Item: INTEGRATING METEOROLOGICAL UNCERTAINTY INTO OPERATIONAL SNOWPACK AND AVALANCHE HAZARD MODELLING: INITIAL EXPLORATIONS IN ENSEMBLE FORECASTING
-
-
Title: INTEGRATING METEOROLOGICAL UNCERTAINTY INTO OPERATIONAL SNOWPACK AND AVALANCHE HAZARD MODELLING: INITIAL EXPLORATIONS IN ENSEMBLE FORECASTING
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Jacopo Borsotti [ Simon Fraser University, Burnaby, BC, Canada ]
- Martin Metzger [ Alpsolut S.r.l., Livigno, Italy ]
- Pascal Haegeli [ Simon Fraser University, Burnaby, BC, Canada ] [ WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland ]
- Simon Horton [ Avalanche Canada, Revelstoke, BC, Canada ]
- Martin Perfler [ Simon Fraser University, Burnaby, BC, Canada ]
- Florian Herla [ Simon Fraser University, Burnaby, BC, Canada ]
- Fabiano Monti [ Alpsolut S.r.l., Livigno, Italy ]
Date: 2026-09-28
Abstract: Snowpack modelling has established itself as a valuable additional source of information for avalanche forecasting. However, translating uncertainty in meteorological inputs into tangible outputs for practitioners remains a challenge. In most operational uses, the required input data is taken from one weather forecast model run (e.g., the HRDPS model in Canada), and the simulations are run in a deterministic mode, producing a single snowpack realization that does not reflect the range of possible snowpack conditions arising from uncertainty in meteorological forcing. Ensemble approaches represent a promising way to address this limitation, as they involve running the model multiple times with slightly perturbed meteorological inputs to explore the range of possible snowpack conditions. While ensemble modelling is widely used in weather forecasting, its application to avalanche forecasting remains limited, in part because the simulations produce immense datasets that are difficult to interpret for operational use. In this pilot study, we explore the use of ensemble approaches for operational avalanche forecasting by examining the effects of perturbations in temperature and precipitation on simulated snowpack stratigraphy and stability. We focus on the 2021/22 winter season at three locations across Canada, where we identify situations involving weak layers to investigate the effects of such perturbations. Our results indicate that even small perturbations can substantially influence the formation and evolution of weak layers, leading in some cases to a wide spread of possible effects that persist throughout the season, while in others their impact remains limited, resulting in more consistent outcomes. Overall, this study highlights how accounting for meteorological input uncertainty can enhance both the interpretation and operational use of snowpack simulations, supporting more robust avalanche forecasting practices. Our approach allows us to estimate uncertainty margins in snowpack stability forecasts, providing an explicit quantification of forecast uncertainty. Moreover, ensemble modeling opens the door to best-member selection strategies, whereby multiple simulations with perturbed meteorological forcing are used to identify the scenario that best aligns with field observations.
Object ID: ISSW2026_FS4.5.pdf
DOI: https://doi.org/10.15788/1790098461
Language of Article: English
Presenter(s): Jacopo Borsotti
Keywords: snowpack modelling, meteorological uncertainty, ensemble modelling, snow stratigraphy, skier triggering
Page Number(s): 1634 - 1641
-