Dec 16th, Kelso Mountain Avalanche - questions

[justiner]

This is not going to be a satisfying answer, but it's good to be reminded that "forecast" is used for a reason when giving out the data about conditions. To try to distilled down a highly complicated and chaotic reality for an entire State into a simple system of colored charts and rose graphs tied to a map is unrealistic. And the experts get it wrong all the time (just like the weather).

But the avy forecast is just one piece of the decision making process on what terrain to travel on, right? Taking your own local, specific and timely measurements is another, as is trying to understand your own subjective feelings and psychology. There's no perfect decision making tree that will make you safe 100% of the time and no avalanche forecaster will say this is the case in Colorado.

The other problem is that hikers aren't always properly prepared to deal with a slope that may slide, either in gear or techniques. I'm no expert either, so I'm usually in the, "stay the hell away" school of thought. Avy transponders are a thing, but they have to be used properly. How to cross a suspicious slope is another, perhaps using technical climber gear and perhaps one by one, so if the slope does slide, you can rescue your own party.

And part of traveling on this terrain is accepting the risk of death and all the knowledge and training in the world doesn't remove the risk and oftentimes will raise it.

It's very similar to climbing technical terrain. I feel that people think that if they can just unlock some key piece of intellectual insight that they can outwit reality. The reality though is that there is risk that you have to accept the consequences of. The only intellectual piece of advice I would have for anyone is that you refresh the need to accept this risk every time you go out,, or your own complacency will add to the danger you have put yourself in.

[cottonmountaineering]

This is not going to be a satisfying answer, but it's good to be reminded that "forecast" is used for a reason when giving out the data about conditions. To try to distilled down a highly complicated and chaotic reality for an entire State into a simple system of colored charts and rose graphs tied to a map is unrealistic. And the experts get it wrong all the time (just like the weather).

But the avy forecast is just one piece of the decision making process on what terrain to travel on, right? Taking your own local, specific and timely measurements is another, as is trying to understand your own subjective feelings and psychology. There's no perfect decision making tree that will make you safe 100% of the time and no avalanche forecaster will say this is the case in Colorado.

The other problem is that hikers aren't always properly prepared to deal with a slope that may slide, either in gear or techniques. I'm no expert either, so I'm usually in the, "stay the hell away" school of thought. Avy transponders are a thing, but they have to be used properly. How to cross a suspicious slope is another, perhaps using technical climber gear and perhaps one by one, so if the slope does slide, you can rescue your own party.

And part of traveling on this terrain is accepting the risk of death and all the knowledge and training in the world doesn't remove the risk and oftentimes will raise it.

It's very similar to climbing technical terrain. I feel that people think that if they can just unlock some key piece of intellectual insight that they can outwit reality. The reality though is that there is risk that you have to accept the consequences of. The only intellectual piece of advice I would have for anyone is that you refresh the need to accept this risk every time you go out,, or your own complacency will add to the danger you have put yourself in.

Yeah I agree with your comments about the forecast, what I have found/experienced over the years is that when CAIC digs pits "Above treeline" it pretty much means wherever they can get their sled to or easily access via a short tour, and in reality its only maybe a few hundred feet above treeline. The snow is so much more variable in the true alpine zone that CAIC does not forecast for. This leaves people thinking that its all good at 12,13,14 thousand feet if the avy forecast says green.

[the_hare]

However, how are we as backcountry users supposed to know when to go?

For me it's not so much a question of when as it is where. Given that most avalanches happen on slope angles between 30-45° and persistent slab problems often rest on NW-SEish aspects, many 14ers take up alternate routes during the winter to avoid these suspect slopes. Including Grays & Torreys from Loveland Pass, others off top of my head are Pikes, Yale, Humboldt, Antero, Princeton... the snowpack is just so notorious in this state that I think most ppl would prefer to avoid the risk by taking routes that head up low-angle slopes towards ridges. Filtering for winter trip reports shows a lot of these more avy-safe routes on various peaks. I've still gone out on these routes on orange days even but even on low-angle stuff I've found considerable-forecasted conditions to make for poor trenching in the snow sometimes.

I also keep heuristic traps in mind in the backcountry and I believe this was one of the main factors behind the most recent fatality on this slope off Kelso. Especially on mtns closer to the front range there can be a lot more traffic from ppl who out of ignorance or otherwise make poor route choices thru avy terrain and get lucky. If I woke up early, drove however long, and have postholed for the nth time thru willows, theres a high chance I'm going to take the more dangerous established trench than keep going down a more taxing safe one. Another good reason i think to become familiar with slope angle tools and planning routes that avoid avy terrain as much as possible altogether.

[Jorts]

Why was there no avalanche on Dec. 14th? It could be that additional wind loading was needed before the slope would have slid. The bump in the avy danger rating on the 16th reflected the additional loading. But it's also plausible that a small avy could have occurred on the 14th, but due to good luck, possible trigger points were avoided on that day. Since we can't know which of these possibilities is correct, I would simply avoid traversing this slope when snow is present (unless I am confident that a stable spring snowpack is present).

Chances are the new snow was not cohesive on the 14th and/or there wasn't enough of it to tip the scales.
From CAIC for the 14th:

After many days of cold, clear weather, the snow surface has become faceted and very weak. Even small amounts of snow could bump the danger as the old snow will support minimal load before it fails. Very soft slabs could result in avalanches, and they may gouge into the facets, creating more significant slides than the slab depth warrants. Strong winds will also contribute to the quick slab development. If you see rounded pillows of new snow beneath ridges and in cross-loaded gullies, go around them or carefully assess them for slab development before committing to steep slopes. Cracking and collapsing are great signs to avoid steep terrain. This new surface problem develops first in the Park Range before spreading east and south.

And for the 16th:

As winds continue to blast over the coming days, slabs will get thicker and more connected, reaching further down into near-treeline terrain. With such weak snow beneath these slabs, you can expect to start triggering avalanches from below and from a distance. They may begin to wrap more extensive terrain features. Although we are keeping Wind Slab avalanches listed for another day, these will start acting more and more like Persistent Slab avalanches, and you should treat them as such.

Although snowfall amounts were minimal over the last week, the old, weak surfaces and newly drifted snow mean you need to reset your thinking. Slopes that were safe to travel on or under last week now may harbor a nasty slab that will lure you out onto a slope before an avalanche breaks around you. If these step down to basal weaknesses, you may trigger a large avalanche similar to the Mines Peak slides from a couple of weeks ago. Time will tell how reactive these new and old layers become, but for now, taking a step back to slopes sheltered from the wind or lower-angle slopes seems like a good move.

And the experts get it wrong all the time (just like the weather).

I realize you probably are not trying to sew distrust of the forecasts put out by CAIC but they are amazingly accurate for forecast zones despite spatial variability. If you read accident reports over the last decade or so, they always include the forecast for that day and the accident reliably corresponds to what was expected based on the forecast. If a person does not know much about avalanches, the CAIC forecast is the best means to figure out what is going on with the hazard in the mountains.

[14ercooper]

I don't remember the exact source, but I saw that somewhere north of 95% of avalanche fatalities occur in the areas that CAIC forecasts avalanche problems.

[SkaredShtles]

<snip> one by one, so if the slope does slide, you can rescue your own party.

This is one of the simplest things that you can do (presuming you're not out solo). And should be SOP for anyone recreating in the winter in the mountains.

[Mel Kel]

I don't remember the exact source, but I saw that somewhere north of 95% of avalanche fatalities occur in the areas that CAIC forecasts avalanche problems.

I'd believe that. Probably influenced by the fact that that's where snow conditions are "good/fun" to ride; no one wants to ski the south crust. (Except me, actually )

I also heard on an avalanche podcast that most accidents happen on Moderate/Yellow days. It's hard to know if that number is inflated because there are more people out recreating on moderate days (vs considerable/high forecast days where people tend to stay in) or because people don't perceive the Moderate/yellow rating to be as serious and therefore get into trouble.

[14ercooper]

I also heard on an avalanche podcast that most accidents happen on Moderate/Yellow days. It's hard to know if that number is inflated because there are more people out recreating on moderate days (vs considerable/high forecast days where people tend to stay in) or because people don't perceive the Moderate/yellow rating to be as serious and therefore get into trouble.

I was actually kinda curious about the accuracy of this statistic, since I'd heard it's highest on Considerable/Orange days, but doing some research on datasets of accidents since 2000 it looks like you're right. In Colorado, filtering down to fatal avalanche accidents it's about ~50% on Moderate days, with only ~30% on Considerable days (looking at some data sets from the last 10 or so years). There's also around 15% on High days, and small percents on Low days (~3-4%) and Extreme days (~1-2% - likely due to nobody really being out). If I don't filter down to only fatal accidents, the numbers skew a little bit differently at around 20%/50%/25%/5%/<1% (low doesn't mean no danger!), which kinda tracks that higher risk days are also more likely to lead to bigger/more dangerous avalanches.
Most accidents (a bit over 50%) happened on skis, with a fairly even spread on the other 50% between snowboard, foot, snowmobile, and snowshoes (with snowmobile being a touch higher and snowshoe being a touch lower). About 2/3rds of avalanche accidents involve some variety of soft slab (mostly storm slabs and softer wind slabs - people love skiing powder!), with the remaining ~1/3rd being mostly hard slabs (some harder wind slabs, but mostly various persistent slabs). There's a touch of various wet slabs, wet loose, loose dry, and even a few ice falls mixed in, but they're much lower.
About 85% of avalanches were either R2 or R3, split about evenly, with the rest being mostly R1 or R4 (also even split) with a few R5 mixed in. Around 60% were D2, with D3 and D2.5 being around 15% each and the remaining ratings also being evenly split in the final 10%.
All elevation bands were fairly evenly split in terms of number of accidents, with "Near Treeline" being just a touch higher and "Below Treeline" being just a touch lower. As far as aspects, about 30% were on E, about 17-18% each on NE and NW, about 10% each on N and SE, around 7-8% each on W and S, and only 1-2% on SW slopes. I didn't really see any correlation between "Aspect" and "Elevation" that implies certain Aspect/Elevation combos are more dangerous, versus just elevations and aspects.
Looking at the same dataset, I also verified my number that ~95% of accidents happened at an elevation and aspect that CAIC listed an active problem - so really good on them for how accurate the forecasts are.

Some interesting numbers, figured I'd share them!

[daway8]

...
Looking at the same dataset, I also verified my number that ~95% of accidents happened at an elevation and aspect that CAIC listed an active problem - so really good on them for how accurate the forecasts are.

Wow, thanks for the super detailed analysis! It's good to see hard numbers verifying the accuracy of the CAIC forecasts and also good to see actual numbers backing up the common refrain that low danger is not the same as no danger...

Obviously, the only way to avoid all avalanche danger is to avoid all avalanche terrain, but it's good to see that the data backs up the idea that those who do recreate in potentially dangerous areas can at least partly mitigate the danger by paying very close attention to all the details of the avy forecast (not just the color coding).

[Jorts]

I was actually kinda curious about the accuracy of this statistic, since I'd heard it's highest on Considerable/Orange days, but doing some research on datasets of accidents since 2000 it looks like you're right. In Colorado...

Did you compile those stats yourself? If so, you should submit them to CAIC. I've never seen a breakdown that thorough but maybe I just never looked in the right place.

With moderate, naturals are unlikely and human triggered are possible. The snowpack talks to you less. You see less slides on similar aspects and elevations. There's just less feedback available to keep you in check. I think late season moderate is the scariest hazard rating.

[14ercooper]

Yeah, I compiled (and de-duplicated) reports myself so I could run some numbers. Depending on the number, it's either drawn from ~150 fatal incidents, ~2000 reports (for things involving the forecast on the day), or ~5500 reports (for things that didn't involve the forecast on the day - just due to limitations in the historical forecast data I was working with). From there, it was just going through and running numbers to pull out some stats for various problem types, avalanche sizes, elevations and aspects, and whatnot that I thought would be interesting/relevant. I tried to pull in a large enough (and varied enough) dataset that any reporting irregularities and errors wouldn't skew the final numbers too much, to hopefully get some useful numbers. I also cross-checked with other people's published analysis (as well as CAIC's own published analysis) just to sanity check that I hadn't accidentally introduced a major skew into my numbers, and my numbers were generally in agreement with published numbers.
Stay safe out there if you're recreating in avalanche terrain!

[the_hare]

Some interesting numbers, figured I'd share them!

Made some graphs of these numbers if I got these figures right

Accidents.png (121.07 KiB) Viewed 5238 times

Fatal.png (116.35 KiB) Viewed 5238 times

Aspect.png (136.88 KiB) Viewed 5238 times