Improving InSAR Tropospheric Delay Correction in Deep Canyon Regions With a DEM Watershed-Based Segmented Linear Model
Accurate atmospheric correction is critical for improving the reliability of InSAR deformation monitoring in mountainous area, such as southeastern edge of the Tibetan Plateau, where rugged topography and complex atmospheric conditions introduce significant tropospheric delays. Traditional correctio...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11062326/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Accurate atmospheric correction is critical for improving the reliability of InSAR deformation monitoring in mountainous area, such as southeastern edge of the Tibetan Plateau, where rugged topography and complex atmospheric conditions introduce significant tropospheric delays. Traditional correction methods, including global linear models (LMs), regular-window segmented linear model (RSLMs), and numerical weather models such as ERA-5 and Generic Atmospheric Correction Online Service (GACOS), often fail to address the spatial heterogeneity of atmospheric signals in such terrains, leaving residual artifacts that obscure surface deformation measurements. To overcome these limitations, this study proposes a watershed-segmented linear model (WSLM) that incorporates vertical atmospheric stratification and lateral watershed boundaries to effectively capture localized atmospheric variability. The performance of WSLM was evaluated using both simulated datasets and real Sentinel-1 data from the Deqin section of the Lancang River and compared with corrections provided by LM, RSLM, GACOS, and ERA-5 corrections. The results show that WSLM effectively reduces atmospheric artifacts, mitigates vertical stratification delays, and improves the recovery of realistic deformation signals. Compared to existing methods, it achieves lower residual phase standard deviations—reducing them by up to 75.78% —weakens phase-elevation correlations, and enhances time-series displacement accuracy. While uncertainties remain in determining the optimal weighting factors and segmentation thresholds, WSLM effectively reduces atmospheric errors and provides valuable insights for deformation monitoring in complex mountainous environments. |
|---|---|
| ISSN: | 1939-1404 2151-1535 |