April 16, 2024

Making large-scale snowpack simulations more useful and relevant for avalanche practitioners

WSL Institute for Snow and Avalanche Research SLF

In this presentation I will give an overview of the research we have conducted at Simon Fraser University (Vancouver, Canada) to make the output of large-scale snowpack simulations more accessible and relevant for operational avalanche forecasting. We first designed a suite of tools to automate the processing of large and complex data sets of simulated snow profiles in an informative way. We then applied these tools and concepts to compare numerical predictions from the Canadian snowpack model chain against operational hazard assessments from forecasters to better understand the capabilities of simulations for regional avalanche forecasting. In addition to revealing encouraging agreements in seasonal summary statistics, the comparison of daily assessments also highlighted the need for operational products that are designed for detailed examination. Such products should empower forecasters to evaluate and challenge the numerical output, thereby enhancing their understanding and trust in the simulated data. Based on these findings, I will present our vision for the further development of snow model post-processing tools that will allow forecasters to benefit more from snowpack simulations and actively learn about their strengths and weaknesses. While the presented tools were designed with a focus on gridded weather and snowpack model chains like those employed in Canada, Austria, and (soon) Norway, they may also support other relevant tasks like automating the training of machine-learning models.

Bio

Florian Herla is an atmospheric scientist and mountain guide. He recently graduated with a PhD from Simon Fraser University in Canada focusing on enhancing the operational value of snowpack simulations and providing insight into how snowpack simulations can address some of the operational challenges of avalanche forecasting. Since bridging the gap between tangible application and academic research is particularly motivating to him, he has now moved his attention to creating collaborations that allow avalanche warning agencies and snowpack model researchers to benefit from each other's efforts more easily. This currently ties Florian to the Norwegian and Tyrolean avalanche warning services where he contributes to developing an operational weather and snowpack model chain, before going back to Canada for a PostDoc in ensemble snowpack modeling.

Zoom-Link: https://wsl.zoom.us/j/65789508973?pwd=Y0UzZm14NUluN3pnbnNWT2JLZVFMUT09

Datum und Zeit:

16. April 2024, 11:00 - 12:00

Ort:

Hörsaal SLF, Davos / Zoom-Webinar

Vortrag:

Florian Herla

Moderation:

Pascal Hägeli

Sprache(n):

English

Veranstaltungstyp:

Vorträge und Kolloquien

Zielpublikum:

Everybody interested in this topic