• temporal decorellation
• robust interpolation
• Correlative Imaging
• 3D SAR Tomography

This thesis deals with the decorrelation problems affecting 3D Tomography techniques when applied to natural scenarios such as cultivated areas, forest and hilly terrain. After a brief introduction about the decorrelation sources, the coherence maps for the examined areas of Rome and Naples, also including some natural areas, are studied. Since C-band data from ERS1/2 spaceborne missions are used, the temporal decorrelation is the main source of decorrelation for our datasets. Thus we try to overcome this issue by applying a classical Tomography processing to a subset of data with a reduced time span. A different kind of approach is also applied, which is potentially more effective, named 3D SAR Correlative Interferometric imaging. This advanced technique exploits the multiple spatial autocovariance data which are less affected by temporal decorrelation to reconstruct the scatterers elevation properties. Experiments with both the techniques are carried out for the two datasets. The first positive results obtained by means of the Correlative imaging pave the way to further developments of this technique.