Multiwavelength InSAR time-series analysis and change detection for improved signal detection quality in open-source radar Earth observing systems

Abstract

Satellite interferometric synthetic aperture radar (InSAR) is a powerful geodetic technique that uses the differences in phase between repeat SAR acquisitions to remotely detect subtle (cm-scale) surface deformation over large (10s to 100s of km) footprints. Over Antarctica, 6-day repeat InSAR from Sentinel-1 provides a powerful tool for measuring short-term changes in Antarctic ice-shelf motion. InSAR measures line of sight displacement from an off-nadir look angle, capturing both horizontal ice flow and vertical ocean tidal displacement in the resulting interferogram, both important effects for understanding ice sheet evolution. Double Difference InSAR (DDInSAR) is a phase-based change detection approach that subtracts two InSAR phase images, removing anything deforming at a constant rate in both InSAR images. DDInSAR can be used to help isolate tidal and ice velocity effects by choosing pairs with net zero or large tidal displacements, commonly producing dense bands of fringes along the edge of an ice shelf where ice shelf response to tides smoothly changes from its hydrostatic maximum to zero. Here, we investigate how changing net tidal displacements affect the resulting interferogram shape by computing DDInSAR images at several modeled tidal heights over Rennick Glacier and Amery Ice Shelf, Antarctica. We compute the pixel-wise angular deviation of DDInSAR images to remove phase wrapping jumps and then take the gradient to algorithmically identify flexure zone extent, exemplifying the ability of InSAR to observe short-term grounding zone changes. This gradient-based analysis offers increased interpretability of flexure zone extent in DDInSAR images and is generalizable to other areas of study.