Early in December I was up on the slopes at work digging snowpits in different locations. I was trying to identify the weak layers in the snowpack that were likely to become reactive with additional snow load. As I was moving from one pit to another, I realized that the structure of the snowpack changed drastically from one place to another. Weak layers that were in one pit did not exist in a different pit only 30 to 45 feet away and vice versa. In other words, the snowpack was spatially variable.
Spatial variability can be very obvious. For instance, the snow in a forested area is very different than in the openings between the tree-covered areas. It can also exist on what seems to be very uniform slopes. In 2004, Cam Campbell and Bruce Jamieson presented a study that looked at Rutschblock test results in avalanche start zones. One of the slopes they tested was a uniform slope, all their tests failed on the same weak layer; however, test results varied from “almost certain to trigger an avalanche” (score 2) to “almost impossible to trigger an avalanche” (score 6) within less than 10 feet.
Depending on its scale and the avalanche problem we deal with, spatial variability can be a friend or an enemy. When our main concerns are large avalanches, spatial variability can limit the size of an avalanche to a relatively small release area. However, when we ski a slope where small avalanches can be deadly, relying on snowpack information from one area to apply to the entire slope can be a bad call.
Regardless if is a friend or enemy, spatial variability is a big deal for avalanche folks. It means uncertainty and we like be as certain as possible when we look for our own and for other people’s safety.
One obvious question that comes to mind is: How do avalanche folks minimize the uncertainty that comes with spatial variability?
The simple answer is: there is no simple easy way of going about it.
Serious avalanche folks work hard to minimize uncertainty. They head to the field and take a hard look at the snowpack. They map the weak layers’ distribution on their start zones and warn the public with meaningful data driven forecast discussion. Ski patrols, and other operations that use explosives, also try to minimize the scale of the spatial variability to where avalanches are very small or areas of instability are too small to produce avalanches. Clearly, the strategies that ski patrollers apply inside a ski area are not applicable for the casual person on a backcountry tour. The hard truth is that there is no silver bullet for managing spatial variability. But there are a few things that can be done to minimize uncertainty in the back country.
• To start, I don’t trust tracks on a slope. Although, it is not that common for the fourth, fifth or sixth person on a slope to trigger an avalanche. It is not very rare either. Tracks on slope are not a guaranty for stability.
• I always pay attention to obvious signs of instability like recent avalanches, shooting cracks and whumpfs (weakness buried in the snowpack which has a lot of air space between grains).
• I dig a lot of quick pits instead of spending my time in one big pit. I fill the information gap between my pits by tracking the layers I saw in my pit by probing with my ski poles.
• I identify the layers of concern, map their likely distribution in my head and tune my goals accordingly. For example, rain, sun and rime crust are usually elevation and aspect dependent. If I know what type of crust it is and the weather when they formed, I can guest where they are likely to be. On the other hand, surface hoar distribution depends on the sky cover, winds, temperatures and relative humidity. Hence, when I know that there is a surface hoar layer buried in the snowpack I get worried, because guessing where it is far more complicated process.
Clearly, none of the techniques I apply on a backcountry tour completely eliminate uncertainty. Some days are just too hard to manage. On these days I chose to ski mellower terrain and avoid avalanche hazard altogether.
Recently, I found out that the email address at the bottom of the column is incorrect. My apologies to the readers that tried to email me. My correct email address is: email@example.com.
• Ron Simenhois is an avalanche forecaster who lives in North Douglas.