Radiometric Calibration and Classification of Airborne SAR Data from the CCRS-SAR580 Italian Campaign

This paper presents some results from end-to-end radiometric calibration of C-band Synthetic-Aper­ ture Radar (SAR) data gathered by the Canada Centre for Remote Sensing (CCRS) SAR580 instrument during a campaign in Southern Italy. The procedures needed to relate the pixel digital number to the backscattered energy are described, from raw data compression to the evaluation of the backscattering coefficient sigma0 on the imaged area, containing 38 triangular trihedral Corner Reflectors (CR). The image quality tests performed on the focused data show satisfactory processor performance in terms of resolutions and sidelobe ratios. The intensity integration method is applied to evaluate the energy terms associated with the point targets and the background clutter and, successively, to calculate the average backscattering from the CR area. The accuracy in the estimate of sigma0 is found to be ±0.52 dB, after removing antenna pattern, range-dependent terms and the gain law used for sensitivity time control. A self-organising clustering algorithm is applied to some of the areas centred around each corner reflector, in order to obtain histograms of scattering properties of bare and ploughed soil. The comparison with radar data taken from a database ofscattering statistics gives results in good agreement with the sigma0-histograms derived from the collected SAR data.