Pits risk vs reward

Back in avi class we learned to dig pits and try to interpret them, with a huge caveat being that unless you match the exact conditions of the slope you are about to ski, it doesn't mean much. But, it seems that to best match the exact conditions of the slope, you have to be on that slope, thus spending more time in the place that has the hazard you are trying to evaluate.

So on Saturday, a stable day and no need to dig pits, but as we went along I was contemplating, if I was going to dig pits, where would they be? I saw places that seemed relatively safe and perhaps matching the main slope. For instance, the same aspect and slope angle but laterally away from the main open chute and bounded by tree cover up and down slope. Is that a good match, or do the presence of the trees make it dissimilar enough to the main open slope that it's not useful?

For reference one of the slopes in question was down to Chain Lakes from Herman Saddle, where you can kind of skip from tree cover to tree cover, but the question is general.

You should re-frame the "risk vs. reward" question. HINT: Consider the risk of not having information vs. the reward of having the information.

The rule for trees is that the distance between the test profile and a tree should be equal to or greater than tree height.

Data acquired from test profiles can have a helpful or harmful effect on your beliefs about the presence of instability and its parameters.

Accurately interpreting the data from a test profile requires understanding the scale and time frame to which the data are relevant.

There is an important relationship between the data from a test profile and your level of uncertainty.

Don't seek answers about instability from a test profile. Seek answers about uncertainty.

You should re-frame the "risk vs. reward" question. HINT: Consider the risk of not having information vs. the reward of having the information.

The rule for trees is that the distance between the test profile and a tree should be equal to or greater than tree height.

The tree rule is extremely helpful. I saw several locations that would meet that criteria and which I would consider a safe place to stand around and dig a pit.

Absent tree cover, what's a typical (good) approach to deciding where/whether to dig a pit when starting at the top of a slope? For a type location, looking south from Herman Saddle there is typically a wind-deposited drop of 3 to 10 feet and then a moderately steep slope going down, with steeper slopes coming down from skier's right (west). It's obviously in the lee of the saddle.

Self-modified 'cuz my post wasn't helpful.

My only helpful point would be, never stop making observations, regardless of the result from the pit(s).

My only helpful point would be, never stop making observations, regardless of the result from the pit(s).

This is excellent advice.

For Travertine's follow-up questions.

1. Dig vs. Don't Dig.

What is your uncertainty?

High uncertainty situations are managed by making conservative decisions, seeking out additional information, or both. Test profiles are a fairly good method of acquiring additional information that can help minimise uncertainty about the presence of instability and its parameters.

2. Choosing a location.

Operational Avalanche Forecasting vs. Backcountry Avalanche Forecasting

This whole topic is actually much simpler than it's made out. Spatial variability is a significant problem if you're limited to a few test locations. In operational avalanche forecasting ( EX: NWAC or a highway area ) the number of observations is often very limited, and the size of the forecast area is usually very large.

Spatial variability becomes a very significant problem when a very limited number of observations are used to issue a forecast for a large area. With limited observations, representative locations are extremely important, and tests generally should be carried out in avalanche starting zones, which can be very dangerous business.

But this is not the case for backcountry avalanche forecasting, where multiple observations are available, and where the size of the forecast area is relatively small. The snowpack is a multi-layer composite material and you are a quality control control engineer who is expected to inspect the material for signs of failure. What is the best strategy for quality control? More observations or fewer observations?

The correct solution to this quality control problem, according to research by Jamieson and Birkeland, is to acquire many observations of the snowpack in the drainage or slope of interest. This allows you to determine the structure of the snowpack over a fairly large area. Simple tests, such as hand shears, allow you to determine the strength of the snowpack, and observations of shear quality allow you to determine the amount of energy being stored in the snowpack.

Therefore, you should choose many locations when developing your map of snowpack observations. Are you observing fairly uniform high energy shears? Time to dial back expectations. Or are you noticing a generally well-bonded pack, with dirty shears that only appear after you pound on your shovel? As with any quality control procedures, you are also expected to evaluate the consequences of material failure.

I wrote a ( far too ) long blog post on this subject.

avalanchesafety.blogspot.com/2009/12/sno...ons-maintaining.html