Kinematic Glacier Thickness Inversion Using 3-D Flow Velocities From Multitrack SAR Images

Accurate estimation of glacier thickness across space and time is essential for understanding glacier dynamics and their response to climate change. Traditional approaches are often constrained by uncertainties arising from dependence on poorly understood physical parameters. In this study, we develop a using remote-sensing–based framework for glacier thickness inversion multi-track synthetic aperture radar observations. Three-dimensional displacement time series of the Jiongpu glacier in High Mountain Asia were generated from Sentinel-1 imagery acquired between 2019 and 2022 using a pixel offset multidimensional small baseline subsets technique. These velocity fields were combined with surface mass balance to invert glacier thickness through a Monte Carlo approach, in which the rheological parameter was adaptively optimized based on terrain slope and horizontal velocity. Sensitivity analyses demonstrate that rheological parameter optimization reduces uncertainty by up to 18.4% in standard deviation and 20.4% in coefficient of variation compared with fixed values, while the inclusion of vertical velocity further decreases variation by 42.1%. Perturbations of slope by plus or minus 2.0 degrees increase thickness uncertainty by as much as 6.08 meters. Thickness change estimates along a transect show strong agreement with elevation change from ICESat-2, with a correlation coefficient of 0.91. These findings confirm that the proposed framework effectively captures the spatiotemporal variability of glacier thickness and reduces inversion uncertainty, providing a robust and scalable approach for monitoring glacier evolution and associated hazards in data-scarce mountainous regions.