Adaptive GNSS/InSAR Fusion for Three-Dimensional Deformation Reconstruction and Mass-Conservation-Based Thickness Inversion of Reservoir Landslides
Applied Sciences · 12.11.2026 · 16 (22) · 11204
DOI: 10.3390/app162211204 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 12.11.2026. Verfasst von Jiacheng Zhang und 3 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Three-dimensional (3D) deformation monitoring of reservoir landslides remains challenging because InSAR observations have limited sensitivity to motion parallel to the satellite flight direction, vegetation-related decorrelation can reduce the availability of coherent measurements, and GNSS stations are often sparse. This study presents a GNSS/InSAR fusion framework for reconstructing 3D surface deformation and estimating landslide thickness. Ascending- and descending-track Sentinel-1 line-of-sight (LOS) deformation measurements are combined with three-dimensional GNSS observations in a joint observation model. A landslide-mask-constrained first-order Tikhonov regularization is used to constrain spatial differences within the mapped landslide, while a benchmark-anchored Helmert variance component estimation (VCE) strategy adaptively updates the relative weights of the ascending- and descending-track InSAR observations, with GNSS observations retained as the reference group. The reconstructed mean velocity field is then used in a mass-conservation-based thickness inversion that accounts for surface-slope advection and combines Laplacian regularization with borehole soft constraints, assuming an approximately stationary basal sliding surface. Synthetic experiments indicate that the framework improves reconstruction stability and recovers the northward component under the tested sparse-GNSS configuration. In the Hongyanzi case study, the northward deformation component yielded an RMSE below 6.5 mm yr−1 when compared with the GNSS observations used as inversion constraints; this is an in-sample residual measure rather than independent validation. At the seven boreholes incorporated as soft constraints, the final thickness field yielded an in-sample RMSE of 0.545 m and absolute residuals of approximately 0.20–0.78 m. These results illustrate the potential of combining GNSS and InSAR observations for joint analysis of surface deformation and landslide thickness, although independent validation and uncertainty assessment are needed to establish the robustness of the thickness estimates.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Zhang, J., Zhou, Y., Li, C., & Sun, Q. (2026). Adaptive GNSS/InSAR Fusion for Three-Dimensional Deformation Reconstruction and Mass-Conservation-Based Thickness Inversion of Reservoir Landslides. Applied Sciences, 16(22), 11204. https://doi.org/10.3390/app162211204
Vancouver
Zhang J, Zhou Y, Li C, Sun Q. Adaptive GNSS/InSAR Fusion for Three-Dimensional Deformation Reconstruction and Mass-Conservation-Based Thickness Inversion of Reservoir Landslides. Applied Sciences. 2026;16(22):11204. doi:10.3390/app162211204
BibTeX
@article{Zhang2026,
author = {Zhang, Jiacheng and Zhou, Yawen and Li, Cong and Sun, Qian},
title = {Adaptive GNSS/InSAR Fusion for Three-Dimensional Deformation Reconstruction and Mass-Conservation-Based Thickness Inversion of Reservoir Landslides},
journal = {Applied Sciences},
year = {2026},
volume = {16},
number = {22},
pages = {11204},
doi = {10.3390/app162211204}
}
EUVIMED