InSAR Phase Unwrapping Error Correction for Rapid Repeat Measurements of Water Level Change in Wetlands

Abstract

Here, we present an enhanced algorithm to correct interferometric synthetic aperture radar (InSAR) phase unwrapping errors by incorporating iterative spatial bridging between islands and phase closure among interferograms. We use rapid repeat airborne synthetic aperture radar acquisitions from NASA's airborne uninhabited aerial vehicle synthetic aperture radar (UAVSAR) instrument to estimate short-term changes in water level within coastal wetlands from a stack of consecutive interferograms acquired with very short temporal separation (∼30 min). The algorithm is applied to six consecutive UAVSAR images collected in tidal wetlands of the Wax Lake Delta, Louisiana, USA. Validation of our water level change retrievals with in situ field observations was conclusive with high correlation and an RMSE generally smaller than 3 cm. Comparison of our algorithm with other phase unwrapping error correction methods shows significant improvement (30%-35% increase in the number of correctly unwrapped pixels) when applied to rapid changes in water level. The set of corrections presented in this work enables measurement of water level change in deltas and other areas where tides drive highly dynamic flooding of inland vegetated areas. Although demonstrated for water level change, the method is applicable to other InSAR datasets with large spatial gradients or observed discontinuities between coherent but spatially isolated areas.

Hillslope‐Channel Transitions and the Role of Water Tracks in a Changing Permafrost Landscape

“…The final mean inter-annual displacement velocity and its standard error were determined by fitting a linear deformation model to the movement time series. We performed quality controls on the data by manually masking areas where the results were deemed unreliable due to unwrapping errors, which manifest as a spatial discontinuity in displacement by integer multiples of half the radar wavelength (Oliver-Cabrera et al, 2022). We noted that errors and subsequent masked areas increased significantly with the inclusion of the 2019 image; thus, we concentrate our analyses here on averaged displacements deriving from the 2015 to 2017 imagery only, and we document data incorporating the 2019 data in Supporting Information S1.…”

SBAS-InSAR based validated landslide susceptibility mapping along the Karakoram Highway: a case study of Gilgit-Baltistan, Pakistan

“…Data preprocessing includes the calculation of time and space baselines between all Sentinel-1A image pairs. After registration and clipping, the DEM data is used to complete image registration, and the relative combination that satisfies a given threshold is selected to produce a differential interferogram set 75 . This study uses a 30 m resolution SRTM DEM to generate interferograms.…”

Spatial Variability in Salt Marsh Drainage Controlled by Small Scale Topography

“…However, its effectivity will be highly dependent on the repeat pass schedule of the observing sensor as well as the type of vegetation observed. If the vegetation has a soft stem, its scattering properties will change rapidly and data acquisition will require a very short repeat observation time, whereas woody vegetation may allow for a longer temporal baseline in the SAR acquisition pattern (Oliver‐Cabrera et al., 2021). Note that the method cannot be applied to submerged vegetation, for example, seagrass.…”