The development of a 3-pass persistent scatterer algorithm using the integer ambiguity search method

Differential interferometric synthetic aperture radar (InSAR) is a well used technique for measuring deformations, but often suffers greatly from effects due to atmospheric differences occurring between different SAR images. Recently a new set of techniques have been developed called Persistent Scatterer techniques, which take advantage of the high number of SAR images available to try and model out the atmospheric effects. To aid this process, all present techniques make use of a Digital Elevation Model (DEM) to remove the interferometric phase due to topography, but as a consequence this contaminates the phase values with an unknown error due to the DEM, which has to be modelled out in the processing chain. In this thesis a new Persistent Scatterer (PS) method is proposed that does not use a DEM to remove the topographic phase component, but rather one of the interferograms used in the study, and hence does not need to calculate topographic height corrections. This is achieved by using the Integer Ambiguity Search (IAS) 3-Pass differential technique. The developed algorithms are then tested and assessed over two test sites in central London, UK. The overall conclusions of the research are summarised below. The IAS 3-pass differential interferometry method gives a differential result that is more consistent with 2-pass results than with traditional 3-pass results. By using the IAS 3-pass method, it is possible to do a PS InSAR analysis without recourse to a DEM or needing to derive any topographic information. The developed IAS PSInSAR algorithms have been tested using simulated data, which has shown that the methods developed can identify small scale target motion over a period of a few years. The IAS PSInSAR algorithms were also tested using real SAR data, the results of which are consistent with GPS results of the test site and previous independent investigations.EThOS - Electronic Theses Online ServiceGBUnited Kingdo