Abstract: This paper presents two methods for calculating the variance of the radar measurements and the target position error in the Cartesian coordinate, based on the standard deviation of the measurement data in the radar polar coordinates. They are referred to as the standard deviation method of transferring the standard deviation of each coordinate component of the target in the uniform rectangular coordinate system to the indirect measurement deviation of the target position, and the statistical method based on the estimation of the straight track line parameters, respectively. The experimental verification combined with some correlation filtering approach shows that the two methods can be successfully applied to obtain the dynamic weight distributions for radar measurement in the multi-radar weighted track fusion model with desirable results, and also have a high engineering value for improving the traditional filters such as the Kalman filter and the multi-radar track correlation algorithm.