Nonuniform Sampling VAD-Based Nonlinear Wind Field Retrieval and Experiment Validation for Weather Radar

Wind profile information can be used to study high-altitude background wind field or to monitor near-surface strong winds, which plays an important role in revealing the characteristics of large-scale weather systems. Velocity-azimuth display (VAD) is the most widely used method for wind profile retrieval in Doppler radar, in which the linear wind field is assumed, and the Doppler velocity information uniformly sampled at equal azimuth intervals is used. However, for severe convective weather systems such as typhoons and tornadoes, the spatial distribution of wind fields is complex and the assumption of a linear wind field may be not satisfied. Besides, when the weather system is not evenly distributed around the radar or concentrated in a certain direction, the uniform sampling at equal azimuth intervals cannot be guaranteed, resulting in the decrease of retrieval accuracy. In this article, a wind profile retrieval method based on a nonuniform sampling VAD (nVAD) technique is proposed. By establishing a nonlinear spatial distribution model of complex wind fields and using the radial Doppler velocity information nonuniformly sampled at azimuth, the traditional VAD can be modified, and the wind profile can be obtained using the numerical analysis where the radar observations are acquired at a set of different elevation angles. The nVAD technique is not limited by the truncated Taylor expansion and the nonlinear wind field can better represent the real complex wind field structure, and the nonuniform sampling method is more suitable for the case with missing measurements. The verification results show that the nVAD technique can effectively retrieve the wind profile for a complex wind field when the Doppler velocity is sampled nonuniformly and has a better performance than the traditional VAD technique.