Climatic controls on global snow surface sublimation based on ERA5-Land

Sublimation of snow represents an often important but poorly constrained component of the hydrological cycle, especially at the global scale. Studies that estimate snow sublimation at point or catchment scales demonstrate a range of uncertainties in the contribution of sublimation to total winter snowfall, ranging from 5% to 90%. Although it is well established that dry, windy and clear-sky conditions favor snow sublimation, a modern, global-scale assessment of the climatic controls and regions where sublimation occurs and is relevant for snowpack evolution, glacier mass balance and water resources is lacking. Existing global efforts are limited by coarse resolution (~250 km) reanalysis data, leaving a critical gap in our understanding of sublimation’s contribution to the water balance across climates and regions. Here we present a global analysis of snow surface sublimation hotspots, using ERA5-Land reanalysis at 0.1° (~10 km) resolution from 1980 to the present. Comparisons with sublimation observations from eddy-covariance flux towers demonstrate that ERA5-Land underestimates sublimation rates, but performs favorably compared to estimates from other reanalysis (GLDAS, GLEAM, MERRA-2) products. Comparisons with station observations also demonstrate that ERA5-Land correctly reproduces global patterns of seasonal snow variability. 

Preliminary results show clear latitudinal, elevation and climatic controls on global surface sublimation. Hotspots of snow sublimation (>80 mm/year) are identified in the higher elevations of South America, North America and Asia, with contributions to total snow ablation ranging mostly from 10 to 20%. Hotspots of lower total annual surface sublimation (30 to 60 mm/year) lie in latitudes between 40 and 60 °N in dry climates, where the contributions to total snow ablation mostly range from 20% to 60%. The strongest surface sublimation hotspots in absolute and relative terms are identified in parts of Greenland and coastal Antarctica, where uncertainty is high as no sublimation observations from flux towers are available to compare with. We also investigate historical (1980-2025) changes in sublimation fluxes in response to warming and changing snow cover patterns. 

Our results highlight regions where surface sublimation may be a significant component of the hydrological cycle, with implications for water resources, glacier mass balance and snow–atmosphere interactions. Important uncertainties remain, particularly in complex mountain regions where the resolution of ERA5-Land data may not fully capture sublimation processes such as boundary layer warming and drying. Furthermore, drifting and blowing snow sublimation are not resolved in ERA5-Land. Future efforts should refine these global estimates by using higher-resolution simulations and improved representations of snow–atmosphere interactions to identify sublimation hotspots over complex terrain.