Item: Computer Assistance in Avalanche Forecasting
Title: Computer Assistance in Avalanche Forecasting
Proceedings: Proceedings of the 1994 International Snow Science Workshop, Snowbird, Utah, USA
- David M. McClung
Abstract: In this paper, I discuss two computerized modules for avalanche forecasting, including results of field testing and experiences with the products. Avalanche forecasting consists of prediction of current and Mure snow stability. Current research at the University of British Columbia is aimed at changing forecasting from an intuitive art to a science. The approach is by developing numerical and expert system modules by using all the data available to a forecaster (numerical and non-numerical). The strategy can presently be defined in three parts: I. An explanation of data classes available and their relation to the scale, character and formulation of the forecasting problem. For a detailed description of data classes in avalanche forecasting based on relevance, ease of interpretation and character (numerical or non-numerical) the reader can refer to Chapters 6 and 7 of McClung and Schaerer (1993). Therefore, I will not discuss this part further in this paper.. II. A description of the numerical portion of the system. This part uses discriminant analysis (both parametric and non-parametric) and Bayesian. statistics to give estimates of the probability of avalanching based on calculations in 6 or 7 dimensional (depending on whether dry or wet avalanches are expected) discriminant space. III. A description of the non-numerical portion of the system. This part consists of an expert system to interpret data from snow profiles with respect to snow stability. The expert system is entirely rule-based in contrast to the numerical algorithm in part II. The data analyzed are nearly independent of the data in part II. The data consist of the information contained in a standard snow profile including comments. See McClung and Schaerer (1993) for more a description of snow profiles and the information used.
Language of Article: English
Keywords: forecasting, snow profiles, snow stability
Digital Abstract Not Available