Avalanche risk in Iceland

In this thesis I present a probabilistic approach to modelling avalanche risk for settlements in Iceland. In particular, two simulation models are developed. These are used to calculate the probability of avalanches travelling a certain distance, and of the flow being a specific width. These two simulation models, in combination with knowledge of the average frequency of avalanche occurrence, permit the calculation of the probability of encountering an avalanche at any point in the terrain (the encounter probability). This may be further combined with knowledge of the proportion of time the location is occupied (the exposure), and the proportion of damage that the avalanche causes (vulnerability) to derive a value for risk. Following their development, the simulation models are validated against the records of avalanching in Iceland and are found to be able to represent the conditions upon many of the avalanche paths satisfactorily. However, it would appear that the models are best applied to those paths in the West Fjords which have a relatively high frequency of avalanching. Paths where the nature of avalanching cannot be adequately represented by the simulation models are identified. A sensitivity analysis shows that both models are fairly robust, being most sensitive to the avalanche sizes for which the data are of the highest quality. Consequently, it would appear that the error introduced from attempting to represent those sizes for which little or no data are available, is minimal. I conclude by providing some example risk simulations for paths in the West Fjords. It is to be hoped that the model outlined in this thesis provides a useful planning tool in Iceland.Arts, Faculty ofGeography, Department ofGraduat