Result & Discussion
Based on the Multi-Criteria Evaluation and the resulting suitability map, four levels of avalanche suitability were classified.
0 -No Avalanche Release Danger (0)
1- Low Avalanche Release Danger (0 to 50)
2- Moderate Avalanche Release Danger (50 to 80)
3- High Avalanche Release Danger (80 to 100)
The results indicate that the most potentially hazardous areas in the North Shore Mountains for Avalanche release are East facing ridgelines and convex hillsides that are above 1200m and are open non-forested areas. Moderate areas include slopes from 30 to 60 degrees that are non-forested at most elevations above 700m and facing most aspects. Low release potential areas are most areas that are forested, at lower to mid elevations including most terrain shapes. This map is a static representation based solely on terrain parameters.
Problems associated with this analysis relate mainly to the nature of avalanche release and its strong dependency on climatic factors such as wind and snow supply. The fact that these factors change temporally and spatially leads to extreme difficulty in any Avalanche Model that attempts to characterize most climatic factors. Most models address this problem by performing statistical analysis on areas where avalanche release has occurred and the parameters have been well documented. The problem with this solution is the lack of transferability of parameter significance between study areas. The aquisition of climatic data on a fine enough resolution to be spatially significant is also a major obstacle preventing the inclusion of these parameters. Terrain based avalanche release models provide a useful guide when constructing an Avalanche hazard map but the nature of Avalanche release has an uncertainty that will usually require more detailed snow conditions and local knowledge to accent the information presented by a model of this nature.
There is also a large potential uncertainty with the information presented due to the possibility of human error both in the data collection stage (DEM production) and during the classification and digitizing process. The situation where the LANDSAT image was classifying forested shadow scores lower than water scores is a good example. The values produced by image transformations are to a certain degree somewhat arbitrary depending on how they are interpreted. The semantics involved in evaluating what constitutes a hazardous slope may be different for me than someone else who would build a similar model. The resulting classification decision produces uncertainty that perpetuates itself at each successive stage as the analysis continues. The person performing the analysis has individual ideas at each stage of the analysis that will in the end affect the appearance and interpretation of the final result. In the case of this analysis, all the suitability scores determined for each factor in the spatial analysis are potential sources of uncertainty that contribute to the final product.