A STATISTICAL-MODEL FOR THE ERROR-BOUNDS OF AN ACTIVE PHASED-ARRAY ANTENNA FOR SAR APPLICATIONS

Radiometric calibration of a SAR-image is achieved by monitoring all the relevant parameters of the radar equation. The precision and accuracy of the instrument are limited by the errors made in monitoring these parameters. Whereas most parameters can quite readily and constantly be monitored, monitoring of the array pattern during the flight is a cumbersome and time-consuming affair, and, therefore, mostly omitted. Random variations in the antenna pattern can thus be expected to have the greatest influence on the uncertainty of the scatter-coefficient estimate. Obviously, the effect of a phased array antenna, of which the antenna pattern is formed by the amplitude and phase of several individual T/R-modules fed to radiating elements, will be even worse. This paper reviews a model for the theoretical error bounds for the radiometric calibration of SAR imagery. The model is then applied utilizing the radar system parameters as will be used in the project PHARUS (acronym for PHased ARray Universal Sar) [1], a Dutch polarimetric airborne C-band universal SAR, which is currently under construction. An error model for the phased array antenna pattern will be presented. This model was applied to a 16x8 phased array antenna to determine the influence of errors in the T/R modules and angle variations of the beam direction.