Abstract. Monitoring linear infrastructures such as railways and highways with Multitemporal Interferometric Synthetic Aperture Radar (MTInSAR) requires to identify spatial displacement gradients to assess related hazards. Estimating the spatial gradients involves the retrieval of the displacement time series in MTInSAR for coherent pixels. However, the algorithms are computationally expensive because pixels outside the linear infrastructure are processed as they are required to aid the phase unwrapping and atmospheric phase filtering at the linear infrastructure. We propose a new approach which makes use of the known location of the linear infrastructure in the SAR images and estimate the differential displacement velocities along the infrastructure from the wrapped interferometric phases. In this way, the effect of incoherent pixels from outside the linear infrastructure and the potential error propagation during spatial phase unwrapping are mitigated. Our experiments based on TerraSAR-X and Sentinel-1 images show good agreement between the estimated spatial gradient velocities from our method and the conventional MTInSAR results. The sensitivity of the choice of grid size is evident in the resultant Root Mean Square Error (RMSE), which is approximately 0.1 cm/year when compared to the conventional MTInSAR results. This is achieved with a grid size larger than 300 m, which smoothed small variations in the differential velocity. Applying our method on Sentinel-1 images enables computationally efficient monitoring of linear infrastructures exploiting the wide area coverage and availability of the SAR images.
Piter et al. (2026) studied this question.