service. The USDA Forest Service declined the request to allow their employees to be interviewed for this investigation.
To my knowledge, this is the first time that the USFS has refused categorically to allow their employees to be interviewed following a serious accident that occurred on a fire.
We reached out to the U.S. Forest Service to ask why they provided no meaningful cooperation to the investigations. Their logic is difficult to follow and involved the Privacy Act and a distinction they tried to make between the recent ADOSH reports and the Serious Accident Investigation Team report which was released September 28. They explained that the ADOSH is a compliance and regulatory agency, while the SAIT report was a safety and accident investigation. The entire USFS statement is below:
USDA Forest Service employees are subject to a variety of laws, such as the Privacy Act of 1974, as amended, 5 USC 55a, designed to protect personal and confidential information. We are legally required to withhold certain information due to the requirements of federal law to protect privacy and confidentiality of our employees. At their request, the Forest Service did provide ADOSH some documents for their investigation and offered to provide written response to remaining questions.
For clarification, Arizona Division of Occupational Safety & Health (ADOSH) is a compliance and regulatory agency.
The Interagency Serious Accident Investigation of the Yarnell Hill Fire was, in contrast, a SAFETY and Accident investigation. While the safety investigators also looked into rule compliance the focus was concentrated on understanding why the accident happened; which may or may not be related to rule compliance.
It is interesting that previous interpretations of the Privacy Act have not resulted in “useless” information from USFS employees in serious accident investigations, although there has been a recent trend to leave out names of people that were involved, which is not necessarily a bad thing.
Refusing to allow the federal government employees to be interviewed by the ADOSH teams is a very disturbing development.
The SAIT report did not include any names nor did it list the people they interviewed, so it is difficult to determine how much if any cooperation they received from the Forest Service. That report basically said no mistakes were made, while the ADOSH reports provided much, much more detail about what happened on the fire.
If this is going to be the policy of the USFS going forward, it can severely disrupt future lessons learned inquiries, and in some cases could make them “useless”. Interfering with the process of learning of how to prevent similar fatalities does a disservice to the dead firefighters.
Cantwell-Hastings law
This ridiculous action by the Forest Service may be one of the unintended consequences of the Cantwell-Hastings legislation which became Public Law 107-203 in 2002. It was sponsored by Senator Maria Cantwell and Representative Doc Hastings, in a knee-jerk reaction to the 2001 Thirtymile Fire. The law requires that in the case of a fatality of a U.S. Forest Service employee ”due to wildfire entrapment or burnover, the Office of Inspector General (OIG) of the Department of Agriculture shall conduct an investigation of the fatality” which would be independent of any investigation conducted by the USFS.
After the trainee wildland fire investigator for the OIG finished looking at the Thirtymile fire, on January 30, 2007 the crew boss of the four firefighters that died was charged with 11 felonies, including four counts of manslaughter. The charges were later reduced to two counts of making false statements to which he pleaded guilty on August 20, 2008. He was sentenced to three years of probation and 90 days of work release.
This law may have made the USFS so fearful of criminal charges and lawsuits that they are refusing to cooperate with fire investigations.
Where do we go from here?
The military has the benefit of a law that is the opposite of the Cantwell-Hastings bill. They have the protection of 10 U.S.C. 2254(d), which states that in the case of an aircraft accident:
Use of Information in Civil Proceedings.—For purposes of any civil or criminal proceeding arising from an aircraft accident, any opinion of the accident investigators as to the cause of, or the factors contributing to, the accident set forth in the accident investigation report may not be considered as evidence in such proceeding, nor may such information be considered an admission of liability by the United States or by any person referred to in those conclusions or statements.
Senator Maria Cantwell and Representative Doc Hastings need to suck it up and admit their knee-jerk reaction to the Thirtymile fire has caused a great deal of unintended harm. In 2001 they thought their ill advised idea might enhance the safety of firefighters, but it has accomplished the reverse. Lessons learned are becoming more difficult to uncover. Mistakes are more likely to be repeated because of their legislation which became Public Law 107-203. They wanted investigations, but investigations have always occurred following serious accidents. Their legislation had zero benefits, and had far-reaching negative consequences.
Senator Cantwell and Representative Hastings should feel a moral obligation to fix the problem they created. They need to craft legislation to protect firefighters, similar to that protecting the military in 10 U.S.C. 2254(d).
Yarnell Hill Fire at 1549 June 30, 2013. Photo by Chris MacKenzie of the Granite Mountain Hotshots. The arrow represents the location where the lookout had been positioned earlier.
(Originally published at 2:46 p.m. MT December 4, 2013; updated at 8:30 p.m., December 4, 2013)
Today the Arizona Division of Occupational Safety and Health (ADOSH) proposed fines totaling $559,000 to be imposed on the Arizona State Forestry Division as a result of the fatalities on the Yarnell Hill Fire near Yarnell, Arizona. Their findings were presented to the Industrial Commission of Arizona during a 1:00 p.m. public meeting in Phoenix. The documents can be found HERE.
On June 30, 19 members of the Granite Mountain Hotshots were entrapped by rapidly spread flames from a brush fire and were killed. One member of the crew who was in a different location serving as a lookout was not injured.
Two citations were proposed, one “willful serious” with a tine of $545,000, and another that was “serious” with a fine of $14,000.
[UPDATE at 6:46 p.m. MT December 4, 2013; The commission approved the fines. The Arizona State Forestry Division has 15 days to appeal the decision.]
The willful serious citation included the following (paraphrased):
Failure to furnish a place of employment which was free from recognized hazards that were causing or likely to cause death or serious physical harm.
Implementation of suppression strategies that prioritized protection of non-defensible structures and pastureland over firefighter safety.
The employer knew the suppression was ineffective, and that the wind would push the fire toward non-defensible structures, but firefighters were not promptly removed from exposure to smoke inhalation, burns, and death.
Thirty-one members of a structure protection group charged with protecting non-defensible structures were exposed to possible smoke inhalation, burns, and death.
A lookout was exposed to the same dangers.
Approximately 30 firefighters working on an indirect fireline in Division Z were exposed to the same dangers.
The Granite Mountain Hotshots continued with suppression activities until 1642 hours on June 30 when they were entrapped by a rapidly progressing wind driven wildland fire.
The serious citation, totaling $14,000:
The employer failed to implement appropriate fire suppression plans in a timely fashion during a life-threatening transition between initial attack and extended attack.
When the fire escaped initial attack none of the following analysis procedures were implemented: Incident Complexity Analysis, Escaped Fire Situational Analysis, Wildland Fire Situation Analysis, Wildland Fire Decision Support System, or Operational Needs Assessment.
On June 29 an Incident Action Plan was not completed for the next operational period prior to transitioning to a more complex management team.
The positions of Safety Officer and Planning Section Chief were not filled on June 30.
On June 30 the Division Z Supervisor (adjacent to the Granite Mountain Hotshots’ Division) departed from his assigned position which left Division Z without supervision during ongoing fire suppression operations.
Today, in addition to the citation information, the following documents were released by the Industrial Commission of Arizona:
We will add to this article later with more details about the investigation report, but below are the conclusions reached by Wildland Fire Associates, the consultants hired by the Arizona Division of Occupational Safety and Health:
Fire behavior was extreme and exacerbated by the outflow boundary associated with the thunderstorm. The Yarnell Hill Fire continually exceeded the expectations of fire and incident managers, as well as the firefighters.
Arizona State Forestry Division failed to implement their own extended attack guidelines and procedures including an extended attack safety checklist and wildland fire decision support system with a complexity analysis.
The incident management decision process failed to recognize that the available resources and chosen administrative strategy of full suppression and associated operational tactics could not succeed. This also remained the case when the strategy changed from full suppression to a combination of point protection and full suppression.
Risk management weighs the risk associated with success against the probability and severity of failure. ASFD failed to adequately update their risk assessment when the fire escaped initial attack leading to the failure of their strategies and tactics that resulted in a life-threatening event.
****
UPDATE at 8:30 p.m. MT, December 4, 2013
We just finished reading the “Inspection Narrative” compiled by AZ OSHA, and the “Granite Mountain IHC Entrapment and Burnover Investigation” report written by Wildland Fire Associates (WFA).
The Inspection Narrative
We noticed a couple of interesting tidbits in the Inspection Narrative that we don’t remember being pointed out in the previous Serious Accident Investigation Team report which was released on September 28.
One was found on page 18. At approximately 1545 hours, one of the the Type 2 Operations Section Chiefs called the Granite Mountain Hotshots and asked if they could spare resources to assist in Yarnell. Either Marsh or GMIHC Captain Steed responded that they were committed to the black and he should contact the Blue Ridge Hotshots.
While the GMIHC said they were not available for the change in assignment, the request from the Ops Chief informed them that they were needed in Yarnell. This may have influenced their decision to move toward the ranch, perhaps with the ultimate goal of assisting in the town. We could not find a mention of this in the WFA report.
One other item in the Narrative (on page 17) we noticed was a disagreement and/or confusion about the break between Divisions A and Z. The Division Z Supervisor didn’t arrive on the fire line until 1 p.m. on June 30. I in addition to the Division break fiasco, he was not clear at all about what tactics in the area could be successful. He left the fire line to head to the Incident Command Post and did not return. Parts of this were also mentioned in the WFA report. The problem with filling the Division Z position was mentioned in the citation.
Below are some quotes from the WFA report:
P. 15: At 1558, ATGS abruptly leaves the fire and goes to Deer Valley. He turned air tactical operations over to ASM2 who was busy dealing with lead plane duties at the time. ASM2 got a very brief update from ATGS that did not include division breaks locations and the location of the on-the-ground firefighters. ASM2 had been ordered as a lead plane because ATGS functions were covered. Continue reading
The Associated Press and the Prescott Daily Courier are reporting that the Arizona version of the federal Occupational Health and Safety Administration will disclose their version of an investigation into the deaths of 19 members of the Granite Mountain Hotshots on Wednesday. The firefighters were entrapped and killed while suppressing the Yarnell Hill Fire in Arizona on June 30. Their requirements are that the report must be complete no later than six months after their investigation was announced, which would make it due no later than the first part of January, 2014.
The Arizona Division of Occupational Safety and Health will formally present their findings to the Industrial Commission of Arizona, their parent agency, at a 1 p.m. public meeting Wednesday, December 4, in the 1st Floor Auditorium, 800 West Washington, Phoenix, Arizona.
Both media organizations said the agency will recommend citations and penalties against state organizations. An agenda that has been released for the meeting said “Discussion & Action of OSHA Proposed Citations & Penalties. Arizona State Forestry Division, State of Arizona”, but does not mention the Yarnell Hill Fire.
The judgments and decisions of the incident management organizations managing this fire were reasonable. Firefighters performed within their scope of duty, as defined by their respective organizations. The Team found no indication of negligence, reckless actions, or violations of policy or protocol.
The objective as stated in the document was to “understand the rationale for the actions/decisions involved in the incident and then, if possible, to learn from them”.
Frequently at Wildfire Today we will write a summary and then our analysis of serious accident reports, but this particular document is very different from the traditional report. It adopts the new paradigm of leaving out conclusions and recommendations, a process that began to be etched into stone in August when the Serious Accident Investigation Guide was revised. This Learning Review claims that “conclusions can sometimes close the door on learning”. I would say on the other hand that they can more frequently open the door to an enhanced safety environment for firefighters. People can sometimes be hit by meteorites, but not often.
And like virtually every research paper, most of the recommendations are for additional studies, ensuring continued employment for academics and researchers.
Call me old school, but this document appears to be more useful for human behavior researchers than firefighters. How did we get to the point where language such as this is used repeatedly in a U.S. Forest Service funded official report about a wildland fire?
“Typical mission flow”
“Synthesis, analysis and sensemaking”
“Margin of maneuver”
“Sensemaking team”
“Single Loop vs. Double Loop Learning”
“Hoberman Sphere”
“Pressures and filters”
“Mind maps”
“Auditory signal”
“Signal detection”
The Learning Review does suggest that two additional products be prepared, one for “the field” and another for “the organization”. Maybe the field document, if produced, will be more useful for firefighters. And presumably the organization version will have conclusions and recommendations that will remain secret if the guidelines revised in August are followed.
I am not sure why the U.S. Forest Service paid the 22 people, plus multiple focus groups, to produce this study if they did not receive for their investment products usable by the field or the organization.
But I am old school when it comes to opportunities for learning lessons.
Little Bear Fire, burning operation on 532 Road, June 13, 2012, Photo by Kari Greer/USFS
There can be a variety of narrative reports written about a wildfire, including 24-hour and 72-hour, After Action Review, Individual Fire Review, Fire and Aviation Safety Team Review, and Lessons Learned Review. But a type that was new to us has been released about the 2012 Little Bear Fire in New Mexico. It is titled “Little Bear Fire Summary Report” which is a vague title for a report that analyzed perceptions — how the fire was viewed by fire managers and local residents.
It could be categorized more as research than a conventional report on a fire. A team of seven social scientists from North Carolina State University, the U.S. Forest Service, University of Colorado, and Ohio State University conducted interviews of stakeholders, with a focus on perceptions of the event itself — communication, evacuation, and wildfire preparedness. They talked with community members, local organizations, and federal agency personnel.
The Little Bear Fire started on June 4, 2012 northwest of Ruidoso, New Mexico and was contained at four acres with a fireline around it during the first five days. On the afternoon of the fifth day a wind event blew embers from a torching tree outside the fireline causing the fire to eventually burn 242 houses and 44,330 acres.
The management of the fire has been a magnet for criticism from politicians, residents, and others. But this new report does not explore in detail the tactics, strategy, or suppression decisions that were made — it concentrates on how the fire was perceived.
“Gordie”, a Wildfire Today reader, in commenting on how the U.S. Forest Service expends time and energy on designating “Honorary Forest Rangers” such as Arnold Schwarzenner and Betty White, wrote in part:
…A public official in Washington state once said (paraphrased): “What we are perceived to do may be more important in our customer’s eyes than our actual accomplishments.” A horrible truth, but for the great unknowing masses, looking good is more important [to] taxpayers than actual functionality.
Applying Gordie’s analogy, the USFS ordered research to determine if they are “looking good”.
We will get to the report’s findings, but first there was one fact about the management of the fire that was new to us. On June 9, the day after the four-acre fire blew up, the New Mexico Governor ordered a second Type 1 Incident Management Team. This decision was made without consulting the existing Type 1 team, which learned of the order hours after it had taken command of the fire. When this was discovered, the second team was assigned to stage at Albuquerque, rather than continue to the fire.
We wrote on September 27 that the media might find it difficult to develop story lines or come up with coherent, introspective, meaningful coverage about yesterday’s release of the Yarnell Hill Fire report if it did not include causes and recommendations. The report provided more information about the deaths near Yarnell, Arizona on June 30 of 19 firefighters, members of the Granite Mountain Hotshots.
The document did not identify causes or contributing factors like we have seen in other fatality reports. It had some conclusions and recommendations, but they were fairly mild and did little toward pointing fingers at specific acts or omissions that caused the accident.
This made it difficult for reporters who in most cases know little about wildland fire to summarize the report in a short news article. Facts about outflow winds, rate of spread, and staying in the black, meant little.
Many of them looked for something that was easy to understand or was measurable, like “radio problems” which was in the headline of some stories, or the number of air tankers. A radio programming mistake, leaving out the tone guard on frequencies, at first made it impossible to use those channels for communication. Some radio systems require not only that the correct frequency be programmed, but that a brief audible tone be added. If the tone is not included when transmitting, the receiving radio will ignore the transmission. The report said crews developed “workarounds so they could communicate using their radios”. Apparently this problem was solved or at least partially mitigated. The report did not elaborate on the “workaround”.
Here are some of the headlines and the first points mentioned about the causes, in articles we found about the release of the report:
Washington Post: In the first paragraph mentions an “unpredictable desert thunderstorm” and “confusing radio communications”.
Huffington Post’s headline: “Yarnell Fire Radio Problems Cited In Deaths Of 19 Firefighters, According To New Investigation”.
Associated Press headline at Firehouse.com: “Video: Yarnell Hill Fire Report Indicates Radio Issues”.
AZCentral.com, at the top of the article is a short video of lead investigator Jim Karels mentioning the radio programming issue.
LA Times cites “problems with radio communication”.
Associated Press at ABC15.com, in the first paragraph, said the report “…cites poor communication between the men and support staff, and reveals that an airtanker carrying flame retardant was hovering overhead as the men died.” (I would like to see a video of that “hovering” air tanker, which was a DC-10.)
NPR Blog cited “weather reports that may have been misunderstood [and] radio communications that the investigators deem ‘challenging.’ “
ABC7 news in Denver: “…poor communication between the men and support staff, and reveals that an airtanker carrying flame retardant was hovering overhead as the men died.”
New York Times: “…it outlined several problems, like radios that sometimes did not work properly, updates that did not give a precise sense of the crew’s movements, and the 33-minute period of radio silence.”
BBC: “…inadequate communication played a role in their fate…The report authors describe radio communications as ‘challenging throughout the incident’.”
In most of these articles citing radio issues, they are referring to the programming mistake, but some go on to discuss a failure of people to adequately communicate their thoughts to one another, which at times was an issue and led to confusion about the location of the Granite Mountain Hotshots. Other related problems were too much radio traffic on some frequencies and the workload of the Aerial Supervision Module which resulted in them missing some incoming radio calls from the 19 trapped firefighters.
(Originally published at 11:19 MDT, September 28, 2013; updated at 6 p.m. September 28, 2013. Observations after reading the report are at the bottom of this article.)
The Arizona State Forestry Division has released the Serious Accident Investigation report of the Yarnell Hill Fire, which on June 30, 2013, killed 19 members of the Granite Mountain Hotshots. It was produced by a very large cast of characters, 18 core Team Members, 17 Support Team Members, and 19 Subject Matter Experts, for a total of 54 people.
The report found:
The judgments and decisions of the incident management organizations managing this fire were reasonable. Firefighters performed within their scope of duty, as defined by their respective organizations. The Team found no indication of negligence, reckless actions, or violations of policy or protocol.
Air Attack’s photo of the Yarnell Hill fire at 7:24 p.m. June 29, 2013
A news conference about the report was live-streamed by at least two Phoenix area television stations. In the question and answer period several national news organizations as well as local media asked questions of the five-person panel which consisted of the Arizona State Forester, two people from the investigation team, and two officers from the Prescott Fire Department.
Below is a 21-minute video released by the investigation team today, which they described as a “A brief overview of the Yarnell Hill Fire Investigation report.” Much of it comes word for word from the report but it makes effective use of Google Earth to provide an overview of the geography of the fire.
Granite Mountain Hotshot Christopher MacKenzie shot the two video clips below shortly after 4:00 p.m. on June 30, 2013. These are the last images of the hotshots before they died. The video was unexpectedly made available today for the first time by the Prescott Daily Courier, which has an article about how the video and other photos of the fire were found.
Our observations after reading the report and viewing the press conference and the question and answer session.
The official report commissioned by the Arizona State Forestry Division, a case of them investigating themselves, did not break much new ground. There was little of a negative nature written about the crew or their employer, the Prescott Fire Department, which was barely mentioned. The Granite Mountain Hotshots were fully qualified, staffed, and trained and they were on day 13 of a permitted 14 days in a row of fighting fire. And, there was “no indication of negligence, recklessness actions, or violations of policy or protocol”.
Why did the Granite Mountain 19 leave the “black”?
The investigators emphasized that they were unable to answer one of the most-asked questions about the fatalities — why the crew left the safety of the already burned area, the black, to attempt to walk 1.6 miles mostly through unburned brush to another safety zone, the Boulder Springs Ranch. They came to within 0.38 miles of their destination when they encountered one of the heads of the fire that had wrapped around the ridge to their left in the box canyon and was headed toward them, cutting off their path probably much to their surprise. Click the map below to see a larger version of the wind at the deployment site.
No one knew where the crew was in relation to the fire
There was confusion about the location of the crew. Other firefighters thought they had either remained safely in the black where they had been for a while, or they had headed north to another safety zone. But instead, they traveled south. When they reported that they were entrapped and were deploying their fire shelters, no one knew where they were. Finally they told Air Attack they were on the “south side”, but even though a DC-10 air tanker was orbiting and ready to drop on them, airborne personnel could not find them, either due to heavy smoke or because they were looking in the wrong place. But under the extreme wind and fire conditions, it is unlikely that air support would have helped the firefighters very much.
Improving situational awareness
This is another fire, like the Esperanza Fire, where if the fire overhead, such as a Division Supervisor, Operations Section Chief, or Safety Officer, had known the location of the personnel on the fire in relation to the real-time spread of the fire, it could have saved lives — 24 on these two fires alone.
It is irresponsible for the wildland fire agencies to continue to do nothing to improve the situational awareness of firefighters, which has proved fatal to too many of them.
We have written about this several times before. Many local fire departments, EMS divisions, and police units have the ability to send location data to dispatchers. If the analog or digital ground-based radio systems being used today can’t handle this task in remote areas, then use a satellite-based system. The U.S. Forest Service asked for proposals to purchase thousands of little location devices last year, and adding high tech video systems to air attack ships could help. We have also written about a device we called a Firefighter’s Emergency Situational Awareness Device, a FESAD.
One of the recommendations in the report was to “review current technology that could increase resource tracking, communications, real time weather, etc.” The Q&A panel today said, in response to a question, that the surviving family members of the 19 Hotshots strongly suggested while being briefed this morning that tracking systems for firefighters be utilized.
Very Large Air Tanker not ordered because of “steep terrain”
The information that the state of Arizona released on July 16 about the resources deployed on the fire said a DC-10 Very Large Air Tanker (VLAT) was in Albuquerque and available on June 29, but was not ordered due to Air Attack’s concern about its effectiveness in steep terrain and inability to deliver retardant before cut-off time. The way this was addressed in today’s report was “ICT4 declines the VLAT offer at 1750 [June 29] based on fire conditions.” There was nothing about “steep terrain”, which didn’t exist on the fire to the extent that it would severely limit the effectiveness of a DC-10 VLAT. In fact, the next day, June 30, they used the hell out of both DC-10s, dropping over 88,000 gallons in 8 flights. A recommendation in today’s report was to “…develop a brief technical tip for fire supervisors/agency administrators on the effective use of VLATs.”
Air tanker drops on Yarnell Hill Fire, June 29 and 30, 2013.
The DC-10s may have been effective on June 29 when the fire was still small, but by the time they both arrived on June 30, the day of the entrapment, the wind event was making it difficult for anything dropped from the air to slow down the fire — too much heat, and too much wind blowing the retardant away before it hit the target.
Aerial Supervision Module taking on too many roles?
During the time of the entrapment the roles of Air Attack and Lead Plane were filled by a single aircraft called an Aerial Supervision Module (ASM), coordinating all of the aerial firefighting, directing air traffic, preventing aircraft from bumping into each other, developing tactics, AND serving as Lead Plane, physically leading the air tankers into their targets about 200 feet above the ground. The Lead Plane duties limited their ability to perform full Air Attack responsibilities over the fire at the same time. The report said, “ASM was too busy handling multiple duties to communicate with the crew just prior to the deployment”.
One of the recommendations in the report is to request the National Wildfire Coordinating Group to develop guidance to identify at what point is it necessary to separate the ASM and Air Attack roles to carry out required responsibilities for each platform.
No overwhelming force
The ordering and use of ground and aerial firefighting resources was less than aggressive on June 29, the day before the tragedy when the fire was still small. The only air tankers used that day were two single engine air tankers, and for only part of the day, dropping a total of 7,626 gallons. After being released, they were requested again by Air Attack, but dispatch only allowed one to respond to the fire, wanting to keep one in reserve in case there were other fires. General Norman Schwarzkopf’s philosophy when confronting the enemy was to use “overwhelming force”. This strategy also is effective when confronting a wildfire. Overwhelming force for a short amount of time can prevent megafires burning for weeks, consuming many acres, dollars, and sometimes homes and lives.
