Unpacking uncertainty in regional avalanche forecasting: A quantitative case study of uncertainty in forecasting regional avalanche danger

Abstract

When assessing a risk, a future event, uncertainty is vital as we cannot know what will happen. Having an active relationship to uncertainty can reduce it and help the decision-makers make informed decisions. Avalanche forecasting is widely used by outdoor recreational and local authorities for emergency preparedness use. Therefore, an additional focus on uncertainty in this field benefits people making decisions to manage avalanche danger. This thesis aims to answer the question: What are ways to reduce uncertainty in regional avalanche forecasting? Through a survey that collected data on how the forecasters assessed the uncertainty they met in the forecasting process, this thesis has identified situations with high uncertainty, the source of these uncertainties and how forecasters face these situations and sources, as well as how they can manage it in the future. The thesis concludes that high uncertainty does not follow the forecasted danger level. However, it follows certain avalanche problems. The uncertain avalanche problems are connected to different sources of uncertainty, and they can be categorised as reducible or non-reducible. Uncertainty from weather forecasts cannot be easily reduced, but sources of uncertainty like snowpack uncertainty or few field observations can be reduced. Some uncertainties can be reduced by collecting more or better data. Other ways to deal with uncertainty is to make it explicit in the bulletins or try to understand it better with discussing with fellow experts.