Published Wed., Oct. 16 2013 at 12:00 PM
|Illustration by Justin Renteria.|
Former Yarnell Fire Chief Peter Andersen sat under a tree in his front yard having his morning coffee on Sunday, June 30, when the Granite Mountain Hotshots drove past his Glen Ilah home.
“At 8:03, [their] two buggies went by,” Andersen says. “Right after they went by, the leaves started to blow. I shook my head. [The state] didn’t listen to me.”
Andersen, who resigned as Yarnell chief in 2011 after 12 years of service, was aggravated because he had warned an Arizona Forestry Division fire manager the night before that it was crucial to attack the steadily expanding fire in the hills above Yarnell at dawn, before prevailing southwesterly winds picked up about 8 in the morning.
|Courtesy of Joy Collura|
|Hikers took this photo of the Granite Mountain Hotshots marching up a trail at 9:18 a.m. on June 30.|
“I said, this being summertime, it will give you three hours . . . without wind at your backs to be able to get this thing under control,” Andersen says he told a fire manager.
Seeing the hotshots roll past so late on Sunday morning was yet another signal to Andersen that the Forestry Division was failing to aggressively attack a wildfire that started two days earlier.
All the ingredients for fire disaster were present: It was peak summer-burning season, the area had just sustained record-high temperatures, and the landscape was overgrown with chaparral, creating a tinderbox poised to explode.
Andersen says the lack of urgency to put out the fire caused him to wonder whether the state was content to let it burn through the dense chaparral that choked the gaps between massive granite boulders strewn across the Weaver Mountains that flank Yarnell to the west. The east slope of the mountains hadn’t experienced a wildfire since 1966.
As the two white Granite Mountain vans drove on a narrow road that led to the base of the mountains, Andersen thought that fire managers were “doing a sloppy job” handling what he and other firefighters knew was a “volatile” situation.
The state’s underwhelming effort to control the wildfire collapsed late Sunday afternoon when the prevailing southwesterly winds were replaced by powerful downdrafts from a thunderstorm approaching from the northeast. Weather forecasters issued a warning about the approaching storm to fire managers at 3:26 p.m. The warning was relayed to the Granite Mountain crew.
The 50 mile-per-hour downdrafts from the thunderstorm blew up a fire that had burned about 4,000 acres by 3 p.m. into an 8,000-acre conflagration a few hours later.
The three key environmental factors affecting wildfire behavior fell into perfect alignment: wind, fuel, and topography. The drought-stricken desert scrub, combined with the thunderstorm’s powerful winds, generated a wall of flame that surged across relatively flat ground at about 12 miles per hour — extraordinarily fast for a fire.
The powerful wind bent the 80-foot-high flames nearly parallel to the ground as the fire approached the base of the Weavers. The intensity and speed of the fire accelerated as it entered several box canyons that served as funnels, further amplifying its fury.
For reasons that remain unknown, the Granite Mountain Hotshots left their safe spot in a burned-over area on a ridge sometime after 4 p.m. and dropped down the side of the mountain. About 4:40 p.m., they hiked through dense chaparral at the base of one of the canyons, apparently attempting to reach Boulder Springs Ranch, which had been designated as a safety zone because the owners had cleared a wide swath of vegetation from around the property.
Suddenly, the fire swept around the northern flank of the canyon’s wall and surged toward the 19 men, covering the last 100 yards in 19 seconds. The crew had less than two minutes to react to the 2,000-degree firestorm that quickly engulfed their position. There was no chance of survival.
Like tens of thousands of people who’ve closely examined the circumstances leading up to the hotshots’ deaths, Anderson doesn’t understand why the crew was in the box canyon in the first place, much less at a time of day when wildfires typically display their greatest intensity and when thunderstorm warnings had been issued.
“Anybody who has ever taken a wild-lands class is warned about box canyons,” Andersen says. “You might as well be standing in a fireplace with the flue open.”
The question of why the men were there haunts Andersen. And, he says, the lack of substantive conclusions in a report issued September 28 after a state-commissioned investigation into their deaths has left him unsatisfied.
“I think it’s a big cover-up, a big snow job,” he says. “It tries to take any semblance of blame off anybody.”
Andersen evacuated from his home about the same time that the Granite Mountain crew deployed their fire shelters designed to withstand temperatures of about 300 degrees.
“The heat was so intense that it was choking me,” Andersen says. “I could see [the fire] coming over the ridge . . . and you couldn’t see the top of the column of smoke. And it was starting to slowly spin . . . like a slow tornado, throwing embers everywhere.”
|Photo by John Dougherty|
|Expert Doug Campbell says, “They knew the rules were against them when they were going downhill in the green.”|
Three days after the disaster, the Arizona Forestry Division commissioned a “Serious Accident Investigation Team” to review events leading up to the fatal entrapment that inflicted the worst blow to an Interagency Hotshot Crew since such forest-firefighting units were formed nationally in the 1940s.
When the investigation team, headed by Florida State Forester Jim Karels, released its report to the public three weeks ago, the 116-page document’s astonishing conclusion was that everybody involved in the Yarnell Hill Fire did everything right — despite the incineration of the 19 hotshots by flames so hellish that granite boulders fractured.
|The Forestry Division’s Russ Shumate.|
“The judgments and decisions of the incident management organizations managing this fire were reasonable,” the report states. “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 report concludes that, because of a lack of communication between the Granite Mountain crew and fire managers and other firefighters in the 30 minutes leading up to the hotshots’ entrapment, it’s impossible to determine why the crew decided to leave the safe burned-over area and descend into the chaparral-choked box canyon — an action that violated numerous firefighting safety protocols.
“We cannot fully know how they made their decisions prior to their entrapment and fire shelter deployment,” the report states. “No crewmembers from the deployment site survived to tell why the crew took the actions they took.”
The state Forestry Division’s Roy Hall, incident commander in charge of the fire at the time the hotshots perished, praised the report for finding “no smoking gun,” according to published news reports.
The inherent contradictions in a report that assigns no blame for the deaths of so many young men reveal what murky standards wild-land firefighters must work under.
On one hand, the National Interagency Fire Center website instructs them always to obey the 10 “Standard Fire Orders” and 18 “Watch Out” situations developed to protect firefighters in the field. Indeed, the center states that the rules are to be strictly followed: “We don’t bend them, we don’t break them.”
On the other, the same agency adopted a policy earlier this year that states the “10 and the 18″ are merely “guidelines” that should be incorporated into decision making by experienced wildfire leaders making split-second judgments in an environment with many variables that could change suddenly.
The result is a system that makes it virtually impossible to hold anyone accountable for fatal accidents, that leaves firefighters with no clear directives about how to operate in the field.
“Everybody’s lawyering up … That’s why the report’s written that way.” — Doug Campbell, a retired Forest Service fire-management officer who’s widely respected for developing a wildfire-prediction system used in more than 20 European countries.
“There appears to be a kinder, gentler, and softer approach” to enforcing the 10 Standard Fire Orders, says Dick Mangan, a retired wildfire accident investigator who has participated in many high-profile, wildfire-fatality reviews — including ones concerning the 1990 Dude Fire near Payson that killed six firefighters and the 1994 South Canyon fire in Colorado that claimed 14 lives.
“I have a hard time understanding that everybody did everything right, and 19 people died,” he says.
Regardless of whether the Granite Mountain crew violated orders/guidelines, leading wildfire experts say the state-commissioned investigation report primarily is a diversionary tactic to protect the Forestry Division and other government entities, including the city of Prescott. They say the report seeks to innoculate wildfire managers overseeing Yarnell Hill operations from potential liability from lawsuits and possible criminal charges.
“Everybody’s lawyering up,” says Doug Campbell, a retired Forest Service fire-management officer who’s widely respected for developing a wildfire-prediction system used in more than 20 European countries but not formally adopted in the United States. “That’s why the report’s written that way.”
Mangan, who had hoped before the report was released that it would “let the chips fall where they may,” says the Yarnell Hill investigation fails to deliver clear lessons that could be used to prevent future fatal accidents. The report, he says, didn’t analyze adequately the state’s management of a complex series of events leading to the fatal incident to determine factors that contributed to it.
“There’s usually a chain of events — things that happened that shouldn’t have happened” — that contribute to fatal wildfire incidents, he says. “If you break the chain of events, then the accident doesn’t happen.”
Wildfire experts interviewed for this story identified key inadequately analyzed factors in the investigation that may have contributed to the tragedy, including:
• The state’s failed initial attack on the fire created a situation that later placed hundreds of firefighters at risk to put out a fire that could’ve been controlled easily.
• Once the initial attack failed, the state dispatched a skeleton management team to direct firefighting operations, but the team didn’t have sufficient resources to adequately fight the blaze. When it assumed control, the state’s “Type 2 Short” incident-management team lacked “safety officers” and “division supervisors” whose absence may have contributed to a breakdown in communications during the crucial 30 minutes before the hotshots died.
• The investigation report didn’t thoroughly examine the mental and physical condition of the Granite Mountain crew on the day it was dispatched to Yarnell — its scheduled day off and the 28th day it had worked in June.
Wildfire experts say it’s essential that firefighters and fire-management teams have an acute understanding of environmental forces that can affect the intensity, rate of growth, and direction of wildfires. Failure to understand and recognize these forces, they stress, can lead to catastrophic results. Among the most important factors are topography, fuel temperature, and wind speed and direction.
The art of understanding how a wildfire behaves and predicting what environmental factors can cause it to change suddenly is defined by professionals as “situational awareness.”
Experts, including Doug Campbell, say a woeful lack of basic training in wildfire behavior has led to a lack of such awareness among firefighters on the front lines, as well as among managers directing operations.
“If firefighters can make accurate predictions as to the specific time and place where fire-behavior changes will occur, then no attack should fail — no firefighter should lose [his] life or be injured by fire,” Campbell states in his book The Campbell Prediction System.
The Yarnell Hill investigation report, Campbell and other experts say, fails to adequately address what was a clear lack of situational awareness by the state management team from the start of the fire. Instead, it focuses almost exclusively on dead firefighters whose actions and decisions cannot be explained.
The report “is a shell game in so many ways that it does a disservice to what we know about fire management,” says Paul Orozco, a retired U.S. Forest Service fire-management officer who participated in the investigation into the deaths of four firefighters in the 2001 Thirtymile Fire near Winthrop, Washington.
There still was more than two hours of daylight after the fire first was reported at 5:41 p.m. on Friday, June 28 — more than enough time for a quick strike by firefighters before nightfall, wildfire experts say.
Former Yarnell Fire Chief Andersen is convinced that the fire could’ve been extinguished the evening it started if the Yarnell Hill Fire Department and the state had quickly responded.
“There’s a jeep road that goes up there to it,” he says. “There’s no reason for that [fire] not to have been put out.”
But Yarnell had only two firefighters, including a volunteer more than 70 years old, on duty that evening. Dispatch records suggest that the department gave no serious consideration to going up the mountain to fight the fire. No one even answered a state dispatcher’s 6:09 p.m. phone call to the Yarnell department.
Yarnell’s fire chief at the time, Jim Koile, isn’t a Yarnell resident and wasn’t in town the night the fire started. Koile resigned October 8 under mounting pressure from outraged residents who believe the department could’ve done far more to protect their community.
Firefighters from the nearby community of Congress were poised to at least attempt to engage the fire with Yarnell’s department but were directed by state officials not to respond.
“They basically told us they were working on it,” Congress Fire Chief Virgil Suitor says of state Forestry Division representatives. “They told us to stand down because it was up on the rock pile.”
There were substantial firefighting resources available nearby in the Prescott National Forest, where the Doce Fire that burned more than 6,800 acres northwest of Prescott was nearly contained. The Granite Mountain crew worked on the Doce Fire from June 18 to 25 and on the West Spruce Fire, near the Doce burn, on June 28 and 29.
A helicopter with a water bucket and handful of firefighters flown to the ridge top could’ve knocked down the then-half-acre Yarnell Hill Fire on Friday evening, experienced wild-land firefighters and former hotshot crew leaders say.
The investigation report doesn’t address why the state didn’t ask Prescott National Forest officials or other federal agencies with substantial firefighter resources for immediate assistance to put out the fire.
The Arizona Dispatch Center alerted state Forestry Division assistant fire management officer Russ Shumate about the lightning-caused fire at 5:45 p.m. Shumate has worked in central Arizona since 1995 and is familiar with the Weaver Mountains.
The report doesn’t mention Shumate’s lackadaisical response to the start of the fire, even though state dispatchers seemed prepared to ramp up the attack immediately.
“Do we need to order additional resources?” the state dispatcher asked Shumate after notifying him about the fire.
“Negative. We might get a crew in there tomorrow,” Shumate responded.
Ten minutes later, a Prescott wildfire dispatcher contacted the Arizona Dispatch Center and asked whether the state was “sending any air resources to Yarnell.”
“Negative,” replied the state dispatcher. Instead, the state requested that an “air-attack” plane still working on the Doce Fire fly over Yarnell and provide an assessment. Air-attack aircraft serve as airborne command centers.
At 6:03 p.m., Arizona dispatch notified Shumate that the Congress Fire Department had been advised to “stand by.” Shumate, for the second time, responded that the state “may have a couple of crew(s) ordered for tomorrow.”
Nearly an hour after the fire was first reported, a crew member aboard a federal Bureau of Land Management fire engine notified state dispatch that his crew was in contact with a local Yarnell rancher who saw the lightning strike, and “we are going to follow him up to where it is.”
The state dispatch log, however, provided no additional information on what happened to the BLM fire engine, stating that radio contact was lost.
There was no other mention of the BLM engine in state dispatch logs until 9:35 p.m., when Shumate stated that it was returning to its base near Wickenburg.
The Yarnell investigation report doesn’t provide information on the failed attempt by the BLM engine to reach the fire. Instead, it focuses on a dispatch from the air-attack plane’s fire officer, who flew over the fire and reported that it was in a boulder field with no vehicular access and showing little smoke.
According to the investigation report and dispatch logs, Shumate, after talking to the air-attack plane, determined it was “less than a half-acre in size, 80 percent out, active only in one corner with low spread potential and no structures or people at risk.”
Shumate told Arizona dispatchers at 7:19 p.m. that the fire was “inactive, not much of a threat” and that he was “not taking action tonight.” Shumate also told state dispatch that he was “at [his Prescott] office until further notice.”
Shumate, the investigative report states, didn’t dispatch firefighters to the scene because he was concerned for their safety moving at night across rugged terrain, where they could be exposed to lightning strikes.
It’s unclear how Shumate came to his assessment on Friday evening. State dispatch logs mention that Shumate didn’t arrive “on scene” in Yarnell until 6:51 a.m. on Saturday, June 29.
State Forestry Division public information officer Carrie Dennett didn’t respond to e-mails sent on October 9 and 12 and to voice messages on October 10 and 12 asking about Shumate’s location on June 28.
It would constitute a serious management failure if Shumate wasn’t on scene to personally assess the fire on the night it started and then made decisions on how and when to deploy resources, says wildfire expert Paul Orozco.
“It [would be] negligent not to physically see the fire and make a call on it,” Orozco says. “This isn’t about guessing . . . If you don’t get it right, it could kill people.”
Once Friday slipped away, former Yarnell Fire Chief Andersen says, the state had another opportunity to put out the fire early Saturday morning — by deploying effective firefighters and air tankers to drop fire retardant immediately after sunrise.
But that didn’t happen, either.
Rather than hitting the fire at dawn (5:20 a.m.), Shumate didn’t arrive in Yarnell until shortly before 7 a.m. His late arrival was a precursor to a series of mistakes, mechanical failures, weather changes, and other delays that would plague firefighting efforts throughout the day.
The investigation report lays out many of the problems as a series of isolated events. It never mentions the state’s failure to quickly apply adequate resources to control what initially was a very small blaze.
More than 17 hours passed from the initial lightning strike before the state had firefighters helicoptered to within a quarter-mile of the fire at 10:48 a.m. on Saturday, June 29. By this time, the fire was reported at eight acres. Its size was reduced to about two acres shortly after noon.
Not only was the state slow to respond, it sent in an ill-equipped, six-man crew from Lewis State Prison that wasn’t trained to handle initial attacks on wildfires (such crews generally are used for mop-up operations), wildfire experts say. But dispatching $1-an-hour-per-man prison crews rather than deploying more experienced firefighters, such as hotshot crews — which cost about $800 an hour — is a way for the state to save money fighting forest fires.
Shumate’s first mistake was failing to place an order for an air-attack-command plane until 8:33 a.m. A little later, the aircraft broke down before takeoff. A replacement was ordered at 9:26 a.m. The second air-attack plane operated only until noon, when it developed an oil leak and was taken out of service. The plane was repaired a couple of hours later, but Shumate released it from the Yarnell Hill Fire.
In addition to the air-attack plane, Shumate ordered two single-engine air tankers at 8:33 a.m., but they were hundreds of miles away. One of the planes, holding 450 gallons of retardant, took off from the Marana Regional Airport at 9:24 a.m. and another, with 750 gallons of retardant, departed from Wilcox at 9:43 a.m.
The long flight time to Yarnell limited the state’s ability to apply retardant on and around the fire early in the morning. The planes each made two drops of retardant near the fire before noon and then landed in Prescott to refuel and for the crew to have lunch. The planes remained on the ground in Prescott for another three hours before flying to an airfield in Wickenburg just after 3 p.m.
Shumate had planned for these planes to use the reloading base in Wickenburg, 41 miles closer than the Prescott base. This would’ve allowed the state to apply more retardant at quicker intervals at less expense. But the Wickenburg retardant-reloading base wasn’t operational until after 3 p.m.
Without the air-attack plane, which Shumate earlier had released from the fire, to report on conditions, he decided to remove two fire-engine crews at 3:40 p.m.
After these reductions in resources deployed to the fire, Shumate then reversed tactics at 4 p.m. and recalled the air-attack plane to do reconnaissance. The plane, however, had flown out of the area and wasn’t available for another 40 minutes.
At this point, conditions on the ground were deteriorating rapidly.
Shumate notified state dispatch at 4:15 p.m. that ground crews were “still having trouble catching” the fire, by then estimated at two to four acres with “creeping factors.” The recalled air-attack plane, meanwhile, still was more than 30 minutes out — which meant that a detailed aerial assessment of the blaze couldn’t yet be made and that the single-engine air tankers, standing by in Wickenburg, couldn’t yet be told where to drop retardant.
Minutes after the air-attack plane finally arrived over the fire, Shumate requested at 4:55 p.m. that a helicopter tanker be sent to Yarnell. This was an indication that ground crews were having serious trouble controlling the fire.
But the order for the “heli-tanker” wasn’t immediately processed because of “confusion” between the Arizona dispatcher, Shumate, and the federal regional dispatch center in Albuquerque, the investigation report states.
The state’s window of opportunity to put out the Yarnell Hill Fire was closing rapidly.
With each passing minute, the fire gained ground, as wildfires tend to do in the late afternoon. The fire gained energy as it steadily backed down the mountain and moved into a location where topography, available fuels, and favorable winds increased its intensity.
The fire got the upper hand at 5:18 p.m. when it jumped a two-track road that served as a firebreak on the eastern flank and quickly spread over two acres. “That should’ve never happened,” says a former Arizona hotshot crew superintendent who continues to work with federal wildfire crews and asked not to be named.
Shumate knew this meant trouble. At 5:38 p.m., he alerted Arizona dispatch that the fire now posed a threat to Yarnell and Peeples Valley in the next 24 to 48 hours.
Shumate’s problems continued to mount. At 5:42 p.m., he learned that the heli-tanker and a large airplane tanker had turned down requests to fly to Yarnell because of dangerous weather conditions.
Three minutes later, Shumate was faced with a crucial decision.
Federal dispatchers offered Shumate the services of one of only two Very Large Air Tankers in the United States. The converted DC-10s are capable of carrying up to 11,400 gallons of retardant. The aircraft, however, are very expensive to operate, costing $12,500 an hour for a full load of retardant. The plane would take off from Albuquerque and would require about two hours of flight time.
The fire’s location on a high ridge was an “ideal” target for the big jet because it could’ve avoided dangerous maneuvers in the mountainous area, says a wildfire air-operations supervisor who asked not to be identified. The big plane could’ve provided the chance to knock down the fire or at least given ground crews an opportunity to contain the rapidly growing blaze.
Shumate turned down the opportunity to call in the big DC-10 tanker at 5:50 p.m. The investigation report states that he made the decision based on “fire conditions.”
Instead of calling in the big jet, Shumate continued to use the single-engine planes, whose effectiveness diminishes as winds pick up. The smaller planes dropped another 5,400 gallons of retardant by the time they were grounded by darkness at 8 p.m.
Shumate estimated in his daily “Incident Status Summary” that the state spent about $20,000 on June 29 in its failed initial attack on the Yarnell Hill Fire. After Shumate lost control of the fire, his Forestry Division bosses called in more than a dozen fire-management specialists as part of a “Type 2 Short” team to take over fire operations the next day.
But wildfire experts say this unit was woefully inadequate to handle the explosive situation it soon would face. One state firefighter derisively called these state fire managers the “B” team.
“It just didn’t have the horses,” wildfire expert Orozco says of the Type 2 Short squad.
|Left: Ex-Fire Chief Peter Andersen says the state’s response was extremely slow. Right: Florida State Forester Jim Karels, whose investigators found that nobody was at fault in the hotshots’ deaths.|
Some members of the state’s Type 2 Short team arrived in Yarnell on Sunday, June 30, in time for a 7 a.m. briefing. The fire had grown to an estimated 300 to 500 acres overnight. The team didn’t include all the division supervisors needed to oversee specific geographic areas of the fire — or safety officers.
Shumate met with incoming incident commander Roy Hall about 9:30 a.m. to discuss fire conditions. Transitions of management teams ideally take about a day, wildfire experts say. But, in this case, Hall formally took over operation of the fire from Shumate at 10:22 a.m., less than an hour after their briefing began.
The transition was so rushed that the state didn’t prepare a written “Incident Action Plan” or accompanying fire maps for the June 30 operations period. These are federal requirements for a Type 2 fire.
The clumsy transition, experts say, left little time for the new team to understand the complexities of the rapidly changing fire. The fire investigation report ignored the shortcomings of the state’s Type 2 Short team and the uneven transition.
Less than an hour after taking over, Hall discussed bringing in a full Type 2 team with state Fire Management Officer David Geyer. At 1 p.m., the state belatedly finished a “fire complexity analysis,” a critical report that should’ve been completed immediately after the state’s initial attack failed the previous afternoon, Orozco says.
The state then requested a federally managed full Type 2 team. An hour later, it scrapped the plan and requested a federal Type 1 team, the highest-level incident-management unit used for the most complex and dangerous fires. But that team couldn’t be on the ground until late the next day.
It was up to Hall’s understaffed Type 2 Short team to handle a fire that was rapidly overwhelming available resources. “They were behind the curve,” says expert Doug Campbell.
The state’s scramble for personnel required it to make a crucial decision that directly affected the Granite Mountain Hotshots.
The lack of sufficient management personnel forced the state to assign Granite Mountain Superintendent Eric Marsh as division supervisor for the southwest flank of the fire. Granite Mountain Captain Jesse Steed assumed immediate command of the crew.
Wildfire accident-investigation expert Mangan believes this was a pivotal mistake that weakened fire managers’ control and understanding of the hotshots’ actions. Marsh, Mangan says, still was in direct charge of the crew as the Division A supervisor and wouldn’t need to report to an independent division supervisor who may have challenged his decisions.
“You have taken one link out of the chain of command,” Mangan says.
This later would prove critical to when the Granite Mountain crew descended from the safe, burned-over area on top of the mountain into the box canyon filled with live fuel. The crew apparently was headed to the Boulder Springs Ranch, which had been designated during a morning briefing as a safety zone.
During the frantic moments leading up to the crew’s entrapment, fire managers were overwhelmed. They were making split-second decisions deploying aircraft, evacuating Yarnell, and calling off firefighters and equipment as the firestorm rapidly intensified. Granite Mountain’s location, the investigation report states, wasn’t a primary concern because managers believed the crew was in the safe, charred-black area near the top of the mountain.
About 4 p.m., an aircraft crew member got worried after hearing a comment on the plane’s radio referencing a crew and a safety zone. The fire officer on the plane asked an operations chief whether radio traffic should be suspended to determine what was happening. The operations chief stated that Granite Mountain was the crew mentioned and that “they’re in a good place.”
The investigation report states that, soon after, Marsh announced on the radio, “We’re going down our escape route to our safety zone.”
The airborne fire manager asked, “Is everything okay?” Marsh replied, “Yes, we’re just moving.”
This was an important moment in the sequence of events that would claim 19 lives. No one from the state’s management team followed up on Marsh’s comment about moving the crew, according to the investigation report.
There are at least two possible reasons for the lack of response from fire managers: Marsh didn’t need to report to a division supervisor since he had assumed that role. And it’s uncertain whether a safety officer was on duty at this critical moment, state dispatch records show.
A state dispatcher contacted Marty Cole at 2:24 p.m. and requested that he report to the fire as a safety officer. Cole lives in Chino Valley and faced at least an hourlong drive to reach the incident command center in Peeples Valley.
As with New Times’ other request of her, Forestry Division public information officer Dennett didn’t respond to e-mails and voicemails requesting information on when Cole and two other safety officers, also requested on the afternoon of June 30, arrived and assumed duties in Yarnell.
Safety officers are principal advisers to incident commanders in fire-management operations.
Among safety officers’ primary concerns are extreme fire behavior, escape routes, and safe zones — the exact issues that Granite Mountain discussed but operational chiefs ignored or misunderstood.
It’s vital to note that a safety officer has authority to override chain of command when an immediate threat to life or risk of serious injury is evident.
Expert Mangan, who offers training courses for safety officers, says Marsh’s announcement that Granite Mountain was moving from its safe zone in the charred area should’ve prompted a safety officer, if one was present, to request that Marsh provide more information and possibly stop the crew from moving off the ridge.
“The more people you have involved in a decision like that, the better chance you are going to come up with a better decision,” Mangan says.
The intensity and speed of the wildfire as it stormed toward Yarnell stunned the state’s management team: “The fire way-outperformed our expectations and surpassed any thoughts we had about our trigger points,” one of the two operations chiefs told investigators.
Granite Mountain also didn’t predict the fire’s eventual path. The hotshots had parked their two vans at the base of the mountain in an area that turned out to be directly in front of the fire after the thunderstorm reversed the inferno’s direction 180 degrees. It’s possible that the crew’s decision to leave its safety zone atop the mountain was related to the threat on their vehicles.
The investigation report states that at 3:50 p.m., an air-attack officer notified Marsh that the fire had reversed direction, was heading quickly toward Yarnell, and could arrive in one to two hours. The air-attack officer also told Marsh that the crew’s vehicles may be in the path of the fire.
Marsh told this officer that he had a plan to address the issue. The investigation report, however, doesn’t elaborate on what Marsh’s plan was. The air-attack plane then left the area because the pilot was approaching the limit of hours he could fly legally in one day.
The exchange between the air-attack officer and Marsh was another opportunity where a safety officer or an independent division supervisor could’ve played a key role by clarifying Marsh’s intentions and advising him to keep the crew in the charred zone.
That Granite Mountain parked its vehicles on ground that later burned and the operations chief’s admission that the fire “outperformed our expectations” show that Arizona’ wildfire managers failed to anticipate the Yarnell fire’s potential intensity and direction during a time of year when monsoon storms are frequent, critics believe.
Basic “situational awareness” of wildfire behavior, Orozco says, didn’t occur in regard to the Yarnell Hill Fire.
Sonny “Tex” Gilligan and Joy Collura began their hike up the Weaver Mountains at 4 a.m. on Sunday, June 30. The avid hikers and part-time cave dwellers wanted to get a close look at the fire atop the mountain. They knew the backcountry inside out and were very familiar with the difficulty of hiking through dense desert shrubs.
On their way up the mountain, they bushwhacked through the box canyon where the Granite Mountain crew later perished. The hikers already were at the top of the mountain when they saw the Granite Mountain Hotshots coming up a two-track trail about 9:18 a.m.
Gilligan, an experienced outdoorsman and former cowboy and miner, was shocked at the hotshot crew’s condition.
“What I saw was a group of men [who] were totally spent. They looked like they were tired. They weren’t somebody you would want to fight a fire,” Gilligan says. “They needed rest.”
The hikers stayed on the mountain until about 2 p.m. with temperatures hovering about 103 degrees. They observed the crew from time to time throughout the day. The crew, they said, didn’t appear to be doing much active work.
Gilligan says their inactivity led him to believe that the fire was a “controlled burn.” It appeared “they were actually trying to let it go, and they just wanted to clear this brush off this mountain,” he says.
Gilligan and Collura saw the fire take off about 12:30 p.m. as it swept over a hill below the mountain in about 14 minutes. Gilligan estimates that it covered about 300 acres in just a few minutes.
“We were looking at . . . rolls of fire, fire jumping up 40, 50 feet in the air,” Gilligan recalls. “No way are we were going to hang around there.”
Throughout the morning, the hikers watched thunderstorms building to the northeast, near Prescott. Gilligan knew the storms could affect the fire. “When there’s a thunderstorm in an area like this, that wind can change quickly, and it can change fast,” Gilligan says. “That’s where the danger is.”
The investigation report doesn’t mention what Gilligan and Collura observed about the fire’s behavior or about the crew’s condition, even though the hikers were the last people to see the Granite Mountain Hotshots alive. Nor does the report provide any details of the crew’s workload the previous month, mention that June 30 was the crew’s scheduled day off, and that the crew had worked 28 days in June.
“We don’t know the condition of the crew [from the report],” says wildfire expert Campbell, noting that this is a crucial missing element in the investigation.
Lead investigator Jim Karels, in an interview after releasing the report, dismissed a statement by federal dispatchers at the Southwest Coordination Center in Albuquerque that the only Hotshot crew available on June 30 initially was the Blue Ridge Hotshots (who indeed were deployed to Yarnell).
Karels insisted that the SWCC never stated that only one crew would be available. “No, absolutely not,” he said. (The SWCC has declined to comment.)
The state Forestry Division’s dispatch log, however, shows that an Arizona dispatcher requested at 6:21 p.m. on June 29 that the SWCC send two hotshot crews to Yarnell by 6 a.m. the next day. A SWCC dispatcher responded four minutes later, stating, “I can fill one with Blue Ridge. That will be the only [hotshot crew] I have for tomorrow, though.”
Karels says the SWCC never turned down a request for Granite Mountain to be sent to the fire, but instead the SWCC “kicked it back” to the state and instructed the state to fill it “internally” with Granite Mountain.
Yet no such exchange between the SWCC and the state Forestry Division appears in Arizona dispatch logs.
Former hotshot supervisors suggest that one reason the SWCC initially stated that only Blue Ridge was available was because Granite Mountain would be working its 13th consecutive day on its scheduled day off. By doing this, the crew would’ve been unavailable later in the week for an assignment out of the area.
In any case, there’s no question that Granite Mountain had only two days off in June and that the Yarnell Hill fire was its 26th day in the month on a fire line. The hotshots spent two days working at the crew station or on “fuels reduction.” The crew often worked 16-hour shifts, SWCC records state.
Campbell believes fatigue may have been a major factor in the crew’s decision to come off the mountain rather than remain in “the black.” Campbell suggests that Marsh and Steed knew that the crew was tired, hungry, and low on water. The option of staying on the mountain all night wasn’t appealing, nor was following the long trail down to Yarnell that the two hikers had taken safely a few hours earlier.
Campbell believes the Granite Mountain crew concluded that its best course of action — one that would allow members to rest and be ready to re-engage the fire the next day — was to get off the mountain as soon as possible by hiking through the box canyon to the ranch safety zone.
“They knew the rules were against them when they were going downhill in the green,” Campbell believes.
But, he says, rules don’t always stop hotshots from attempting to accomplish a mission.
“The culture of a hotshot crew is a problem,” Campbell says. “They aren’t one to hold back. They are braver than they ought to be.”