Use of geometrical SAR simulation for visibility prediction: application to mission planning and urban study

The specification of future radar satellites and many remote-sensing applications need information about the geometrical distortions seen in a SAR image. The main geometrical distortions are those that are called “shadow” and “lay-over”; it is very difficult to quantify them with a lone SAR image, but the use of many SAR images allows a numerical evaluation of these distortions. Unfortunately not only is this procedure very expensive, but also, before the launch of RADARSAT only SIR C produced satellite multiincidence radar images. On the contrary, digital elevation models are easy to access, for a large variety of countries and landscapes. The authors show the interest of geometrical SAR simulations in order to estimate the probability of geometrical distortions and extract practical information concerning visibility. They describe their approach of SAR simulation, based on an adaptation of ray-tracing to radar geometry, and expose two specific applications. The first one is used to specify the best incidence angle for the simulation with a representative DEM base (100000 km²), and the second one is used to predict the visibility of urban objects with specific satellites. This work is based on incidence angle specification for RADARSAT III, and emphasizes the radar potential for urban study