Validation of Wind Fields Retrieved by Dual-Doppler Techniques Using a Vertically Pointing Radar

Dual-Doppler techniques for 3D wind retrievals have been widely used for some time, but very little effort has been dedicated to their validation. In particular, the vertical component of the wind field w is not well sampled by the radars due to the scanning strategies used for weather radars, and relies heavily on imposed conditions such as the mass continuity equation and the vorticity equation. Therefore, validation of the retrieved w with independent observations remains important before using the retrievals for the study of weather systems dynamics. The polarimetric Doppler radars JuXPol and BoXPol perform volume scans followed by a Range Height Indicator (RHI) scan and a vertically pointing scan. In this study we use the latter to derive w above the BoXPol radar and to compare it to the vertical component of the wind field retrieved by the Dual-Doppler technique. Both profiles of w show similar overall shapes with minima and maxima situated at about the same heights. The values of w obtained from the vertical scan are within the range of the profiles obtained with the Dual-Doppler method, except above an altitude of 4km where the vertical scan shows higher upward velocities than the Dual-Doppler. This can be caused by a tilted updraft which can only be partly sampled by the vertical scan, and results from different sampling times. Bias and root mean square error indicate a relative good agreement between both estimates especially for the median and high values of w, but an underestimation of the Dual-Doppler w at the lowest values.

1. Introduction

Dual- and Multi-Doppler techniques have been long used for the retrieval of 3D wind fields in regions covered by at least 2 radars (e.g. [1], [2]). The principle of Dual-Doppler wind retrievals involves geometry consideration as radial velocities from different radars are combined, but also makes use of the mass continuity equation and/or the vorticity equation. The latter constraints are needed because in the case of two radars the problem of retrieving the three spatial components of the wind vector is underdetermined with only two measurements. In the case of three or more radars, these constraints help in particular for the retrieval of the vertical wind component that is poorly sampled due to the relatively low elevation angles (typically lower than 20°) that most weather radars use for their volumetric scans.