Global vs local glacier modelling: a comparison in the Tien Shan

Abstract. Glaciers in the Tien Shan are vital for freshwater supply, emphasising the importance of modelling their future evolution. While detailed 3D models are suitable for well-studied glaciers, regional and global assessments rely on simplified approaches. However, their accuracy remains understudied. Here, we compare the evolution of six glaciers in the Tien Shan using (i) a 3D higher-order ice flow model and (ii) a global glacier model (GloGEMflow). Additionally, we explore the impact of using in-situ measurements of mass balance and ice thickness, as opposed to relying on globally available data. Our findings reveal that the choice of mass balance model complexity and calibration has a minimal impact on aggregated volume projections, with less than 3 % variation by 2050 and less than 1 % thereafter. The use of a detailed versus a simplified ice flow model results in some noticeable discrepancies in the first half of the century, with an 8 % variation in aggregated volume change by 2050. These disparities primarily stem from calibration, while the glacier evolution pattern remains consistent, showing good agreement between the detailed and simplified model. In general, our results demonstrate that the initial ice thickness estimation has the largest effect on the future remaining ice volume, potentially resulting in 2 to 3, and even up to 4 times, more ice mass remaining. Our findings thus suggest that when modelling small to medium-sized glaciers the emphasis should be on having a reliable reconstruction of the glacier geometry rather than focusing on a detailed representation of ice flow and mass balance processes.