Assessing the impact of ice thickness uncertainty on future glacier evolution in the Himalayas using a higher-order glacier flow model

• Differences among thickness datasets lead to a standard deviation of 1.5%–6.2% in projected volume loss. • Central Himalayas glaciers show highest sensitivity to ice thickness uncertainties. • High-order flowline model amplifies glacier response to initial ice geometry variations. Understanding future glacier evolution in the Himalayas is essential for anticipating changes in regional water resources. However, significant uncertainty remains due to differences in the glacier ice thickness datasets used to initialize projection models. This study investigates the influence of ice thickness uncertainty on modeled glacier changes using a higher-order flowline model. We simulate the future evolution of 30 glaciers across the western, central, and eastern Himalayas under a committed climate scenario, assuming constant climate conditions from 1991 to 2020. Seven widely used global ice thickness datasets are applied to assess their impact on projections of glacier thickness, volume, and surface velocity. By 2100, glaciers are projected to lose an average of 23.4 ± 3.7 m in thickness, 0.10 ± 0.02 km 3 in volume, and 0.67 ± 0.54 m a −1 in surface velocity. Differences among thickness datasets lead to a standard deviation of 1.5%–6.2% in projected volume loss, with larger divergence in the central Himalaya. The inter-dataset variability decreases with glacier elevation and size, whereas smaller and lower-elevation glaciers remain more sensitive to initial thickness differences. Although the overall magnitude of uncertainty is moderate, its influence persists throughout the century and is comparable to, that from climate model resolution. The use of a higher-order model further amplifies these differences due to its sensitivity to initial geometry. These findings highlight the need for improved thickness observations and emphasize the importance of incorporating multiple thickness datasets in regional glacier modeling.