The Python Energy Balance model for Snow and Ice (PEBSI): Application and tradeoff analysis on Gulkana Glacier, Alaska

• Published in: Journal of Glaciology
• Main Authors: Claire V. Wilson, David R. Rounce, Louis Sass, Albin Wells, Emily Baker, Mark Flanner, S. McKenzie Skiles
• Format: Article
• Language: English
• Published: Cambridge University Press 2026-01-01
• Subjects: energy balance; glacier meteorology; glacier modelling; mountain glaciers; snow/ice surface processes
• Online Access: https://www.cambridge.org/core/product/identifier/S0022143026101543/type/journal_article

Glacier energy-balance models offer mechanistic insights into glacier mass balance under a changing climate, yet their considerable data requirements hinder large-scale applications. Here we present the open-source Python Energy Balance model for Snow and Ice (PEBSI), which includes physically based albedo evolution using the Snow, Ice and Aerosol Radiative (SNICAR) model. PEBSI is calibrated and validated using robust in situ data from Gulkana Glacier, Alaska from 2000 to 2024. Simulations forced with original and bias-corrected climate reanalysis data show that statistically downscaling reanalysis data with in situ observations is necessary to reproduce summer mass balance (mean absolute error [MAE] = 0.75 m w.e. vs 0.22 m w.e., respectively). A grid search across two parameters, a precipitation factor and a densification parameter, reveals tradeoffs in performance compared to seasonal mass balance and end-of-winter snow density and depth. No single combination of parameters minimizes all errors, underscoring the inherent overparameterization of energy-balance models and challenges with translating coarse climate data to the glacier scale. The calibrated model successfully simulates the 2024 melt season, agreeing with surface-height change (MAE = 0.48 m) and albedo (MAE = 0.066) observations. Moving forward, PEBSI provides unique opportunities to quantify albedo feedbacks and their impact on present and future glacier mass loss.