In this Record we present a case study where we use high resolution satellite radar interferometry (InSAR) for measuring ground surface movements in a dense urban area of Australia. InSAR is a space-based satellite radar imaging method that can measure changes in surface movement across an imaged region over time. It is complementary to other surveying techniques, such as ground levelling surveys or leveraging Global Navigation Satellite Systems. InSAR is being increasingly applied to monitor tunnelling and other civil infrastructure construction projects around the world, including in Australia. Our case study was undertaken as an InSAR monitoring pilot project in 2017 as part of the Metro Tunnel Project, prior to the commencement of tunnelling beneath central Melbourne, Australia. Persistent Scatterer InSAR analysis of Melbourne was conducted on a set of 52 high resolution (1 metre) COSMO-SkyMed Spotlight images captured from ascending and descending orbital passes over a period of approximately eight months between March and November 2017. The accuracy of ground displacement measurements derived from the InSAR analysis (quantified by the 2-sigma uncertainty), is ±1.6 millimetres. A measurement point density exceeding 36,000 points per square kilometre was achieved for both ascending and descending InSAR displacement datasets. The vertical displacement has been derived by combining the independent ascending and descending InSAR data by interpolation to a regular grid of points at 10 metre intervals. This interpolated data set of InSAR-derived vertical displacements indicates that the vast majority (greater than 99%) of InSAR-detected measurement points in the project’s area of interest experienced upward or downward movement of less than 5 millimetres during the eight month monitoring period. We validated the interpolated vertical displacement values derived from InSAR with in-situ levelling measurements on 29 benchmark arrays where there were sufficient InSAR data points to make the comparison. We found that the mean absolute difference in vertical displacements derived from levelling and InSAR was less than 3 millimetres for 26 of these 29 arrays. Our case study leveraged one of the best possible InSAR datasets for ground movement monitoring in an urban area, yet it still has limitations. For instance, we found that the coverage of InSAR data points in wide open vegetated spaces such as parks is low. As a consequence, InSAR should always be used in conjunction with a suite of monitoring techniques and tools.
Australia) et al. (Thu,) studied this question.