Technical information, news, research, and opinion on avalanches, snow safety, and winter backcountry travel.

Wednesday, February 23, 2011

The Scientific Method

It's been a slow winter here in Washington, with a lot of rain, and not a lot of snow. This means that avalanche danger has been low for a long time, and it's easy to get out of the avalanche safety habit, especially if your avalanche safety habits are inconsistent in the first place.

What forms the basis of our avalanche safety habits?

On Monday I skiied Chair Peak with two local characters. Had an incredibly fun day out and about in the snow. At one point during the day, one of my ski partners asked me if I had ever thought about teaching avalanche education courses. Yes, I've thought about it. But, unfortunately, while I am allegedly qualified to teach avalanche courses, I don't have enough free time.

But it got me thinking. What would I even teach?

I would teach the scientific method, because it's what I use, and based on my current understanding of avalanche safety, the scientific method is the only realistic approach. But I wouldn't start by assuming that my students couldn't learn. Avalanche safety is a blend of complex sciences, and yes, there is a lot of complicated material. That said, learning how to approach the challenge of avalanche safety is the single most important thing you can ever learn about avalanche safety.

The scientific method provides a fantastically simple framework for teaching avalanche safety, and the primary observation categories are already determined: terrain, weather, snowpack, and humans. You start out with nothing and gradually collect information. This information helps you form inferences that eventually lead you to a conclusion. These conclusions tell you where and when to travel. Yes, as with all forms of science, it is incredibly important to avoid bias.

It really comes down to learning how to conduct observations that tell you whether or not it's safe to travel at a specific time and place. The where and when of it, if you will. The Avalanche Handbook clearly states that applying the scientific method is a good first step. If I taught avalanche classes, I wouldn't tell my students that avalanche safety is complicated and difficult, I would tell them that avalanche safety involves learning how to apply the scientific method to the problem. In short, the scientific method of avalanche safety involves observations, inferences, and conclusions that help us decide.

Avalanche safety is all about where and when can I travel based on the available data? Much of what we refer to as safety rituals or good habits really refers to the consistency with which we apply the scientific method to the problem. All the huzz about human factors just refers to the avoidance of bias in order to ensure that what we do is actually science.

Scientists have to avoid bias or their work isn't scientific. I know how this feels, because there have been days when I've stared at data on my computer screen, wishing that I could find a way for the data to provide the conclusion that I wanted it to provide.

The problem is that wishes and wants aren't science, and in the backcountry, they'll get you killed.

Further Reading:
Addendum
The following is an excerpt from my contribution to an article in The Avalanche Review. This article discussed the relationship between large avalanche cycles and backcountry avalanche forecasting. You'll see that the discussion is centred around the scientific answer to the question how does an avalanche cycle change instability in the short, medium, and long-term.

The size of the forecast region is the most important factor, with precise answers only available for very small areas. However, even for small areas, the chaotic interaction between terrain and weather makes it difficult to predict the effects of widespread avalanching on future snowpack instability. The following scenario, which is just one possibility out of many, hints at the overall complexity of this forecasting problem.

Instability will persist when a bed surface composed of faceted crystals is immediately reloaded during a storm. On the other hand, future snowpack instability on that slope will be very different if the faceted crystals exposed by avalanching are subjected to multiple melt/freeze cycles prior to the next storm. Melt/freeze activity is often limited by aspect, so it is possible for the faceting process to continue on cold aspects, while faceted crystals on warm aspects undergo rounding as a result of melt/freeze metamorphism. In this highly general scenario, the weather builds new patterns of snowpack instability that are difficult to uncover without careful observations.

Therefore, for most recreational skiers, knowledge of a recent avalanche cycle is a very general and imprecise piece of information. General information often has a dangerous and unwarranted influence on individual or group beliefs about the presence of instability and its parameters. Without abundant information, expert knowledge, and significant experience ( Randy and Nick provide great examples ), a recent avalanche cycle should not exert undue influence on recreational travel choices and decision-making at any operational scale. More than anything, incremental changes to the snowpack caused by synoptic scale weather events will alter the characteristics of the danger but won't eliminate it.

  • The chaotic relationship between terrain and weather is a primary source of uncertainty.
  • Incremental changes to the snowpack are a primary source of uncertainty.
  • Avalanches remove weak snow from some, but not all, slopes.
  • Avalanches may or may not remove all the weak snow from a specific slope.
  • Use multiple sources of information to determine the likelihood of avalanche formation.
  • An avalanche cycle over a large area certainly does not mean a specific slope is safe.
Proactively managing uncertainty is essential to safe decisions.