High resolution PolInSAR with the ground-based SAR (GB-SAR) System: measurement and modelling

Abstract

Ground-based work is necessary for a comprehensive assessment of the operational potential and limitations of PolInSAR in airborne and satellite SAR applications. A study is made of the performance and usefulness of the UK’s Ground-Based SAR (GB-SAR) Outdoor System in high-resolution PolInSAR studies of vegetation using modeling results. The facility provides fully-polarimetric L- through X-band imagery down to a resolution of several wavelengths. However, the measurement process is slow in relation to pulsed systems as it requires the antenna head to be mechanically scanned across an aperture. The PolInSAR technique requires high coherence between interferometric image pairs, and the long data acquisition times raise the question of temporal decorrelation. We developed two models incorporating motion, a physics-based model and a signal processing model. The former incorporates a PolInSAR crop simulator employing the distorted Born approximation, applied to a simulated canopy of wheat plants based on field-collected physiological measurements. GB-SAR simulations of mature wheat canopies suffering a range of wind-blown disturbances are examined for coherence stability. These calculations permit the analysis of the behaviour of coherence with system and canopy descriptive parameters, such to quantify the suitability and performance of measurement environments for PolInSAR analysis. The models indicate that clutter motion will degrade interferometric performance both during aperture formation, and between repeat-pass observation. However, we conclude that the GB-SAR system will be robust to small amounts of clutter motion and will serve as a suitable tool for PolInSAR experimental studies.