Monitoring Coastal Reclamation Subsidence in Hong Kong with Distributed Scatterer Interferometry

Hui Lin,Qishi Sun,Hansheng Wang,Liming Jiang,Mi Jiang,Peifeng Ma

doi:10.3390/rs10111738

Abstract

Land subsidence has been a significant problem in land reclaimed from the sea, and it is usually characterized by a differential settlement pattern due to locally unconsolidated marine sediments and fill materials. Time series Synthetic Aperture Radar Interferometry (InSAR) techniques based on distributed scatterers (DS), which can identify sufficient measurement points (MPs) when point-wise radar targets are lacking, have great potential to measure such differential reclamation settlement. However, the computational time cost has been the main drawback of current distributed scatterer interferometry(DSI) for its applications compared to the standard PSI analysis. In this paper, we adopted an improved DSI processing strategy for a fast and robust analysis of land subsidence in reclaimed regions, which is characterized by an integration of fast statistically homogeneous pixel selection based (FaSHPS-based) DS detection and eigendecomposition phase optimization. We demonstrate the advantages of the proposed DSI strategy in computational efficiency and deformation estimation reliability by applying it to two TerraSAR-X image data stacks from 2008 to 2009 to retrieve land subsidence over two typical reclaimed regions of Hong Kong International Airport (HKIA) and Hong Kong Science Park (HKSP). Compared with the state-of-the-art DSI methods, the proposed strategy significantly improves the computational efficiency, which is enhanced approximately 30 times in DS identification and 20 times in phase optimization. On average, the DSI strategy results in 7.8 and 3.7 times the detected number of MPs for HKIA and HKSP with respect to persistent scatter interferometry (PSI), which enables a very detailed characterization of locally differential settlement patterns. Moreover, the DSI-derived results agree well with the levelling survey measurements at HKIA, with a mean difference of 1.87 mm/yr and a standard deviation of 2.08 mm/yr. The results demonstrate that the proposed DSI strategy is effective at improving target density, accuracy and efficiency in monitoring ground deformation, particularly over reclaimed coastal areas.

Highlights

Reclamation of land from the shallow sea to create new land has been a common practice to meet the increasing demand for urbanization and population growth in coastal regions worldwide [1,2]
For further analysis between observed settlements and levelling survey measurements, we focused on the residual settlements at the southern runway of the airport, which is characterized by complicated reclamation geology due to variations in alluvial deposit thicknesses and lithologies [42]
Performance Evaluation of statistically homogeneous pixels (SHPs) Identification and Optimal Phase Estimation To assess the reliability of our results in terms of performance of SHPs over areas with multiple types of land cover and highly complex scattering characteristics, we select the Hong Kong Science Park (HKSP) reclaimed area for this purpose, since the reclamation land at this site was underdeveloped during the TerraSAR-X data acquisition period, and the land contained various types of terrain targets such as bare soils, buildings, roads, sparse vegetation and mixed land covers

Summary

Introduction

Reclamation of land from the shallow sea to create new land has been a common practice to meet the increasing demand for urbanization and population growth in coastal regions worldwide [1,2]. PSI techniques only focus on point-wise radar targets with high reflectivity and stable phase characteristics over time, most of which belong to angular, man-made objects such as building facades and corners This PSI technique limitation usually leads to a low spatial density of PSs in coastal reclamation land characterized with natural targets or low reflectivity regions, especially for newly developed reclamation or some reclaimed civil infrastructures (e.g., highways, ports and airfields) [3,14,15,16]

Objectives

Results

Conclusion