Study On Error Analysis And DEM Precision Improvement Methods Of Spaceborne Distributed InSAR

Spaceborne distributed interferometric synthetic aperture radar (InSAR) is a novel remote sensing system, which is an integration of the the spaceborne SAR and satellite formation flying technology. The system fulfills the three-dimensional terrain mapping by formation flying and cooperation of several satellites. Because of its advantages such as high vertical accuracy, strong survival capability and antijamming ability, short scanning cycle and low cost, the novel InSAR system has been one of the research hotspot around the world, especially the successful launch of the first spaceborne distributed InSAR system– TanDEM-X, makes the new system a new upsurge. Aiming at the novel system, the three-dimensional localization principles, the total error model, the DEM precision improvement methods are studied systematically. Here the chapter 2 and chapter 3 analyze the transference and the influence of parameter error, the chapter 4 and chapter 5 analyzes the DEM precision improvement methods. The work demonstrated in this thesis is expected to support the realization of future InSAR system. The research in each chapter is arranged as following:The three-dimensional localization and altimetry principles and the error sensitivity of the spaceborne distributed InSAR system are studied in chapter 2. The localization and altimetry principles of the new system are compared with the height measurement principles of two traditional InSAR systems as well as the error sensitivity. The equivalence among three model's error sensitivity is also proved in this chapter.The error problem of the spaceborne distributed InSAR system is studied in chapter 3. Based on the three-dimensional localization equations, the first order error sources of the new system are derived, and the property of every single error is analyzed in detail, then the total error model of the new system is set up. Computer simulation experiment is carried out to demonstrate the validity of the total error model. Key error of the new system are discussed here, the influences introduced by the interferometric baseline error and the illumination synchronization error are analyzed theoretically and demonstrated by computer simulation.Chapter 4 focuses on the calibration problems of the new system's DEM product. Firstly, the reasons that conventional parameter calibration algorithms for InSAR system has a low accuracy are analyzed, four ways which can improve the calibration accuracy is presented, then an optimized parameter calibration algorithm is proposed, and the basic principles of disposing calibrators are given. Secondly, three typical DEM calibration algorithms based on the ground control points (GCPs) are analyzed, a new DEM calibration algorithm based on the total error model is brought forward, because of the using of the total error model, the algorithm can eliminate the systematic error and suppress the height error of GCPs efficiently. Finally, considering the urgently requirement of the high accuracy global mosaic DEM, a joint calibration method which uses the superposed areas of the adjacent DEMs information is proposed to calibrate several DEMs simultaneously, its product is a mosaic DEM of the whole big scene. The method can calibrate the DEM of scene which has no GCPs, the computer simulation results demonstrate the availability of the method.The optimization design of the system is studied in chapter 5. To minimize the influences of the interferometric parameters error, the limitation of the conventional concept of the optimal baseline is analyzed, and the relationships between the interferometric parameters error and the vertical accuracy, the perpendicular effective baseline and the vertical accuracy as well as the difficulty of the signal processing, the along track baseline and the total coherence are studied, then a interferometric baseline optimization method is proposed. Focusing on the estimation error of the doppler centriod, an elliptic orbit total zero doppler steering method is proposed, furthermore a novel yaw steering method is proposed, those two methods can decrease the doppler centroid to zero by adjusting the attitude of the satellite, the former must adjust the yaw and the pitch attitude at one time, while the latter only need adjusting the yaw attitude, the two methods can restrain the estimation error of the doppler centriod by reducing the doppler centriod to zero. Focusing on the illumination synchronization error, based on the two methods presented ahead, a maximum coherence illumination synchronization method is proposed. Compared with the traditional method, the new method steers the two satellites respectively and decreas the difficulty of attitude and beam-steering cooperation between the two satellites. The new method can enhance the coherence and the number of the independent looks for interferometric phase estimation, and eliminate the influences introduced by the error of interferometric coregistration time and improve the processing efficiency. Because of the zero doppler centroid, at last, the new method can upgrade the DEM precision.