| Differential Interferometric Synthetic Aperture Radar(InSAR)is a fast-developing space-to-ground observation technology in the past 30 years.Compared to the traditional survey technology,DInSAR has a wide range of coverage,high precision,weather independence and other advantages.However,the incoherence and atmospheric delay are the inherent error sources of the DInSAR method,which constrains the further improvement of its accuracy.Based on the multiple SAR images covering the same area,time series InSAR(TS-InSAR)technology is an effective method developed over the past two decades.Through the selection of high coherent point and modeling the specific deformation(such as linear deformation,periodic deformation),elevation error and atmospheric delay with redundant observation,this technology has reached an high accuracy,showing great potential in regional deformation monitoring(such as urban deformation monitoring),disaster monitoring and evaluation(such as volcanic and earthquake deformation measurement,landslide activity monitoring),energy and resource exploration(such as oil and gas field exploration,mineral resource exploitation),global environmental changes monitoring(such as glaciers movement monitoring,polar ice changes)and other broad areas.InSAR technology is mainly used for volcanic deformation monitoring,seismic deformation measurement,regional deformation measurement and other large-scale application in the early stage,and the accuracy of them are in the centimeter level.One main factors that constrain its improvement in accuracy and range of applications are the quality of the SAR images.The resolution of most early SAR images is above 10 m and the revisiting cycle is 30 days or more.Since 2007,some countries have developed and launched a new generation SAR satellite with significantly improved performance.Such as German TerraSAR-X and Italian COSMO-SkyMed satellite with 2 m spatial resolution SAR images,Sentinel-1 satellite with wide imaging TOPS mode.E.Sansosti,P.Berardino and their team defined such satellites as a new generation satellites systems.The major improvements of this new generation SAR satellites include shorter revisiting cycles,higher resolution,higher orbit accuracy,higher image quality and so on.These new generation SAR satellites have brought unprecedented high-quality SAR data,providing strong support to improve the accuracy of previous research.These high-resolution data,with large-scale coverage and short revisiting cycle,make cross fusion of InSAR technology and geophysics,Glaciers,hydrology and other disciplines become easier,and make many research that are never considered before become possible.However,as a result of the improvement of these new generation SAR imaging mode,the accuracy and efficiency of the image processing algorithm need to be further analyzed and verified.In the new field research and application,what kind of role it can play and how accurate it can reach need more experiment and exploration.Based on the integration of new generation SAR datasets,besides the research on the image processing of new imaging mode,this paper focus on the modeling and methodology to extract the topography and displacement information,and explores the possibility of improving the precision in previous research and the potential new application brought by the new generation SAR images.Aiming at the problem of extraction of topography with high accuracy and on time by InSAR technology,this paper performs related research based on the TerraSAR-X/TanDEM-X(TSX/TDX)datasets.In this paper,the digital elevation model of Tianjin(as flat area)and Yanliang(as hilly area)was extracted by TSX/TDX data.Compared with SRTM DEM,the ability of TSX/TDX data to extract detailed elevation information is confirmed.The Daguangbao landslide area is selected as the mountainous study area.The traditional method is subject to incoherence and overly dense topographic changes in this area.This paper proposes a re-flattening iterative method to deal with these problem in the data processing,and the precise TSX/TDX DEM is generated.The TSX/TDX datasets provide us timely,high-precision and large-scale coverage acquisition of terrain information,bringing new methods and possibilities for our post-earthquake evaluation or other work.In this paper,the TSX/TDX high-precision DEM and pre-earthquake SRTM DEM are compared.As a result,the landslides spatial distribution map and the cumulative volume of this landslide are obtained,which is estimated to be 1.189x109 m3 through this method.This estimation is close to the results(1.159×109 m3)from five-years engineering geological survey and its accuracy is higher than other DEM comparison results.The European Space Agency(ESA)launched Sentinel-1 A satellite in 2014,which utilizes the new TOPS imaging mode.Its great changes of Doppler center frequency in the whole image bring some difficulty in the data processing of TOPS mode.Aiming at the problem that may occur in the TOPS data processing process,this paper proposes a iterative processing chain including cross-correlation method and spectra diversity method,which can eliminate the potential phase jump problem during the burst mosaic Taking the Daguangbao landslide as case study,this paper integrate the TSX/TDX DEM and perform the time series InSAR analysis with the Sentinel-1 dataset.As a result,the activity of Daguangbao landslide in 2015?2016 is successfully monitored.The results show that although it is eight years after Wenchuan earthquake(i.e.the time Daguangbao landslide formed),there are still four active area,with a maximum deformation rate of 8 cm/year.Germany TerraSAR-X and Italy COSMO-SkyMed satellite can provide 2-meter high-resolution SAR images,providing a viable way to study specific objectives in the small-scale area.This paper uses the multi-platform data to monitor the surface deformation in Shanghai from 2009 to 2010.During the calculation with traditional methods,we find that the coincidence of the results from the same orbit(with descending orbit)is much higher than that of the results from the different orbit(even with the same satellite).It is proved that the horizontal displacement is the factor we miss in the vertical deformation calculation.In this paper,we propose a method which consider the horizontal displacement based on multi-platform data to extract the two-dimension displacement in Shanghai.The results show that the vertical displacement rate in Shanghai from 2009 to 2010 is 2.3 cm/year.At the same time,combining the groundwater exploitation/recharge data and underground hydrological structure,we analyze the deformation reasons and affected buildings in subsidence bowl in detail.Monitoring the subsidence(including infrastructure instability)in frozen soil areas with X band SAR data is often a challenge due to temporal decoherence.This paper proposes a Stacking InSAR approach based on temporary scatterers to extract road subsidence in frozen soil areas with TerraSAR-X high resolution images.In the case of less coherent interference,based on this method,we have successfully extracted the road deformation and analyze the result together with terrain and geomorphology information.It is found that the middle section of the road is Thermokarst area,and the maximum accumulated subsidence reach to 10 cm during the monitoring period. |
PDF Full Text Request