Snowpack factors associated with strength changes of buried surface hoar layers

Abstract

Every winter in North America, failures in layers of buried surface hoar (frost) release many slab avalanches, some of which kill recreationists. Some surface hoar layers stabilize within a week of burial and others require a month or more. Little is known about whether snowpack factors such as crystal size, snowpack depth, slab thickness, load, temperature and temperature gradient are associated with strength changes of these layers. We tested buried surface hoar layers once or twice per week at study sites in the Columbia Mountains from 1994 to 1998, and measured over 300 changes in shear strength. We assess the factors associated with the rate of strength change using rank correlations. The factors are ranked to identify which are most relevant for forecasting changes in strength. Useful predictors include the total snowpack depth, the maximum grain size, and slab depth. The correlations between these predictors and the measured rate of strength change are discussed in terms of physical processes. We illustrate the predictive potential of combined factors by comparing measured values of the rate of strength change with fitted values from a regression tree.