Application of the Zhang–Gal-Chen Single-Doppler Velocity Retrieval to a Deep Convective Storm

Abstract

The Zhang?Gal-Chen single-Doppler velocity retrieval (SDVR) technique is applied to a multicell storm observed by three radars near the Orlando, Florida, airport on 9 August 1991. This dataset is unique in that 3-min volume scans at very high spatial resolution (200 m) are available during a 24-min period. The retrieved (unobserved) wind, determined using only the radial wind and reflectivity from one of the radars, is compared to the (observed) winds obtained from a hybrid three-dimensional wind synthesis. Error statistics demonstrate that the retrievals perform best when applied in a reference frame moving with the storm; however, the results also show that the specification of this frame is problematic. The findings also indicate that, in an environment where the mean flow has a critical layer, the moving reference frame is best defined as a function of height rather than a volume mean. The benefit of such a reference frame is case dependent and is best realized in regions such as a surface cold pool or upper-level divergence at storm top. Error statistics demonstrate that the SDVR technique recovers the horizontal wind with greater accuracy than it does the vertical velocity?suggesting that for deep convection, the absence of dynamical constraints is critical. The kinematic and O?Brien techniques and a new variational technique, in which the solution to a second-order ordinary differential equation for the vertical velocity is expressed in terms of Bessel functions, are tested as possible alternatives to the SDVR vertical velocity. Results indicate that this new technique yields vertical velocities significantly better than those using the other three methods.