Detecting ground settlement of megacities using InSAR techniques

Abstract

Subsidence phenomenon in Shanghai has been monitored by interferometric synthetic aperture radar (InSAR) techniques in this study. The research employed different InSAR techniques to study both the spatial and the temporal variations of the land subsidence phenomenon in Shanghai, using datasets acquired by four SAR systems - the L-band JERS-1 SAR system, the C-band ERS-1 and ERS-2 systems, and the C-band ENVISAT ASAR system. The behaviour of scatterers in Shanghai has been assessed with time by means of coherence analysis using L-band JERS-1 and C-band ERS-1/2 SAR data. Results obtained showed that InSAR pairs in Shanghai do not maintain coherence after a period of 7 months, and that the L-band JERS-1 SAR is more suitable than the C-band ERS-1/2 SAR for mapping land subsidence phenomenon in Shanghai that is characterised as slow. Residual phases from differential interferograms were successfully minimised using a higher order polynomial and the least squares method. Accumulative land subsidence maps have been produced for Shanghai at three different periods: 1992-1998, 1999-2000 and 2003-2007, using different SAR datasets. Results obtained showed that not all places in Shanghai are subsiding, but the city centre and its surroundings are under the threat of land subsidence. Quantitative measurements indicated that the downtown area subsided at a rate of 30 mm/year from 1992-1998 (L-band JERS-1 SAR data); 20 mm/year from 1999-2000 (C-band ERS-1/2 SAR data); and 10 mm/year from 2003-2007 (C-band ENVISAT ASAR data). Qualitative analysis of L-band JERS-1 differential interferograms has revealed three stages of land subsidence from 1992 to 1998. The urban areas in Shanghai have been successfully mapped by using interferometric coherence maps and point targets. Areas mapped as urban match those that were mapped as areas experiencing land subsidence. The later finding confirmed the assertion by some researchers (Gong et al., 2005; Xue et al., 2005; Zhang and Wei, 2005; Zhang et al, 2002) that massive engineering constructions and large number of high-rise buildings in Shanghai contribute to the land subsidence problem. Empirical models have been developed, both in time and spatial domains, for the prediction of the amount of land subsidence using 10 JERS-1 deformation maps. The models were developed using the method of least squares and higher order polynomials. Results obtained indicate that such an approach could be employed to predict the amount of land subsidence, in both spatial and time domains. The study was limited by the fact that some auxiliary data, such as location of wells or underground water pumping stations in the city, elevation data, number of construction units, and building loads, that could be useful for analyses and validation of results were not available. However, the main success of this study is that, it demonstrated the power of using InSAR techniques for ground surface deformation monitoring over large areas such as Shanghai.