Getting Caught: The Avalanche In Kittelfjäll
A look at the first-hand account of how a strong group of Swedish skiers got caught in a slide
“On February 16th 2014, Martin, Maria and I remotely triggered an avalanche in our “home” mountain range in Kittelfjäll, Sweden. The avalanche broke about 100 m above us and caught and partly buried us all. The crown was roughly 300m wide and the path was about 400–500 m. In some miraculous way, we all survived and got away with minor physical injuries — I broke both of my legs, got a crack in my chin and a hell of a black eye, Maria got a dislocated femur fracture, while Martin strained his groin and got a small bruise. In time, our bodies and minds will heal, but I suspect that the process will be long and winding. There are so many questions banging against my forehead: How could we let this happen? Where did we go wrong? So many emotions swirling around in my chest; guilt, fear, shame, and an immense gratefulness of being alive”
— Andrea Mannberg*
Case studies are a great way to learn from the experience of others. Often, there isn’t a smoking gun that points to an obvious error of judgement; more complex incidents that involve experienced and skilled parties are generally much more nuanced in their causation and combine a number of factors and small errors of judgement that lead to an incident.
The Kittelfjäll case study explained below is one such event. Could you find yourself in a similar situation?
A Casual Day
On Sunday 16th February 2014 three close friends made the long five-hour drive from their home town to the Kittelfjäll backcountry in Sweden for a day of backcountry skiing. They were a solid team that skied together frequently, both in the local ski areas and on further afield trips to world class backcountry ski destinations. They all had good avalanche skills and even some formal training in the form of a three-day avalanche fundamentals course taken the year before in Jackson Hole.
As a team they knew each other well, they’d talked through the current snowpack issues together and they carried and knew how to use all the right avalanche equipment: transceiver, shovel and probe, as well as ‘Air Bag’ packs.
They hadn’t really consulted the avalanche forecast for the area, but they knew it was Considerable (3), and had been firmly stuck at that hazard rating for several days. In the words of one of the group:
“The avalanche forecast for the area was always posted a considerable hazard rating, so we never gave it much attention”.
If they’d been tracking the forecast from the nearby ski resort posted three days before their ski tour, they’d have been alerted to a persistent weak layer (PWL) some 20 cm below the surface that was reactive and providing clean, planer failures with energy when subjected to testing as the video shows.
Hidden Dangers
The visibility was poor as the group set off. This isn’t uncommon in Scandinavian countries and Scotland where the culture is to get out into the mountains in flat light and high winds, otherwise you could go weeks without skiing anything.
In the flat light the team made the obvious decision to stay below the high alpine part of the mountain, in the sparsely spaced birch forests that are endemic in this part of the world in order to provide them with the contrast needed to ski.
Visibility is also a key factor in assessing the gradient of a slope, and the birch assisted them in staying on slopes that angle didn’t go above 30 degree’s. Nor was the terrain they had in mind particularly complicated. No glaciers, no overlapping avalanche run out zones, and no obvious terrain traps such as cliffs or river gullies. These felt like safe slopes.
What Happened?
This is how things unfolded in the words of one of the party who wrote candidly about the incident:
“It had snowed an additional 10 cm, pretty good for Sweden and the snow was colder than the day before. It looked like we were going to have a good day. We easily found the tree’d small and low incline ridge that would safely take us to our run. On the flat ground, we heard several “whumps”, some of them big. To get a sense of the instability, we jumped on a test slope on the side of the ridge — about 35 degrees. This produced a settlement but no movement in the snow.
When the slope got a bit steeper the whumps stopped. Since we knew that the snowpack was a bit unstable, and since visibility was poor, we had no intention of skiing the big alpine terrain. However, we also wanted a run of decent length, so we aimed for the tree line. As we approached it, we all agreed that this was not the day when we wanted to have big snow fields above us, as the current snow condition implied that there was a heightened risk to remotely trigger an avalanche. Since we needed to traverse to get to our run, we stopped to discuss the way forward. We had a small open snowfield above and below us that we needed to cross. After a lengthy discussion, where everyone had the right to say ‘no’ we decided to skin one at a time over to the other side. Martin went first.
While we waited, we isolated two columns of snow with our poles and did hand shear tests. It was very, very hard to get the snow to go anywhere. It seemed to have bonded really well with the old snow, however, when we put our full weight on the snow and pulled hard it came out with a very clean shear. Martin’s traverse went well, and we followed one at a time. Once at our run that we planned we decided that it was too convex and steep, and too uncertain as to where we would end up, so we decided to turn back. One at a time to the place where we had started. All good. We decided that we would try to get to our usual run. It would be a bit of a slog, but we would at least get one good run. We discussed how to get there, if we should try to stay high or go through the trees and skin back up. Since we didn’t have that big snow field above us, we choose the latter. Better safe than sorry.
And then we got greedy. Going down the way we got up would basically mean that we would just skin down, that seemed a bit boring. Instead we choose to traverse just a little bit to the west, where a sweet little snow field with a mellow incline opened up and give us a good run. Visibility was relatively poor, so we really couldn’t see what we had above us. I saw something that looked like a cornice that worried me, but since we were heading back I didn’t really worry about it.
Martin had only gone a few metres in front of us onto the snow field when we heard the whump. I remember Martin saying “this is us turning back, this is no place to be”, and starting to come back towards me. I did the same. Then I saw him running towards me. “WTF”. Then, from the back of my eye, I saw a huge wave if snow charging down the mountain. When I realised what was happening I tried to grab my ABS, but before I could, I was knocked over by the avalanche. I reached for the trigger again, and this time I managed to pull it. However, the balloons were of little use. The snow threw me right into the trees. I felt it slamming against my head and ripping the skis off my feet. All I could think was “protect your head” as the snow pushed me forward. Then as soon as it had started, everything went still. My back against a mountain birch and my head above the snow facing uphill.
I heard Martin shouting my name, “I’m Ok”. I tried to get up but I was buried to my chest, stuck with my butt and back on one side of the mountain birch and my legs folded on the other side. I tried to dig with my hands but I was shaking too hard. Martin was surrealistically fast at digging himself out, he came running asking if I was okay, then ran forward to dig up Maria who was half buried face down. Once he checked that she could breathe he dialled 112 for the police and tried to give directions to our position, then called a group of friends who we knew were skiing nearby. I could hear Maria scream, blood freezing in my veins. “Is she okay?” — “she has a femur fracture — her leg is dislocated. I don’t know if I should reposition it or not? Then the fear of a new avalanche. I dug frantically with one hand and it became apparent too that my leg was also broken…
Learning Points
Persistent weak layers are the monsters that lurk beneath the surface of the snow in the backcountry. These weak layers are often associated with continental climates such as the interior states of the United States or central Asia, but all mountain ranges have the ability to develop these layers and do so — regularly. Through a process known as temperature metamorphism, the snowpack changes over time due to temperature variations. Crystals in the snowpack called ‘facets’ form when the air temperature is particularly cold and the snowpack is thin, setting up an unstable structure in the snowpack.
Added to this lurking layer of facets an ever increasing snow load being deposited onto the lee slope from the alpine winds above them, the snows strength on the slope was stretched to its limit — like an elastic band stretched to breaking point. Martin’s weight as he stepped onto the connected slope beneath the wind loaded slope above was the straw that broke the camel’s back of the slope. Once the fragile weak layer was initiated, it then propagated upslope releasing a large slab avalanche on top of the entire group.
Snowpack & Weather
Whilst avalanche aware skiers in continental climates are familiar with PWL’s and track them routinely, they often slip past skiers from maritime climates who often fail to appreciate how dangerous they can be.
If the snowpack was so unstable, why was it so difficult for them to trigger the faceted layer during their stability tests? Stability tests are great at telling us when a snowpack is dangerous. There’re not so good at telling us when a snowpack is safe. This is because of something called spatial variability. Spatial variability is a key factor in avalanche accidents for skiers who have avalanche skills and training. Although the weak layer was present in their pit, it presented as being very strong when tested. The snowpack is never uniform but changes from place to place due to the effects of wind, angle and aspect. Even very subtle differences on the same slope will often hide a PWL from us, or as in this case disguise its weakness. Basing our decision to ski a slope on the results of just a couple of targeted information points such as a snow pit can lull us into a false sense of security about slope stability. Just because the snow is strong at points A and B, it doesn’t necessarily follow that it will be strong at point C. Use the results of a stability test such as a compression test or an extended column test (ECT) to say no, never to say go!
The weather was a significant factor in the accident. It’s easy to make good route choices when the sun is shining. Biting cold arctic winds and flat light turn easy decisions into hard decisions. Communication in a team is compromised by poor weather. Route planning the night before is a key way to provide a margin of safety, providing a considered passage through avalanche terrain and pointing us towards the safe aspect and elevation zones away from the current avalanche problem. But often, such as in this case study, the reality is that we don’t operate in the mountains like that. On short cold winter days, we often play ‘mini golf’ in areas that we are vaguely familiar with and chalk up short laps of powder, or we alter our original plans depending on the weather conditions and snow quality we find. Fatmap can help here, the ability to instantly locate ourselves live in a 3D terrain map whilst in poor visibility by simply pulling out a phone is a great ace to have tucked in your back pocket. In this way we can not only see the angle of the slope we’re on, but equally as important, the size and angle of the slopes that we’re connected to and threatened by from above.
Human Factors
Add to these snowpack and weather influences a few human factors such as confirmation bias — we all want to get a stable snowpack test result, right? So even when the pit gives us a strong clue: “It seemed to have bonded really well with the old snow, however, when we put our full weight on the snow and pulled hard it came out with a very clean shear.” We have a tendency to focus on the hard to release part of the test as opposed the clean shear that’s just occurred.
Other human factors that would have been pressing on the group that day were both commitment and familiarity. After a five-hour drive, like any of us, they were keen to get a few decent turns from their day. In an area that they knew well and had always felt pretty benign in the past, they felt confident to push their margin of safety — just a little.
What about the good stuff? What did they do right? They were a team not a group, they knew each other’s strong and weak points. They communicated and made decisions together in an open and flat hierarchy. More often than not that kind of facilitated communication will expose a bad idea. They had the right equipment and just as importantly knew how to use it. They carried a mobile phone to contact nearby friends and alert the rescue services, and were able to accurately provide their location. They were all physically strong and capable with a positive mind-set, so when things did go badly wrong they had the capacity to minimise its impact.
There’s no such thing as a textbook avalanche accident rescue. Of the dozens of rescue scenario’s that we run every season they all are pretty chaotic. That is the nature of unexpected high stress situations in hostile environments. But what is worth highlighting from this incident is that the rescue isn’t over once the digging is finished. In many ways, it’s just the beginning. Do you carry appropriate warm clothing, group shelter and first aid kit? Do you have appropriate first aid skills? We often don’t think about such things, but this is a great demonstration of how vital these skills are just for a short day’s back country skiing in an area close to a ski resort.
This case study teaches us to be vigilant about persistent weak layers in the snowpack and to fully examine the avalanche report for their existence, as opposed to just noting the headline figure of ‘category 3’ as in this case. It teaches us the dangers of placing too much emphasis on our own snowpack tests due to spatial variability. It’s also a stark reminder that we should repeatedly practice our rescue skills and carry the right equipment for such a rescue including spare clothing, group shelter, first aid kit, GPS and a communication device.
Mike Austin is the co-owner of Avalanche Geeks, an avalanche education company that provides courses and training for the public and mountain professionals in the European Alps & Scotland.
