Transferring avalanches between paths

Abstract

Estimates of exceedance probabilities of runout lengths of avalanches for a specific path can rarely be based on measured avalanches in that slope alone, if they are to become statistically reliable. Thus one has, directly or indirectly, to include information of known runout lengths in other paths, i.e. to transfer runout lengths between paths. An attempt is made to classify such transfer methods, including both topographical methods that only make use of information on the shape of the path, such as methods based on runout ratios and a/ß-models, as well as physical methods which also make use of physical models, simulating the avalanche as it runs down the path. By introducing a specific standard slope all avalanches in a given dataset can be transferred to that slope. The length of the transferred avalanche in the standard slope then becomes a slope-independent measure of its length. Using an Icelandic dataset of 196 avalanches we demonstrate how estimates of exceedance probabilities of run out lengths may vary with the choice of transfer method and how the order of the slope-independent lengths of the avalanches in the dataset will vary. The implication for avalanche risk assessment is briefly discussed.