The Energy Balance Behind Glacier Retreat: A Regional Reconstruction for Western Canada (1980–2022)

Abstract Glacier mass loss has accelerated globally in recent decades, yet many regional and global glacier models still rely on temperature‐index melt approaches that limit physical realism and process attribution. Here, we apply a glacier evolution model incorporating a physically based surface energy balance (SEB) melt component to reconstruct glacier mass changes across western Canada from 1980 to 2022. We assess how well simulated mass changes reproduce observed glacier changes across spatial scales and identify the dominant climatic drivers of glacier mass loss. The model is forced by ERA5 reanalysis data with no downscaling and minimal bias correction and incorporates machine‐learning‐based albedo fields trained on MODIS observations. We find that glacier mass changes across western Canada are successfully reconstructed at both regional and individual‐glacier scales, with a root‐mean‐square error of 0.24 m yr −1 in regionally averaged thinning rates for 1985–1999. A long‐term increase in melt energy (+0.48 W m −2 yr −1 ) emerges as the dominant driver of regional glacier loss, primarily linked to declining summer albedo and increasing longwave radiation and sensible heat fluxes. Sensitivity experiments show that reliable reconstructions at both regional and glacier scales require precipitation corrections, specifically a multiplicative precipitation factor and an elevation‐dependent precipitation gradient calibrated against glacier‐specific geodetic mass change trends. Introducing an additional calibration‐dependent albedo bias correction further improves seasonal mass balance estimates at the glacier scale. These results demonstrate that simplified SEB‐based models, when combined with physically informed and spatially resolved calibration, can robustly capture glacier mass changes and provide mechanistic insight into the climatic drivers of glacier change.