Old snow as a human problem

It was one of many days in a mediocre winter. The base was lean, it had snowed a good 10 centimetres overnight with a lot of wind. Nobody wanted to go freeriding in the area, nor did they want to go on a longer tour. The situation report spoke of a combined drift and old snow problem, danger level Considerable. The old snow problem had been in the text section of the bulletin every day for weeks, a kind of background noise to the season. The fresh drift snow was much more present in our minds that day. We considered a variant with a short ascent from the lift to a longer descent over flat, comfortable alpine terrain.

The only critical point was a short, somewhat steeper slope area at the very beginning of the descent: bordered on the left by a few tufts of grass and stones peeking out, the top 15 meters perhaps 30-32° steep, then quickly becoming significantly flatter. We stood together at the top and discussed whether it would be better to go around the outside and cross to the next flat section with some pushing. On the other hand, it was really only the first few meters that were a little steeper. And going over the flat section with a bit of momentum would definitely be the more pleasant option. We agreed: at most, the first turn would take off the little bound drift snow on the surface, but even if it did - that shouldn't be a problem in this terrain.

One of my two touring colleagues set off on the descent while we watched him from the entrance. A few quick turns and he was down on the flat, the snow even looked pretty good. We watched as he crossed the flat section - whether it would work out without pushing? - and got ready to leave.

My colleague was just past the next crest, maybe 300m away, when a crack opened up at the top of the entry slope. The slope seemed to detach as a whole, initially without breaking into smaller slabs. A few seconds later, the slope next to ours also broke away, and a few more seconds later the one next to it. In the meantime, our colleague was standing much further down on another crest - long outside the danger zone. Our position at the start was not at risk either. The avalanche didn't travel far thanks to the flat terrain, but the avalanche edge was around 150m long. The trigger point was probably in the first flat section in one of the many places with little snow. The break continued flat up the slope until the terrain became steep enough for a descent.

The steeper slope at the beginning was borderline that day, according to Munter, but not completely beyond good and evil. And if we had bypassed the slope and crossed directly into the flat section - a "permitted", not even borderline, variant according to Munter? Of course, I'll never know, but I suspect we would have triggered the avalanche anyway.

Even black swans have teeth

Like everyone who is out and about, I have made mistakes from time to time. Some, but certainly not all, I am aware of. I do my best not to repeat them. Of all the events in my personal catalog of stupid decisions, up until that day three years ago, nothing was surprising in retrospect, and in some cases not even in advance. It was all part of my understanding of familiar, "normal" dangers. Because the mistakes were obvious and clearly defined, the ways to avoid them are also obvious and clearly defined. The lessons I learned from these experiences are primarily to do with self-awareness, not with knowledge that affects my understanding of snow and avalanches.

The old snow release back then was different. Nobody was hurt and it wasn't even particularly close, but the experience was still more drastic than other mistakes, some of which had more unpleasant consequences. Our initial reaction was a closed "What the fuck", as the modern freerider says. And not, as usual: "We should have known that".

(continue on the next page)

The fact that there was an old snow problem and that old snow problems are insidious and difficult to assess was already clear to all of us in theory. But as humans are like that: sometimes you have to see things first in order to understand them in practice.

What McCammon describes as the trap of habit is also described in psychology as "normalcy bias". As a species, we seem to have an innate blinkered mentality that causes us to underestimate the dangers that lie outside our understanding of normality - both in terms of the probability of occurrence and the possible consequences.

If you are used to reasonably "predictable" avalanche problems, such as drifting snow or wet snow, more unpredictable avalanche problems lie outside what you perceive as "normal". What we cannot imagine does not exist. Just as I tend to underestimate the avalanche danger on my local mountain precisely because I know the slope, I may underestimate the "non-normal" danger precisely because I have an accurate idea of the "normal" danger.

It's basically the classic induction problem (the one with the swans and David Hume, not the one with the burnt tomato sauce on the induction hob): I've only ever seen white swans (drift snow avalanches). Consequently, I assume that there are no black swans (larger, more unpredictable old snow avalanches). If a black swan bites my finger while I'm feeding the ducks, it's not only painful, but also upsets my view of the world - it shouldn't have existed in the first place!

Strategic fear

Since the incident three years ago, I have become much more sensitive to the background noise of old snow in the situation report. I can now imagine much better what a worst-case scenario on a particular slope would mean under certain circumstances. This idea is very scary.

Sometimes, however, I lose my fear: The danger level drops, nothing has happened for a long time, friends ski all the beautiful, challenging lines in terrain that is actually at risk of old snow because the danger level "allows" it. Whenever I'm tempted - which is quite often - I try to scare myself. I put on a mental horror movie, imagine huge avalanches and picture in detail how the whole slope will break into slabs and swallow me and my touring colleagues up for good. The desired effect is usually achieved quickly.

Strategic risk minimization according to Munter, or with the usual Munter follow-up methods (Stop or Go, Snowcard), comes up against certain system-related limits, residual risk or not, in the case of an old snow problem. At least if you use these methods to calculate more critical slopes based on the gradient and danger level, which is what they tempt you to do.

The avalanche problem of wet snow has always been treated as a special case at Munter, as some reduction factors do not apply here - long before the term "avalanche problems" made it into general avalanche linguistic usage. If the special rules for wet snow did not exist, the whole thing would not work in a spring situation. There are no special rules for other avalanche problems. Perhaps it's time to rethink this for the avalanche problem of old snow. This would certainly not take anything away from the tried-and-tested strategic methods of risk management, to which we owe a great deal.

Formalized behavioural recommendations for a latent old snow problem that are built into the strategic methods in a concrete way could look different depending on the method (certain reduction factors of the PRM not permitted, expansion of the spatial scope with Stop or Go, etc.), but the result would ultimately remain the same: large-scale avoidance of the problem areas.

Another recommended bergundsteigen article (The avalanche situation report, a misunderstood love?) gives understandable, insightful advice on how the situation report can and should be used (summary: take note of the danger level, but be sure to read the rest as well) and suggests adopting a more holistic workflow, as recommended by the Austrian Board of Trustees for Alpine Safety. Not instead of strategic methods, but in addition.

Meanwhile, I wouldn't want to miss out on the fear of old snow, because fear as a form of risk management is very effective: nothing will happen to me where I don't dare to go. And I don't dare to go because it has become impressively clear to me that mediocre chess skills are not enough to win at Russian roulette.