Integrating a laminar-flow inversion with random-cell interpolation for glacier ice thickness: An application to the Western Kunlun Mountains

Glacier ice thickness is a critical parameter for simulating glacier dynamics and estimating glacier volume. This study proposes a novel approach that integrates the laminar flow model with random cell interpolation to estimate glacier ice thickness with high efficiency and accuracy. A primary advantage of this method is that it avoids 0ver-reliance on central flowlines, a significant constraint in many traditional models that is difficult to accurately delineate glaciers with complex flow patterns. Furthermore, the integration of random cell interpolation allows the model to generate glacier cross-sections that are more consistent with idealized, physically plausible shapes, refining the raw thickness estimates. The effectiveness and accuracy of this approach were validated in the Western Kunlun main peak region. The results demonstrate the model's reliability, with the total estimated glacier volume falling between the results of the established H&F and GlabTop2 models. The RCGTI calibration, a key component of our method, is proved to be highly effective, reducing the overall volume overestimation by 8.9% from 490.02 to 446.57 km³. Crucially, when compared to ground penetrating radar (GPR) measurements, our model shows a better match than other mainstream inversion methods. This study demonstrates that the proposed integrated model provides a robust and effective method for large-scale assessments of glacier storage.