The Avalanche Review, VOL. 6, NO. 4, MARCH 15, 1988
Copyright © All Rights Reserved; AAA

Hazard Ratings May be Hazardous
by Jill Fredston

The interaction of three main variables; the snowpack, weather, and terrain determine whether an avalanche hazard exists, a fourth, human, variable must be added. Without people or property, there is no hazard.

As leader of the now defunct Alaska Avalanche Forecast Center, I was frustrated by the four basic avalanche hazard definitions used by the majority of the U.S. forecast centers in public backcountry forecasts. I felt strongly that we could produce accurate snow stability forecasts but could not predict what level of hazard backcountry travelers would create for themselves. We all know that it is very possible to go out during a period of low instability and create a high hazard for ourselves through poor route-finding. Similarly, it is possible to travel safely during periods of high instability by careful choice of slope angles and routes.

The basic rationale for the use of hazard categories in backountry forecast is that they allow users with diverse avalanche experience to readily perceive the level of danger, identify trends (i.e. whether the hazard is expected to stay the same, increase, or decrease), and compare conditions over time and between areas. Also, they help to standardize the formats of the various forecast center products and thus are felt to minimize the public's confusion. Generally, the public forecasts state the overall hazard category for each forecast area and then are backed up by site specific information.

Over a period of several years, I became convinced that the hazard levels were potentially hazardous in themselves. First, interviews revealed that users tended to focus on the stated hazard levels and ignore or filter the rest of the message. Secondly, the use of hazard levels seems to inadvertently encourage people to accept the forecast as an end product--an assumption that has repeatedly gotten victims into trouble is: "the recording said the hazard was low-moderate, so we didn't expect to get caught".

Ideally, users should be encouraged to use the forecast as one of many critical pieces of information in THEIR OWN ongoing decision-making process and the forecast should provide the basic wherewithal so that even a user with limited avalanche awareness can do this. The importance of this is underscored by the fact that backcountry avalanche forecasts are regional in nature but the snowpack is highly variable. It is even further underscored by the fact that of the various activity groups, backountry recreationalists lead the nation in avalanche fatalities and the majority of avalanche accidents are human-triggered. In Alaska, over 95 percent of the avalanche victims in the last decade triggered the slides that caught them.

I began to experiment. I moved the statement of hazard levels from the beginning of the message to the end in an attempt to focus the listener's attention on the content of the message. That seemed to help except during those uncomfortable periods when the snowpack was mostly stable but had a few nasty, isolated areas just waiting for human triggers. When the hazard was "moderate" or as some would gave it, "low-moderate", users often perceived any description of hazard as an attempt by the forecast center to, in the words of a user, "cover its tail".

My next step was to eliminate the use of hazard levels altogether and just produce a detailed snow stability assessment/forecast. This worked well for users with some avalanche knowledge but was more difficult, at least initially, for the inexperienced because they had no ready measure by which to compare conditions yesterday with those of today--or the snowpack in one location with that in another. I then knew that I had to use some kind of rating system.

So I switched from avalanche hazard ratings to snow instability ratings. These ratings were an obvious take-off on the accepted hazard terminology. The definitions are as follows:

LOW INSTABILITY: Mostly stable snow. Natural and human-triggered avalanches are unlikely except in isolated locations on steep, snow-covered slopes and gullies.

MODERATE INSTABILITY: Areas of unstable snow. Natural and human-triggered avalanches are possible on steep snow-covered slopes and gullies.

HIGH INSTABILITY: Mostly unstable snow. Natural and human-triggered avalanches are likely on steep snow-covered slopes and gullies.

EXTREME INSTABILITY: Widespread unstable snow. Avalanches are certain on some steep, snowcovered slopes and gullies. Large destructive slides are possible.

These ratings worked well! Users of all experience levels seemed to be hanging up the phone with a clear understanding of the dominant factors influencing snow instability in each forecast area and were alert to clues indicating such instability. They had a good feel for the likelihood of natural or human-triggered releases during the forecast period. They were also aware of expected trends in snow instability and the reasons for these changes.

We as forecasters were no longer putting a misleading label on the product we could offer. Avalanche hazard is relative to where the traveler is standing and under what conditions. Snow instability is relative to the strength of the snow and the forces applied to it. We were forecasting snow instability based on available snowpack, weather, and terrain data but were no longer presuming to forecast the highly unpredictable variable of human decisionmaking.

You may argue that all this is a question of semantics. Certainly, the most important part of any forecast is the content. a forecast must be as specific as possible, accurate, and concise yet educational. Semantics ARE important though-the misleading hazard language equates to false advertising and thus inadvertently undermines the credibility of a forecast.

Hazard ratings may work for roads, railroads, and ski areas where public access can be regulated and avalanche control is an option. In the backcountry, however, the decision-making so critical to a good hazard evaluation is ultimately the responsibility of the traveler. Whether or not an avalanche accident occurs depends, to a large extent, on whether the backcountry traveler allows it to happen.

In the backcountry avalanche forecasts, the variable being forecasted is snow instability and therefore the terminology used should reflect this. The use of snow instability ratings rather than hazard levels encourages listeners to use the forecast in the way in which it is intended--as a tool. While in the backcountry, travelers must then continually seek clues and key information so that they can evaluate the hazard, weigh their options, and choose the safest alternative.

The Avalanche Review, VOL. 6, NO. 4, MARCH 15, 1988
Copyright © All Rights Reserved; AAA