Glacier mass changes in the Western Pamirs during 2003-2025 from high-resolution stereo satellite images

The Pamir Mountains are a key focus of research on the response of glaciers to climate change in High Mountain Asia. They constitute an important mountain water tower that is highly vulnerable to future climatic and environmental change, and they host glaciers that experienced limited but increasing mass loss during 2000-2019. Few studies have constrained mass balance in detail within the region, and these have highlighted strong spatio-temporal variability. Quantification of glacier mass changes and their drivers is challenged by sparse in situ data, strong East-West gradients of temperature and precipitation, and complex glacier processes including debris cover, surging behavior, and collapse features. In this study, we focus on 41 glaciers in the Sangvor region of the western Pamir for 2003-2025 to: i) evaluate the robustness of previous estimates of mass losses for 2000-2019, ii) identify glacier trajectories beyond 2019, and iii) understand the influence of surging glaciers on the observed mass changes. We adopted a stereo image-processing workflow from the Ames Stereo Pipeline and generated DEMs from SPOT5, SPOT6, and Pléiades stereo satellite imagery. Glacier mass balance for each time series was derived using current best practices for DEM co-registration, bias correction, and uncertainty propagation. Additionally, we applied new algorithms for jitter correction, seasonal snow correction based on in-situ data, and a duration-dependent correction scheme for volume-to-mass conversion uncertainties.

 

Our results provide a time series of high-resolution geodetic surface height changes and glacier mass balance over seven sub-periods spanning 2003 to 2025: -0.19 ± 0.04 m w.e. a^-1 for 2003-2019, -0.63 ± 0.05 m w.e. a^-1 for 2019-2025, and -0.28 ± 0.02 m w.e. a^-1 for the entire period of 2003-2025 for three key intervals. Our 2003-2019 results agree with the mean mass balance measured by ASTER for 2000-2019 (-0.21 m w.e. a^-1) and with the temporal trend. We highlight a sharp mass gain (0.16 ± 0.03 m w.e. a^-1) between 2014 and 2019, followed by a pronounced progressive shift toward negative mass balances. During 2019-2025, characterized by exceptionally warm and dry conditions, the mass balance has been increasingly negative despite higher uncertainty in our annual estimates, reaching its maximum loss of -1.45 (+0.51/-0.19) m w.e. a^-1 in 2024-2025. We also find no statistically significant difference between the mass balance of surging glaciers and the regional mean for the periods of our study. Overall, our results reveal strong temporal variability in glacier mass balance in the region. While long-term mean mass balances agree with previous ASTER-based estimates, our high-resolution geodetic time series resolves their short-term variability, illuminating a complex evolution that includes a marked mass gain for 2014-2019 and a rapid shift toward strongly negative balances thereafter. Comparison with reanalysis data suggests that this variability is more closely linked to precipitation anomalies than to temperature, suggesting a dominant role of mass accumulation and snowfall variability. These results demonstrate the value of high-resolution stereo imagery and motivate extension to the wider Pamir region and can form a new, high-resolution baseline for modelling projections of future glacier changes in the region.