Mixed Models Analysis of Radar Residuals Data

Mixed models provide a novel approach to the analysis of radar tracking residuals by considering randomness from different sources. Through properly considering randomness, mixed models can provide greater power to determine the statistical significance of various parameters needed for radar calibration. This paper applies a mixed models approach to the analysis of radar tracking residuals from calibration satellites observed by the Cobra Dane radar and finds a time dependent bias in the azimuth residuals.