Northern California Avalanche Data Science

–Models for Avalanche Prediction and Forecasting
Published

December 1, 2023

Project Overview

Backcountry skiing and snowboarding are becoming popular ways to recreate on snow outside of a ski resort setting. However, there are additional dangers that come from skiing on ungroomed snow outside of a ski resort. Avalanche forecasters integrate daily snow observations with weather forecasts and snowpack history to issue daily estimates of how likely avalanches are. For this project, I am collaborating with the Mount Shasta Avalanche Center and the US Forest Service to visualize historic avalanche data and develop probabilistic models to help avalanche forecasters better predict avalanches. The project involves webscraping, database design, probabilistic models, and human decision making in the backcountry. All the code will be open-source and you can follow along on the blog here

The key collaborators on this project are Avalanche Forecasters Nick Meyers, Casey Glaubman, Eric Falconer, Sam Clairemont, and Corey Beattie of the Mount Shasta Avalanche Center.

Rolling average avalanche forecast

Three day rolling average model of snow accumulation versus avalanche danger forecast across recent avalanche forecasting seasons.