Study On TOPS DInSAR Process And Its Application To Monitor Ground Deformation Over The Jing-Jin-Ji Area

Synthetic aperture radar(SAR)is a type of active microwave remote imaging system developed in the late 1950 s.It is capable of monitoring ground surface with its characteristics of all-weather and all-time.Differential synthetic aperture radar interferometry(DInSAR)technique exploits the phase difference between two temporally separated SAR images acquired from the same orbit,or slightly different orbits,to generate a differential interferogram that contains the signature of the ground deformation with high resolution and wide coverage,which makes it one of the most promising measuring technologies in the application of surface deformation monitoring.In April 2014,ESA successfully launched a new generation of remote sensing satellite,Sentinel-1,and provided SAR observation images to the society for free.The Sentinel-1 satellite supports four imaging modes providing different resolution and coverage,the default mode for terrain observation is the Interferometric Wide Swath(IW)mode.IW mode captures three sub-swaths using Terrain Observation with Progressive Scans(TOPS)SAR.Compared to Strip mode,the TOPS mode is more advantageous in terms of achieving a wider image swath and a shorter revisit time to the Earth's surface.Therefore,using data captured in TOPS mode is more likely to obtain interferograms with high coherence in DInSAR process.However,the unique features of TOPS mode lead to the conventional InSAR processing strategy being not able to be directly applied.There are two problems in TOPS DInSAR processing: 1)the minimum imaging unit in TOPS mode is a Burst,those Bursts should be merged first to obtain wide-swath images;2)in the applications of mapping large scale ground displacement(e.g.earthquakes),traditional coregistration method is hard to satisfy the strict accuracy requirement of TOPS processing.The cause of the second problem is that the Doppler centroid varies during the observation,slight misregistration in the azimuth direction can cause the presence of severe phase jump.Usually the geometrical coregistration and the enhanced spectral diversity(ESD)algorithms are adopted to resolve the problem of TOPS data coregistration.However,for the geometrical coregistration method may not completely compensate the phase ambiguity,this strategy is only suitable for the TOPS data acquired in stationary areas where no or little deformation is expected in azimuth direction.In this dissertation,the basic theory of TOPS imaging mode is thoroughly studied,the difficulties in TOPS DInSAR processing are systematically analyzed,and the corresponding solutions are explored.The main research work in this dissertation is listed below:(1)Compared to conventional DInSAR processing strategy suited for Strip mode,the difficulties in TOPS mode DInSAR processing are minutely analyzed based on the operational principle of TOPS mode.(2)A universal TOPS DInSAR processing scheme is designed for mapping nonstationary scenarios.This processing scheme is separated into five modules: digital elevation model pre-processing module(DPM),preliminary coregistration module(PCM),pixel offset coregistration module(OCM),spectral diversity coregistration module(SCM)and enhanced spectral diversity coregistration module(ECM).For validation purposes,the scheme is applied to the Sentinel-1 data for monitoring the coseismic deformation field of the 2016 Kumamoto earthquake.The results demonstrate that it is capable of effectively extracting large scale deformation from TOPS data.(3)TOPS SAR data are applied to the multi-baseline DInSAR ground deformation monitoring.We utilized 54 Sentinel-1 SAR data acquired over the the Jing-Jin-Ji Area and obtained the corresponding LOS deformation field with large coverage.