Integrated coastal subsidence analysis using InSAR, LiDAR, and land cover data

Abstract

Land subsidence is an important cause of relative sea-level rise along the Gulf Coast. There is a lack of effective monitoring of coastal subsidence with high accuracy and high spatial resolution for improving coastal risk assessment and mitigation. This study is the first attempt to integrate satellite interferometric synthetic aperture radar (InSAR) and airborne light detection and ranging (LiDAR) methods to investigate the spatiotemporal pattern of coastal subsidence. The study area is around Eagle Point, Texas, a region known for its fast rate of relative sea-level rise in recent decades. From 2006 to 2011, the line-of-sight velocities were up to −33 mm/year based on ascending ALOS-1 PALSAR-1 images. From 2016 to 2021, the vertical velocities were up to −34 mm/year based on ascending and descending Sentinel-1 images. Additional details of the subsidence pattern were revealed by incorporating the surface difference derived from 1-m airborne LiDAR results. Comparisons of the InSAR-derived velocities from image time series and the LiDAR-derived surface changes from time-lapsed observations were conducted at different spatial levels with linkages to land cover patterns and topography. The results showed that local subsidence rates could vary significantly below the spatial resolution of InSAR results, indicating a valuable role of airborne LiDAR results in extending InSAR results to parcel and building levels and explaining subpixel uncertainties. Also, subsidence appeared to be stronger in vegetated areas than in developed areas and negatively correlated with surface imperviousness. The magnitude of subsidence was not correlated with elevation along selected transect lines. Overall, this study demonstrated the benefits of combining InSAR results with other geospatial datasets to characterize coastal subsidence. In particular, the high vertical accuracy of InSAR results and the high spatial resolution of airborne LiDAR results could be complementary, highlighting the necessity of multi-resolution data fusion to support studies on coastal flood vulnerability, infrastructure reliability, and erosion control.