SNOW PITS... good or good for nothing?

[summit]

This tangent discussion seems worthy of its own thread:

2) One of the things I keep seeing in these reports is the detailed pit scores. While I think that there is a place for pits (especially for highly trained, experienced, and daily professional practitioners like ski patrol or avalanche reduction experts), the subjective nature of shear/compression tests combined with the spacial variability in snowpack to me means they are of dubious value. Too much opportunity for erroneous conclusions and analysis

Thank you for reemphasizing that point. Reiterating my earlier comment: "The snowpack is NOT your patient. It is a population. If you forget that for a moment, you will be too confident about a decision made with insufficient or non-representative data."

Like all decision making data, pit data is subject to all the confounding types of errors found in any statistical study. Yes, it is an art and a science, and handing a complex tool to an unprepared user won't yield good results.



Am I saying, "Don't bother with a pit!"? NO NO NO!

Am I saying, "Understand what you are doing and examine enough variables to make this ONE datapoint a useful part of your dataset!"? YES YES YES!



SO WHAT IS THE POINT OF A RECREATIONIST'S SNOW PIT?

When the goal is to stay alive and ski, the purpose of a pit is to TEST your EDUCATED assumptions with varying levels of confidence, not to tell you absolute truths.

What a snowpit is NOT good for:

1. Primary reason to ski a slope
2. Full and certain picture of another slope, or even the current slope
3. Identifying all concerns

For any backcountry traveler, a snow pit IS good for:

1. Examining snowpack structure for variations from existing knowledge
2. Identifying otherwise unrecognized concerns
3. Looking for reasons to say NO

Pits are another tool in the toolchest. They are an important tool, but not the primary tool. Anyone who takes a Level I should be able to get something out of a pit.



Remember the limits of the tool and the artisan:

1. Keep pits representative of your slope in question
2. Use tests within your skillset and knowledge
3. Base your conclusions on actual data and recognize what is just an assumption
4. Many quick pits and tests are better than a single complex one.
5. Practice improves speed, interpretation, and reduces user error
6. Get a saw! (You already carry a slope meter, right?)

You get many chances to gather the right information and analyze your slope correctly. You only get one chance to ski it. When gambling with your life, make an informed decision.



So... discuss!

[Bunion 2020]

Great post Summit. When I have a minute I will add a bit from my currently addled brain.

[JoeStrummer]

Nice post, this is something I contemplate.

My feeling is people see what they want to see in a pit. If you think the slope is going to kill you, you will probably find evidence of that. If you think the slope is good to ski, you will probably find evidence of that, as well. As you say, it's just another data point but I have skied enough in Colorado that I just assume there are nasties buried in there somewhere. That combined with some other paranoia means I'm sure I miss some awesome winter ski descents.

I skied 50 days in the backcountry last year in Colorado and dug 7 pits. Three of those were for fun because I like playing in snow and studying its metamorphosis throughout the season. Maybe another 20 bc days in Europe and dug no pits. I take self-preservation and partner safety very seriously but my own math proves I don't really include digging pits as part of that drill.

I guess I dig them more for a history lesson of the winter to date than I use them to test assumptions. I also ski various places every year, and digging pits in Colorado reminds me how truly awful the snowpack is compared to other destinations in Europe or Canada.

[goldenboy]

My feeling is people see what they want to see in a pit. If you think the slope is going to kill you, you will probably find evidence of that. If you think the slope is good to ski, you will probably find evidence of that, as well.

This is where I'm at with pits. If you read Avalanche center forecasts religiously, you already know what you're going to find, at least as a snowpack-wide generalized snapshot. I don't need to dig a pit to know that I'll find a hoar'n'crust sandwich 6" off the ground and some surface hoar layers in the upper snowpack right now in my local area.

Traveling to a different area... that's a different story and I might dig a pit to see what's going on. Or I might not. On a recent hut trip the snowpack told me more than any pit would have by repeatedly whoomphing on the flat skintrack. So I didn't dig.

[moose pit]

This is where I'm at with pits. If you read Avalanche center forecasts religiously, you already know what you're going to find, at least as a snowpack-wide generalized snapshot. I don't need to dig a pit to know that I'll find a hoar'n'crust sandwich 6" off the ground and some surface hoar layers in the upper snowpack right now in my local area.

Traveling to a different area... that's a different story and I might dig a pit to see what's going on. Or I might not. On a recent hut trip the snowpack told me more than any pit would have by repeatedly whoomphing on the flat skintrack. So I didn't dig.

You know what you expect to find.

I dug a hand full of pits last year, mostly in areas I didn't know or when I was curious about a known instability's consolidation. For me, since there is very little forecasting done in the areas I ski, I tend to dig a full pit once in a while but several quick pits every day I'm out. Coupled with observation and watching the weather I think it gives me a good idea as to what is going on.

I always hate to ski with people that think digging a pit can give them a green light to ski something that just doesn't feel right; hoping that one good test will absolve a bunch of questionable observations.


(good post summit - type two error lol)

[Ermine]

Some mountains don't have detailed Avalanche reports. You're on your own then and interpolating from the adjacent reporting regions.

[smitchell333]

Glad it struck a cord. I can't help but get a feeling from the 2nd Creek slide as well as others where the parties involved start reciting the pit scores like they are some sort of magical runes to tell you whether the snowpack is safe to ski, while the reality should be much more obvious from general observations about the time of year, snowpack structure, windloading, and terrain/slope.

Despite the fact that I've done maybe one or two hundred pits over 15 years of continuous b/c skiing myself, I just dont personally feel like it is that often that I get a whole lot out of them because;

1) Typically I feel like I already know most of what I end up seeing in a pit, either from CAIC or other observations like numerous pole probing, hasty pits, small test slopes, observations of natural activity, etc.

2) I don't trust one pit as truly representative of any slope that I'd ski because of issues of spatial variability. All too often there is no way to safely dig a pit on a representative slope - say if all of the slopes of similar aspect and elevation are 40 deg and windloaded. It has been documented that false stable results are somewhat common. http://www.fsavalanche.org/NAC/techP...alseStable.pdf

That said I do think pits have a place, especially for those trained and experienced in their execution and those that do them on a daily or weekly basis as one of their tools to continuously study the snowpack.

This is where I'm at with pits. If you read Avalanche center forecasts religiously, you already know what you're going to find, at least as a snowpack-wide generalized snapshot. I don't need to dig a pit to know that I'll find a hoar'n'crust sandwich 6" off the ground and some surface hoar layers in the upper snowpack right now in my local area.

Traveling to a different area... that's a different story and I might dig a pit to see what's going on. Or I might not. On a recent hut trip the snowpack told me more than any pit would have by repeatedly whoomphing on the flat skintrack. So I didn't dig.

Well put. This is pretty close to my thinking and modus operandi. Getting shooting cracks on the side of the skintrack when jumping on small test drift will tell you more and faster than any snowpit.

[Meadow Skipper]

There is no avy forecasting in my area, so I often (though not always) dig, but it's just a piece in the overall evaluation. Weather patterns and familiarity with the terrain and conditions play a part.

[DasBlunt]

One thing not mentioned here and most people I read about or see, are doing compression tests. According to Tremper, Extended Column Test is superior because of the ability of the ECT test in "indicating propagation potential as well as initiation". Most of the other tests, Tremper says, "only indicate the propensity for fracture".

In Colorado, from the little I know, propagation seems to be very important.

[jon turner]

I rarely ever dig a pit simply for the fact that a pit should never make one feel more confident in skiing a slope. It is used by a recreational skier to reinforce observations I've already been making, i.e. if I see a lot of signs that the snowpack is weak (cracks, slabbing, whoompfing), I might dig a pit to see what layers are causing this. I probably wasn't going to ski a high consequence slope when observing these conditions, so why did I need to dig a pit to reinforce this decision?

The most useful reason to dig a pit is to observe the evolution of a snowpack over a span of time. So, an avalanche forecaster is busy digging pits regularly on similar aspects in the same zone on likely a weekly basis to observe how the snowpack is changing, what's going on with the weak layers, etc. I can glean this info from the CAIC report, so it's not as necessary for me to do it personally. Now, if there was a zone that I skied on a regular basis, I might be digging pits every time I go to confirm what I'm reading in the avi reports, but it's not necessary.

My decisions to ski/not ski a particular slope are more related to what I'm seeing holistically as I'm travelling through the backcountry, skiing low risk test slopes, etc., knowledge of how the snow reacts during different parts of our season (i.e. usually less stable early season and becoming more safe in late March-May), which aspects stabilize sooner during which parts of the season, etc. Pits rarely come into play in making day to day decisions.

Edit to add that I might also dig a pit if I'm expecting to see signs of instability but am not. For example, in the early season, with a shallow snowpack, I usually expect to see lots of signs of instability and stay off of high consequence terrain. Two weeks ago, I was out and the slope seemed very stable, and I observed very few signs of instability. In this case, I very likely might dig a pit searching for any evidence of instability that I could find since I really am expecting it at this time of year. However, if I don't find anything, it doesn't mean I'm going to jump on a steep line. I know that with the shallower snowpack in the early season, that even though I might not observe weaknesses, it is by definition weak due to being relatively shallow.

[montanaskier]

another great post by Summit. I don't dig very often either. Like GB and others have said, you mostly know what you are going to find. I dig sometimes just to see and feel but never as a go/no go decision maker. I know I am overly cautious, but I just don't expose myself to highly dangerous terrain in the winter in Colorado. I know what is possible everytime I tour and I try to minimize those risks by meadow skipping alot in the winter. The risk simplly isn't worth the reward to me at this point in my life. I'll save the big steep lines for the spring.

[::: :::]

after all the proselytizing for being avy educated that i've read on this site, i'm surprised to read that folks are just eyeballing their lines and going with their personal experience.

even tho the various pit tests are ultimately touchy-feely in nature (ie they aren't hard data & don't translate to other pit tests), it seems they are still are valuable because they indicate what's going on under the snow where you are right now - for whatever that's worth, it's more than just eyeballing a line or reading a regional forecast

i've only been out in the BC a few times and i'm relatively new to all this, but isn't this exactly what we're all warned about in Tremper or AVY I class: experiential guessing? human error?

[RootSkier]

No one is saying that. They are saying that a pit is but one measure of overall snowpack stability and that it is only representative of that one specific spot.

Without digging a fullblown pit you should still be well aware of the general nature of the snowpack if you have been in that area and reading the forecasts. Also, what is the weather like? How long since the last snow? Is it warming up? Is there any evidence of natural slides? Is there any evidence of the snow settling in a good way? Is there wind transport now or in the recent past? Whumping? Shooting cracks?

I would be surprised if there are many people caught in slides where all of the obvious warning signs were completely missing.

[Rip'nStick]

Great post Summit.

The two things I agree with most are that the digging a pit is simply to confirm what you already derived from other sources, including forecasts, experience and personal observations; and that it is but one observation of many, useful for revealing surprises that would lead you to reconsider why you are there in the first place. A few hasty pits seem better at this than one Reushblock test, though that is useful for overall knowledge and is good when you are in unfamiliar, isolated terrain like some hut trips.

I mostly ski PNW maritime or Sierra snowpack so I have it relatively easy compared to those in the middle but the one thing I always do is make sure I am feeling good about my choices before I get myself in the terrain. I'll go out in sketchier conditions just to fart around and watch the snow fall, but that would be in safe terrain. Unless you feel good about things, why put yourself in a place that is questionable? Even for experts, reading conditions is a crap shoot.

Sounds like the conditions at 2nd Creek were questionable. And by watching the video buddy got way to up n it for my tastes. Jeebus.

I need to get a saw.

[CookieMonster]

Absolutely excellent post Summit! It's really important to point out the difference between unfounded assumptions and reasonable conclusions.

I am not sure how many recreational backcountry skiers have formal training in the scientific method, but scientific data must be acquired and handled carefully.

When data are thin, one must consider what Occam's Razor can teach us. ( The principle of parsimony. ) If you see signs of instability, even if they're somewhat unclear, the most reasonable conclusion is that the snowpack is unstable.

It is not scientifically sound to use snowpack data to construct complex reasoning about why the snow might be stable instead. This is called wishful thinking, or the search for supportive evidence, or stupidity.

The accident that inspired this thread is a good example of how poor terrain selection can be a direct result of failing to accurately interpret snowpack tests as unstable ( multiple Q2 results from compression tests = unstable ).

The Avalanche Handbook suggests extrapolating to the "worst case scenario". In this accident, the original, unstable test results meant that there was a significant chance of unstable snow nearby. In the presence of unstable snow, triggering is most likely around stress concentrators.

Instead, as Summit neatly outlines, the OP's assumptions were completely unfounded. Many people in this thread have exactly the right idea about interpreting the results of snowpack tests. Instead of thinking "how big can I go?", the question is "how conservative do I go?".

[DasBlunt]

Thanks Cookie Monster. I read up on parsimony, and it can also be the "mirror of prejudice" in highly favorable but highly dangerous conditions.

[summit]

This is where I'm at with pits. If you read Avalanche center forecasts religiously, you already know what you're going to find, at least as a snowpack-wide generalized snapshot. I don't need to dig a pit to know that I'll find a hoar'n'crust sandwich 6" off the ground and some surface hoar layers in the upper snowpack right now in my local area.

You've been at this a lot so I know you've experienced the following at some point: The avalanche forecast says it is one danger level, but at some point during the day you said "wow... I feel like it should be higher danger!"

Was that forecast wrong????? Not as such.

An avalanche center is limited to the data it gets from observers. They have a lot more data than you can collect, but they are forecasting for an AREA, not a SLOPE. You must decide whether to ski a particular slope in the area.

Spatial variability is not limited to occurring within a slope. It happens within a mountain, a drainage, a mountain range, etc! We all know there are microclimates. Pits are a great way to spot check personal observations and forecast products to see if the the snowpack matches assumptions.

I've dug pits where the results changed my plan because the findings weren't what was expected based on forecasts and previous observations. I've found layers that weren't in the forecast. If I had not dug, I wouldn't have known.

[FiveLarge]

^^^
While I don't always dig a pit for reasons mentioned, I agree with Summit's thoughts that it's often a good idea to dig to see if there is something "missing" from all other data sources.

I'm curious as to what all of you personally define as a "hasty/quick" pit. I don't want the textbook definition - describe with some detail what the "hasty/quick" pit looks like and what the process of evaluating that "pit" looks like for you. I agree a higher sample set is better than just one really thorough pit, but I wonder how good/thorough the data is that comes from these quick pits. That's why I wonder what they look like for others. I'm not looking to critique, but rather learn.

And I pole probe, too, but what does that really tell us?

Thanks for starting the discussion, A.

[Hugh Conway]

after all the proselytizing for being avy educated that i've read on this site, i'm surprised to read that folks are just eyeballing their lines and going with their personal experience.

digging enough pits isn't particularly practical in touring some terrain; other observations of the snowpack can be.

as said above pits can be useful in regions where there aren't useful, or any, avalanche forecasts or reliable snow observations available. Outside of classes/fucking around the useful pits I've dug are where I didn't know fuck all about the snowpack.

as a practical matter the end result of a conservative approach is you do fuck all in the mountains and watch other, luckier, people have fun.

[goldenboy]

You've been at this a lot so I know you've experienced the following at some point: The avalanche forecast says it is one danger level, but at some point during the day you said "wow... I feel like it should be higher danger!"

Was that forecast wrong????? Not as such.

An avalanche center is limited to the data it gets from observers. They have a lot more data than you can collect, but they are forecasting for an AREA, not a SLOPE. You must decide whether to ski a particular slope in the area.

Spatial variability is not limited to occurring within a slope. It happens within a mountain, a drainage, a mountain range, etc! We all know there are microclimates. Pits are a great way to spot check personal observations and forecast products to see if the the snowpack matches assumptions.

I've dug pits where the results changed my plan because the findings weren't what was expected based on forecasts and previous observations. I've found layers that weren't in the forecast. If I had not dug, I wouldn't have known.

True Summit. I am lucky enough to get a lot of info here. Not only do I have the CAIC report to read, I also have our local CBAC forecast to read, and unlike most centers, the CBAC puts all their obs out there for everyone to see. So I might be able to read a whole slew of people remarking on how stable things are in a given area, or maybe there's one random report of a small slide, and that's something to take note of, too. I'm just making the point that I have a lot of info to work with before I even leave the door.

So now I'm out there and I'm not liking what I'm seeing. Maybe the wind moved a bunch of snow since the report came out. Maybe I'm in an area that got a bunch more snow than the rest of the forecast zone. Whatever the case may be. If I'm not feeling good about it, regardless of the forecast, I'm more likely to just call it than anything else, including digging a pit.

Let's say you dig a pit (one pit!) and it actually looks pretty good, and the forecast was good, even though what you're seeing looks bad. So now you're skiing it? That seems like the scenario that smitchell was complaining about that got this thread going in the first place.

I guess bottom line is that I start with as much info as I possibly can, then try to add to it w/ visual obs, hand pits, pole probing, etc. If things seem worse than they originally appeared, I'm probably out. If things seem even better, then great!, but I probably already put myself in a conservative line choice if that's what I was expecting when I left the house.

I guess that's the short version of my take on it. (And as has been mentioned, that all flies out the window in a new area or one that has no avy forecast.)

[Joey Joe Joe Junior Shabadoo]

I don't know if this was mentioned in the other thread where this started but there was a paper at ISSW this year that basically echoed this discussion and talked about the reliability and usefulness of pits and at what point you should even bother to dig a pit. IIRC it basically said that pits should mainly be used as a last double-check only after you've taken into account the avi report/forecast and all the first hand observation you make when you get out there. Furthermore less experienced bc travelers rely too heavily on pits in all the ways you guys have pointed out and that it can turn into a case of just enough knowledge to be dangerous.

[summit]

I guess bottom line is that I start with as much info as I possibly can, then try to add to it w/ visual obs, hand pits, pole probing, etc. If things seem worse than they originally appeared, I'm probably out. If things seem even better, then great!, but I probably already put myself in a conservative line choice if that's what I was expecting when I left the house.

Sounds like a good approach!

Let's say you dig a pit (one pit!) and it actually looks pretty good, and the forecast was good, even though what you're seeing looks bad. So now you're skiing it? That seems like the scenario that smitchell was complaining about that got this thread going in the first place.

Very true, but that is NOT what I was implying. If I already thought things looked bad, I don't *need* the pit to confirm it. As I and others mentioned:

If you see signs of instability, even if they're somewhat unclear, the most reasonable conclusion is that the snowpack is unstable.

It is not scientifically sound to use snowpack data to construct complex reasoning about why the snow might be stable instead. This is called wishful thinking, or the search for supportive evidence, or stupidity.
...
the right idea about interpreting the results of snowpack tests: Instead of thinking "how big can I go?", the question is "how conservative do I go?".

CookieMonster sums it up well! Continuing in response to Goldenboy:

I might still dig the pit to see if I can figure out WHY things are looking bad so I know what to look for in the future. If my pit shows good, then maybe I hit a random good pocket, dug in the wrong place, did my tests wrong...

But the problem is the clues that your slope is "different" are not obvious on the surface. Maybe this valley has a thin mid-storm facet layer or a surface hoar layer that wasn't prevalent in other valleys, or even other slopes in that valley. Sometimes the forecast may even warn "in some areas layer X may be present in a few areas."

Bottom line: Some things you might not find without digging... or having a pucker-factor-moment... or worse. Digging gives early warning.

[CookieMonster]

If I already thought things looked bad, I don't *need* the pit to confirm it.

Just for clarification: I wasn't writing in reference to your post, nor the concepts and methods outlined therein ... which were excellent ( as I noted ).

I was writing in reference to the general theme that I frequently see here and elsewhere ( including the original post about the 2nd Creek slide ) where recreational snowpack obs are often "bolstered" with a lot of highly unscientific noise ( speculation, justification, etc ) in an attempt to make the shit sandwich look or taste better.

Thanks Cookie Monster. I read up on parsimony, and it can also be the "mirror of prejudice" in highly favorable but highly dangerous conditions.

You're welcome DasBlunt. It's good to point out that parsimony can be abused, but the general principle is useful, especially since we're using it as a technique to select between different conclusions: stable, unknown, unstable.

as a practical matter the end result of a conservative approach is you do fuck all in the mountains and watch other, luckier, people have fun.

That's true, but conservatism is considered best practise when uncertainty is high. As usual, people should not feel limited by best practises, and if this thread were the appropriate place, I would write about the dangers of excessive conservatism.

I'm curious as to what all of you personally define as a "hasty/quick" pit.

My Procedure: Excavate some of the surface snow with gloved hands, maybe a foot deep. Maybe 2 feet deep. Use my hand to form a wall and then gently brush off the loose snow. Look at the profile. Maybe do hand shears or quick hardness tests.

I use hasty pits to 1. ) identify storm snow instabilities, 2. ) check the state of the snow crystals and pore space, 3. ) check for uniform increases in hardness with depth.

And I pole probe, too, but what does that really tell us?

It can tell you something if there are thick layers and significant changes in resistance. This would indicate changes in hardness that may be important. More than anything, it's a useful tool for getting an "index of curiosity" ... as in ... "am I curious about this location?". Over a wider area, you can get some idea about consistency.

According to research by Jamieson, pole probing is a poor method of identifying thin weak layers.

[FrontRangeDrummer]

I've dug pits where the results changed my plan because the findings weren't what was expected based on forecasts and previous observations. I've found layers that weren't in the forecast. If I had not dug, I wouldn't have known.

I had an experience like this on the same day as the Dry Gultch death. CAIC report looked good, my partner's experience on similar terrain the day before was good, it was a nice day, the run looked really fun, it was sunny, and morale and confidence was high. Then two pits revealed scary things we didn't expect, and we backed off.

After the first pit, I actually caught myself thinking "Lets dig another pit and see if it looks any better." This sort of logic leads to poor decisions.

Pits can never guarantee stability, but they can bring instability to your attention, filling in missing information about the snowpack.

[PNWbrit]

Good thread.