| Introduction | Methodology | Results | Discussion | References |
Contents
The Multi-Criteria Evaluation (MCE) Results
The Least-Cost Path Results
The Multi-Criteria Evaluation (MCE) Results
The Least-Cost Path Results
 |
Image Courtesy of Lee Lau
The Multi-Criteria Evaluation (MCE) Results
A Terrain-Specific Avalanche Forecast Model
A Terrain-Specific Avalanche Forecast Model
 |
| Fig. 32 |
Click to enlarge
This map (Fig. 32) is the result of combining the effects on avalanche hazard of five different criteria: slope, aspect as influenced by insolation, aspect as influenced by windloading, general terrain variation (ridgelines), and terrain traps. The hazard levels represented on my map are the sum total of each of these criteria for each map cell. All criteria have a direct effect on avalanche hazard except for terrain traps. While terrain traps do not directly affect avalanche danger, it is nonetheless sensible to avoid them in case an avalanche occurs. That is, it is more hazardous to travel in a terrain trap than not to.
Recall that the purpose of this analysis is to improve upon the standard hazard rating model by producing a terrain-specific, image-based avalanche forecast model. I assert that Fig. 32 represents just such a model. It is far less vague than the current hazard rating system discussed in the introduction, because it reflects the fact that avalanche hazard varies with terrain. It even improves upon Bruce Tremper's practice of providing additional information in his daily forecast. While such information is incontestably useful, it has limited utility for many backcountry users. This is because many users do not possess sufficient knowledge and skill to evaluate the terrain around them and compare it to the information they received in the forecast - or at least, they are unable to do so while their attention is monopolized by whatever activity it is they are doing. This is why Tremper expressed concern that a 'perfect' forecasting format would be obvious enough for neophytes to understand. However, while some inexperienced backcountry users may lack the requisite ability to identify that they are travelling on a 35º slope on a North-facing aspect, they might well be able to identify the slope they are travelling on as 'that dangerous-looking dark-red area on the map at the trailhead.'
Additionally, Tremper stipulated that the same forecast format must provide sufficient detail for 'hard-core' skiers to make route-finding decisions. I believe that this map facilitates this endeavor better, even, than the standard topo (elevation contour) map. First of all, Digital Elevation Models (DEMs) are a much more accurate representation of terrain than contour maps. More information enables better decision-making, as long as it is not an overwhelming amount of information. Furthermore, even 'hard-core' skiers would benefit from an image-based hazard model such as this one. Because it is terrain-specific, they can get a decent estimate of the stability of their destination before they even leave the parking lot.
The Least-Cost Path Results
The Least Hazardous Path Along the Spearhead Traverse
Click to enlarge
The Least Hazardous Path Along the Spearhead Traverse
 |  | |
| Fig. 33 | Fig. 34 |
The above maps represent the least hazardous route through avalanche terrain based upon the combination of my selected criteria. The result of my MCE - the terrain-specific avalanche forecast model (Fig. 32) - was used as a friction surface to compute a least-cost path along the Spearhead Route. The route avoids areas of high hazard rating where possible, and takes a direct route through them when it is not possible (Fig. 33). This does not always mean sticking to high elevation ridgelines (Fig. 34); where the route must deviate from those areas, it selects south-facing aspects for descents to areas of low hazard. Recall also that the least-cost path accounts for anisotropy.
The deviations from the Spearhead Route are of little analytical consequence. I drew the Route on the map myself, and it is not signposted. The representation of it on the map serves only to provide context for the terrain. However, it is worth noting that when I drew it, I attempted to draw it in such a way that the Route facilitated easy travel; though it is likely not the 'easiest' route to follow in reality, it represents my best effort to represent low friction with respect to the anisotropic surface only, and not hazard. That is, it is designed to minimize unnecessary elevation loss or gain. In other words, it represents a route that a backcountry traveler might plan were they not concerned about avalanche danger at all. If you look at Fig. 34, there doesn't appear to be any qualitative difference between the two lines at all; they look like two equally valid paths for travelling through the terrain. There is not even much variation between them. However, if you look at Fig. 33, it is immediately obvious that the blue line encounters significantly less hazard, even if you know nothing whatsoever about avalanche prediction. This illustrates the efficacy of the model to communicate terrain-specific hazard variability.
Such a tool could be useful for backcountry trip planning, regardless of one's ability or experience.