Friday, April 13, 2018

A Bad Day, That Could Have Been Worse

February 16th was the kind of bad day that could have been a lot worse.

The short of the day, was that Aaron and I were skiing the backside of Microdot in Hatcher. Climbing back up, we overtook another group of two on what appeared to be a low consequence slope, and triggered a short, wide, 3-foot deep slab above complicated terrain that took three of us for a ride and left one us buried neck deep.

For a few seconds, I was certain I was going into a hole and would be buried and maybe crushed. Luck is all that kept me on the surface in the end. Like all days, there were a lot of decisions made, most were small, and seemingly inconsequential. This is my attempt to look back at them, with the advantage of hindsight, and consider what went right, and what went wrong.


After weeks without snow, Hatcher Pass had been walloped early in the week, first by a windy and warm storm that dropped around 2 feet of snow. I initially expected the snow to be very reactive due to the old and wind hammered base.

Data on how the new snow what reacting on the old base was sparse. The road to the Pass was left unplowed. Skiers struggled to get up to Hatcher Monday or Tuesday. Mid-week, a second, colder system passed through dropping another foot of lighter snow with minimal wind, burying clues from slides that occurred during or after the first storm.

The road remained unplowed Wednesday and Thursday, and what little info was coming out of the Pass offered mixed signals. Considering the wind-hammered base the initial storm had fallen on, natural avalanches were sparse, when it seemed they should have been rampant. There was fairly ample evidence however, that the new snow was readily reacting to skiers.

The slides being reported were running out in typical storm slab configurations: generally narrow, slowly, and softer in composition. These are the types of slides one might expect as the storm abates and the snow begins to settle down. 

On Friday, with the road plowed, Aaron and I decided to go check things out


Decision 1: Self-assigned danger rating of considerable. Synopsis: Correct

Hatcher has a once a week av forecast issued on Saturdays, which was not yet out. Observation info was limited, and mixed. That being said, the base/bed surface the snow had fallen on warranted suspicion. Based on what I knew, I assigned a danger rating of considerable. The rating plays a role in my personal decision making. A professional, or personal assignment of “considerable” equals an automatic one “flag” for me. Three flags, and the day is over.

Decision 2: Flat light delay. Synopsis: correct.

Friday dawned with high clouds, but a forecast to clear. Given the self-assigned considerable rating, we delayed start until noon when it was apparent that skies were actually clearing. Hatcher is a difficult place to ski when you can’t see. When conditions are good, you can continue to ski therein low light, but the point of this ski was to get info. Not being able to see or having obscured vision, was both a risk, and negated much of the purpose of going at all.

Decision 3: Observing a fresh skier-triggered slide and not marking it as a flag. Synopsis: incorrect.

As we arrived, we passed Marmot’s west face, and observed a very fresh (less than 5 minutes old) skier-triggered storm slide. It appeared the skier had skied past a trigger point and was some ways down the slope before they realized the slope was sliding, and safely exited the relatively narrow slide path. The skier’s two partners skied next to the slide with no additional reaction. I asked the ranger if he had witnessed the slide but he said he had missed it. I suspected it was similar to the skier-triggered slides that had been reported earlier during the week: slow-starting, deep, able to run full path, but non-cohesive. This was concerning, but I did not consider it a flag at the time. The slide was clearly slow, and the snow seemed somewhat stubborn, and the slide had triggered where the skier passed by a classic trigger point: a rock and convex protrusion.

In hindsight, these were some of, if not the first skiers on this slope, and they had nearly instantly triggered a slide. Regardless of whether the slide indicated a “healing” interface this should have been flag 2 for the day.

Decision 4: Pit test – no clear decision rendered.

On the way up Microdot, we passed a pit dug by the avalanche center. I reused the back of the pit to isolate a new column. I would not call this a formal test, however, it was far more formal than any previous pole pits conducted to this point. We observed a moderate force trigger, easy leverage, and a clean shear. The pit, to me, indicated typical spatial variability, we exposed a potential, though not widely existent, smooth bed surface. We knew this reaction was spatially variable based on pole pits. This confirmed a suspicion that the slides were triggering in specific locations, and were generally sliding more due to the mass of available loose snow/inertia, but were failing to propagate on a wide scale due to a lack of a cohesive slab structure or the existence of a wide spread consistent bed surface and weak layer combo. Basically, hit the right spot, you would get something to slide, and in steep terrain it would carry enough snow to slide down the fall line, but it would do so fairly slowly. In general, it’s not likely I’d ever commit a flag to a single pit result in any situation, unless the pit revealed an unexpected condition, in which case, I’d be looking to dig more pits. Pits are good for getting a good look at layers, but are indicative of a small spatial area, especially in Hatcher. In hindsight, this was evidence that should have been accounted for, but not compelling.

Decision 5: Subtle collapsing on established skin track – no clear decision rendered.

We observed subtle collapsing on an established skin track on Micro’s rounded west side. Aaron seemed to notice this more than I did, but Aaron is a good bit taller than me, and heavier as a result. When he commented on this, I was able to get some very deep, muffled collapsing by jumping up and down. The aspect where this occurred is underlain by large (and now buried) boulders that are typically exposed; and is a wind-hammered ridge. Collapsing was not a surprise given the sub-surface, but what was surprising, was that the skin track was a day or more old and had been traversed by at least 8-12 skiers. Again, Similar to the pit results, this reaction, on this slope, did not (and likely still would not) warrant a “flag,” but it should have increased the level of suspicion, and been a clear indication thigs were not “alright.”

Decision 6: False positive observation of skiers in Rae Wallace. Synopsis: incorrect.

This was one biggest mistakes of the day in my view. While climbing, we observed two skiers drop into a steep and finned chute on the western side of Rae Wallace. The skiers were cutting hard, and passed over numerous potential trigger points, but failed to initiate anything more than sloughing. I gave this observation immense weight, more than any other observation to this point. Why? Because it showed what I wanted to see. I wanted conditions to be stable, and this appeared to show it. Instead of considering this a neutral observation (maybe they got lucky, may steep north-facing had already slid, etc.) I overshadowed the previous observations made.

Decision 7: Skiing the north side and deviating from plan. Synopsis: wrong, no safe exit strategy for conditions.

The plan had been to ski the sunny south side, however, it was well tracked. As we topped out, a group of two descended the unskied north side. I heard a loud collapse as the first skier dropped in, but observed no cracks or other reactivity. The second skier, a snowboarder, dropped in next, with no collapse. I felt good about the slope as it is concave in the center, and punctured with micro features that break the slope and can act as snowpack anchors. The skiers had initiated a collapse, but it was right off the ridge, not a surprise, and there were no visible cracks. Aaron also felt good after watching the two skiers drop in, so we followed a few minutes later. We did not get any reactions. Skiing this slope would not cause us any problems, but we did not have a good exit strategy, and that’s what would bite us.

Decision 8: Climbing north side of Microdot. Synopsis, incorrect.

The snow was incredible, we were leaning toward the idea that conditions were stable. We had the choice to either climb back up the north side, or exit the bowl. The latter option chews up a lot of time (less skiing), but the former would either put us on a steep slope we’d just skied when other skiers were likely to descend on us. We could also put in a more circuitous climb northeast of the main north side run. The two skiers in front of us were a few minutes ahead of us, and began to break a trail, climbing this circuitous north east route. We did not communicate with them about route selection prior, but later confirmed that they chose this route to avoid skinning directly up the main run, which they also assumed would have skiers dropping in.

At the time, this seemed like a prudent decision, I was familiar with the terrain, I had recently used this route, and felt good. In the face of no better evidence, this might have been an OK decision. In the face of the actual evidence observed, and having two flags already up, we likely should have just exited the bowl.

Decision 9: Catching and passing on a steep slope. Synopsis: too close, avalanche/incident.

This is where it all came to head. Trail breaking was slow and deep. Aaron and I were able to catch up with the two other skiers, who I will refer to as skiers 1a and 1b, in a sub bowl approximately 200 vertical feet above Murphy Lake.

Skiers 1a and 1b were approaching an approximately 100-foot northerly facing slope above the sub bowl that lead into the next, smaller bowl. This was the incident slope. Skiers 1a and 1b were approximately 25 yards apart; 1a was entering the incident slope, setting a skin track across the slope using a narrow natural bench that cross-cut the slope.

In a non-judgmental way, I did not like this route, and would have preferred to stay in the flats of the sub-bowl and use a short concave gully on the climber’s left of the incident slope. I’d used this small gully feature before, but, with deep snow, and a small and unintimidating incident slope, I opted to stay the course and follow the broken track to try to catch skier 1a and relieve them of the arduous trail breaking.

I caught up with 1a approximately mid-slope on the incident slope, and was about 10 feet behind 1a.

Everything I’d observed to this point, barring one false positive, should have had told me I had 2 flags raised, and the antennae should have been on high. Getting this close to another skier on a steep (greater than 35 degrees), albeit short slope, that I did not like to begin with, over terrain traps, was just stupid, and about to be a big mistake.


At this time, skiers 1b and Aaron were just entering the incident slope about 25 and 27 yards respectively behind me and 1a.

As 1a and I passed a mostly buried rock outcrop, (the area featured numerous micro features including buried boulders and micro channels), the snowpack settled 2-3 inches.

All 4 skiers felt the settlement.

The slope began to fracture and broke an apex crown about 30 feet above skier 1a and I, and fractured approximately 150 yards across, at a depth of 2.5-3 feet. Slope angle across the slope varied due to the micro terrain.

Aaron was the farthest back, and on the edge of the fracture. By turning around he was able to escape the slab.

Skier 1b deployed their airbag and was knocked down. They were carried an estimated 25 vertical feet or more, sliding through a channel. Skier 1b said they felt their airbag slow them down and keep them above the moving snow.

For myself, a memory kicked in as the snow slowly crumbled and I felt myself sink and lurch down. I turned downhill and was able to take a stride or two before being knocked backward by the acceleration. I was in a sitting position, upright, facing downhill, leaning back over the tails of my skis. I could tell immediately that I was being carried directly toward a number of potential terrain traps at the base of the slope, likely small tarns.

As the slide initiated, I in line with a low, snow-covered ridge of rock, this feature played a role in keeping me safe. By constantly moving my ski tips upward, while using my poles laid flat behind me (no wrist straps attached) I was able to provide some directional steering and float above the moving snow and ride on the higher micro terrain. I could occasionally feel the bed surface with my fists, likely buried snow-covered rocks.

As the rock ridge narrowed and ended, the snow pulled me to the right, toward a channel that lead into a terrain trap that I could see was going to fill.

At this moment, I realized I had to fight with everything I had to stay on the little ridge, or I would end up in the hole.

I plunged my left hand into the snow and planted the handle of my left pole as hard as I could into the bed surface. The snow pushed past me for a second and I was able to wheel leftward, redirecting myself at the last second, to instead stay straight and ride over a 3-4 foot snow covered ledge. The torque from the maneuver tore the pole from my hand, which was planted with enough force it was still pointing outward after the slide.

As I spilled over the ledge, the slide lost momentum, and I could see I was on a small alluvial feature, and was going to be OK. I was carried an estimated 50 vertical feet according to a GPS track.

Skier 1a also turned down hill as the slide initiated, however, 1a was directly above several terrain traps located at the base of the slope. Skier 1a deployed their airbag. I heard the bag deploy, but never saw if it inflated. After the incident, 1a said they believed the bag deflated during the slide or never immediately inflated, and when I discovered 1a, there was no evidence the bag had been inflated at the time of burial.

Skier 1a was carried right, and traveled an estimated 75 vertical feet. I was able to see skier 1a get dragged right, but lost sight of them due to my own struggle and to a raised terrain feature.

The slide lasted an estimated 10-15 seconds, and had effectively slipped the entire incident slope.

It was unlike any of the skier triggered slides yet observed, and carried a classic slab pattern, including a few step downs.


Aaron was off slope as the slide stopped; Skier 1b was on top of the snow but partially hidden from view of Aaron and I as they were in a small channeled terrain feature; I was upright at the base of the slope below the ledge, my left leg was buried shin deep, but the tail of that ski was at least 2 feet deep, my right ski was on the surface. Skier 1a was not visible to the rest of the group and was buried to their neck, facing down slope, slightly reclined, and surrounded above all but their right hand.

Of note, skiers 1a and 1b also had a dog with them. It is not clear to any of us where the dog was at the time of the slide’s initiation, nor where the dog was after the slide. It is possible the dog was buried or partially buried in the slide, and extricated itself, as we all distinctly remember it appearing after the slide during the extraction of skier 1a.

I immediately called out that I was up, and OK as the slide stopped, I could see Aaron. Aaron confirmed he was ok. About 5 seconds later, skier 1b called that they were OK and stood up.

The three of us were now visible to each other, but could not see skier 1a. We began to shout and call for skier 1a. No response was heard, and there were no visible indicators of their location.

I initiated the search, and called for all beacons to be switched to search mode.

I had the last visual of 1a, and was closest at the time of the slide. My beacon indicated an initial distance of about 25 meters to skier 1a; Aaron, who was farthest away, had an initial reading of 65 meters.

Based on the fact that I suspected skier 1a was fully buried in a terrain trap, and was relatively close, I chose to dig down and release my foot from my buried left ski and abandon it rather than excavating it first. I believed I could reach 1a as, or more quickly on foot.

I unclicked from my right ski and used it, and my remaining pole as aids to scramble through the debris while heading in the direction I had seen 1a carried, checking my beacon for reference.

Aaron still had both skis and skins on and was beginning his search, however, he was father away and had to traverse the base of the debris initially. Aaron remarked later that he was surprised how slow and challenging it was to travel through the still-soft debris, and that he felt like he was “going through mud.”

I called 5 meter increments, which also seemed to come slowly, and was lead on a trajectory over the small ridged terrain feature that divided where I had been sent and where 1a had been sent. I also initially felt that progress was slow, however, not having skis on made it easier to scramble a direct route across the debris and micro features since the distance was short.

Upon climbing the small ridge feature, my readings dropped instantly from 20 to 15 meters.

That’s when I first heard a muffled call for help.

Below me, 10 meters away or less, I spotted a glove protruding from the snow.

I yelled “I have a hand.”

Later, 1a would say this was the first thing they heard, despite the yelling the three of us had been shouting.

I was slightly uphill of the hand, which I still did not know was not just a glove on the surface. I began to yell to 1a that I heard them.

As I scrambled down, I was able to make sight contact with 1a, whose head was just barely above the surface, and yelled this to the group.

1a was buried upright, facing downhill, leaning slightly back, buried with compact snow to their neck, loose snow to their mouth, their right arm stuck up, their left arm buried. Though loose snow had collected around 1a’s mouth, they had an unobstructed airway. Snow was piled about 1.5 feet above 1a, only their hand was visible on the surface, their head was sunken into the depression.

As the slide had stopped, 1a knew it would be their last chance to create an air pocket, and shoved their arm upward and swung it across their face. The snow settled and instantly froze them in that position.

The time between the end of the slide and my contact with 1a was 3-4 minutes. This was verified by a GPS track.

I assessed that 1a was breathing, conscious, able to make eye contact, and had color in their face. At face level, I brushed loose snow away from 1a’s mouth, and asked 1a if they were hurt, or felt any pain or injury. 1a said they were struggling to breath; 1a’s breathing and speech was audibly labored, this being due to the weight of snow on their chest.

I chose to forgo any further potential injury assessment, and began to excavate snow away from 1a’s chest.

I carry my shovel with the handle in place so it can be deployed in a single motion. If you don’t do this, you’re wasting time. Adrenaline will be on high in these moments, stupid things like sliding a shovel shaft into the blade will be stupidly slow. Everything you do to make the search faster are an aid, whether it’s easy access to your beacon, or rescue equipment.

Despite the adrenaline, I was aware that I kept a high situational awareness, watching for near or far hang fire.

Even as the urge to want to start digging furiously set in, I made sure to tuck my beacon back into my jacket, closed the zipper, and situated my pack uphill and away from where I’d be excavating snow. These were all conscious decisions that happened. I was aware they were happening, but they all felt automatic. I can only attribute this to both practicing for this, and for rehearsing it mentally.

Approximately 1 minute after I arrived, Aaron arrived. Aaron deployed his shovel and began to assist removing accumulated snow, forming a shovel train.

Here’s where I made a mistake, albeit, a small one: Aaron, like I, was relieved to see skier 1a above the snow, conscious, etc. I never communicated to Aaron that 1a was struggling to breath though. To this point, communication had been good, but lacking that knowledge, Arron did not know there was still urgency to the situation. This came up after the incident, but it was worth noting.

Skier 1b arrived about 1 minute after Aaron. 1b was emotionally upset from the incident, and relieved to find their partner well.

After a few minutes of excavating 1a down to their waist, I slowed and asked 1a to re-assess for potential injuries. 1a said they felt good, other than a sore knee. 1a was now able to breath easily and to communicate fully. I continued to excavate behind 1a, as well as to dig to 1a’s feet so I could release 1a’s skis, which were both still attached. I did my best to avoid knocking more snow on 1a, and communicated clearly when I was moving blocks of snow away from and behind their head, knowing that the sensation of reburial can cause panic in these situations. 1a was calm and collected the entire time.

Once free of 1a’s skis, I used 1a’s backpack straps to pull 1a fully free.


The group spent some time after the response discussing the incident and assessing conditions that had lead up to the slide. Both groups agreed that while there was evidence of instability, there were also not as many typical flags, nor ample evidence of flags. Skiers 1a and 1b had reported they had been snow machine skiing in the area the day prior, and reported that they had spent the day easing onto progressively steeper slopes with no results or reactivity. Aaron and I had noted what we saw as inconsistent activity, indicative of spatial variability. None of the skiers in the groups had conducted their own formal stability tests. Experience levels in the group were high. All skiers involved in the incident had 10 or more years of backcountry skiing experience.

All skiers involved agreed that spacing and slope angle where a problem, particularly given that there were signs of instability.  

At the time of the incident, I had approached within 10 feet of skier 1a. Who knows whether better spacing could have prevented the incident, or whether it still may have caused a slide as 1a passed, or delayed the slide until myself, 1b, or Aaron passed.

Skiers 1a and 1b were not upset, they, like myself, did not suspect the slope to be a threat. To the point, not knowing whether the slide would have triggered because of spacing, they were thankful we had decided to follow them. No other skiers would ski the north side of Microdot that day.

Another factor that contributed to the slide, these bowls are in a sheltered, northerly facing terrain feature. With additional protection provided by the small micro features. They were likely points to harbor buried surface facets over the wind hammered base that had been proving to be a reactive layer over the buried bed surface elsewhere. Our previous travels had been on aspects that likely would have seen the facets wiped away.

On first glance, the relatively small size of the slope, and apparent “slope anchoring” provided by these micro terrain features, might have given the impression that while steep, the slope was not high risk. Over a flat run out, or with a smaller snow load, the slope would not have necessarily been problematic, but given the snow load, the two flags that were up, the evidence that conditions were still unsettled and variable, and existence of terrain traps, this slope deserved better travel protocol, or avoidance entirely. Interestingly, we had intentionally avoided the main north side slope expecting skiers to enter it, but no one ever did.

After the slide, we all skied back down our skin track to Murphy Lake, then across the lake, and back down to Independence Mine.

Root Causes

Personal: I want to see what I want to see. I want stable snow so I can go ski the lines that engage me. Lacking snapping red flags, I chose to focus on evidence that leant itself toward stability, and ignored or dismissed the evidence to the contrary.

Protocol: had I been willing to see the evidence for what it was, I would have been more concerned with both route selection and spacing. While my intentions were good to catch up, I should have waited before crossing the incident slope, and caught up with 1a on the flat above.

It sucks to know you messed up. The biggest take away for me, was the reminder that I need to ease up. I have a habit of seeing what I want. There’s a balance. This sport is inherently risky, it is uncontrolled, and thus risk management is subjective and personal.
View from below. T is the trigger point, black line was the skin track, 1a is where skier 1a was buried

View of the approach slope.

Burial location. Good example of how a relatively small slide can pile a lot of snow in the right features

Not great shots of two skier-triggered slides on Marmot. The slide to the right was triggered as we arrived, the slide to the left was triggered moments before we left.


1 comment:

William Finley said...

That was a really good analyst. Really liked the way you broke down the decisions - especially the "it showed what I wanted to see" comment. That happens too often.