Mapping Vertical Land Movement in Singapore using INSAR and GPS
Catalao, Joao¹; Raju, Durairaju²; Fernandes, RMS^3
1IDL, University of Lisbon, PORTUGAL; ²Tropical Marine Science Institute, National University of Singapore, Singapore, SINGAPORE; ³Universidade Beira Interior, IDL, Covilhã, Portugal, PORTUGAL

Global mean sea level rise associated with global warming has a major impact on coastal areas and represents one of the significant natural hazards. The Asia-Pacific region, which has the highest concentration of human population in the world, represents one of the larger areas on Earth being threatened by the rise of sea level. Recent studies indicate a global sea level of 3.2mm/yr measured from 20 years of satellite altimetry. Combined effect of sea level rise with local land subsidence, can be overwhelming for coastal areas. Singapore has undergone relative subsidence which created many small islands. Submergence also accelerates coastal erosion because it facilitates greater inland penetration of storm waves e.g. the observed accelerated erosion in East Coast of Singapore, a reclaimed land area with 185 hectares

In this study we investigate the spatial variation in vertical land motion (VLM) along the coast of Singapore over the past two decades and we examine the impact of spatially variable VLM on relative sea level trends. For that, we use an integrated approach based on the merging of terrain displacement velocities estimated by time series of interferometric synthetic aperture radar (InSAR) data acquired along ascending and descending orbits and repeated GPS measurements.

We processed 39 single look complex images from ERS2 satellite (ascending and descending passes) spanning from October 1995 until January 2000. A master was selected for each pass and two sets of time series differential interferograms were created. The Permanent Scatter (PS) technique was used to determine spatial and temporal patterns of surface deformation as well as their average rate. The vertical velocities of 5 continuous GPS stations of the Singapore Satellite Positioning Reference Network (SiReNT) were used as constraints in the PS solutions to further enhance the solution quality. GPS observations are not expected to have significant systematic errors and would filter out PS' long wavelength residual unmodeled errors due to unwrapping, orbital errors, and uncorrected atmospheric artifacts. Existing bias and tilts between PSs and GPS solutions are estimated minimizing a cost function and then removed from the PS estimates, resulting in a map of vertical land motion with increased spatial density resulting from the merging of ascending and descending PSs.

The results show that Singapore mainland has experienced subsidence phenomena from 1995 to 2000 with a mean subsidence rate of -1.5 mm/yr. On some localized areas, a significant subsidence with rates of -7 mm/yr was detected. Highest subsiding rates are near the shore on low flat land, above 5m, associated with reclaimed areas or built areas in the past years. The result seems to be reliable and consistent with the geologic setting. The scatter pattern of the deformation indicates an anthropogenic cause, related with compaction of built areas or reclaimed areas, rather than a natural cause.

These images were acquired within thin the Framework of the ESA Cat-1 (n. 6940) project