Multi-beam Raindrop Size Distribution Retrievals on the Doppler Spectra: Influence of Averaging and Mean Horizontal Wind Correction

Acquiring the raindrop size distribution (DSD) from radar data is still a challenge. For profiling radar, this distribution can be estimated from the Doppler spectra. However the Doppler spectrum is not a direct measure of the DSD. The radial component of the wind shifts the Doppler spectrum related to the raindrop size distribution along the Doppler velocity interval. Furthermore, the Doppler spectrum may be broadened by turbulence effect. The Doppler spectra of rain are modelled using Rayleigh scattering, the gamma distribution for the raindrop size, a size-shape relationship and a Gaussian kernel for the turbulence. Comparing the measured Doppler spectrum with the modelled one, a non-linear least-square fit technique is employed to obtain the parameters of the raindrop size distribution (D0, μ) and the turbulence broadening factor of the raindrop size spectrum (σ0). The intercept parameter (Nw) and the radial wind component (v0) are estimated by scaling the broadened DSD along the spectral reflectivity axis and the Doppler velocity axis respectively. This approach has been proposed in [1]. So far, this technique has not been further developed, exploited and validated. It is our intention to investigate this methodology using the Sband Doppler-polarimetric TARA radar to retrieve raindrop size distribution from drizzle to heavy precipitation with high spatial and time resolution. Since TARA can profile in three directions, three raindrop size distribution profiles are estimated, which can give insight in the microphysical homogeneity of the precipitation. Because TARA can be used as wind profiler, the mean horizontal wind is estimated and a correction to remove the effect of the radial mean horizontal wind is implemented in the retrieval procedure. Comparison with and without correction is done. Another issue is averaging. Should the Doppler spectra be averaged before the non-linear least-square fit? How should they be averaged and how many averages should be carried out? These questions are discussed. High resolution multi-beams raindrop size distribution retrievals are illustrated and discussed. 1. MICROPHYSICAL MODEL OF RAINDROPS The Doppler spectrum, or spectral reflectivity sZHH(v), can be expressed as the Doppler spectrum resulting from the raindrop size distribution, sZHH( ), convoluted by a Gaussian-shaped kernel of width σ0 modelling spectral broadening. Spectral broadening has several causes, among them turbulence and wind shear in the radar resolution volume. v