Machine Learning Approach for Snow Depth Estimation from Temperature Sensors Demonstrates Effectiveness

Research conducted at Syracuse University has employed a machine learning approach to estimate snow depth from temperature sensors in a remote forested watershed in the Adirondack Mountains, New York. The study found that vertical temperature sensor profiles can accurately infer snow depth with a root mean squared error (RMSE) between 1.8 and 6.5 cm. This method demonstrated effectiveness in an area with a shallow snowpack and frequent midwinter melt events, showing little sensitivity to sensor mounting method, vertical sensor spacing, or time of day.

Key Takeaways:

• The research used temperature sensor profiles to estimate snow depth for monitoring multiple winter seasons in a 1.3 km2 (130 ha) forested watershed.
• The study employed a random forest machine learning model to estimate snow depth from snow temperature profiles and landscape properties.
• The model showed an RMSE between 1.8 and 6.5 cm, which is lower than or comparable to existing methods.
• The research found that the method was effective in areas with a shallow snowpack and frequent midwinter melt events.
• The study did not find significant sensitivity to sensor mounting method, vertical sensor spacing, or time of day.
• Eight profiles of iButton temperature sensors were paired with trail cameras and snow stakes for daily snow depth estimation.
• An additional four temperature profiles with sensors exposed directly to the snow were added in November 2022.

Statistics:

• 1.3 km2 (130 ha) forested watershed in the Adirondack Mountains, New York
• 20 cm vertical spacing between temperature sensors
• 8 temperature profiles paired with trail cameras and snow stakes for daily snow depth estimation
• 4 additional temperature profiles with sensors exposed directly to the snow at 10 cm vertical sensor spacing
• RMSE between 1.8 and 6.5 cm for snow depth estimation
• 2019: Vertical temperature sensor profiles were installed in a grid pattern
• 2021: Daily snow depth estimation started in November
• 2022: Additional 4 temperature profiles with sensors exposed directly to the snow were added

Sources:

• A Machine Learning Approach for Snow Depth Estimation From Temperature Sensors. Hydrological Processes, 2025; 39(9).
• Journal of Engineering, October 20, 2025; p 999 (by NewsRx).
• Wiley-Blackwell - www.wiley.com/; Hydrological Processes - onlinelibrary.wiley.com/journal/10.1002/(ISSN)1099-1085.
• Madision Gunn, Syracuse University, Syracuse, NY 13244, United States.