On the Radial-Based Estimation of Colored Noise Power in Weather Radars

A radar antenna captures thermal radiation emitted from diverse sources such as the ground, the Sun, the sky, precipitation, and man-made radiators. In the radar receiver, this external radiation generates noise that combines with the receiver’s internal noise, contributing to the overall system noise. Consequently, the power of the system noise is contingent on the antenna beam’s position. Inaccurate measurements of noise power can result in reduced coverage if the power is overestimated or in radar data images cluttered by noise speckles if the power is underestimated. In addition, employing an erroneous noise power in estimators at low to moderate signal-to-noise ratios (SNRs) can introduce bias in meteorological variables. Thus, to ensure the highest quality radar products, it is essential to compute meteorological variables using noise power measured from data collected at each scanning direction. A technique that achieves this has been developed. It employs a set of probability-based criteria to accomplish this goal. However, a critical assumption of this technique is that noise-only time series at adjacent range locations are independent. In radar systems, this assumption may be flawed due to the characteristics of digital signal processing in the receive path resulting in colored noise. This article introduces an extension to the existing technique, which facilitates an accurate estimation of colored noise power.