Statistical analysis of snow-pack stability in the northern Sierra Nevada, California

Abstract

Quantitative relationships were required to help understand avalanche phenomena and create an avalanche forecasting model at Alpine Meadows Ski Resort, California. Currently avalanche forecasting relies entirely on the avalanche forecaster's previous experience. Time series relationships between meteorologic conditions and avalanche activity were determined using correlation coefficients for controlled avalanche paths. The depth of snowfall, wind speed and snow settlement were determined to be the most important factors in determining avalanche activity. Temperature and moisture content data demonstrated less consistent correlation coefficients. Forecasting models were developed using least squares multiple regression techniques to forecast avalanche activity based on the U.S. Forest Service avalanche size classification, the crown fracture height and the percentage of the avalanche path that slid. The regression models were able to correctly predict the avalanche activity on 50 to 88 percent of the potential avalanche days based on the above criteria. To improve the model, methods for accurately measuring the bedsurface shear strength and estimating snowfall accumulation in the various avalanche snow loading zones is necessary.