After several weeks of quiet thanks to an unseasonably stable snowpack, three men were killed in two avalanches on Feb. 25.

The avalanches – each very different from the other – sent forecasters with the Colorado Avalanche Information Center’s team in the southern mountains scrambling. Two went way south, to La Manga Pass west of Antonito, near the New Mexico border. Two went to a low-elevation hill just above Vallecito Reservoir northeast of Durango.

The two teams had contrasting experiences as they studied the avalanche scenes. All four of the avalanche experts would analyze snowpack, weather patterns and piece together clues to create detailed reports outlining the events that led up to the avalanche. They interviewed survivors and mourning family members to help them craft a narrative as well as an analysis of what went wrong on that Saturday.

The reports are the cornerstones of avalanche safety research and offer insights that other backcountry travelers might be able to use to better inform their navigation through avalanche terrain. The forecasters at the avalanche center have penned six accident reports so far this winter, outlining the snow, weather and decisions of 12 backcountry travelers caught in avalanches that buried nine people, killing seven of them.

Decades of those avalanche reports have identified trends and common missteps that have helped countless backcountry skiers, snowmobilers and hikers recognize hazards in both terrain and decision-making.

Those potential lessons are always in the backs of the minds of CAIC forecasters as they investigate accidents. Their reports carefully detail what went wrong, including the decisions and actions of people who paid terrible costs for fleeting lapses.

“It’s a personal, delicate task and one we take very seriously,” said Ethan Greene, the avalanche center’s director since 2005. “We are trying to juggle all these things and do the best job we can to serve all these different needs but part of that is to do no more additional harm. I mean we’ve all had close calls where some lucky bounce kept us from being killed. We’ve all had friends who have been killed. We try really hard to relay the facts and keep our commentary productive rather than pointing fingers.”

The call came into CAIC forecaster Chris Bilbrey’s phone around 4 a.m. on Feb. 26. Rescuers had just pulled the bodies of two skiers from avalanche debris on a low-elevation slope near Vallecito Reservoir. Bilbrey wondered if he had missed something in his Saturday forecast.

“I thought, what did I blow in the forecast to cause this to happen?” said Bilbrey, who scoured his forecast before dawn. “I was asking myself, ‘Did I miss something and lull them into thinking things were different than they were?’”

He called Greene. The two concluded the forecast had been on point and declined to issue an update or change the forecast for Sunday.

“Self-critique is a common thing and we do make changes when we think we’ve made a mistake,” Greene said.

Bilbrey’s forecast for the region on that Saturday ranked avalanche hazard as considerable above and near treeline with moderate danger below treeline. (CAIC uses five levels of danger to rank avalanche threats: low, moderate, considerable, high and extreme.) About 40% of the fatal accidents CAIC investigates occur when forecasters have ranked the hazard as moderate. It’s a very broad range.

The forecasters internally refer to hazards ranging from “mellow yellow moderate” to “spicy moderate,” with “moderate moderate” in the middle. For Saturday, Bilbrey warned that “the most dangerous areas are wind-loaded slopes.” Southwest winds had drifted snow onto northerly- and easterly-facing slopes creating wind slabs in unexpected locations.

“Strong and erratic winds built dangerous drifts lower in the terrain than usual, and on wide open slopes below the treeline,” Bilbrey wrote in the Saturday forecast.

The Sunday forecast eased the hazard to moderate in all elevations but still emphasized the threat of wind-loaded slopes. Even without adjusting the Sunday forecast in light of the deaths, Greene was able to deploy additional warnings in the hour before dawn on Sunday.

He posted news of the fatal slide on CAIC’s Instagram account early on Feb. 26, with reference to another investigation underway further south and a nonfatal incident involving a snowmobiler near Wolf Creek Pass. The post warned that “heavy snowfall and very strong winds battered the Colorado mountains last week” and the avalanche danger was receding, but slowly.

“We do have the ability to get people’s attention and even though we didn’t see a need to change the (forecast) product, we were able to emphasize a need to be heads-up,” Greene said.

Bilbrey and fellow forecaster Jeremy Yanko spent early Sunday preparing to visit the site near Vallecito Reservoir. Meanwhile, Rebecca Hodgetts, the lead forecaster for the center’s southern mountains, started driving to the southern end of the San Luis Valley with forecaster Matt Huber.

The two teams would have very different investigations.

The wife of James Sutton had called La Plata County Search and Rescue on Saturday night, saying her husband and his friend Jurgen Montgomery had not returned to their home in Durango after leaving to ski that morning with plans to be back around noon. The two were very experienced. Rescuers were able to access the men’s phone accounts to see they had taken photos of maps of terrain around the southern end of Vallecito Reservoir. The searchers found the men’s car at a trailhead in the area.

The helicopter crew from Flight for Life Lifeguard 5 in Durango flew over the area Saturday just before midnight and, using night vision tools and a spotlight, saw ski tracks going into avalanche debris and none coming out. The crew switched their avalanche beacons to search mode and scanned the debris from 200 feet up and got intermittent signals.

Searchers with La Plata Search and Rescue’s winter response team found a single ski in the debris pile at 3:30 a.m. Using their transceivers, they found Sutton, 67, buried under three feet of debris. Montgomery, 47, was buried just below him. They extricated the men’s bodies by 5 a.m.

Just before dawn, Ron Corkish, the incident commander with La Plata County Search and Rescue, called Bilbrey. Corkish was worried about the windy storm coming in. He was concerned that his volunteers would need to cut down aspens and willows to haul the bodies of the men out and wanted to delay the last phase of the extrication until after the storm had passed.

Yanko and Bilbrey were driving to the reservoir when Corkish suspended operations around 7 a.m. They passed a parade of search and rescue trucks leaving the trailhead as they drove the snow-packed County Road 501A. The two forecasters were in a hurry.

“There’s a storm coming and if we can get there before the evidence gets covered, it lets us write and paint a better picture of what we saw and what we think happened,” Bilbrey said.

When they arrived at the trailhead, the rescue team was gone. There were some people at the trailhead though. One of the men’s spouses and her friend were there. Yanko and Bilbrey spent almost an hour talking with the women.

“That added an interesting tone to the start of the whole day,” Yanko said. “It personalized it for us.”

The women were stoic. They recognized the risks the men took in pursuit of their passion, Yanko said.

“I was amazed at how strong they were and accepting that this was what these guys had signed up for and these were things they loved to do,” Yanko said. “This is a game, by and large, of statistics. The more you are exposing yourself to these risks, the more likely one of these instances is going to happen.”

Bilbrey said the stories he hears from friends and families about the victims of the avalanches he investigates “can really hit home internally for me.”

“I try to take that and add it into the pride and quality of what we put out,” Bilbrey said. “It can be a motivation to stay focused and try to provide the best information I can. Not all encounters touch you internally but this one did for sure.”

Yanko and Bilbrey had GPS coordinates of the slide but were largely unfamiliar with the area.

“It’s not a place you typically go skiing, but with this productive snow year and lower elevations holding snow, people are exploring,” Yanko said. “These guys were going against the grain a little bit, trying some new places out. And they lived close by.”

With a grim task ahead and the words of mourning family bouncing around in their heads, Bilbrey and Yanko followed skin tracks they suspected were set by the two men 24 hours earlier.

The tracks told a story. There was a spot where it looked like the two skiers had tried to make a kick turn in the snow – a 90-degree turn while climbing uphill – but the snow was rotten and they tried a different direction. There was another spot, blasted by the wind, where they appeared to be unable to get a purchase with their skins. At that spot, it looked like the two skiers had sidestepped up the wind-scoured knob. Their side steps, etched in hardpack, eventually reached the debris field.

The investigators began what Bilbrey described as “a tedious, methodical process.” Using apps on their phones, they measured the length of the avalanche – 650 feet – and its width – 200 feet. They ranked its destructive force: D2, which means big enough to bury, injure or kill a person. They measured the crown of the avalanche where the snowfield had broken, which was about two feet deep. They dug a pit, looking for the weak layer that shed the wall of snow.

The layer they found matched what they had seen in other areas. Winds had wiped snow from exposed areas on Valentine’s Day and new snow had loaded onto those hardened layers in sustained winds.

They wrote down everything they saw. They took dozens of photos. They visited the bodies of the men, noting that the triggers for their avalanche airbag packs were still zipped inside their shoulder straps. Their skins were on their skis and their boots were in walk mode. All that information, combined with the tracks, supported Bilbrey’s and Yanko’s conclusion that the men were climbing when the avalanche released.

They noted what Yanko called “really junky terrain,” with burned stumps, bedrock, willows and young aspens.

The area was burned in the 2002 Missionary Ridge Fire, which sparked in early June and burned 73,000 acres in 39 days. The lack of trees on the slope left it exposed to wind, which had deposited deep drifts on the northwest-facing slope at about 8,300 feet. It looks like the two skiers had been traveling in a non-wind-loaded area when they reached that slope and were navigating through the suddenly complex terrain.

Bilbrey said he thought back to countless times when he’s been climbing and encountered wind-blasted snow that made it difficult for skins to get a grip, forcing him to change directions and seek out better climbing conditions.

Greene calls it “a detective process.” This avalanche was a bit less complicated because the investigators did not need to interview any survivors or witnesses. That can be one of the most challenging parts of the job. Survivors are mere hours out of one of the most traumatic experiences of their lives when CAIC investigators reach out for information. Days later there are secondary interviews and sometimes the investigators need to talk with survivors a third time.

They also talk with family members who can be unfamiliar with the risks of backcountry travel. Those can be the toughest interviews. Sometimes those family members are worried their loved ones did something wrong. And chances are they did. That takes a delicate touch by avalanche investigators.

“People get involved in bad situations all the time. Whenever someone dies, there is something that went wrong. But that does not mean they made a huge mistake or did something reckless or egregious,” Greene said. “Most of those experiences are a close call and a good story. A few times it’s a fatal accident. The difference between the two is very thin.”

The day after their tour, Bilbrey and Yanko started compiling their information into a report form. They divided up the seven sections and planned to have a rough draft ready for review in a couple days by CAIC bosses like Greene and Spencer Logan.

Most every avalanche investigation has CAIC forecasters on the ground within 24 hours of the slide and a full report within seven days.

Greene gets defensive when he hears complaints that a week is too long. His team is personally invested in every report and they need time to get it right, he said.

“These are not esoteric events. We try very hard to make sure our reports are fair and accurate and they have some value to the people who are reading them,” Greene said. “I’ve had kids ask me about reports I’ve written about their parents being killed. We are careful because we think about that husband or wife or child.”

As Yanko and Bilbrey quietly skinned into a remote zone, Hodgetts and Huber were pulling into a hectic scene.

There were dozens of trucks with snowmobile trailers lining La Manga Pass in the southern San Juans near the New Mexico border when they arrived the morning of Feb. 26. Hodgetts and Huber jumped on Forest Service snowmobiles and shuttled about a mile into the site of an avalanche. More than 30 people – search and rescue members with the Conejos County Sheriff’s Office, Forest Service personnel and many locals – were helping to search for Kevin Gray, a husband and father who had been buried in an avalanche at around 10,800 feet while snowmobiling near the Red Lake Trail on Feb. 25.

Normally, CAIC investigators arrive at a debris field for solemn study.

“This situation was a bit different,” said Hodgetts, whose Manga Pass investigation included several unusual tasks.

A storm was rolling in. The Conejos County sheriff had never managed an avalanche rescue. Hodgetts and Huber helped organize the throng of volunteers. They divided the searchers into lines, each carefully probing the snow with poles, hoping to find the missing 45-year-old Gray.

The slope was timbered, so there were a lot of strikes on trees in the dense debris. A lot of digging that didn’t yield anything.

Gray was born in La Jara and grew up in the southern end of the San Luis Valley. He had been snowmobiling in that area all his life. On that Saturday, he had been with five other friends when he was caught in the slide. He was the only one in his party not wearing an avalanche beacon. More than a dozen sledders had searched through Saturday night for Gray. That number nearly tripled on Sunday as ranchers, neighbors, friends and family rallied.

Hodgetts was in close contact with Greene back in the CAIC Boulder office. He was working with the Colorado Search and Rescue Association to organize more avalanche-savvy rescue teams. He connected with Wolf Creek ski area patrollers, who arrived midday Sunday with their avalanche rescue dogs.

The search was suspended for an hour or so Sunday afternoon as the dogs worked, but the scene was so hectic the dogs were unable to locate Gray, said Hodgetts, who used the downtime with the dogs to conduct her investigation of the slide, measuring and studying the snowpack.

As the storm arrived and snow started falling, the sheriff suspended the search Sunday afternoon. Hodgetts and Huber were worried about working in the zone the next day with the new snow on the steep slope above the debris field.

A friend of Gray’s wanted to bring an excavator into the debris field. The sheriff agreed.

Hodgetts and Huber suggested using explosives to test the slope before rescuers returned Monday with that excavator.

Greene, who was providing Hodgetts and Huber with spot weather forecasts for the zone, worked with the Conejos County sheriff, Colorado Emergency Management Office and the Colorado Department of Transportation to arrange delivery of a cache of explosives. Hodgetts and Huber are former ski patrollers and licensed to handle explosives. (About half the CAIC staff are licensed blasters under CDOT’s explosives permit, which is issued by the Colorado Department of Oil and Public Safety.)

After navigating those bureaucratic hurdles and securing permission from the sheriff and Forest Service with help from the Colorado Search and Rescue Association, Hodgetts and Huber hiked to the top of the peak and threw six 5-pound charges – CIL Cast Boosters – into the hanging snowfield above the debris. Nothing slid.

Rescue teams from Alamosa and Rio Grande and Mineral counties showed up Monday to assist the Conejos County searchers. And, again, friends and family were there.

They found Gray buried about two-and-a-half feet under the snow. They needed chain saws to clear buried trees to reach his body.

Hodgetts spent the first day visiting with Gray’s fellow riders and his family at the scene, including his wife, son and daughter and other family members. The family was part of the search effort and Hodgetts often checked in and made sure they were OK. When they reached Gray, the family sat with him, alone, for about an hour. His brother towed his body out, followed by a somber procession of snowmobiling mourners.

“This gentleman had been snowmobiling there all his life,” Hodgetts said.

She and Huber spoke a lot during the rescue about how they might word the analysis part of their report and “how to reach this community and use the momentum of this tragedy to try to educate people in this community more,” she said.

In their final report, Hodgetts and Huber detailed three things “you can do … before leaving your house to stay safer in the backcountry.” They advised reading avalanche forecasts and making a plan to avoid areas with increased hazards. And, of course, they reminded report readers to always carry avalanche rescue equipment and know how to use it.

The delicate task of education is something all of CAIC’s avalanche investigators think about. For Hodgetts, one thing that made it easier on La Manga Pass was the family’s acknowledgment that their husband and dad should have been carrying avalanche safety equipment. (“His wife was obviously furious that he didn’t have any equipment,” Hodgetts said.) It’s impossible to know if his friends could have reached him, but a beeping avalanche beacon would have given them a chance.

It is Spencer Logan’s job to hone that message. He’s been with CAIC since 2004 and he works with avalanche investigators across North America and Europe to compile reports and analyze trends in avalanche accidents.

“It’s really, really hard to find something new in accidents,” Logan said. “There’s a reason that our human factors and decision-making are so important. We can focus on different aspects of those but it’s a similar pattern of mistakes we have been talking about since the ’70s. Yes, each event plays out in a different way but step back far enough and they all look really similar.”

Avalanche reports, much like newspapers, have distinct sections for the facts and analysis. Each of CAIC’s reports include summaries of weather, snowpack, events leading up to the accident and the rescue. The analysis falls under the heading “Comments” and always starts with words like this: “All of the fatal avalanche accidents we investigate are tragic events. We do our best to describe each accident to help the people involved, and the community as a whole better understand them. We offer the following comments in the hope that they will help people avoid future avalanche accidents.”

Avalanche report writing was pure facts a couple decades ago. The trend toward including commentary and analysis is new in the past 15 years or so. The CAIC investigators are careful not to “blame and shame,” Logan said.

“We are trying to understand the decisions being made in the moment by constructing a factual narrative to figure out what was going on and what we can learn,” Logan said.

Logan edits first drafts of the reports and works with avalanche investigators on finalizing comments. He also sends close-to-finished drafts to families of avalanche victims for final review before publishing.

Logan sees a couple trends emerging this season. One is that four of the seven fatal avalanches this winter have involved multiple skiers or snowmobilers caught at the same time, which suggests that most of the travelers were not aware they were in terrain or on a slope that could slide. If they were aware of the danger, they likely would have been traveling one-at-a-time on the slope.

“From the evidence we see at Vallectio, for example, they did not seem to be focusing on that terrain as an avalanche path because they did not take the standard precautions we would expect them to,” Logan said.

“Why did this happen to two experienced and well-equipped backcountry travelers?” Bilbrey and Yanko asked in their final report on the Vallecito slide. “Unfortunately, we will never know the details of their decision-making, however from the evidence we collected we can speculate that they did not think they were in a dangerous area. If they had, they probably would have approached the situation differently. They probably would have chosen a different route and they would have had their airbags ready to deploy.”

No matter how hard the forecasters try to avoid making judgments, condemnation can come from online readers of those reports. The strategy for CAIC when social media posts turn mean is to hope that someone else in the community steps up and reminds readers that even the slightest mistake in the backcountry can be disastrous. And everyone makes mistakes.

“People take away what they want to take away,” Greene said. “So most of the time we just sort of hope someone else will jump in there and clarify or start a discussion.”

Another challenge, which is more recent, is the chance that CAIC avalanche reports – and the commentary by forecasters illustrating missteps – could become part of litigation. That happened in a 2021 wrongful death lawsuit filed by the family of a man killed in a slide in 2019 in the San Juan Mountains’ Sen. Beck Basin while he was part of an avalanche safety class. And again in 2020 when two backcountry snowboarders were charged criminally for triggering an avalanche that buried a service road near Interstate 70.

When their reports end up in a courtroom, Greene said, “it reinforces our processes and our focus on being factual.”

“Having accident reports picked apart in a legal setting can make you much more sensitive to what you are writing and making sure it’s supported factually and you are not mixing facts and opinion,” Logan said.