Item: PATROLLING THE SNOW SURFACE ENERGY BALANCE
-
-
Title: PATROLLING THE SNOW SURFACE ENERGY BALANCE
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Austin Finnell [ Keystone Ski Patrol, Summit County Rescue Group, Dillon, CO, USA ]
Date: 2026-09-28
Abstract: Avalanche professionals usually build snowpack intuition from weather cues they can directly observe, including sun, cloud cover, air temperature, and wind. This study asks whether those cues account for what actually controls snow surface temperature. A minimal one dimensional snow surface energy balance model was developed and forced with observations from the Senator Beck Basin Study Plot in Colorado from 1 January to 28 February 2026. The model was tuned with measured station data to create a baseline simulation. The baseline model reproduced observed snow surface temperature well, with MAE = 1.05 °C, RMSE = 1.53 °C, and r = 0.962. To test the model's response, downwelling longwave, absorbed shortwave, sensible heat, and latent heat were independently changed by ±10%, ±20%, and ±30% in separate runs. At every perturbation level, scaling downwelling longwave produced the largest increase in RMSE. A 10% reduction in downwelling longwave increased RMSE by 1.09 °C, compared with 0.05 °C for sensible heat, 0.04 °C for absorbed shortwave, and 0.01 °C for latent heat. Within this model and record, snow surface temperature RMSE was most responsive to proportional changes in downwelling longwave radiation. A simple air temperature proxy for snow surface temperature performed far worse than the energy balance model, with RMSE = 7.21 °C, about 4.7 times the baseline error. These results suggest that practitioners may underestimate the contribution of longwave radiation to snow surface conditions, not because other variables are unimportant, but because longwave acts continuously without a visible signal and is therefore easy to omit from a mental model. In this two month record at one station, changing LW↓ by a given percentage produced a larger response than changing absorbed shortwave, sensible heat, or latent heat by the same percentage. As snow modeling becomes more common in avalanche forecasting, understanding longwave influence can help users interpret model results and reconcile differences between snow surface temperature, air temperature, and visible weather conditions.
Object ID: ISSW2026_P4.2.pdf
DOI: https://doi.org/10.15788/1790099289
Language of Article: English
Presenter(s): Austin Finnell
Keywords: Snowpack Properties, Instrumentation and Remote Sensing, Weather Forecasting
Page Number(s): 260 - 267
-