A statistical model for prediction of precision and accuracies of radar scattering coefficient ( sigma degrees ) measurements derived from SAR data

Abstract

A model is presented to estimate a relative error-bound associated with radiometric calibration of the scattering coefficient derived from synthetic-aperture radar (SAR) data. This error bound is based on a statistical coefficient-of-variation error model. The error model was exercised parametrically to determine what factors most significantly influence radiometric errors from SAR systems. It was found that errors in the measurement of the antenna elevation gain-pattern resulted in the most dramatic increases in the statistical error bound. The results from this analysis can be utilized in both the design of SAR systems and measurements programs needed to calibrate SAR systems to minimize the error bounds associated with radiometric calibration.<>