avalanche case histories

[deliberate]

http://www.avalanche.ca/cac/library/avalanche-accidents

4 volumes, on the right hand side. Each volume covers a time period, Volume 4 is the most recent.

a lot of fatailities, so not exactly a light summer read. However, they are well written, and include data/figures. I took an avalanche class, and what was missing was empirical data (from an engineering background), these results produced an avalanche on this degree slope. I even said to the instructor that the only way we get feed back on our test interpretation is if the slope fails, but that was brushed aside.

The whole 15 min to dig up a person stat you hear thrown around is also not supported by this. I skimmed it, and it seemed like a lot of fatailities were recovered within 15 minutes but died to asphyxiation (not the trauma ones). Recovered in 7 mins, not breatihing, revived with CPR. Saying you need CPR training is pretty obvious though. One guy survived 55 minutes because he made an air pocket. One dog surived and dug himself out in an 18-20 hour period.

On Page 69, Volume 4, it recounts 1 heli-skier fataility, caues "isolated weakness". They skiied all day on the slope, and it was tracked up when it slid. That is terrifying.

I'm sharing this because I found it interesting.

[CookieMonster]

I'll second the recommendation.

For Deliberate:

The 15 minute figure is simply designed to give an idea of your survival chance past 15 minutes, and isn't really designed to give you a survival benchmark for burials of less than 15 minutes. There are a variety of statistics on the survival rates when recovery occurs within 15 minutes, but this simply raises a different question: is it wise to search for a way to construe the statistics in your favour? You can clearly see that the statistics are highly unfavourable ... so perhaps that is the take-away.

With respect to the positive feedback loop described in your post, which went unaddressed at the avalanche course you attended: it might be useful to remember that you can't always know what you did right, nor can you always know what you did wrong. You collect data during each backcountry outing, right? If you know that the data are subject to bias, or if you know that the data from one trip, or even one decision, are not representative, then what is the correct response?

This is why The Avalanche Handbook discusses the importance of using a mix of distributional data and singular data when constructing a forecast ( with the assumption, of course, that the data are accurate. ) If you haven't gotten a copy of The Avalanche Handbook, you might be extremely interested in Chapter 6, which presents an advanced, theoretical framework for avalanche forecasting. It will fascinate you.

Here is my own take on backcountry avalanche forecasting, in 4 parts:

Goal. Align your perception of instability with reality.
People. Manage thoughts, actions, and beliefs that could affect the forecast.
Awareness. Maintain awareness of instability by observing terrain, snowpack, and weather.
Uncertainty. Acknowledge, identify, and reduce uncertainty to make safer decisions.

http://avalanchesafety.blogspot.com/...avalanche.html

[PNWbrit]

[url]I took an avalanche class, and what was missing was empirical data (from an engineering background), these results produced an avalanche on this degree slope.

Unfortunately that's not the way it works.

And that's a common take away from taking an Avalanche class.

[Hugh Conway]

Unfortunately that's not the way it works.

And that's a common take away from taking an Avalanche class.

+1000

case histories are great for teaching examples. If you are the rare person who can synthesize that bulk of knowledge into some useful package more power to you; for most people reading them is probably more of a hindrance than a help especially given the substantial differences in ski style, terrain, equipment, snow training, and snow preference between <1996 and now.

[deliberate]

I guess I thought it would be different coming from my background in slope stability, reliability, and a bayesian instead of frequentist approach to dealing with the representativeness of small samples.

From what I've seen, the state of armchairing is: no individual tests mean anything, until an avalanche occurs, and then there is always that 1 test that meant it was going to happen.

[Hugh Conway]

From what I've seen, the state of armchairing is: no individual tests mean anything, until an avalanche occurs, and then there is always that 1 test that meant it was going to happen.

that would be bayesian, no?

[CookieMonster]

The missing element of this discussion is the relationship between your forecast, your perception of instability, and how your beliefs/attitudes might exert a positive or negative influence on the forecast.

Therefore, as a hedge against chaos, and errors:

Data -> Forecast -> Decision

* Use multiple data sources.
* Issue an ensemble forecast.
* Make an ensemble decision.

All of the above is null and void without significant self-awareness.

[PNWbrit]

no individual tests mean anything

Pretty much sums it up.

Stop concentrating on the test rsults and numbers. When you get out there I think you'll be surprised how little people rely on them for decison making. Mostly they're for observation of what we expected and noticing trends or deviations away from the expected.

There isn't a magic CT, RB or Q number that you go out hunting for like a truffle hound. You're certainly not going to find a magic answer by looking in other peoples accidents. They are worth reading for lots of other reasons though.

[Alaskan Rover]

I've only been caught by two major slides in my lifetime (and I'm hoping that three is NOT the charm)....once with skis (they came off right away), and another time strapped into my Sherpa snowshoes. By far the worst and scariest was when I was wearing snowshoes...I seriously thought my number was up. Luckily it was in a timbered area and a smaller slide to boot, the trees I think saved my ass. I happened to be at the tail-end of the slide, and so the trees made a break for me, instead of coming down on me. It took a while, but I was able to wiggle out of my snowshoe bindings. Until then, I was effectively STUCK.

The second slide was an open-area slide and larger. Because my skis came off when it came, I was able to maintain some maneuverability (at least more than with the skis still on!)...and was able to maintain an upward position. I don't quite know how I didn't get buried in that slide, but I was lucky enough to have remain mostly on top.

I ended up with one arm free, with which I freed my other arm and then slowly the rest of me. A lot of people who haven't been in a slide think the snow must be light and shit. Fuck no, it's dense and heavy. And when your stuck in it, your body heat will actually melt the snow (not always) and then re-freeze and make self-extrication harder.

I was lucky enough to not be buried by both slides. Both times I was solo in the Chugach Range. If I would have been buried in either, I wouldn't be here to write this...just like a lot of shit I've done. Been lucky so far...but the unlucky and the stupid aren't alive to post second thoughts.

The point about if you think you might be buried...to make air space...is a good one. But when you're caught in this tumbling mass of snow, what's logical isn't always at the forefront of your mind.

[DasBlunt]

your point-

"On Page 69, Volume 4, it recounts 1 heli-skier fataility, caues "isolated weakness". They skiied all day on the slope, and it was tracked up when it slid. That is terrifying."

-shows that all the data points in the world cannot give certainty and 100% results.


the book, "surviving in avalanche terrain" is a good read and humbling.

[plasticman]

And when your stuck in it, your body heat will actually melt the snow (not always) and then re-freeze and make self-extrication harder.

Your body heat has something to affect the situation, but the primary source of heat in a slide comes from the quick conversion of potential energy into kinetic energy when a slide is triggered.

The energy is stored on the hillside when the snow falls as potential energy and is quickly converted to kinetic energy when the snowpack fallis or slides down the hill. This sudden release of energy is enough to actually melt the snow crystals and get them to rebond. As the slide comes to a rest the kinetic energy leaves the system (the heat is taken away) and this is the "concrete effect" that is often noted by victims as the snow quickly refreezes.

On a fresh and large enough slide the snowpack will actually be steaming from the debris pile enough to be visibly observed and to fog up your goggles.

Never fun...

[ISL]

Thanks for the link! .

I took an avalanche class, and what was missing was empirical data (from an engineering background), these results produced an avalanche on this degree slope. I even said to the instructor that the only way we get feed back on our test interpretation is if the slope fails, but that was brushed aside.

It's true, what you say. That's why I jump on lots of test slopes.

I guess I thought it would be different coming from my background in slope stability, reliability, and a bayesian instead of frequentist approach to dealing with the representativeness of small samples.

From what I've seen, the state of armchairing is: no individual tests mean anything, until an avalanche occurs, and then there is always that 1 test that meant it was going to happen.

The trouble is that we're so very information-starved that no matter how much we know about the snowpack; small perturbations can change the game completely. A correct and complete physical model, should one ever be developed, will accurately predict the behavior of the snow under stress. A few minutes of sunbreak yesterday, a temporary shift in wind during a storm three weeks ago, or the passage of a low foggy cloud two nights ago, if not included in the model, can yield wildly varying outcomes.

If "isolated weakness" freaks you out: For an oft-cited survey of spatial variability, check out this paper by Campbell and Jamieson. For more avy reading material, check out the free archives of The Avalanche Review (and consider becoming a member of the AAA).

It's true that no one test means anything. An ensemble of tests, however can start to tell you something. As important as the results you get from testing is the sense of consistency - "Is my picture of the snowpack a consistent one? Is there lots of spatial variability? These observations have been way too consistent - what am I missing?"

I approach pit tests as hypothesis testing. I most often do one when I have a testable prediction in about the snow - "Is that weak layer I noticed an hour ago present and reactive on this aspect at this elevation?" They're also a good way to partially catch up on the history of the snowpack in an unfamiliar location.

As a skier in Washington, I view testing as adding only a little bit more information in concert with what you know from telemetry, the weather forecast, your in-head picture of the entire history of the year's snowpack, the avalanche forecast, and recent skier reports.

Instability can be predicted, sometimes with certainty; stability rarely, if ever, can. There will be days with warm rain atop a thick cold snowpack shot-through with hoar layers on crust that you can be sure will generate easy slides. People, especially the smart ones, seem to get killed in "considerable" conditions; it seems to be a time when lethal instability is hard to predict.

Simple is good - lots of factors can complicate a snowpack. Be willing to walk away when it gets complicated, and always leave a way to do so. Some factors can temporarily simplify things - they're your fickle friends. 3-6" of well-bonded fluffy fresh on the bed surface of a slope that slid huge yesterday? It's almost certainly go time.

Your only two handles on the situation are terrain/route selection and timing.

When it comes to stability, snow dislikes change.

The only way to learn about instability is to go play in it. Find very safe places on "High" and "Extreme" days and make small amounts of snow move. When snow sluffs naturally off your roof, try to figure out why it happened just then and there.

I'm a physical scientist - it took me a couple of years to realize and come to terms with the fact that the answer to every question in the avalanche game is "maybe". I try to keep my integrated risk exposure down to one death every 30,000 days; doing so is extremely difficult.

[big mike]

this winter was a tricky one in b.c. i skied around fernie alpine resort all winter and prob was outside area more than in it. I was involved in some slides and it scared the shit out of me. im no rookie started touring 20 yrs ago in montana and last 5 winters in canada. ive read many books on avy stuff and was mildly obsessive about it last 2 winters because i was touring so much, maybe 70 days last winter. i followed the cac forecast, read about accidents, and dug snowpits(not all the time). problem is i love to ski untracked, but i dont have a death wish. safety is important to me but i do ski with an aggresive crew who like to get after it. some stuff we ski is close to the hill and we hit it a lot and it has heavy timber. i guess my question is when every book says you should not be on steep terrain with new snow for a few days hoping that things will stabilize, this is exactly where i want to be when it is puking. i like to break skintrack when the snow is up to my knees then ski a great line. im sure im not alone in that. that being said i do make calls on where i want to be try to limit my risk, stay in trees and out of large runout areas. im no pro but i think i have as much field experience as any layman. what do other people do when faced with this... i know there are any # of factors involved with every decision but would like some feedback. also remember skiing a south facing slope mid winter that had a foot of fresh on what we suspected was a layer of surface hoar from a cold snap. we hit it and had a great run tho i was a little hesitant things seemed fine in the way up we dug a pit(i know.. after we skied it) nd got an easy release 6 from the wrist in a ct real smooth and planer, maybe 30 cms down. were we f ing idiots for being there?

[daveg]

Hot off the press: Avalanche Accidents in Canada Volume 5. Review of fatal avalanche accidents 1996-2007. http://www.avalanche.ca/cac/store/books