At 23 years old, Buchanan had her dad’s keen eyes and her mom’s adorable squirrel cheeks, irrepressible smile and golden ringlets. A sense of adventure ran through both sides of the Buchanan family; the endearing inclination for mischief and humor was all hers. Born and raised in Durango, she’d spent her whole life traipsing through mountains and canyons: hiking, climbing, riding fat-tire bikes, skiing, boating, four-wheeling, sitting around campfires with cold beers, singing and dancing under the stars, sleeping on the ground or in the back of her pickup truck in ski area parking lots. She was now in her third year at Montana State University, where she’d been exploring a different set of mountains and also earning an undergraduate degree in snow science.

The plan was to drive to Taos, New Mexico, with her dad on Wednesday. Then it was back to Bozeman for the start of spring term. The trip to Taos had become something of a tradition, not just because Purgatory employees get to ski there for free, but also because, when she was 18, on a whim she’d entered a Red Bull Freeride event there and her dad had gone with her. Because of her age, she had to compete in the women’s pro category. She was the first one on the course, her dad remembers, and of the first four competitors she was the only one who made it down.

Buchanan had been back in Durango for most of December on winter break, staying at her parents’ classic 1940s stucco bungalow downtown, the house where she’d grown up sharing a room with her little sister. Picking up holiday shifts at Purgatory’s rental shop, where her father had been outfitting skiers for 30 years, and where she’d been working and hanging out since she was a little girl, she met Ryan Moomey. Mooms, as some of his friends called him, was a full-time ski tech in the rental shop. In the summer he worked in the bike shop. He’d also grown up in Durango, but was four years older. He’d heard of Buchanan, but their paths had never crossed.

They became fast friends. They took to riding the employee shuttle up and back every day together and getting out for ski breaks whenever possible. (Buchanan’s boyfriend, Garrett Reigan, was on a climbing trip near Cerro Torre, in South America, for six weeks.) Moomey—thoughtful, soft-spoken, even-keeled—was taken by Buchanan’s energy and passion. He chased her all over the mountain and for the most part kept up. They talked about local backcountry lines and the stuff she was studying: snow crystals, weather, heuristic traps. “She would just be in this realm when she was talking about the classes she was taking,” he said. She told him how eventually the plan was to end up back in Silverton as an avalanche forecaster. In the meantime, they made plans to get in a ski tour up there when the rush was over, in January, before she shipped out.

Silverton was Buchanan’s default home. Her father, Evan Buchanan, worked for the gold mines in the early ’80s and co-owned a bar, called the Pride of the West saloon, in town. In the summer, he is vice president of operations for the famous narrow-gauge railroad that runs up from Durango three times a day. Her first season out of high school, in 2010–11, she spent the winter there, living in her family’s 1970 Nomad trailer, saving up money and skiing 85 days in the backcountry that season. She devised and talked her way into a joint internship between the Silverton Avalanche School and the Colorado Avalanche Information Center (CAIC). Among many other things, she helped with admin and hut logistics for a Level II avalanche course that Ben Pritchett, education coordinator for the CAIC, was teaching. Pritchett told me she was still green in the backcountry—to truly become an avalanche expert, he said, she would need not only to study and gain knowledge, but also to build field experience—but she was smart and tuned in.

It is hard to imagine a better place to start such a project. Silverton, as Pritchett put it, is “the avalanche hotbed of the U.S.” It was the site of the first avalanche forecasting center in the country, where the storied snow scientist Ed LaChapelle set up shop in 1972. Before then, ski areas and the DOT simply fired howitzers or dropped explosives on potential avalanche terrain to make it slide. What LaChapelle imported from the Swiss Federal Institute for Snow and Avalanche Research (SLF) in Davos, where he’d studied in 1950, was a budding science that promised to pinpoint when and where avalanches might happen, before they happened. “If you live in Silverton, you can’t not be around avalanches,” Pritchett added. “The culture of avalanche professionals in Silverton is an amazing culture to be involved with. And Olivia was deeply steeped in it.”

Four years later, she was that much deeper. “She was savvy. She knew snow really well,” her advisor, Jordy Hendrikx, director of the Snow and Avalanche Lab at Montana State University, told me. An affable, boyish Kiwi with a well-cropped beard and sleek Oakley eyeglasses, Hendrikx was working on a groundbreaking study using GPS and crowd-sourced data to better understand travel behavior in avalanche terrain. I’d missed him by about an hour at the Northwest Snow and Avalanche Workshop in Seattle that fall and so had flown to Bozeman to see what he was up to.

It was late March. We were standing below the south face of Lone Peak at Big Sky, above the Shedhorn lift. As his students—Buchanan’s classmates—wallowed around in their ski boots in the manky spring slush, measuring the alpha and deflection angles of a historic slide in 1996 (there’s still a patrol sled wrapped around the top of a tree there as a reminder of that event), Hendrikx brought up Buchanan. Originally, he said, she’d talked about drifting around a bit after she finished her degree, doing some climbing and skiing. She hadn’t always had the stomach for school. But that fall she’d changed her mind. She told him that after finishing her undergraduate degree, she hoped to join his master’s program in snow science to work specifically on decision-making in avalanche terrain. Hendrikx was pleased. “I have students I’m concerned about in terms of risk-taking,” he told me. “Olivia wasn’t one of them.”

Buchanan’s grandfather, her dad’s dad, Garth Buchanan, was thrilled to see her on that path. He’d grown up climbing and skiing in the Cascades, and after high school he’d served with the 10th Mountain Division in World War II. He was part of the epic, decisive assault on Riva Ridge in northern Italy in February 1945, which involved 2,000 vertical feet of mixed climbing—rock, mud, ice—at night with ropes, pitons, and heavy machine guns. (He would later earn a Bronze Star for another action.) During training at Camp Hale, Colorado, he told me, their method of avalanche protection was to tie a rope around one of the guys and have him jump up and down on the slope. “If it didn’t go,” he explained, “we skied it.”

When he returned to Washington State, Garth met his future wife, Olivia’s future grandmother, a girl by the name of Sue Nygren, who was an accomplished downhill ski racer, climber, and member of the Seattle branch of The Mountaineers, along with the already legendary Fred Beckey. She and Garth skied together frequently at Mount Baker, which was then just a rope tow and some big backcountry. Eventually, he received a Ph.D. in behavioral science and statistics at the University of Colorado Boulder. He and Sue married and moved back East, where his first job was on the research staff at MIT. Garth spent much of his career, before moving his family to Colorado once and for all, doing defense analysis work in Virginia, researching, among other things, human factors in the military and airline industries.

Over the holiday, just in the last few days, Olivia and Garth had had a long conversation about critical incident technique (CIT), a pioneering methodology developed in the postwar years by John Flanagan at the University of Pittsburgh and the nascent American Institutes for Research in D.C. The idea was to use statistical analysis to assess human behavior in defined situations. The earliest of these studies, based on thousands of interviews, were designed to look systematically at the performance of U.S. Air Force pilots and combat leaders: Why were some effective in critical situations while others failed? A half century later, an engineer and backcountry skier by the name of Ian McCammon, whose work Olivia had studied, would use essentially the same methodology, analyzing reports from more than 700 avalanche incidents over a period of 30 years (including 504 fatalities), to show the influence of six heuristic traps (investigated in-depth in The Human Factor 1.0), or psychological shortcuts, in critical decision-making that had led to the triggering of avalanches.

In Bozeman, Hendrikx and his co-author, MSU political scientist Jerry Johnson, took the study one step further. Not only had they begun to collect survey data, with the help of students like Buchanan, on a wide range of demographic and psychographic data—what were the particular characteristics of each group going into the backcountry, and how did they make their decisions?—but they were now also able to layer an objective and highly accurate set of crowd-sourced geographical data based on new technologies. With GPS-enabled smartphones, a downloadable tracking app and sophisticated GIS mapping systems, they began to assess not only what choices were made, but also how those choices were made in relation to specific terrain. And they vastly expanded the sample beyond just the groups who set off (and reported) avalanches to include a much broader majority of groups who went into the backcountry and came back, whether by luck or by skill, without incident.

“Until fairly recently, I was only interested in the geophysical aspects of snow,” Hendrikx told me. He was sitting on the edge of his desk on the ground floor of Traphagen Hall, an old brick edifice built nearly a century ago when the school was still called the Montana College of Agriculture and Mechanic Arts. “If I dig a pit here, how representative is it of the slope? How different are the spectral patterns? How do we do statistical models to develop lead time for forecasting on roads? But more and more I was starting to realize that I could spend the next three years teaching students solely about the minute detail of a snow crystal and they could still make a stupid decision and die. They also needed to understand the human-behavior aspects.”

He stood up then, in his excitement, and with a red dry-erase marker drew an upward curve on the whiteboard that looked very much like a classic avalanche slope with a dangerous convex roll at the summit. “There’s this great curve that’s often used to look at research funding,” he explained, “called the progress-to-funding curve.” He pointed to the top of the slope, where it began to flatten out, to indicate where he felt we stood in terms of what we know about the objective hazards and physical properties of snow. “We could spend a whole bunch more money and a whole bunch more time and make incremental increases in terms of our knowledge and the number of people we’re going to save.” On the other hand, he argued, with regard to what we know about human behavior, “we’re at the bottom of that curve.”

Buchanan touched base with Moomey a couple of times on the phone that Monday afternoon. They both knew conditions were not going to be spectacular; it hadn’t snowed more than three inches since the last decent dump, just before Christmas, and now it was heating up a bit. But he had the day off, and she had one day left, so this was it. The point was just to get out and have a good, mellow day in the backyard. Nothing serious. “That’s kind of part of backcountry skiing,” Moomey told me. “The snow isn’t always going to be amazing, but it’s still gonna be fun just to be out there.”

Buchanan and Moomey had read the forecasts. For several days, the danger rating hovered at moderate. There were serious concerns over a persistent slab deep in the pack, now compounded by new wind-loading over the weekend. Buchanan knew they’d have to be careful. But she also knew the terrain, had skied it countless times over the years. She’d seen nothing through the telescope that indicated instability. She knew others had skied it recently. On top of all that, the danger level, due to the dry conditions, was most likely going down.

She also knew about F.A.C.E.T.S., the acronym that McCammon coined as a tool to remind people that their judgment in these situations is likely under the influence of dangerously common biases: Familiarity (she’d skied this before multiple times in similar conditions); Acceptance (between friends, each wanted the other to have an awesome day); Consistency (this one, she knew, applied more to people who were overly committed to a specific objective; she was happy to bail at any point or to look at alternatives as needed); Tracks/Scarcity (this had to be the day); and Social Facilitation (there were those tracks off the summit!). But as long as they stayed on north-northwest aspects, she figured, as long as they avoided cross-loaded terrain features, dug pits with an open mind, paid attention and watched themselves, they’d be good.

“Let’s do it,” she said to Moomey.

The next morning, well caffeinated and in no particular rush, they left Moomey’s 1997 Jeep Cherokee parked at the road closure near the edge of town and headed up a well-beaten skin track on Lackawanna Road. The weather was calm and clear. They angled for the steep trees on the far side of Swansea Gulch, for the breaking sun and the long ridge climb to the summit. They had the world to themselves. The snow was old but still fairly soft. At tree line they stopped, took a break in the sun, shot some photos, and chatted for a bit. They hadn’t seen any slide activity. They dug a pit so Buchanan could do a compression test. She knew what she was doing; it was a pleasure to watch her work. “She went off about all this technical jargon,” Moomey remembers, “and I just smiled and pretended like I knew what she was talking about. She even got out a little magnifying glass and a snow card. She was really looking at the different crystals. She felt very comfortable with all the results. Nothing was sliding.”

Buchanan never seemed to get tired or hungry. She might’ve pressed on all the way to the summit, but Moomey had to stop and eat something. Around 3 p.m. they finally called it, not far from the top, where there was a decent entry point into Idaho Gulch. The climb had taken longer than expected; the sun was starting to hang low. Still, they had plenty of time—a couple of hours, they figured, before it was pitch black. And anyway, they had headlamps and extra clothes if it came to that. They dug another pit. Buchanan and Moomey did another round of tests and were satisfied with the results. They stripped their skins, buckled boots, and suited up for the ride down. Before she dropped in, she called her friend in town, the one with the telescope, and left him a message that he should look up and check out their turns.

“The snow was actually really nice,” Moomey told me. “Creamy, soft, old but dry.” They were both pleasantly surprised. They followed protocol, skiing down one at a time and watching each other. Everything felt great. When they got back down toward tree line, Buchanan suggested they traverse back across to the closer of the Rabbit Ears rather than continuing down to the far side of town. The Rabbit Ears were a more direct line to where they’d left the car; there was no reason to hike back in the dark. Plus there were more trees that way, more shade, so they figured the snow might be more consistent. And maybe safer, too. “You just feel safer once you’re in the trees,” said Moomey.

Sidehilling a fresh track toward the opening of the first chute, with Buchanan breaking trail, they found themselves walking onto a convex knob, the pitch falling away steeply below them. Moomey didn’t like the way it felt. Neither did Buchanan. They kicked their skis around and backtracked to where they could sidestep up and over the knob. They traversed to the other side, skier’s right, and came back around along the edge of the trees to where the slope wasn’t quite as steep. When they got to where they wanted to jump in, Buchanan was standing just below and a little ahead of Moomey. “She turned back,” he remembers. “She looked right at me and said, ‘OK, it’s my turn, watch me.’ Like nothing. Like we’d been doing all day.”

She turned away, slid her skis forward a few feet, and everything fell apart. “All of a sudden everything went in slow-mo,” Moomey remembers. “The whole area we were in dropped. Everything. I saw her body skiing away from me and then just drop. And then I didn’t see her again. It really still feels like I could’ve just grabbed her, like somebody tripping in front of you and you could reach out and catch them. She was right there.”

It took Moomey a few seconds to process what was happening. He remembers the whooshing sound, like snow sliding off a metal roof, the fracture line right next to his skis, the main crown arching upslope and across the whole mouth of the chute. Then everything sped up. All the snow rushing down, the whole slope moving quickly with Buchanan in it. He’d been caught in little sloughs before, had seen friends tumbled in small slides. This was not that. He started screaming her name: Olivia! Olivia! “I gotta start acting,” he told himself. “This is not some little thing. This thing is massive.”

He fumbled with his jacket, got his beacon out, and switched to receive. There was no signal. Looking down through the trees to where the chute doglegged to the left, he couldn’t see how far the thing had slid. He jumped into the path, below the crown. All he could think of was getting to Buchanan. It was rough going; the snow had pulled out all the way to the ground. As quickly as he could, he sketched his way down over grass and dirt, big chunks of hard snow, branches, rocks, and bits of trees. Eventually, about 600 feet below the fracture line, he got a signal. He tried to zigzag over the debris, still on his skis, around all the little new-growth trees, so he wouldn’t miss anything. The beeps started getting quicker. He was looking for a pole, a ski, anything. Closer… Closer… Then he realized he’d gone too far. He kicked off his skis, dropped his pack. With only his shovel and his beacon, he started charging back uphill. Closer…

Then he saw her.

Her body was wrapped around a tree, her purple ski pants out to one side, her arms and head flopped on the other. She still had her backpack on. Her skis and poles were gone. Her head was buried in the snow. He grabbed her by the backpack, pulled her off the tree and onto her back. He got her helmet and goggles off. Her eyes were open. “She was looking right at me,” he remembers, “like, ‘Help me,’ like she was scared.” She didn’t blink. He cleared the snow from her face and mouth and listened for her breath. Nothing. At which point he realized, “I need help right now. Five minutes ago.” Thus began an ordeal of horrifying, heroic proportions to try to keep the faltering ember that was Buchanan from going out completely.

He pulled out his phone. Amazingly, he had bars and was able to get through to 911. It was 3:50 p.m. The dispatcher set the search-and-rescue wheels in motion, and then, staying on the phone with him, proceeded to coach and encourage him while he started CPR. She counted for him while he did chest compressions and breathed into Buchanan’s mouth. Keep doing it, said the dispatcher. Stop for a short break, then start again. After a while, the dispatcher disconnected so that Kimmet Holland, director of Silverton San Juan County Ambulance Association, could get through. Everybody’s mobilizing, he told Moomey. There’s a Flight for Life helicopter coming from Durango. But it’s gonna take a long time for this process. The only thing you can do in the world for her right now, Holland said, is continue the CPR.

Moomey kept at it. During one pause he stood up and waved his jacket over his head so Holland could locate him with his spotting scope from the parking lot, 2,000 feet below.

, said Holland. Moomey got off the phone and went back to work. He did a set, then took a break. Nothing. Between sets, he talked to Buchanan. Come on, girl, stay with me, he said. You can make it. Time dragged. The sun went down behind the mountains. It got colder. He told Buchanan what was happening, how he could see lights flashing down below, fire trucks and ambulances. There were more cars pulling into the parking lot, a crowd milling around, snowmobiles trying to break a trail up the slide path. The whole town was out. “Olivia was starting to turn blue and purple,” he remembers. “Her skin started to feel really cold.” He took off his jacket and tried to wrap her in it. He tried to hold himself together. “At a lot of points I felt like she was already gone,” he said. “Like she was already dead. Then I felt like she might be there.”

Eventually the helicopter came up for a first pass and managed to drop four rescuers and a sked somewhere above. By the time they got down to Moomey, with the light pretty much gone, he’d been doing CPR for more than an hour. One of the EMTs found a weak pulse in Buchanan’s carotid artery. They wrapped her in blankets and strapped her into the sked. Mark Gober, a CAIC highway forecaster who’d also gotten to know Buchanan during her internship season, and had become a kind of career mentor for her, rode with the helicopter pilot on the second pass. Given the snowpack, he wasn’t comfortable landing the bird, so the rescuers had to break a trail and lower her down the avalanche path in the dark. They reached the staging area at around 7 p.m. Her dad had already driven up from Durango and was waiting for her. Eventually, after some time in the back of an ambulance, she was loaded into the helicopter for a quick flight to Mercy Regional Medical Center in Durango, where, with her mother, her sister, and aunt beside her, she was pronounced dead.