A Method for Retrieving Mean Horizontal Wind Profiles from Single-Doppler Radar Observations Contaminated by Aliasing

Abstract

The velocity–azimuth display (VAD) technique was designed to estimate the areal mean vertical profile of the horizontal wind above a ground-based Doppler radar. The method uses radial velocity observations under the assumption of a linear wind field, though it encounters difficulty when the observations are contaminated by velocity ambiguities, large noise, and when viable data exist only over a restricted azimuthal range. The method suggested in this paper uses gradients of radial velocity, rather than only the velocity itself, to derive wind profiles and thus is termed the gradient velocity–azimuth display (GVAD) technique. Both the VAD and GVAD methods are tested first on simulated data to examine their sensitivity to different type of errors in radial velocity. The retrieved mean wind profiles are shown to be insensitive to random errors in radial velocity, even at large amplitude. However, the VAD method is very sensitive to systematic errors caused by velocity ambiguities. The experiments indicate that if only 3% of a full-volume scan of radial wind data is contaminated by aliasing errors, the relative rms error in the mean wind profile retrieved by VAD can reach 50%. In contrast, GVAD is very robust to such errors. Application of GVAD to Weather Surveillance Radar-1988 Doppler (WSR-88D) data collected during the 3 May 1999 tornado outbreak show that it has the ability to obtain accurate wind profiles even when the observations contain large errors caused by velocity ambiguities and random noise.