Nonlinear Wind Analysis of Single-Doppler Radar Observations within a DVAD Framework

Abstract

AbstractThe application of the distance velocity azimuth display (DVAD) method to the retrieval of vertical wind profiles from single-Doppler radar observations is presented in this study. It was shown that Doppler velocity observations at a constant altitude can be expressed as a single polynomial function for both linear and nonlinear wind fields in DVAD. Only a one-step least squares fitting of a polynomial function is required to obtain the vertical wind profile of a real wind field. The mathematic formulation of DVAD results in two advantages over the traditional nonlinear VAD method used for the nonlinear analysis of single-Doppler observations. First, the requirement of only one-step least squares fitting leads to robust performance when Doppler velocity observations are contaminated by unevenly distributed data noise and voids. Second, the degree of nonlinearity to properly represent a real wind field can be directly estimated in DVAD instead of being empirically determined in the traditional method. A proper nonlinear wind model for approximating the real wind field can be objectively derived using the DVAD method. The merits of DVAD as a quantitative single-Doppler analysis method were compared with the traditional method using both idealized and real datasets. Results show that the simplicity and robust performance of DVAD make it a good candidate for single-Doppler retrieval in operational use.