Aliasing in InSAR and SBAS Time Series

Interferometric synthetic aperture radar (InSAR) is a well-known imaging technique used by geophysicists to measure deformation of the Earth's surface over time. Accurate measurements of these displacements are crucial to correctly interpret subsurface processes. We show that when a closed loop of displacement gradients exceeding π radians occurs in a wrapped interferogram, the unwrapped solution will systematically underestimate the total displacement due to aliasing, even in the absence of noise. Common InSAR practices such as spatial filtering and averaging decrease resolution and thus increase the risk of aliasing. For time-varying processes, interferograms formed over longer time spans are more likely to under-sample the true displacement. Synthetic and real time series analyses using small baseline and subset (SBAS) [1] techniques show that the inclusion of aliased interferograms results in systematic and significant errors. These examples suggest that adaptive filters and varying the subset of interferograms used in SBAS throughout the scene will be important improvements to InSAR analyses.