A technique for measurement of spaceborne SAR antenna patterns using distributed targets

Abstract

A method to measure the antenna elevation pattern of synthetic aperture radar (SAR) using distributed targets is introduced. The features of the method are (1) the antenna pattern model parameters are estimated as the solutions of maximum likelihood estimation; (2) To guarantee the uniformity, a screening process based on a chi-square similarity test is applied to the image data; and (3) Noise generated in the SAR receiver and data processor tends to broaden the estimated antenna pattern. To improve it, the estimated noise level is subtracted from the image data. The authors assume that the scattering coefficient over the evaluated images is unknown yet constant (although this could easily be extended to the case where the variation of /spl gamma/ or /spl sigma//spl deg/ with incidence angle is known), Three types of antenna pattern models are tested, Among them, the best result is given by a fourth order power model. This technique is applied to selected image data sets from the SIR-B mission, including several scenes analyzed previously by Moore from the Amazon rain forest and Illinois farmland. For the Amazon data (which give the best results), the authors found that the technique adopted in the paper gives residual errors on the antenna pattern fit of less than 0.08 dB for a given scene. Applying the antenna pattern estimated from one scene to others, residual errors of less than 0.3 dB are achieved.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>