Item: ENTROPY AND THE ART OF AVALANCHE FORECASTING: AN EXAMINATION OF FACTOR WEIGHTS AND CLASSES
-
-
Title: ENTROPY AND THE ART OF AVALANCHE FORECASTING: AN EXAMINATION OF FACTOR WEIGHTS AND CLASSES
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Dave Gauthier [ Wilderness Geological, Yarker, ON, Canada ]
- Cam Ross [ BGC Engineering Inc., Ottawa, ON, Canada ]
- Caio Stringari [ BGC Engineering Inc., Vancouver, BC, Canada ]
Date: 2026-09-28
Abstract: Every avalanche forecast is based on many factors, such as recent avalanche observations, instability test results, or weather conditions. The art of avalanche forecasting is partly about deciding how much weight to put on each bit of information provided by the factors. As Ed Lachapelle pointed out in 1980, in practice the avalanche forecaster always has an imperfect state of knowledge of avalanche conditions, and by necessity their objective is to each day improve their forecast by iteratively incorporating new information to reduce uncertainty in the expected conditions. This paper is less about how the forecaster would do this, as that has been well covered by others. Rather the objective is to re-examine the groupings of Class I, II, and III forecasting factors as they have been traditionally defined and to estimate quantitative weights for several of the most commonly used, to help the forecaster more systematically and rigorously update forecasts based on their observations of avalanche, snowpack, and weather factors. This paper provides quantitative estimates of weights of the most common forecasting factors, and an evaluation of class membership for each. It proposes a simple framework to allow the forecaster to consider the weighting of information provided by the common factors and to estimate the weights of other factors relevant to a particular operation or location, and to generally consider information entropy in defining new factors and in improving the efficiency of avalanche forecasting workflows. Applications to different scales of decision problems, for example in backcountry travel, are also explored.
Object ID: ISSW2026_P2.43.pdf
DOI: https://doi.org/10.15788/1790099007
Language of Article: English
Presenter(s): Dave Gauthier
Keywords: Forecasting, entropy, factors
Page Number(s): 1546 - 1550
-