Become a Member

Get access to more than 30 brands, premium video, exclusive content, events, mapping, and more.

Already have an account? Sign In

Become a Member

Get access to more than 30 brands, premium video, exclusive content, events, mapping, and more.

Already have an account? Sign In



Perfect Chaos

How did the disaster that left 11 dead on K2 happen under blue skies, without a breath of wind? The complete account of three days of tragedy and heroism.

Lock Icon

Unlock this article and more benefits with 25% off.

Already have an Outside Account? Sign in

Outside+ Logo

25% Off Outside+.
$4.99/month $3.75/month*

Get the one subscription to fuel all your adventures.

  • Map your next adventure with our premium GPS apps: Gaia GPS Premium and Trailforks Pro.
  • Read unlimited digital content from 15+ brands, including Outside Magazine, Triathlete, Ski, Trail Runner, and VeloNews.
  • Watch 600+ hours of endurance challenges, cycling and skiing action, and travel documentaries.
  • Learn from the pros with expert-led online courses.
Join Outside+

*Outside memberships are billed annually. Print subscriptions available to U.S. residents only. You may cancel your membership at anytime, but no refunds will be issued for payments already made. Upon cancellation, you will have access to your membership through the end of your paid year. More Details

Chhiring Dorje could see the pinhole lights of two headlamps below him. As he rappelled the thin 5-millimeter cord, the antiseptic glow of his LED light revealed fresh avalanche debris in the snow around him. At the end of the rope, standing side by side, were two figures. It was Pemba Gyalje and Pasang Lama.

“Big problem,” Pasang said. He had somehow dropped his ice axe earlier on the descent. Now a serac avalanche, occurring about two hours earlier, sometime between 8:00 and 9:00 p.m. on their summit day, August 1, had wiped out more than 600 feet of the rope fixed that morning in the Bottleneck. A slip would quickly lead to oblivion. Pemba started his lonely descent. But without an axe, Pasang was stranded.

The air was calm. Somewhere above, 12 more headlamps were struggling to descend K2.

The annals of high-altitude mountaineering are filled with stories of storm and slaughter. But the tragedy unfolding in the darkness around Chhiring Dorje was different—it occurred under clear, cobalt skies. Less than 24 hours after the serac fall, 11 people would be dead. The mountain killed highly experienced alpinists as well as local porters and guides. It killed those using supplemental oxygen, and those who climbed without it. It killed by avalanche, exhaustion and simple slips on easy ground.

The Sherpa fished a knife from his pack, cut a length off the line above them, and tied one end to his harness. He tethered the other end to his friend. With only one ice axe himself and Pasang hanging from his waist, Chhiring faced into the glacial ice and began front-pointing down.


“K2 is the most beautiful of all the high peaks,” Reinhold Messner once proclaimed. “An artist has made this mountain.” Standing on the Pakistani-Chinese border, with four major ridges and four major faces, K2 is a perfect pyramid of alpine architecture. In 1979, Messner repeated the mountain’s first-ascent route, the southeastern Abruzzi Ridge, declaring a “five-day victory, alpine-style without high camps.” His statement belied a more complex, hybrid strategy. Messner’s team of five had established three camps and fixed lines to 7,400 meters before he and Michl Dacher made a lightweight dash to the summit, bivying once.

Messner understood the benefit of siege tactics, writing in his book All 14 Eight-Thousanders: “Whoever has humped up his own camps, installed his own camps, achieves a far more intimate feel for the mountain. … The difficulty of the mountain, the height, our own personal limitations do not permit us to go up into areas beyond our capabilities without outside help.”


For better or worse, high-altitude mountaineering has changed dramatically in the intervening 29 years. Eight-thousand-meter peaks are a limited resource, and as the popularity of the sport has ballooned in the past two decades, these 14 peaks have become increasingly crowded. Whereas only 100 people had summited K2 by 1994, the 40th anniversary of the mountain’s first ascent, 181 more climbers would reach the top in the next 13 years. Twenty-first century climbers have many logistical tools at their disposal—not only fixed lines and a burgeoning labor force of local guides and porters, but modern weather forecasting, advanced oxygen systems and satellite communication.

Even by recent standards, 2008 seemed destined to be a banner year. In early June, reported 10 expeditions had received permits for K2, making for a total of 50 expeditions and more than 300 people flocking to Pakistan’s five 8,000-meter peaks.

“Remember only a few years ago when K2 expeditions were rare, with several years leaving the mountain empty?” the website asked. “Ten years back only the most hardcore mountaineers went to Pakistan at all. Not anymore: With the recent events in China and Nepal, all of a sudden Pakistan seems the hub of safety and order while Himalaya climbing in general is taking off like never before.”

It’s difficult to describe the expeditions to K2, which involved some 60 people this year, by nationality. For instance, the Netherlands-based Norit team (named for their chief sponsor, a Dutch bottled-water company), which was the first to arrive in basecamp on May 29, had an Irishman and an Australian as well as five Dutch climbers. Likewise, a group led by the French expeditioner Hugues d’Aubarede had a Catalonian, two Germans and an American. International groups are common, as climbers pool resources to pay for permit fees and basecamp expenses. On the mountain, some of the groups would coalesce to work together as a single team, and others would not.

Roughly half of the expeditions planned on using bottled oxygen. Most teams carried at least one satellite phone as well.

Many of the expeditions supplemented their manpower by adding Pakistani and Nepali climbers. The experience levels of these men, and their responsibilities, varied considerably. The Korean expedition employed a large team of Sherpas, who assumed the traditional role of fixing rope and stocking camps on the lower Abruzzi Ridge. Two other Sherpas, Chhiring Dorje of an expedition led by Mike Farris of Sandstone, Minnesota, and Pemba Gyalje from the Norit team, were climbing K2 as equals with their Western teammates. Both had received financial help to pay for their share of the expedition, but were under no professional guiding obligations.

The Italians and Serbians had opted to use local Pakistanis. With no transportation costs or foreign permit fees, they are far cheaper than their Nepali counterparts, but are also considered by many to be less technically proficient. Despite later accusations of commercialism, only one person appears to have employed a guide in the traditional role. Hugues d’Aubarede hired two Pakistanis: Qudrat Ali, an exceptionally experienced Pakistani mountain guide, and Karim Meherban, who had worked on K2 before. The two men would accompany d’Aubarede on each of his forays up the mountain, but shared little in the rope-fixing duties on their route.

In the accounts that would follow, the Nepali and Pakistani climbers would be variously described as “guides,” “Sherpas,” “high-altitude porters” and, simply, “climbers.” The trouble in accurately titling them reflected their uncomfortable and ill-defined purpose on the mountain. Was their role to help shuttle loads down low, or to lead in fixing rope? Were they expected to coordinate efforts among the pluralistic teams, or merely take orders from their employers?

These questions remained unanswered.


Despite differences in tactics, almost everyone was intent upon the same real estate: the sweeping, rocky Abruzzi Ridge, or its more direct variation, the Cesen Route. (Only a French trio eyeing the West Face for an alpine-style ascent appeared to be seriously considering another objective.) Both lines ascend the mountain’s southeast aspect, tackling separate spurs that converge at a broad shoulder just below 8,000 meters. From there, they share the same path to the summit. More than 80 percent of successful K2 summiters have climbed the mountain either via the Abruzzi or Cesen.

As everyone undoubtedly knew, the amount of traffic on these two routes would virtually ensure a continuous line of fixed ropes leading from the Goodwin-Austin Glacier over 6,000 feet of vertical gain to a convergence at Camp IV (at 26,200 feet) on the Shoulder, and beyond, towards the summit. Yet fixing rope and establishing camps requires an enormous amount of labor and equipment.

By mid-June, the Norit team had single-handedly fixed all the way to Camp III (23,100 feet) on the Cesen. The Korean Sherpas, led by their sirdar, Jumic Bhote, were spearheading efforts to equip the Abruzzi Ridge.

Eric Meyer, a doctor on Mike Farris’s expedition, was struck by the demeanor of the Sherpas as they worked on the Abruzzi: “They were always pleasant, helpful, upbeat, never spoke bad about anyone,” he says from his home in Steamboat Springs, Colorado. “You can’t imagine a more classy group of guys doing a dangerous job.”

On the Cesen Route, however, some friction occurred as more expeditions arrived in basecamp and used the lines. Wilco van Rooijen, the Norit team leader, politely visited each camp and asked for $500 per climber as compensation to his team for time and equipment. Rice and D’Aubarede assented. Tension rose when van Rooijen returned a week later to d’Aubarede’s camp, carrying 1,200 feet of rope, and asked if Qudrat and Karim would fix the rest of the Cesen Route to the Shoulder. D’Aubarede responded that his team wasn’t adequately acclimatized, and did not plan on leaving basecamp that day, before new snow on the route could stabilize.

“For us, since we lacked proper acclimatization for a summit push, it made no sense to rush things and take additional risks,” d’Aubarede’s teammate Nick Rice, of Long Beach, California, would later explain on his blog. “(Wilco) became quite abrasive … accusing us of being lazy, and sitting in basecamp while they fixed all the way to the summit. We insisted that it wasn’t our fault that they had decided to arrive at basecamp a month before us and that they couldn’t expect any of us to be properly acclimatized. … People like him widen the rift between independent climbers and big commercial expeditions.”

However, none of the teams was commercial in the sense that they offered guided services; in fact, the only person who referred to himself as a guide was Qudrat Ali, Rice’s teammate, hired by d’Aubarede.

Eventually, the Norit team continued fixing rope to within a few hundred meters of the Shoulder, while Rice, d’Aubarede and others carried gear and acclimatized lower on the route. The Korean Sherpas and Serbian high-altitude porters, meanwhile, had established Camp III on the Abruzzi Ridge. After everyone returned to basecamp, the tension appeared to dissipate. It was only the first week in July, and both routes were almost fixed to 8,000 meters. The acclimatized parties could attempt to summit in the next spell of good weather.

After more discussions with Norit and the Serbians on July 6, Rice reported a spirit of cooperation: “The Serbians have three high-altitude porters and have 24 bottles of oxygen in basecamp. If they use oxygen, they will be able to work very efficiently when fixing the lines in the Bottleneck, and this will help us all.”

Everyone’s summit plans—even those of independent climbers—hinged on the advance work of fixing ropes toward the top.


Expedition politics soon took a back seat to concerns over the forecast. As July slowly passed, swollen bands of gray clouds smothered the Baltoro Glacier, alternately snowing and raining on basecamp. When the mountain cleared, clouds streamed around the summit. Massive avalanches boomed off the South Face, dusting several parties as they carried loads low on the mountain.

The climbers first welcomed a chance to rest, but soon chafed. “Weather Despair,” Gerard McDonnell, the Irish member of the Norit team, titled a report posted on July 16. “Crappy Weather Arrives,” Mike Farris’s expedition blog reported on June 20. Some climbers’ return flights were rapidly approaching; others worried that so much time at basecamp would diminish their strength.

One gray morning, as the snow swirled sideways across the talus outside his mess tent, Hugues d’Aubarede decided that he had had enough. The mountain had been obscured in storm for nearly two weeks, and virtually all the forecasts agreed that it would stay that way for some time. A dip in wind slated for July 26 might signify a high-pressure system, but probably not. He turned on his satellite phone and called his trekking agent to arrange for porters to come to basecamp and help carry down his gear. On July 20 he called his family and told them he was done with K2.


Two days later, the subtle dip in wind had morphed into a staggering void: Good weather was coming. The mere mention of a summit window boosted the sagging basecamp morale. Rice wrote, “Four days would be perfect for a summit push on our route. D’Aubarede, who had already decided to head down, and whose porters were on their way, was not at all pleased.”

While d’Aubarede considered what to do, the other teams recommenced strategizing. On July 25, representatives of all expeditions met to discuss their plans for the summit push. The teams climbing the Abruzzi Ridge would leave first, on July 27, since they needed an extra day to finish fixing rope from Camp III to the Shoulder. The Cesen teams would follow a day behind, so that everyone would converge on the Shoulder on July 30. July 31 would be summit day.

An efficient plan to fix ropes above Camp IV was paramount to group success. Though Dong-jin Hwang of Korea, who was instrumental in organizing the basecamp meetings, would be nominally in charge of the advance party, the de facto leader of the group was Sheehan Baig, an experienced high-altitude porter from the Serbian team.

“Sheehan had relationships with each expedition,” Chris Klinke of Farris’s team, recalls from Unadilla, Minnesota. “He was the guy everyone was turning to for guidance, and he was the only person who had summited K2 before.”

After waiting to see an updated forecast on the day of the meeting, d’Aubarede with difficulty decided to stay. His guide, Qudrat, however, needed to return to Skardu to lead another expedition. At the last minute d’Aubarede found a replacement, another Pakistani, named Jehan Baig. Jehan had arrived as part of the Serbian team but been dismissed under unspecified circumstances.

The meeting, Gerard McDonnell wrote, “had primarily a humorous tone.” The jolly Irishman was a popular figure around basecamp, and an experienced hand in polar expeditioning, having traveled to Antarctica with Rolf Bae, a member of the Norwegian trip. “Hopes are high,” McDonnell wrote. “It was joked that the next meeting would take place at Camp IV.” The climbers raised their glasses in a toast to success.


“Everything was going well to Camp IV,” Wilco van Rooijen later reflected to the Associated Press, “and on the summit attempt everything went wrong.” The ascent up the fixed lines on the Cesen Route to the Shoulder was fairly routine. After nearly a month of waiting, everyone was thrilled to be moving toward the summit. Yet delays and warning signs suggest that the climb was not going as planned.

On July 28, the Cesen climbers moved to Camp II (20,800 feet). They reported strong winds and light snowfall. “All is not well,” the Norit expedition blog read the next morning. Still, though cold and gusty, the conditions were passable. From the Abruzzi Ridge, however, the Koreans reported via radio that winds were too high, reportedly 80 mph, to continue to Camp III; they would wait a day at Camp II instead.

The Norit team faced a difficult decision—push back their summit day until August 1, or skip the rendezvous with the Koreans and the other Abruzzi groups on the Shoulder and continue alone. This second option would mean making their summit bid without the ropes that the Koreans had pledged to bring, or extra manpower from the Sherpas on the Abruzzi. After some debate, the Norit team remained in Camp II.

Everyone expected the winds to decrease the following day, July 30, with the arrival of the forecasted high-pressure system. They only increased. “Knetterharde wind. Strong winds!” the Norit team reported from Camp III. “The team fears they may lose the tents.” The next morning, the winds finally abated.


At 10 p.m. on July 31, a small band of porters and guides clustered outside the tents of Camp IV, fumbling in the darkness as they tried to sort gear by headlamp. Sheehan Baig, though, the experienced K2 summiter who was supposed to lead this first group of Sherpas, Pakistanis and two Italians, was nowhere to be found.

Pemba Gyalje questioned the Serbians and learned that Sheehan and another of their high-altitude porters had suffered intractable vomiting, likely the result of food poisoning, at Camp II. The two porters had descended, though their loads apparently were never redistributed to the others. As he sorted loads, Pemba made an even more unsettling discovery. Of the 2,400 feet of rope the group planned on using, roughly half was missing. He rustled through various backpacks and gear, and woke up others, including Rice and Meyer, who were resting inside their frigid tents, to ask about it.

By now it was midnight, and an unexpected visitor joined the group. The Basque climber Alberto Zerain had left Camp III on the Abruzzi Ridge at 10 p.m. Climbing solo with a light pack, he reached Camp IV in only two hours. Zerain called out to the others still in their tents, trying to cajole them into hurrying up to leave with him. He received few responses.

Pemba, meanwhile, suddenly found himself cast as the new leader of the advance party, though he was under no obligation to guide on K2. How much this new responsibility weighed on Pemba is not known. But at the last minute, Pemba convinced his friend Chhiring Dorje Sherpa, himself an equal member of Mike Farris’s expedition, to come along with him.

After an hour of waiting, Zerain finally continued alone. A few minutes later, Pemba and the advance party gave up their search for the missing equipment, and followed, now nearly three hours behind schedule.


From Camp IV the Shoulder gradually steepens from a broad snowfield into a narrow couloir known as the Bottleneck. Some 600 feet long, the Bottleneck averages between 50 to 60 degrees in angle and is threatened from directly above by a large serac 400 feet high. To avoid this serac, the route makes an icy traverse to the left, below the ice cliff and above the highest band of rock cliffs that form the edge of the South Face. Beyond the traverse, more than 900 feet of steep snow remain to the summit.

Breaking steps toward the start of the Bottleneck, Zerain found difficult conditions, with deep snow that would soon force him to dig for anchors. He climbed the Bottleneck unroped; at times the tattered remnants of old line provided additional security. In his steps, the Sherpas, led by Chhiring and Pemba, followed quickly. It’s unknown who took responsibility for deciding precisely where and when to fix ropes, but it’s difficult to imagine those decisions being a priority to Zerain, who was essentially climbing solo.

Meanwhile, a long string of headlamps steadily left camp. The four Norwegians departed at a little after 2 a.m., followed by d’Aubarede and his two Pakistanis, four members of the Norit team and a group of Serbs. Though these climbers were intended to leave three hours after the advance party, the delays had put them less than an hour behind. Using the kicked-in steps, they quickly began to catch the lead group.

Zerain and the first Sherpas reached the top of the Bottleneck at dawn. Above them loomed the serac: hanging pieces of ice, boxcar-sized, that glowed with ominous clarity in the early-morning sun. According to an account by Zerain later published on Spain’s El Mundo website, the Sherpas hesitated, as there were no fixed ropes on the traverse. Zerain asked for two ice screws, and led across brittle 70-degree serac ice interspersed with rocky sections. He fixed the last of the rope, and waited for the first Sherpa to bring his camera. Then he said that he had to keep moving and would continue on from there unroped. The implicit message was he was also done fixing rope.

By now, the main body of climbers had caught up with the last members of the advance group. In the Bottleneck, they found powder snow over blue ice. There were shouts from above to send up more rope, and climbers cut up pieces of the new and old fixed lines and passed them upward through the line. Chris Klinke, one of the few to delay his departure from Camp IV to give the advance party more time, found a traffic jam at the base of the Bottleneck: “I tried to count the line up of climbers ahead of me and realized that it was over 20 climbers,” he later wrote on his blog. The line, he said, was at a standstill. He waited for almost two hours, and finally turned around.

Chhiring Dorje, who evidently had waited for his American teammate Klinke to catch up, describes the chaotic scene in an e-mail: “While climbing in couloir, I saw one stone come out from the serac and everybody [was] afraid … Wilco was before me; he slipped down and I caught him.”

A moment later, Dren Mandic, a Serbian climber, fell. He apparently had unclipped from the line, possibly to adjust his oxygen system or to pass a climber. “He was behind the Norway girl [Cecilie Skog],” Chhiring recounts. “She shouted and I looked [but] he had already [fallen].”

Climbers in Camp IV saw a small human form tumbling 600 feet down the slope leading to the Bottleneck. He came to rest on the south side of the Shoulder. A group from camp rushed over, but by the time they reached him, he had died.

Also trying to help was Jehan Baig, one of d’Aubarede’s two Pakistani hires. According to Nick Rice, Jehan Baig, the last-minute replacement guide, had suffered nagging altitude sickness on the climb to Camp IV. That morning, it had taken him nearly half an hour to strap on his crampons, and evidently he had decided to turn around at the Bottleneck. Jehan Baig was just beginning to descend when Mandic fell, and he down climbed to where he lay.

The group initially decided to recover the body. But as they prepared a lowering system, Jehan Baig, who wasn’t carrying an ice axe, slipped. He slid towards the South Face, clawing at the wind-hardened snow, falling feet first for 50 or 60 feet; then his crampons caught and flipped him headfirst. He accelerated, and disappeared over a serac. The group abandoned its efforts and sadly returned to camp.


By early afternoon, despite being hours behind schedule and having witnessed two deaths, more than 20 climbers were pushing past the traverse and on toward the final summit slopes. Klinke was descending to Camp III with Roberto Manni, who had also turned around, when he felt a foreboding sense upon his last sight of the Bottleneck. “I stopped to take a picture. ‘This isn’t good,’ I said to Roberto. The people had only moved [150 feet] from when we left Camp IV, and that was an hour and a half ago.” It was 3:00 p.m.

At roughly the same time, Alberto Zerain arrived on the summit of K2. He had started from Camp III, and opened the entire route from Camp IV to the top without oxygen—a singular achievement. He was also the only climber to reach the summit on schedule. As Zerain began to descend, he was surprised to notice such a large group of climbers still heading up. “They were far from me, hours away, but they were not turning around,” he would say on El Mundo. “But then I thought, they know what they are doing.”

At 5:30 p.m., five Koreans and two Sherpas stood on top. Jumic Bhote, the Koreans’ head Sherpa, asked to borrow a sat phone to call his pregnant wife in Nepal. Two of the Norwegians, Lars Nessa and Cecilie Skog, soon arrived. All used supplemental oxygen, though Skog had given her system to her husband, Rolf Bae, who was apparently struggling on the final terrain.

An hour later, Chhiring Dorje summited, becoming the first Sherpa to climb K2 without oxygen. Next was Ireland’s first K2 summiter, Gerard McDonnell, at 7 p.m., followed by Hugues d’Aubarede and his remaining Pakistani porter, Karim Meherban. Last to arrive were two Dutchmen, Wilco van Rooijen and Cas van de Gevel, the Italian climber Marco Confortola, and Pemba, who had been one of the first to leave Camp IV more than 20 hours earlier.

A little after dusk, observers in basecamp could clearly see the lights of several headlamps descending. When somebody in basecamp signaled with a blinking flashlight, the lights responded in kind. Then they passed out of view. Of the 18 climbers who summited, only Zerain had negotiated the traverse and descended the Bottleneck in daylight.

Leading the pack were the Norwegians Nessa, Skog and Rolf Bae. Bae had turned back only 300 feet below the summit, but waited for his wife. The trio began the difficult task of reversing the traverse at dusk, with Bae in the lead. Skog traversed next and heard the sickening roar of a large avalanche in the darkness. A second later, Skog was wrenched off balance as the rope she was clipped to broke somewhere below. Bae’s headlamp disappeared.

Skog called out in the black night for her husband, but got no response. After waiting some minutes, she and Nessa had little choice but to continue down. They pulled out an emergency cord they had carried, fixed one end to the torn end of the fixed line and rappelled into the Bottleneck. Accounts vary, but the avalanche had taken out what the Norwegians believe to be 600 feet of rope, and Chhiring later estimated as 1,500 feet. Carefully, the two remaining Norwegians down climbed toward Camp IV.


Chhiring Dorje, the sole member of Mike Farris’s expedition to summit, was one of the first to arrive at the traverse after the avalanche. In the Bottleneck, only the Norwegians’ 50-meter rope remained, just enough to rappel into the top of the couloir. At the end of the line, Chhiring met Pemba Gyalje and Pasang Lama, also called “Little Pasang” to distinguish himself from his cousin “Big Pasang” Bhote. He had lost his ice axe.

“[Little] Pasang Lama was worried, but I said don’t worry,” Chhiring recounts in his e-mail. “We have only two options—one is staying here, which is very dangerous under the serac. The other option is to descend down with one ice axe, which may lead us to Camp IV … if we don’t slip.”


It’s difficult to imagine the confusion and panic when the other climbers above learned what had happened. Night had fully descended, and the new moon provided precious little light. Everyone had been on the go for more than 20 hours. Those on oxygen were running out. Furthermore, the climbers were spread out over a wide swath of terrain; they never had a chance to congregate.

“People were running down but didn’t know where to go, so a lot of people were lost on the mountain on the wrong side, wrong route,” van Rooijen would tell Reuters. “They were thinking of using my gas, my rope. So actually everybody was fighting for himself and I still do not understand why everybody were leaving each other.” It’s unclear exactly who van Rooijen was referring to, but the statement speaks to the chaos on the mountain. Any sense of group teamwork had completely broken down.

The climbers faced a bleak choice—to attempt to down climb the Bottleneck unroped by headlamp, or sit out the night at 8,300 meters. Most carried only one ice axe, and the traverse into the Bottleneck featured short stretches of glacial ice approaching 70 degrees. It is unknown if anyone had an extra rope, as the Norwegians did. On the other hand, waiting until daybreak meant enduring temperatures that plunged to minus-40 degrees amid the thin air perpetually weakening their bodies.

After the Sherpas, Cas van de Gevel and two of the Koreans, Go Mi Sun and Kim Jae Soo, down climbed the Bottleneck that night.

Gerard McDonnell, Marco Confortola and Wilco van Rooijen bivouacked together above the traverse at 8,300 meters, in bucket seats carved into the steep snow. Confortola, speaking later to a reporter in Islamabad, described the frigid night: “Since Gerard was having a difficult time, I made his hole bigger to help him lie down for a little bit. Gerard was very cold. I was also cold and began to shiver.” The three remaining Koreans (Hyo-gyung Kim, Kyeong-Hyo Park  and Don-jin Hwang) and Jumic Bhote bivied as well, slightly lower on the mountain. Nobody knew where Hugues d’Aubarede and Karim Meherban were.


At 4 a.m. August 2, Roeland van Oss at the Norit basecamp received a call from Camp IV. Eight people had succeeded in down climbing the Bottleneck that night after the avalanche. That left nine unaccounted for.

One ominous possibility, however, presented itself. According to Chris Klinke, van de Gevel later said he saw a shape fly past him as he was down climbing the couloir. It was only for a moment, it was the dead of the night and he was exhausted, but van de Gevel thought he glimpsed d’Aubarede’s backpack. An in-depth piece in The Independent of London states that he told colleagues he saw d’Aubarede fall, and the article reports that Karim fell shortly after, but later accounts also suggest other possibilities.

Roeland immediately began to mobilize a rescue effort. Support staff and other climbers pitched in to provide fresh batteries for the communication equipment and prepare a medical tent in basecamp—though realistically the only people in position to respond quickly enough to rescue any survivors were those exhausted climbers already on the upper mountain.


At dawn, the three lost Koreans and Jumic Bhote could clearly be seen from Camp IV. They were just above the traverse, bivouacked at 8,350 meters, moving little. Two rescue teams left Camp IV at 8 a.m. Pemba Gyalje, accompanied by Cas van de Gevel, was looking for van Rooijen. Tsering Bhote and Big Pasang Bhote went to rescue their brother and cousin Jumic, and the Koreans.

A bit later, probably between 9 and 10 a.m., according to Eric Meyer, observers in Camp IV noticed another human form, this one silhouetted at the very top of the serac. The person was clearly off route. “They seemed to be pacing back and forth, like they were trying to find a way down,” Meyer remembers. Lacking binoculars or optical scopes, nobody could be sure who it was. Chris Klinke speculates that, if d’Aubarede fell down the Bottleneck in the night, the only person unaccounted for was Karim Meherban. The question remains unresolved, but Karim never returned.

Soon after, clouds began to roll in. The weather, which had been perfect for the last 48 hours, was changing. By noon, an ominous cloud capped the upper mountain, shrouding any survivors in shadows.


Meanwhile, Confortola, van Rooijen and McDonnell continued their descent, with van Rooijen evidently moving a little ahead. Having survived an open bivy at nearly 8,400 meters, they were functioning at the extreme limits of consciousness. They came upon the three Koreans, lying roped together in a state of total exhaustion (accounts do not mention Jumic). The three ambulatory climbers tried to revive them, but eventually were forced to move on. But as Confortola resumed his descent, McDonnell abruptly turned around and began climbing back up the mountain, according to The Independent. No one knows what came over McDonnell, but his friends reason that he was still trying to help the Koreans.

Pemba and van de Gevel returned to their tent after searching in vain for van Rooijen. Then the radio crackled to life. It was Tsering and Big Pasang Bhote. They had found a man passed out in the snow at the edge of a fresh avalanche cone. “I said what is the color of [the] suit, they said green and black [and] then I understood that was Marco,” Pemba states in an e-mail.

Tsering and Big Pasang Bhote were still intent on finding their teammates above the Bottleneck, so Pemba set out again, to find Confortola. A few minutes before locating him, he asked his Sherpa friends if they saw anyone else. They radioed that they had just seen someone in a red suit fall from the traverse.

Marco Confortola would recall passing out in the snow after down climbing the Bottleneck. He claims to have been awakened by the roar of an avalanche, and opened his eyes in time to see McDonnell’s boots tumbling down amid the debris. The Italian continued his descent, only to pass out again. It was then that he was found by the two Sherpas. “If Marco was right, then Gerard was swept by avalanche a few hours before,” Pemba writes, saying that in that case, the red suit “could be Karim.” He, McDonnell and Karim all wore red.

Reaching Confortola, Pemba administered oxygen from an emergency bottle at three liters per minute. Confortola revived, and after a few minutes the diminutive Sherpa was helping the once-strapping Italian down the mountain. Then the radio squawked from underneath his down suit.

Tsering Bhote and Big Pasang Bhote had re-climbed the Bottleneck. Tsering Bhote was at 8,200 meters and Pasang Bhote had reached the Koreans and his cousin Jumic at their bivy at 8,350 meters. Somehow, Big Pasang had managed to revive Jumic and two of the Koreans. They were all descending high in the Bottleneck.

“I said OK, go fast as soon as possible and safely,” Pemba recalls. And then, with brutal precision, the serac let loose again. Pemba heard the noise first, a deep, inevitable rumbling. Several seconds later, a fusillade of serac blocks tumbled toward him. Gravity steered the main flow of the avalanche to his left. Smaller pieces of ice landed around him and Confortola.

Pemba saw bits of color rolling down in the stream of ice. The broken forms of Big Pasang Bhote, Jumic and the two Koreans swept past. Some accounts put McDonnell’s body in this avalanche.

A small piece of ice, as Pemba recounts, drove into the back of Confortola’s head. Confortola stumbled, and began falling. From just behind him, Pemba groped at the bubbly plastic down suit, and he held on.

In the sudden silence following the maelstrom, a lone voice called from above. It was Tsering, somehow still in the Bottleneck.

“Come down quickly!” Pemba shouted. Below, he saw Pasang Lama slowly ascending with two of the Koreans to help the effort, and called out to him. They waited until Tsering safely crossed the debris field to join them, then turned their backs on the Bottleneck, and started down.


Van Rooijen was still unaccounted for. Evidently, the Dutchman had down climbed the Bottleneck ahead of Confortola, but became lost as he tried to find Camp IV. “I was on the wrong side of the mountain,” he later told the National Geographic Adventure webpage. “People at basecamp saw me go over the wrong side of the ridge and they radioed people in Camp IV. I had to sit out a whiteout because I couldn’t see anything and I knew I couldn’t go down any further.”

By late afternoon, Pemba, Tsering Bhote and Little Pasang Lama had returned to Camp IV. While Confortola received first-aid treatment, a call came in from the Norit basecamp. Amazingly, the climbers there had received several satellite phone calls from van Rooijen, who reported that he was lost on a south-facing wall. Because of the cloud cover, he couldn’t determine his location, nor could basecamp find him. Then, at 5:45 p.m., the skies cleared. Through a telescope, Chris Klinke spotted an orange dot below Camp IV but well to the left of the Cesen Route, about a third of a mile left of and slightly above Camp III.

Pemba and Cas van de Gevel left Camp IV late that evening, hoping to signal the lone survivor back to the fixed ropes. Descending by headlamp, Pemba reached Camp III at around 2 a.m., having become separated from van de Gevel sometime during their descent. Strangely, Pemba thought he could hear van Rooijen’s satellite phone ringing from somewhere in the darkness. But the Sherpa was beyond spent, and could not search in the night. He crawled into a tent to wait until morning.

At dawn, basecamp located both van de Gevel, who was descending the fixed ropes to Camp III, and van Rooijen, who had traversed across the south face to within 500 feet left of it. But his position was still hidden from Camp III. When van de Gevel reached camp, he woke up Pemba, and the two radioed basecamp. After receiving directions, they set out after van Rooijen, and spotted a forlorn figure front-pointing towards them, driving his fists into the snow for security. His face was burned and blistered, but they recognized their teammate.


Of the nine climbers who had spent the night above the Bottleneck, only two survived: Confortola and van Rooijen. As the badly frostbitten men hobbled off the mountain and were shuffled into helicopters to be flown to Skardu, they related the piecemeal imagery and incoherent visions of their survival.

How had K2 killed 11 people so quickly?

The survivors initially pointed to logistical errors and breakdowns in communication. “The biggest mistake we made was that we tried to make agreements,” Wilco van Rooijen told Reuters. “Everybody had his own responsibility and then some people did not do what they promised.” In another widely quoted interview, this one with the Italian news agency ANSA, Marco Confortola agreed: “A 656-foot rope, very light but strong … was not brought by a somewhat sloppy porter, which was just the beginning of the problems,” he said, in an apparent reference to the equipment left at Camp II when Sheehan Baig descended ill.

Yet the lack of equipment is symptomatic of a greater problem. The teams tried to cooperate by pooling their resources, but no one appears to have taken command of overall strategy: “You had a summit team, but there was no clear leader,” says Meyer. The critical lack of organization can even be traced to a simple, guiltless event that occurred days earlier: Sheehan Baig falling sick with a stomach bug at Camp II. Many were counting on the personable, experienced Pakistani guide from the Serbian team to orchestrate the efforts of the advance group on summit day. When Sheehan Baig was forced to descend, not only was important material left behind, but a capable leader was lost.

The sheer number of people crowding the route exacerbated communication problems. With more than 20 people ascending the Abruzzi Ridge, and another 10 on the Cesen Route, few even noticed Baig’s absence.

“I didn’t know Sheehan had turned around until Camp IV. It wasn’t communicated to us,” Klinke recalls.

Another contributing factor appears to be language. Though they communicated in English, none of the advance party was native speakers. This seems critical in explaining why Sheehan Baig’s absence was never conveyed to the rest of the group, and how half of the fixed line was left behind. “There were five or six different languages being spoken,” recalls Klinke. “Korean, Urdu, Nepali, Italian, English, Dutch—no wonder people were confused.”

Last, the masses created delays on summit day. “The group was so tightly spaced in the couloir,” Eric Meyer says. “And they were moving so slowly, one on top of the other.”


Another interpretation of the tragedy, one that received little attention in the mainstream media but has been quietly discussed within climbing circles, is whether it is best to fix ropes on summit day at all. Some argue that anyone attempting K2 should be capable of climbing through the Bottleneck unroped, or moving in self-sufficient teams. This was the accepted strategy until the mid-1990s.

“There were no lines above high camp nor any sign of lines in 1993,” says Phil Powers of Boulder, Colorado, who climbed K2 on July 7 of that year. “We climbed without ropes on our summit bid.” Powers remembers the Bottleneck and traverse as being straightforward snow climbing of perhaps 50 degrees.

Thor Kieser of Denver, who journeyed to K2 in 1992, agrees. “Sure, people get killed by seracs and avalanches all the time—but a bunch of people stuck above the Bottleneck, incapable of down climbing? That’s unbelievable.” On his attempt, Kieser climbed solo the entire summit day. “At 2 p.m. I was 300 feet from the summit, breaking trail through knee-deep breakable crust … so I turned around.” His decision is a humbling example of true mountain judgment and control.

Indeed, Alberto Zerain, climbing solo this year, was able to summit and descend before the serac avalanche

Yet Chris Warner of Annapolis, Maryland, who climbed the mountain in 2007, disagrees with the notion that climbers should not use fixed ropes. “We needed the resource of people to break trail and fix ropes. We had it worked out so that different groups left Camp IV three hours apart. The first group fixed to the Bottleneck, and then we made sure we were in charge of fixing the hardest part: the Bottleneck and traverse. … Wouldn’t you want to be in control on the most dangerous climbing day of your life?”

Regardless of the equipment shortages, the breakdown in leadership, and whether fixed ropes should even be used, each climber was ultimately responsible for the decision to continue or turn around. Many climbers wisely abandoned their summit push in light of the delays. Why did others continue?

With so many going for the summit at once, a herd mentality may have developed. Perhaps some people lost track of time. Perhaps others were aware of the late hour, but rationalized that with fixed ropes, they could feasibly descend in darkness. The weather was perfect, without a breath of wind. If there were ever a time to gamble the summit of K2 against a nighttime descent, this would have been it.


Why the serac became suddenly and incredibly active is a final question. Meyers points to the weather. “The week earlier, winds were 80 to 100 miles an hour on the summit, so it was a big weather change,” he says, noting that there was no debris or signs of recent activity when climbers arrived at Camp IV. The next morning, summit day, became eerily warm. “It was intense, really intense. … Everyone peeled off their down suits in Camp IV. It was obvious the serac was starting to get irradiated.”

Other survivors are more dismissive of the conditions and the serac: “You know if you’re going to climb K2 that you are willing to face these risks,” van Rooijen said.

Yet at least six—Jumic Bhote, Tsering Bhote, Pasang Bhote, Pasang Lama, Karim Meherban and Jehan Baig—of those who took those risks were paid professionals, not private adventure seekers. Four paid with their lives.

If any greater understanding can come from a season that killed so many, it must come from the heroic actions of the Sherpas. Though members of different teams with different responsibilities, they decisively came to the aid of those around them.

“The only people who acted like a team,” Chris Klinke says, “weren’t a team at all.”

The more that is revealed about what happened, the less any one cause seems to have been singularly decisive. The true tragedy of K2, the lasting meaning of the disaster, lies hidden in the indecipherable human motives that led each person towards the summit.

Consider d’Aubarede, who changed his ticket and decided to leave, only to reconsider and stay to give K2 one more try. Or Jehan Baig, who had been fired by the Serbian expedition only to be hired at the last minute by d’Aubarede. Rolf Bae turned around short of the summit, but waited for his wife and teammate to catch up so that they could descend together.

Jumic Bhote, after speaking to his pregnant wife from the summit, apparently refused to leave his clients in the night. By the time the last survivors limped into basecamp, he too was dead, and his wife had given birth to a baby boy.

K2 killed, as it always has, by indiscriminate acts of violence.

Freddie Wilkinson is a writer and guide who lives in Madison, New Hampshire.

This article appeared in Rock and Ice issue 173 (December 2008).