A sparse parameter mode for MT-InSAR deformation retrieval and uncertainty assessment

Multitemporal InSAR is a widely used geodetic technique for measuring ground deformation. However, assessing the accuracy of InSAR deformation results is challenging, especially when field measurements such as leveling are limited in coverage or unavailable. While many studies have attempted to calculate the uncertainty of deformation using a priori InSAR stochastic models to assess the deformation uncertainty, these models are often biased by various factors. In this letter, we propose a new method called the sparse parameter model (SPM) for InSAR deformation retrieval and uncertainty assessment when instantaneous deformation is not the focus. The method estimates the sparser deformation time series and leverages redundant SAR observations for the deformation uncertainty assessment and decorrelation noise suppression. The proposed model is tested by both simulated and real Sentinel-1 datasets and the derived deformation was validated with GPS measurements in the real application. The results demonstrated that the overall uncertainty of InSAR deformation, as estimated by the SPM, is 5.4 mm, falling well within the expected range of uncertainty, which highlights the effectiveness of the SPM in retrieving InSAR deformation and assessing uncertainty.