Improved InSAR Atmospheric Corrections Using Variable Tropospheric Integration Heights and Multi-Source Global Ionospheric Maps

Atmospheric effects significantly impact the accuracy of Interferometric Synthetic Aperture Radar (InSAR) measurements of ground deformation, necessitating precise corrections for atmospheric phase artifacts. These signals primarily arise from spatial and temporal variations of the troposphere and ionosphere, and are typically partially mitigated using global atmospheric models and global ionospheric maps. Here, we present new methods for estimating more accurate tropospheric and ionospheric phase, which reduce phase error significantly. The tropospheric correction method dynamically identifies the upper height up to which the tropospheric model is reliable for the hydrostatic and wet components, and uses these parameters for individual SAR acquisitions to estimate the tropospheric phase in the radar line-of-sight (LOS) direction for the corresponding interferograms. For the ionospheric correction, we combined vertical total electron content (vTEC) products from eight ionospheric analysis centres, which differ in spatial/temporal resolution and estimation strategies. In addition, multiple mapping functions were tested to project the vTEC values into the LOS direction. We validate the proposed corrections by applying them to hundreds of Sentinel-1 small-baseline interferograms over southeastern Iran (Makran), central Türkiye and northern Syria. Results show that incorporating variable tropospheric heights and multi-source ionospheric models leads to substantial improvements over existing methods for mitigating tropospheric and ionospheric artifacts. For the study areas, the proposed tropospheric correction reduced the standard deviation of unwrapped interferograms, achieving improvements of approximately 4–20% relative to existing model-based approaches (D-LOS and GACOS). The new ionospheric correction achieved a fourfold greater improvement compared to the CODE correction. The new methods yields improvements in a greater proportion of interferograms, alongside markedly fewer and weaker instances of deterioration. Furthermore, the new correction method achieves a 25.5% improvement in the GNSS–InSAR RMSE compared with data corrected jointly by GACOS and CODE.