A Novel Method of Stacking InSAR Interferograms for Measuring Localized Earthquake Deformation

Surface displacements caused by small-to-moderate earthquakes are generally localized within small regions with magnitudes of only several centimeters. Interferometric synthetic aperture radar (InSAR) measurements of such deformation are prone to be contaminated by localized turbulent atmospheric delays, which are difficult to be mitigated by current methods. Here, we develop a new method to generate an optimal combination of coseismic interferograms for stacking to measure the deformation. The interferograms are weighted during the stacking process based on their signal-to-noise ratios (SNRs). To verify the performance of mitigating the tropospheric delays, we conduct both synthetic data experiments and practical applications with three different types of small-magnitude earthquakes to compare our method with current methods of differential InSAR, time series analysis, and multi-interferogram stacking. The experiment results demonstrate that our method can mitigate not only turbulent atmospheric delays but also other nonseismic deformation signals better than previous methods. The optimal combinations for measuring localized earthquake deformation consist of interferograms with both small and large standard deviation values. Further inversion tests reveal that accurate measurements of surface deformation obtained by our method are favorable to determine source faults and slip distributions of small-magnitude earthquakes without surface rupture. Although the experiments are conducted with only consideration of earthquake deformation, our method is also expected to be applicable for other tectonic and nontectonic activities, such as volcanoes and landslides, that occur within localized regions and over short time periods around several days.