Avalanche Discussion

It doesn't really matter if this year's snowpack is set up worse than last year's snowpack because no one here, myself included, is going to stop skiing. Therefore, the first line of defense is guarding against perceptual errors because such errors are always more dangerous than the state of the snowpack. Sorry to get metaphysical but the snowpack just "is". I mention this because there have been some pretty insane posts in the Trip Reports with respect to perception of instability ( along with some interesting errors and misconceptions being passed around in this thread. ) Perception of instability has not been mentioned once in this discussion so far. NWAC produces incredible forecasts but those forecasts are simply not applicable at the slope scale.

Can you explain this more?  I'm not sure what you're getting at...

It absolutely matters if the snowpack is set up badly and will 100% effect my decisions when I go skiing.  Stop skiing?  Certainly not, but it's not a binary choice.  If the snowpack is fragile, I'll hunt out trees and low angle slopes and approach anything steep with more respect and caution than I would if we had a predictable, bombproof base.  How does the pack's instability not matter?  And we're skiing the slopes that NWAC is discussing in their forecasts... slope scale?  

Maybe I'm not understanding what you're getting at.  Can you elaborate?

EDIT -- looks like you explained it some -- I see where you're going with it now. Though I know I make go-no go decisions from time to time based on the forecasts. Usually it's not "no go skiing", but it influences the destination without a doubt.

Just a few additions to BirdDog's post and a few other posts here. In 2003 the problem was layers that *weren't* deeply buried. ( BirdDog, maybe you meant "very deeply buried" relative to the total height of snow? ) The snowpack was much thinner than usual in 2003. As most people here know, the large fatality count in 2003 ( in BC ) resulted from two incidents in which a large number of people were simultaneously placed in harm's way. Either of these events could have happened in any year ( with any snowpack ) as recent events have shown.

It doesn't really matter if this year's snowpack is set up worse than last year's snowpack because no one here, myself included, is going to stop skiing. Therefore, the first line of defense is guarding against perceptual errors because such errors are always more dangerous than the state of the snowpack. Sorry to get metaphysical but the snowpack just "is". I mention this because there have been some pretty insane posts in the Trip Reports with respect to perception of instability ( along with some interesting errors and misconceptions being passed around in this thread. ) Perception of instability has not been mentioned once in this discussion so far. NWAC produces incredible forecasts but those forecasts are simply not applicable at the slope scale.

You are correct, the large number of fatalities was in part due to the fact that a large number of people were placed in harms way. Another factor was the size of the releases. You write about the perception of instability. In the 02/03 Kooteney snowpack, a weak layer formed in late November. This weak layer persisted throughout the season, until it caused releases, some as late as March. The perception of instability (in late season) was in part masked by the fact that the weak layer was deeply buried; meaning the weak layer was at the bottom of the snowpack; and thus often overlooked, especially when digging pits which generally are not dug to near ground level (at least not while skiing).

You are also correct in stating avi forecast are to be applied in a general sense and are not to be taken as gospel for any one slope at any given time. My concern this year is the current weak layer will become deeply buried as we receive more snow and may persist for some time, possibly into spring. NWAC has stated this.

So, how do we judge the perception of stability/instability later in the season with a weak layer near the bottom of the snowpack? How do we adjust our decision making to account for this?

We'll certainly be skiing, but we may not get onto the steeps at all...

Then how will you create the momentum necessary to go down the hill and make turns?

Then how will you create the momentum necessary to go down the hill and make turns?

Seriously? :

I don't have any answers on how to account for perceptual errors ( i.e. the persistent forms are associated with higher levels of perceptual errors because they're out of sight and often out of mind and harder to trigger when buried deeply, etc. ). Perception of instability is much better for new snow because the instabilities are often near the surface. Some of my thoughts on this snowpack discussion are: we're getting some older information ( now the prior ) and some newer information ( the current ) and I'm not really sure how to sort out what is hard snowpack data and what is speculation. No problem with speculation or forecasts in this thread but I don't really know how to make sense of the information. Second, the data in this thread doesn't really refer to the spatial scale of weak layers noted by observers, which makes it difficult for me to understand where the weak layer might be a concern. Over a wide area? A small area? A single drainage?

I guess my approach, for my own tours, is really trying to be extremely objective about the snow over which I'm traveling, rather than thinking about what the bulletin said. I may choose one area over another ( or I might stay home as I did this weekend ) based on the bulletin but I don't choose aspect or specific slopes based on the bulletin. Also don't tend to pay attention to the instability rating in the bulletin because it can be much higher or much lower than wherever I'm skiing.

When I say that the snowpack doesn't matter, I am referring to my earlier comment: "the snowpack just is". Everyone, including myself, is going to ski this snowpack, so the state of the snowpack is simply something to accept. So I guess this is my way of saying that I've accepted the instability and prefer to plan around it rather than trying to outsmart it or whatever. For example, orographic snow can be a highly localized phenomenon. Therefore the NWAC bulletin might forecast "moderate" danger or even "considerable". However a small ridge might receive much more snow ( 10x ) than the surrounding terrain. In that case, danger at that area could very well be extreme, even though the forecast said "moderate".

To address an earlier question: Avalanche forecasting has several scales. Synoptic, meso, micro ( and possibly slope depending on your training and interpretation. ) NWAC bulletins apply to the synoptic scale and meso scale but do not include evaluation of instability for specific slopes. Similarly, avalanche forecasting has two primary scales: space and time. Forecasting difficulty ( and error rate ) increase as spatial size increases and as length of time increase. It's much more accurate to forecast instability for a single slope for a very short length of time than to forecast instability for an entire mountain range for three or four days. In any case, the NWAC bulletin includes information about aspects and elevation bands that can be used as part of a slope scale forecast but this information should *never* be used alone.

I suppose my real question is as follows: since data sampling forms much of the basis for perception of instability, if we are supplying observations and discussion here, how are we going to supply the information so that it's most useful? Just knowing that weak layers exist ( or that the snowpack is upside down ) doesn't really help much. Furthermore, I'm not entirely convinced that ( set up poorly or not ) the snowpack is primarily unsafe. According to The Avalanche Handbook, the seasonal snowpack is stable most of the time; the snowpack is seriously unstable for only a few hours each winter.

It's the prevailing state ( conditional instability ) that concerns me. This is where the snowpack is mostly stable, except maybe triggering level is a little lower than usual, and maybe pockets of instability aren't quite as rare as one would like. This problem doesn't require widespread instability across an entire mountain range ( which the "poor set up of the snowpack" portion of this thread seems to discuss. ). Conditional instability just requires a patchwork of time-dependent critical flaws with (n) areal size and (m) distribution ( both at random ) to create a nice mess for everyone. Personally, I think it would be fantastic if instability were found everywhere across the Cascades because then a nice big cycle could clean things out. It's secretive instability that worries me. ( Perception wise ).

Second, the data in this thread doesn't really refer to the spatial scale of weak layers noted by observers, which makes it difficult for me to understand where the weak layer might be a concern. Over a wide area? A small area? A single drainage?

Every pit I've dug this new season I've seen one commonality - a persistent faceted crust at or right near the ground.  This goes back to early snows in November as well, and on just about every aspect you can imagine.  From North Cascades to South.  Everything above this layer has been varied and hasn't carried over as "true" from one aspect to another, or from one area to another.  We've found safer areas to ski using data we get from the layers above this faceted crust.

To address an earlier question: Avalanche forecasting has several scales. Synoptic, meso, micro ( and possibly slope depending on your training and interpretation. ) NWAC bulletins apply to the synoptic scale and meso scale but do not include evaluation of instability for specific slopes. Similarly, avalanche forecasting has two primary scales: space and time. Forecasting difficulty ( and error rate ) increase as spatial size increases and as length of time increase. It's much more accurate to forecast instability for a single slope for a very short length of time than to forecast instability for an entire mountain range for three or four days. In any case, the NWAC bulletin includes information about aspects and elevation bands that can be used as part of a slope scale forecast but this information should *never* be used alone.

I suppose my real question is as follows: since data sampling forms much of the basis for perception of instability, if we are supplying observations and discussion here, how are we going to supply the information so that it's most useful? Just knowing that weak layers exist ( or that the snowpack is upside down ) doesn't really help much. Furthermore, I'm not entirely convinced that ( set up poorly or not ) the snowpack is primarily unsafe. According to The Avalanche Handbook, the seasonal snowpack is stable most of the time; the snowpack is seriously unstable for only a few hours each winter.

Agreed that this information should not be used alone.  That's why regular bc travelers' pit info and observations can be so important to friends who travel the bc, TAY, or NWAC - because we take into account the micro scale as you say.  Pits, rouchblocks, ski cuts, turning around - it all helps to evaluate conditions, and sometimes can be pertinent enough for NWAC to put up a specific orographic warning.  Hence, bc travelers in communication with each other and organizations like NWAC and TAY are able to share this info with one another.  Knowing weak layers are present IS important and does help.  I think it helps us to analyze possible scenarios that could occur in present conditions, that might not be considered if the snowpack was laid out differently.  As far as the avalanche guidebook's comment on snowpack stability, I would say that fits into the "information that should never be used alone."  The first weekend in December last year proved that wrong as the snowpack was incredibly unstable for almost 2 days straight.  NWAC posted increasing danger from high to extreme, and even in the areas we thought were relatively safe, we had telltale signs of instability.  Enough to make me turn around and wait a a few days to go back out.

_{Edited for grammer.}

Hey kids watch out tomorrow - the word EXTREME isn't used very lightly.


NWAC:
THURSDAY, THURSDAY NIGHT
Increasingly heavy snow or rain is expected Thursday, along
with very strong winds and significantly rising freezing
levels. This weather should produce a substantial increase
in the danger, with extreme danger expected to develop
Thursday afternoon and evening before slowly decreasing
Thursday night. Widespread natural or human triggered
avalanches should become certain Thursday afternoon, along
with an increasing potential for some large destructive
avalanches. Back country travel in avalanche terrain should
be avoided, and travel confined to lower angle terrain well
away from avalanche path runouts.


No matter what you may think - perceptual errors and scale aside - please use caution and be safe.

cd

I did a pole test in Gold creek valley at 3250 ft  on Tue. and found 2 ft settled on top of a 3in rain crust , to 12in  of air in the bushes, Maybe it was just the spot I checked but there was this void where I could see brush. my poles were out about 3ft for touring. May be 4.5 feet of snow on the ground. I stayed clear of slide areas and pooed around in the trees and by the creek.

FYI, skiing inbound at Snoq today, and in the lesser-traveled areas where you can still find virgin powder* I found eighteen inches of light fluffy over about an inch and a half of breakable rain crust, over another foot of light and fluffy. I can't speak for any other areas around, as I was pretty set on staying inbounds, but those layers are setting up to be pretty nasty. The rain crust is IMO the result of heavy rains from last weekend; anything that didn't slide in the last cycle should be considered extremely hairy. The crust was easy to detect with a pole test on gentle terrain, but thinned on slopes above 30 degrees, and on really steep slopes was really hard to discern. I didn't see anything similar to that while skiing Baker on Thursday, so I think it's just a result of the lower-elevation rain from last weekend.

Early in the day it seemed real stable, but late in the day I was setting off sluffalanches large enough bury a small poodle, and after dark the layer above the crust starting slabbing up enough for a few small cracks to appear. If that's the inbounds terrain at West and Central, I'd be real careful about heading into the BC, even if NWAC had dropped the dreaded "E" word from the forecast.


* I spend a lot of time at Snoq. Yes, even Summit West and Summit Central have stashes. No, I won't tell you where.

minor correction, hedgie...the frz rain crust was formed New Year's Day....which seemed like a weekend.

Oh, yeah. I spent New Year's Day getting the snow on Baker, but you're probably right; that was likely rain at Snoq.

NWAC

Looking like some dicey stuff in the next couple days...

Forgive me if this seems like a digression, but looking at this article about the recent financial catastrophes, I find it interesting that this sounds exactly like snow science:

www.nytimes.com/2009/01/04/magazine/04risk-t.html

Here's a quote from page four:

What will cause you to lose billions instead of millions? Something rare, something you’ve never considered a possibility. Taleb calls these events “fat tails” or “black swans,” and he is convinced that they take place far more frequently than most human beings are willing to contemplate. Groz has his own way of illustrating the problem: he showed me a slide he made of a curve with the letters “T.B.D.” at the extreme ends of the curve. I thought the letters stood for “To Be Determined,” but that wasn’t what Groz meant. “T.B.D. stands for ‘There Be Dragons,’ ” he told me.

There be dragons!

As grad student I attended a seminar where the speaker argued that the climate sensitivity was log-normally distributed.  Loosely speaking this means it has a longer, fatter tail at the high end of the range then a normal distribution.  Thus while the expected sensitivity might still be relatively small (e.g. a few degrees C), there exists a significant chance the outcome will be much worse.  Years later, it's still the thing that worries me most about climate change.

In the mid 1980's a few traders realized that the actual daily changes in foreign currency exchange rates had a fatter tail then the commonly assumed log-normal distribution.  They made a lot of money.

I have no idea whether the tail is fat or skinny on any of the relevant snow science distributions or even what they look like.  I think this type of argument is most relevant to persistent instabilities where there is a low probability of triggering something, but grave consequences if it is.  Other then our propensity to fall into heuristic traps, these tail events are the most worrysome to me because they are the most unpredictable for experienced, avi savvy skiers.

Too bad you can't apply the principles of triangular arbitrage to evaluation of instability.

Yeah, especially the idea that in theory triangular arbitrage is supposed to be risk free. I could go for some risk free bc!

Today's quote from NWAC doesn't lead me to think of the word "risk-free":

...extreme danger does not adequately emphasize the extent of the anticipated avalanche potential.

NWAC forecast

I was thinking more in terms of lemons and expert halos. Now we're in a more classical northwest cleansing period and the risks are rather obvious.

It's kinda looking like this upside-down snowpack is on it's way to being rightside-up given the recent weather patterns, and forecasted dry period over the next weeks... Cleansing period? Any thoughts/testimonials?

With the freezing level hovering about 10-12,000 ft. the snowpack has consolidated over the past number of days.

In the future we should only have to worry about the new snow on top and wind slab build up.

Update (my comments are for Central Washington to Southern Oregon)...with the understanding that conditions may be different at elevations above 8-9,000 ft. As soon as it snows (any amount) and the wind starts to blow surface slabs with once again be an issue.

(I keep forgetting their are others outside of Oregon and Washington that read TAY...)

It really depends on what area you're going to be skiing. Down here in Oregon, the the snowpack has defitinetly consolidated and the avy danger has gone down, but we haven't seen as much rain as up north. I'm still concerned about the buried depth hoar, especially at the higher elevations, and will be keeping an eye out as spring time approaches.

Also, I'm headed to the Whistler area next week, and the danger is still very high in that area due to the continental like snowpack. This winter's upside down snowpack has made me appreciate living in the northwest (I wouldn't like to see this stuff every stinking year...........)

We should really be preparing ourselves for the likely eventuality of a bad persistent weak layer once there is sufficient load and cohesion overlying the existing surface structure. One of the surfaces that exists is an amazingly slick ice glaze. I saw this surface near Paradise in a few isolated spots but also noted that one TAY trip report mentioned an ice glaze near Baker. If it is the same surface layer at Baker it will definitely be one to watch for in the weeks ahead. In general the bond of eventual new snow is not likely to be very good even absent this ice glaze.

I disagree with some earlier posts regarding the application of NWAC or other avalanche forecasts on a slope scale. Avalanche forecasts are used in the planning phase of a trip and NWAC (or other forecasts) combined with telemetry obs are the primary sources I use for planning a trip down to the slope scale. I then use the power of observation once at a destination to modify the original trip plan if necessary or desireable. Initially, most of this modification is done visually with wind effects and visible avalanches (if any) playing a major role. The keys in being responsive to the potential for trip changes are to remain alert and to pay attention to observations and secondly to make sure to ski with partners who are open to communication. Any time unexpected observations can be made and also at any terrain-critical decision point it is the best practice to regroup and to communicate.

The problem with observations are really twofold: 1) Visual observations (primarily of wind effects with respect to dry snow decisionmaking) are limited by visibility. By definition then, decisionmaking and route scale (perhaps meso is a good word here) choices are limited by poor feedback from observations with poor visibility. 2)Snowpack observations based on test results, particularly in the case of persistent weak layers, may be unreliable for anything greater than the precise location of the snowpack test. New information on the ECT may provide some greater confidence but the test itself is very new. With persistent weakness I would personally rely to a greater extent on the avalanche forecast (which is based in part on multiple observations like the professional INFO-EX in Canada). It's worth noting that CMH in Canada ruled out by fiat 2/3 of it's heli-skiing terrain in the winter of 2002-3; a winter dominated by unpredictable and multiple persistent weaknesses. Even just two weeks ago both skiers near Washington Pass and Canadian forecasters warned of unpredictable and highly variable snowpacks, to the extent that at Glacier National Park forecasters maintained a high rating for nearly a ten day period due to such weaknesses. Natural slides were still coming down and results from control work were definitive.

In snowpacks with sufficiently buried weak layers I think it wisest not to try to forecast stability on most slopes (perhaps with a few exceptions) but rather to turn one's attention to the consequences of triggering a slide. Persistent weak layers are easier to trigger when deeply buried than are most storm snow layers and triggering a deep weak layer obviously opens the door for the worst of potential consequences. The ultimate key to safety is wise use of terrain not trying to be a snowpack guru, for, as La Chapelle said, "Don't try to outguess the snow". This is especially so with snowpacks dominated by persistent weakness. Everyone over time will make mistakes in evaluating snowpack stability on all scales, slope or otherwise.

Nice treatise on our current snowpack, Gary.
Thanks.

1. The bulletin is an outstanding, high-quality resource that is simply not applicable at the slope scale. It is dangerous to use a synoptic scale or mesoscale forecast product as the sole source of information for go/no go decisions. From trip reports, it seems most TAY members pay attention and evaluate instability in the field - regardless of the bulletin. This is definitely the right course of action. Relying on the bulletin is foolish.

2. The three principle difficulties in forecasting are a.) spatial and temporal variability, b.) incremental changes to weather and snowpack, and c.) variations in human perception. The NWAC forecast specifically discusses how to use their forecasts on their web site. I don't think NWAC wants anyone to use their forecast at the slope scale.

For Gary_Brill. I'm sure you know all of this already but I'll explain my thinking.

1. Your post disagrees with my thoughts on using the bulletin at the slope scale. I'll explain my reasoning.

The bulletin is a great tool for choosing which mountains to go ski ( or whether to go out at all ) but the scale of the information ( synoptic scale and sometimes mesoscale ) doesn't match the scale of an individual slope ( the slope scale ).

It is unwise to forecast stability for an individual slope with Class III information. Class I and Class II information more closely match the scale of the problem. I understand that information from specific slopes, and even reports on TAY, is integrated into the bulletin but even then, correct forecasting procedure says that information from each data class must be weighed according to the degree to which the information reveals instability.

For slopeside forecasts, this information must be relevant to the current place and time and must be derived from the conditions at hand. Not the bulletin.

2. Avalanche forecasting is a Bayesian activity. This requires the prior ( data about the history of the terrain, weather, and snowpack ) and the likelihood ( data about the current terrain, weather, and snowpack. ) Finally, the posterior ( the prediction ) is assembled.
One important characteristic of Bayesian revision is that a single piece of data ( cracking in the snowcover ) has the power to completely revise the entire forecast.

According to The Avalanche Handbook, you can't form an accurate prediction without the prior and the likelihood. The bulletin does not include information on instability for specific slopes ( the likelihood ). Furthermore, The Avalanche Handbook specifically states that over-reliance on the bulletin is probably the source of many backcountry accidents.

3. Your post doesn't actually address the key problem with observations.

Observations and data sampling, especially snowpits, form much of the basis for perception of instability. Problems arise when these observations foster "false stable perception". It's like the proverb: "absence of evidence is not evidence of absence".

This is especially important during conditional instability - which is the prevailing state of the winter snowpack. Perception is always poorer during conditional instability than absolute instability and low instability. Data sampling can be a helpful, or harmful, part of the search for instability.

Anyway, poor visibility and the well-known difficulties of extrapolating slopeside instability tests across a wider area, aren't nearly as important as how the results of observations alter the recreationist's perception of instability.

The "human factors" of course come into play if you start to examine what the recreationst "wants" to perceive about instability. That's really the key issue with respect to observations: how observations affect perception of instability.

The correct procedure during poor visibility, or any other high uncertainty situation, is to implement a very wide margin of safety.

4. Quote: "Persistent weak layers are easier to trigger when deeply buried than are most storm snow layers and triggering a deep weak layer obviously opens the door for the worst of potential consequences."

I profoundly disagree with your statement. Skier triggering of instabilities below 1 meter is rare.
If you want to talk about deeply buried weaknesses, you have to discuss the problem of perception. The persistent forms, when deeply buried, present a very serious perception problem, regardless of the energy required to release the slope.

Deeply buried persistent forms are not easy to trigger when skiing. Shallowly buried persistent forms can be skier-triggered and the fracture can propagate to areas where the snowpack is very deep.

I'm surprised you would even say that "deeply buried weaknesses are easier to trigger" because it's not true. Storm snow instabilities are much easier to release than deep instabilities. Soft slab avalanches that run in storm snow are responsible for many fatalities.

Once again, winter 2003 is discussed. Those famous accidents occurred because a large number of people were in harm's way. I'm not sure I see the parallel between the snowpack that is hundreds of miles away ( and in a different snow climate ) and the snowpack that is "right here, right now"?

I've heard forecasters in Revelstoke compare the current snowpack in their area to the winter of 2003 but I haven't heard any forecasters in Washington compare the snowpack here to the snowpack in the southern interior of British Columbia. I'm not sure how/why the avalanche bulletin up at Rogers Pass ( I take it that's the Glacier Park to which you refer ) has any relevance to conditions here in the Cascades?

But all this aside, it is true that deep instabilities have a well-deserved reputation for producing very dangerous avalanches. However, I'd hate to have someone read this thread and think that storm snow instabilities are harder to trigger than deep instabilities.

6. Quote: "We should really be preparing ourselves for the likely eventuality of a bad persistent weak layer once there is sufficient load and cohesion overlying the existing surface structure."

What does this mean? Did you mean to say "We should prepare ourselves for another avalanche cycle once there is a cohesive slab overlaying a persistent weak layer."?

Otherwise, can you clarify your statement?

Load and cohesion are good for stability in the long term. In fact, the persistent forms are dangerous because they resist strength gains from load ( due to their anisotropy ) which means the weak layers often have very poor cohesion ( which would be evidenced by hand tests and a low number of bonds per unit volume ).

7. Ice glaze.

Sure an ice glaze exists in places. However the temperature of the snow that falls onto the ice glaze, and the temperature difference between the ice glaze and the new snow crystals, is what determines how the new snow bonds to the ice glaze.

Does it seem reasonable to expect a nice soft slab cycle to run on this glaze? Sure. In some places. Does it seem reasonable to expect a nice soft slab cycle to run in places where this glaze isn't found? Sure. In some places.

Is this ice glaze the layer to watch? Possibly. But with the general powder hound culture in these parts, I'd probably be more concerned about instabilities in new snow ( soft slab avalanches ).

8. Quote: "New information on the ECT may provide some greater confidence but the test itself is very new."

Did you mean propagation saw test? Because as far as I'm aware, the ECT has been validated. Doesn't the ECT result in fewer "false stable" results ... and more "false unstable" results? The ECT is specifically designed to balance fracture propensity with propagation propensity - very helpful for instabilities in new snow.

9. Quote: "With persistent weakness I would personally rely to a greater extent on the avalanche forecast (which is based in part on multiple observations like the professional INFO-EX in Canada)."

My thoughts are that persistent weaknesses call for conservative terrain choices. I bet most avalanche professionals would agree. Relying on the bulletin is not the same as making conservative terrain choices.

1. The bulletin is an outstanding, high-quality resource that is simply not applicable at the slope scale. It is dangerous to use a synoptic scale or mesoscale forecast product as the sole source of information for go/no go decisions. From trip reports, it seems most TAY members pay attention and evaluate instability in the field - regardless of the bulletin. This is definitely the right course of action. Relying on the bulletin is foolish.

1. Your post disagrees with my thoughts on using the bulletin at the slope scale. I'll explain my reasoning.

The bulletin is a great tool for choosing which mountains to go ski ( or whether to go out at all ) but the scale of the information
( synoptic scale and sometimes mesoscale ) doesn't match the scale of an individual slope ( the slope scale ).

Nobody that I know is relying on the bulletin as the sole source of go/no go decisions. What I was saying is that this thread has over-emphasized slope scale, etc. arguments at the disadvantage of losing track of the value of avalanche bulletins as a planning tool. That is one of their chief values. Along with telemetry and trip reports the bulletins do provide the best source for trip planning - call this macro routefinding decisions - planning where one will ski. I obviously then went on to say that observations, visual and other, allow one to modify one's initial plan at a meso scale - different slope or specific route. Extremely small scale - micro scale modifications of route - are then made based on visual and other observational cues.

The value of the bulletin is enormous in choosing an initial plan and has slope specific value with respect to aspect and for choosing tests and locations for such tests.

2. Avalanche forecasting is a Bayesian activity. This requires the prior ( data about the history of the terrain, weather, and snowpack ) and the likelihood ( data about the current terrain, weather, and snowpack. ) Finally, the posterior ( the prediction ) is assembled.
One important characteristic of Bayesian revision is that a single piece of data ( cracking in the snowcover ) has the power to completely revise the entire forecast.

I don't disagree. Any observation allows one a decision point for modifying one's plan or route.

According to The Avalanche Handbook, you can't form an accurate prediction without the prior and the likelihood. The bulletin does not include information on instability for specific slopes ( the likelihood ). Furthermore, The Avalanche Handbook specifically states that over-reliance on the bulletin is probably the source of many backcountry accidents.

I think you are over reaching here. The bulletin speaks to aspect, often specific regions, and specific snowpack problems such as surface hoar or a faceted suncrust in the worst case scenario. Likewise being aware of a glaze or freezing drizzle crust apparently near the volcanoes (at least) is very pertinent - although not slope specific.

3. Your post doesn't actually address the key problem with observations.

Observations and data sampling, especially snowpits, form much of the basis for perception of instability. Problems arise when these observations foster "false stable perception". It's like the proverb: "absence of evidence is not evidence of absence".

This is especially important during conditional instability - which is the prevailing state of the winter snowpack. Perception is always poorer during conditional instability than absolute instability and low instability. Data sampling can be a helpful, or harmful, part of the search for instability.

Nor was I trying to address the issue of stability tests in a general sense since making observations is a no brainer.

Anyway, poor visibility and the well-known difficulties of extrapolating slopeside instability tests across a wider area, aren't nearly as important as how the results of observations alter the recreationist's perception of instability.

Visual observations of wind effects are very critical in forming an initial expectation for new snow instabilities on certain slopes and can be very valuable in wet snow instabilities.

The correct procedure during poor visibility, or any other high uncertainty situation, is to implement a very wide margin of safety.

And not to travel above treeline in periods of higher hazard with poor visibility.

4. Quote: "Persistent weak layers are easier to trigger when deeply buried than are most storm snow layers and triggering a deep weak layer obviously opens the door for the worst of potential consequences."

I profoundly disagree with your statement.

You misread it. Deeply buried weak layers are easier to trigger than are storm snow weak layers of the same depth because persistently weak layers have very low bond densities and are with a few exceptions weaker than are storm snow layers by the time such layers are deeply buried. Although one could argue that storm snow layers could also be very weak, by the time they are buried more than a couple of feet storm snow interfaces tend to have gained some strength.

Skier triggering of instabilities below 1 meter is rare.

If you want to talk about deeply buried weaknesses, you have to discuss the problem of perception. The persistent forms, when deeply buried, present a very serious perception problem, regardless of the energy required to release the slope.

Deeply buried persistent forms are not easy to trigger when skiing. Shallowly buried persistent forms can be skier-triggered and the fracture can propagate to areas where the snowpack is very deep.

I'm surprised you would even say that "deeply buried weaknesses are easier to trigger" because it's not true.

Disagree. Rare but less so with persistent weaknesses. I've triggered 3' and 4' slabs myself. Also, deeply buried weaknesses are often triggered from areas where overlying burden is shallower (which can be difficult to know - as is true apparently in the northern Rockies right now due to wind effects) or can be triggered when shallower weaknesses step down. In any case more deeply buried persistent weaknesses when they fail- as at Washington Pass and in Canada according to reports within the last couple of weeks - are likely to be catastrophic. Therefore, the fact that they are a meter or more buried is more than offset by the potential consequences when triggered.

Soft slab avalanches that run in storm snow are responsible for many fatalities.

True, but easier to manage.

Once again, winter 2003 is discussed. Those famous accidents occurred because a large number of people were in harm's way. I'm not sure I see the parallel between the snowpack that is hundreds of miles away ( and in a different snow climate ) and the snowpack that is "right here, right now"?

Bad weak layers are often regional. The Xmas weak layer this year was present from Oregon to Washington, from BC to Alberta to Montana and likely farther SE into the US Rockies. The high freezing levels and wet snow/ rain that mitigated that problem did so from Oregon and southern Washington initially and then to a lesser degree into Northern Washington and coastal BC. The extremely high freezing levels of late affected most of Western North America resulting initially in greater instabilities and thenceforth in slow strength gain of snowpack upper layers.

I've heard forecasters in Revelstoke compare the current snowpack in their area to the winter of 2003 but I haven't heard any forecasters in Washington compare the snowpack here to the snowpack in the southern interior of British Columbia. I'm not sure how/why the avalanche bulletin up at Rogers Pass ( I take it that's the Glacier Park to which you refer ) has any relevance to conditions here in the Cascades?

That they didn't mention it doesn't mean it wasn't the case. I am sure that digging a pit to the base of the snowpack at Washington Pass would yield the same weak layer as at Rogers Pass, the only difference being some consolidation at Washington Pass due to overburden pressure from very heavy snows the first week of January.

But all this aside, it is true that deep instabilities have a well-deserved reputation for producing very dangerous avalanches. However, I'd hate to have someone read this thread and think that storm snow instabilities are harder to trigger than deep instabilities.

Yet buried at the same depth that is the case. Nothing can be weaker than buried surface hoar or an advanced layer of depth hoar so it only makes sense that the weaker the layer the easier it is to trigger. Note for evidence that avalanches routinely slide on lower slope angles (poorer bonding) with persistent weaknesses.

6. Quote: "We should really be preparing ourselves for the likely eventuality of a bad persistent weak layer once there is sufficient load and cohesion overlying the existing surface structure."

What does this mean? Did you mean to say "We should prepare ourselves for another avalanche cycle once there is a cohesive slab overlaying a persistent weak layer."?

Otherwise, can you clarify your statement?

Load and cohesion are good for stability in the long term.

It will take an initial thin snowfall, subsequent faceting and loading, especially as cohesion is gained to lead to a highly unstable snowpack. By long term I assume you are talking about over time, whereas instabilities will arise with a gain in cohesiveness sufficient for propagation.

7. Ice glaze.

Sure an ice glaze exists in places. However the temperature of the snow that falls onto the ice glaze, and the temperature difference between the ice glaze and the new snow crystals, is what determines how the new snow bonds to the ice glaze.

Does it seem reasonable to expect a nice soft slab cycle to run on this glaze? Sure. In some places. Does it seem reasonable to expect a nice soft slab cycle to run in places where this glaze isn't found? Sure. In some places.

Is this ice glaze the layer to watch? Possibly.

I'll be worried about an ice glaze so slick that a fall on a 20 degree angle probably couldn't be stopped. I'll also be worried about what looks like it will be a very poor bond at most elevations and on most aspects to a very different old snow surface than the soon to come new snowfall.

But with the general powder hound culture in these parts, I'd probably be more concerned about instabilities in new snow ( soft slab avalanches ).

With sufficent snowfall and wind, but the weakness that will be of most concern for the longest period of time will be the bond to the existing surface and a high likelihood of near surface faceting of the first snowfall in the next cycle. Concern for storm snow instabilities will (with sufficient load and cohesiveness) will soon be outweighed by concern for dangerous deeper slabs.

8. Quote: "New information on the ECT may provide some greater confidence but the test itself is very new."

Did you mean propagation saw test? Because as far as I'm aware, the ECT has been validated. Doesn't the ECT result in fewer "false stable" results ... and more "false unstable" results? The ECT is specifically designed to balance fracture propensity with propagation propensity - very helpful for instabilities in new snow.

Validated by one or two sets of observations. It's still new.

9. Quote: "With persistent weakness I would personally rely to a greater extent on the avalanche forecast (which is based in part on multiple observations like the professional INFO-EX in Canada)."

My thoughts are that persistent weaknesses call for conservative terrain choices. I bet most avalanche professionals would agree. Relying on the bulletin is not the same as making conservative terrain choices.

It is ultimately all about terrain. The bulletin, especially when it calls attention to very weak persistent layers tells me that my terrain choices have to be all the more conservative - that also means that I have question whether I should be trusting of my test results if they lead me to believe that a particular steep slope is skiable . I have to consider (as always) the consequences of triggering.

Thanks for the discourse Cookie Monster.

I'll try to keep this simple... I take into consideration every bit of information and don't consider any of it gospel... Since I'm probably going to ski something most of the time, I take responsibility for my decision making --and since I'm probably wrong about how stable things are anyway (don't know anybody that can out smart the snowpack), I try to play it conservative and think of what slope would have the lesser of evil consequences (including the route back out from the bottom...) if I was to screw up on my evaluation--and hope for the best (or not ski anything sometimes!!!).

A few weeks back I skied Tumalo Mountain down here in Central Oregon (alone...a rare thing even if one wanted to..). Did three laps on the back as it seemed stable. A friend of mine said "oh it's stable..I'll head up tomorrow" after giving him my details of my ski day. I told him not to take my word about stability as it was just my "take". Point here is that I would not want anybody to think something was stable just because I thought my day was a stable ski day. Things change of course and so we all have to make our own final "evaluation" and decision--not rely on somebody elses.

Years ago the phrase "all the experts are dead" (grim, sorry...) was sort of circulating (regarding avalanche science...) I guess meaning you can't outsmart the snowpack.... Rumour has we are suppose to get 6" of new by Sun!!!! Fingers crossed for freshies......

Maybe we can try this again. This thread, especially posts by prominent local experts, is going to influence perception of instability in the Cascade Mountains. Therefore, I really think we should get this right.

Rather than argue with a local expert, I've formulated some questions for Gary_Brill. The only reason I'm asking these questions is because Gary_Brill is a local expert who has made some pretty strong statements about what to expect in the Cascade snowpack for winter 2009. I think, as you are an expert and a local educator of some reknown, that I would personally find it very useful if you could provide some additional clarification.

1. Quote: "We should really be preparing ourselves for the likely eventuality of a bad persistent weak layer once there is sufficient load and cohesion overlying the existing surface structure."

---- If correct, this is an important prediction. Do you mind explaining your basis?
---- Do you have evidence? Snow profiles from a variety of areas. That sort of thing?

2. Quote: "In general the bond of eventual new snow is not likely to be very good even absent this ice glaze."

---- If correct, this is an important prediction. Do you mind explaining your basis?

3. Quote: "It's worth noting that CMH in Canada ruled out by fiat 2/3 of it's heli-skiing terrain in the winter of 2002-3; a winter dominated by unpredictable and multiple persistent weaknesses."

---- Does Canadian Mountain Holidays have any ski tenures in the Cascade Mountains or are all their ski tenures in the Columbia Mountains?

---- How is the decision making process at Canadian Mountain Holidays during the winter of 2003 relevant to snow conditions in the Cascade Mountains for winter 2009?

4. Quote: "Glacier National Park forecasters maintained a high rating for nearly a ten day period due to such weaknesses. Natural slides were still coming down and results from control work were definitive."

---- How are snow conditions and avalanche conditions at Rogers Pass, which is the interior mountains of British Columbia, relevant to snow conditions in the Cascade Mountains for winter 2009? Your previous answer "regional phenomenon" is fine but I'd like to see some evidence in support of this rather than speculation. Otherwise, it's just speculation.

5. Quote: "In snowpacks with sufficiently buried weak layers I think it wisest not to try to forecast stability on most slopes (perhaps with a few exceptions) but rather to turn one's attention to the consequences of triggering a slide."

---- On what are you basing this advice? I only ask because this statement is a rule of thumb. Have you considered that some people might take your advice and forgo instability tests on slopes? Is that your advice?

6. Quote: "This is especially so with snowpacks dominated by persistent weakness."

---- Where is the evidence that the Cascade snowpack for winter 2009 is dominated by persistent weaknesses? Do you mean now or in the future? If you mean the future, do you mind explaining the basis for this prediction?

7. Quote: "Deeply buried weak layers are easier to trigger than are storm snow weak layers of the same depth because persistently weak layers have very low bond densities and are with a few exceptions weaker than are storm snow layers by the time such layers are deeply buried."

---- Do you have evidence to support this statement? Do you think a quick review of avalanche accidents in Washington State from 1995-2008 supports this statement?

8. Quote: "The extremely high freezing levels of late affected most of Western North America resulting initially in greater instabilities and thenceforth in slow strength gain of snowpack upper layers."

---- Are you absolutely sure that high freezing levels result in slow strength gains in the upper layers of the snowpack? Does instability persist longer in colder temperatures or warmer temperatures?

9. Quote: "Note for evidence that avalanches routinely slide on lower slope angles (poorer bonding) with persistent weaknesses."

---- Are you sure that poor bonding is the reason? Is it possible that there are other reasons that might be more important? Is it possible that numerous factors, such as the strength of the weak layer itself and the stiffness of the slab at the interface might be more important?

10. Quote: "Concern for storm snow instabilities will (with sufficient load and cohesiveness) will soon be outweighed by concern for dangerous deeper slabs."

---- If correct, this is an important prediction. Do you mind explaining your basis?

11. Quote: "Validated by one or two sets of observations. It's still new."

----  Are you sure the ECT has only been validated by one or two sets of observations?

I'm not trying to be snarky. Gary_Brill has made some interesting predictions and statements. I'm curious.

Maybe we can try this again. This thread, especially posts by prominent local experts, is going to influence perception of instability in the Cascade Mountains. Therefore, I really think we should get this right.

Rather than argue with a local expert, I've formulated some questions for Gary_Brill. The only reason I'm asking these questions is because Gary_Brill is a local expert who has made some pretty strong statements about what to expect in the Cascade snowpack for winter 2009. I think, as you are an expert and a local educator of some reknown, that I would personally find it very useful if you could provide some additional clarification.

1. Quote: "We should really be preparing ourselves for the likely eventuality of a bad persistent weak layer once there is sufficient load and cohesion overlying the existing surface structure."

---- If correct, this is an important prediction. Do you mind explaining your basis?
---- Do you have evidence? Snow profiles from a variety of areas. That sort of thing?

2. Quote: "In general the bond of eventual new snow is not likely to be very good even absent this ice glaze."

---- If correct, this is an important prediction. Do you mind explaining your basis?

My evidence is based on observation and I am trying to weave in experience from previous snowpack and weather scenarios that I believe are about to play out once again. From the current avalanche forecast this morning:

"The main dangers from the current snowpack are not
avalanche related currently, but stem primarily from
potential damage due to loss of edge and a resulting long
slide on the crust or possible injury from unsuccessful
negotiation of the semi-breakable crust. However, the
currently stable snowpack structure is becoming well
primed for a significant increase whenever appreciable
future new snowfall arrives, as the bond of new snow to
either the crust or surface hoar should be quite weak."

There is more information based on what appears to be snowpack evaluation in the current forecast.

3. Quote: "It's worth noting that CMH in Canada ruled out by fiat 2/3 of it's heli-skiing terrain in the winter of 2002-3; a winter dominated by unpredictable and multiple persistent weaknesses."

---- Does Canadian Mountain Holidays have any ski tenures in the Cascade Mountains or are all their ski tenures in the Columbia Mountains?

---- How is the decision making process at Canadian Mountain Holidays during the winter of 2003 relevant to snow conditions in the Cascade Mountains for winter 2009?

The point I was trying to make is that at CMH where one can safely say that the level of skill and experience is unsurpassed, in the presence of a very bad snowpack, their experts chose to simply rule out many runs rather than to try to evaluate each on the basis of snowpack evaluative skills. In otherwords they did the very same thing that a backcountry skier might do in the planning phase of a trip after viewing a concerning avalanche forecast. The difference is that their decision was based on their own perception (from multiple observations) of a scary snowpack combined with confirmation from other operators and entities through INFO-EX which is more or less the professional version of the snowpack exchange and avalanche forecasts we as recreationalists get.

4. Quote: "Glacier National Park forecasters maintained a high rating for nearly a ten day period due to such weaknesses. Natural slides were still coming down and results from control work were definitive."

---- How are snow conditions and avalanche conditions at Rogers Pass, which is the interior mountains of British Columbia, relevant to snow conditions in the Cascade Mountains for winter 2009? Your previous answer "regional phenomenon" is fine but I'd like to see some evidence in support of this rather than speculation. Otherwise, it's just speculation.

As I mentioned previously extremely weak layers are often regional, but vary in degree. Roger's Pass was likely very similar to Washington Pass until the large snowfall at the latter the first week of January. Remeber that our own forecasts warned of very similar structure to Rogers Pass throughout the Cascades until the major avalanche cycles of earlier January. The avalanche cycle at Rogers Pass began and lasted later there because it didn't warm up as quickly.

5. Quote: "In snowpacks with sufficiently buried weak layers I think it wisest not to try to forecast stability on most slopes (perhaps with a few exceptions) but rather to turn one's attention to the consequences of triggering a slide."

---- On what are you basing this advice? I only ask because this statement is a rule of thumb. Have you considered that some people might take your advice and forgo instability tests on slopes? Is that your advice?

I'm not giving advice but merely stating what I do. A backcountry skier only need be wrong once. When the snowpack is very weak or unpredictable in strength and variability I would sooner look at the consequences and punt rather than try to prove my expertise and then bet my life on it. Conservative terrain choices are something you and I agree on Cookie Monster. I believe all the more so when known significant weak layers or weak bonds exist.

The way I look at backcountry skiing decisionmaking is sort of a decisionmaking tree that works like this for me:

What are the concerns in the snowpack? This is what Tremper refers to as the "AVALANCHE DRAGON". Recent snows absent weak layers? > Choose terrain away from the worst problems which are usually wind deposition until there has been a sufficient period of stabilization. Strong winds? > Stay below treeline when possible. Still snowing heavily or warming during snowfall? > Same choices but be wary of locations where natural slides may come into play. In all of these cases I rely on observations both visual and snowpack related and constantly engage myself in making observations of all types so that I am not surprised.

Still for aggressive terrain choices I think of it as though all the stars must be lined up. In otherwords the snowpack must be homogeneous and well consolidated by time and strength gain or the amount of snow that can slide has to be definitively limited by a very strong basal layer. I have more confidence in the latter of the two cases although it must remain cold in the latter case for reasonable stability. This usually means that there is a bombproof layer (like today's surface) with a limited amount of new snow reasonably bonded to it. This is something that is far more likely to happen in a maritime snowpack. Even in this situation there are slopes I usually avoid along with terrain traps.

Recent snows that have fallen onto one or more weak layers such as surface hoar, depth hoar, faceted crusts, near surface facets, or an ice layer? Watch out! Continue to make observations but don't rely on stability tests to justify marginal terrain choices. Rather choose less agressive terrain. Pay attention to the consequences of triggering, realize that remote triggering and extensive propagation are possible.

Spingtime diurnal instability? > Is the snowpack well set up (consolidated)? Aggressive skiing on sunny aspects while watching for increasing penetration or crustal weakening is likely OK. I am aware of the possibility of long falls on certain slopes. Is the crust marginal (barely supportive or not supportive on foot)? > This case is more problematic and either is representative of a poor freeze or of recent less than well consolidated snowfalls. One has to be quite careful and observant here as with minimal warming deeper slides may be possible.

6. Quote: "This is especially so with snowpacks dominated by persistent weakness."

---- Where is the evidence that the Cascade snowpack for winter 2009 is dominated by persistent weaknesses? Do you mean now or in the future? If you mean the future, do you mind explaining the basis for this prediction?

I wouldn't speculate on the whole winter although we recently went through such a period. But it does seem likely that for a period of time we will be dealing with a snowpack with a variety of poor bonding surfaces that may remain lasting concerns - i.e. will become persistent weaknesses - which really just means that new snow layers will not bond as well or as quickly to these surfaces.

7. Quote: "Deeply buried weak layers are easier to trigger than are storm snow weak layers of the same depth because persistently weak layers have very low bond densities and are with a few exceptions weaker than are storm snow layers by the time such layers are deeply buried."

---- Do you have evidence to support this statement? Do you think a quick review of avalanche accidents in Washington State from 1995-2008 supports this statement?

My own personal experience with close calls supports this statement. I can deal with storm snow weaknesses but find that when I have been surprised it is almost always with persistent weak layers. The best supporting evidence as I mentioned earlier is that avalanches typically release at lower angles (may release) on persistent weaknesses such as buried surface hoar, depth hoar, faceted crusts, near surface crusts, or ice layers. One fatal avalanche with depth hoar in the Rockies alledgedly released on a 15 degree slope. Surface hoar often releases down to angles in the low 20's but even lower angles are possible under rare conditions. I've personally released faceted crusts on low 30 degree angle slopes more than once. Corollary evidence comes from avalanche accidents where with these weak layers those involved have obviously had difficulty in evaluating slopes because the slopes that fail fall outside the range of the victims previous experience in terms of both angle and scale (propagation).

8. Quote: "The extremely high freezing levels of late affected most of Western North America resulting initially in greater instabilities and thenceforth in slow strength gain of snowpack upper layers."

---- Are you absolutely sure that high freezing levels result in slow strength gains in the upper layers of the snowpack? Does instability persist longer in colder temperatures or warmer temperatures?

I could try to answer the first part of that question but not the second because it is conditional. High freezing levels result in more rapid rates of settlement, but not necessarily on shady slopes with crust surfaces. Initially high settlement rates (flow rates) may redistribute stress in the snowpack in such a way that underlying weaknesses receive greater stress and could fail if sufficiently weak. In the long term the warming, if not at an increasing rate, causes the recently settled layers to densify and begin to bond more rapidly than they would otherwise. Confidence in a snowpack that has undergone warming is for me not a certain thing until subsequent cooling has taken place for a period of time. Crusts may inhibit settlement rates.

9. Quote: "Note for evidence that avalanches routinely slide on lower slope angles (poorer bonding) with persistent weaknesses."

---- Are you sure that poor bonding is the reason? Is it possible that there are other reasons that might be more important? Is it possible that numerous factors, such as the strength of the weak layer itself and the stiffness of the slab at the interface might be more important?

I gave my understanding on this earlier here, but:

There are always multiple factors. but probably it is worthwhile discussing static and dynamic friction. Weak layers are weaker and fail more easily because of low bond densities (with load) at lower angles. Once failed persistent weak layers are often made up of plate-like or sugary grains. Both of these structures are "slippery" - they slide easily. So anything on top of them also slides easily, at lower angles. The harder and smoother the bed surface - the lower it's capacity for dynamic friction, the easier is it for snow to slide.

10. Quote: "Concern for storm snow instabilities will (with sufficient load and cohesiveness) soon be outweighed by concern for dangerous deeper slabs."

---- If correct, this is an important prediction. Do you mind explaining your basis?

I believe the current snow surface will not allow for good bonding in many places and on a variety of aspects and will become persistent for some unknown period of time.

11. Quote: "Validated by one or two sets of observations. It's still new."

----  Are you sure the ECT has only been validated by one or two sets of observations?

I know of a couple of groups that have done validation tests for the ECT but only in the past 2-3 years.

I'm not trying to be snarky. Gary_Brill has made some interesting predictions and statements. I'm curious.

I hope this is helpful.

So, if I read through this whole thread, is that like taking Avy 1? Just kidding...

Gary, wow, thanks for sharing a lot of expertise here. I know you teach avy classes and are a reknowned expert and teacher. Have you written/published any books about evaluating snowpack or, I guess more specifically studying the snowpack in the cascades (maritime snowpack)? If so, should I just look on Amazon?

So, if I read through this whole thread, is that like taking Avy 1?   Just kidding...

Gary, wow, thanks for sharing a lot of expertise here.  I know you teach avy classes and are a reknowned expert and teacher.  Have you written/published any books about evaluating snowpack or, I guess more specifically studying the snowpack in the cascades (maritime snowpack)?  If so, should I just look on Amazon?

I wrote a book for my Level I class (which I am no longer teaching), but have never published the book for general consumption - although it is something I've occasionally considered.