3D-mapping from TerraSAR-X staring spotlight data

Radargrammetry is a well-established technique for deriving digital surface models (DSMs) from synthetic aperture radar images. This work focuses on the novel Staring Spotlight mode of the TerraSAR-X satellite which delivers images with a GSD down to 16 cm. Employing such images, the 2D and 3D geo-location accuracy of the physical sensor models - integrated into SAR range/Doppler equations - is assessed and an automated workflow for DSM extraction is presented. An analysis of the mapping potential completes this study by evaluating the DSM accuracies w.r.t. reference LiDAR data.