CorPhU: an algorithm based on phase closure for the correction of unwrapping errors in SAR interferometry

Abstract

SUMMARY Interferometric Synthetic Aperture Radar (InSAR) is commonly used in Earth Sciences to study surface displacements or construct high resolution topographic maps. Recent satellites such as those of the Sentinel-1 constellation allow to derive dense deformation maps with millimetric precision with high revisit frequency. However, InSAR is still limited by interferometric coherence. Interferometric phase noise resulting from a loss of coherence, due to changes in scattering properties between repeated SAR acquisitions, may lead to unwrapping errors, which then in turn lead to centimetric errors in time-series reconstruction. We present an algorithm based on interferometric phase closure to automatically correct unwrapping errors. We describe the algorithm and highlight its performances with two case studies, in Lebanon with Envisat satellite data and in Central Turkey with Sentinel-1 data. The first data set is particularly affected by unwrapping errors because of long spatial (500 m) and temporal baseline interferograms (6 yr) and decorrelation due, in particular, to vegetation. The second data set contains unwrapping errors because of temporal changes in the scattering properties of the ground. For these two examples, the algorithm allows the correction of almost all detectable unwrapping errors, without requiring visual inspection or manual deletions. Our algorithm is efficient especially on large data sets, such as with Sentinel-1 constellation, where interferometric phase is redundant and improves eventually the reconstruction of time-series.

Autonomous extraction of millimeter-scale deformation in InSAR time series using deep learning

“…We apply a phase-preserving filter and multilooking (i.e., averaging of adjacent pixels) so that the final pixel size is about 70 m in range and azimuth 71 . Potential unwrapping errors are corrected using CorPhu 72 . Interferograms are unwrapped using a branch-cut method 73 in areas for which coherence exceeds 0.5.…”

Permafrost Dynamics Observatory—Part I: Postprocessing and Calibration Methods of UAVSAR L‐Band InSAR Data for Seasonal Subsidence Estimation

“…manuscript submitted to Earth and Space Science phase unwrapping errors can yield physically unrealistic, spatially isolated estimates of deformation. Where three or more coregistered interferograms are available, phase closure constraints can be used to correct for unwrapping errors (as in (Yunjun et al, 2019;Benoit et al, 2019)). With only a single interferogram per study site, such an analysis is not possible in this study.…”

“…UAVSAR interferometric products tend to exhibit interferometric coherence much higher than analogous spaceborne InSAR products; further, masking out of low‐coherence pixels is a straightforward method of culling interferometric data that are corrupted by significant decorrelation noise. Individual interferograms that contain discontinuous regions of coherent pixels separated by decorrelated pixels can often exhibit phase unwrapping ambiguities and local phase unwrapping errors, which can be mitigated by empirical unwrapping correction algorithms and masking out of statistical outliers, respectively (Benoit et al., 2019 ; López‐Quiroz et al., 2009 ; Yunjun et al., 2019 ).…”

A Kalman Filter Time Series Analysis Method for InSAR

“…In practice, ϵ ij only results from how we construct interferograms and could be directly estimated by quantifying the effect of multilooking and filtering during the processing of each interferogram. Alternatively, it could be measured either before time series analysis by forming triplets of interferograms (Benoit et al., 2020; De Zan et al., 2015) or a posteriori by looking at the discrepancy between real and reconstituted interferograms from time series (Cavalié et al., 2007).…”