Ground Clutter and Noise Mitigation Based on Range–Doppler Spectral Processing for Polarimetric Weather Radar

Spectral polarization filtering in the range–Doppler domain plays an important role in weather radar clutter mitigation. However, when ground clutter and precipitation overlap, these methods tend to filter out clutter-contaminated precipitation, leading to estimation errors. To address this problem, this article proposes a ground clutter and noise mitigation method based on range–Doppler spectral processing for polarimetric Doppler weather radars. The proposed method can filter out clutter and noise and retain precipitation overlapped with clutter by analyzing the property differences between precipitation and clutter and noise in the range–Doppler spectrogram. Specifically, due to the spatial continuity of precipitation in the range–Doppler domain, the spectral moments (e.g., velocity and spectral width) are also continuous. In addition, polynomial fitting is used to compensate for the missing spectral moments, and the missing precipitation is recovered by Gaussian fitting using the computed spectral moments. The results demonstrate performance improvements after applying the proposed method to radar data collected by Chinese operational weather radars. The proposed method is compared with several other algorithms, and the comparison results show that the proposed method performs best in clutter and noise suppression and precipitation retention performance.