A Dealiasing Scheme of Doppler Radar Velocity Based on Polynomial Fitting

A new 2-D dealiasing scheme has been developed to correct the aliased velocities in weather radar data. The new dealiasing scheme first chooses a reference range circle (i.e., the initial reference circle) based on a polynomial fitting of the radial velocities along the azimuth. The fit velocities in the initial reference circle are used as the reference velocities to correct velocities in the two neighboring circles. Subsequently, starting from the three reference circles, the new algorithm checks for continuities among adjacent gates and corrects the gates with a velocity difference of more than 0.8 x (Nyquist velocity). Finally, an overall dealiasing is performed to correct isolated velocities and residual unprocessed velocities. The performance of the new dealiasing scheme is compared with that of the old algorithm (i.e., dealiasing starting from the selected initial reference radial) in several complex volume scans. The results show that the new algorithm performs better, especially in isolated aliased velocities, in several large velocity-missing gaps in the azimuth, and in reducing the propagation of erroneous dealiasing. A statistical survey of 2630 volume scans measured by the radar in Longyan, China from May to August 2016 demonstrated that the new algorithm failed to recover the aliased velocities in only 23 cases (0.87%), while the old algorithm failed to correct the aliased velocities in 38 cases (1.4%), further confirming the improved performance of the new algorithm.