Núria Virginia Devanthéry Arasa
Persistent Scatterer Interferometry (PSI) is a group of advanced differential interferometric SAR techniques that are used to measure and monitor terrain deformation. Different PSI techniques have been proposed in the last two decades. In this thesis, the two PSI chains implemented and used at the Geomatics division of CTTC are described: the local area PSI and the PSIG chains. The first part of the thesis is devoted to the local area PSI chain, used to analyse the deformations over small areas. The chain includes a linear deformation model to directly deal with interferometric wrapped phases. Moreover, it does not directly involve the estimation of the APS, thus simplifying the procedure and its computational cost. The chain has been tested using different types of SAR data. The availability of high resolution X-band SAR data has led to an improvement of the PSI results with respect to C-band data. The higher image resolution and phase quality implies an increase of the PS density, an improvement in the estimation precision of the residual topographic error and a higher sensibility to very small deformations, including the displacements caused by thermal dilation. An extension of the classical PSI linear deformation model has been proposed, to account for the thermal dilation effects. This allows obtaining a new PSI outcome, the thermal dilation parameter, which opens new interesting applications since it provides information on the physical properties of single objects, i.e. the coefficient of thermal expansion, and the static structures of the same objects. The second part of the thesis describes the PSIG chain, whose aim was to extend the interferometric processing to wider areas. The ability to cover wide areas is essential to obtain a unique and consistent deformation monitoring for the available SAR image full scenes, i.e. typically 30 by 50 km for TerraSAR-X, 40 by 40 km for CosmoSkyMed and 100 by 100 km for ASAR ENVISAT and ERS. This is particularly important for the forthcoming C-band Sentinel SAR data that will cover 250 by 250 km with a single image scene. The key steps of the PSIG procedure include a new selection of candidate PSs based on a phase similitude criteria and a 2+1D phase unwrapping algorithm. The procedure offers different tools to control the quality of the processing steps. It has been successfully tested over urban, rural and vegetated areas using X-band PSI data. The performance of the PSIG chain is illustrated and discussed in detail, analysing the procedure step by step.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados