| Profiling of Clouds | >> |
Abstract
In this contribution we consider Doppler spectra, received with a vertically pointing millimeter-wave radar (35GHz), to retrieve information on microphysical structure of clouds and dynamics of cloud systems.
The Doppler spectra are determined by the velocity and the scattering cross section of all the particles present in the radar sampling volume. The structure of the spectra has a complexity that arises from the variegate collection of particles forming clouds: ice crystals of different habits and size, or ice crystals mixed with cloud droplets (mixed-phase clouds). Moreover, vertical air motions influence the particles falling velocity.
As we assume that every class of cloud particles coexisting in the sampling volume produces a Gaussian spectrum, we develop an algorithm that fits the measured spectrum with a linear superposition of Gaussian curves. The moments of the Doppler spectrum, used to retrieve cloud properties, are then evaluated for every class of cloud particles by associating them with the parameters of the relative Gaussian curve.
Several clouds analysed in this study show Doppler spectra split up into two classes above the melting layer, which indicate the existence of mixed phases or separate habits of ice crystals. The observation of coherent multiple peak structures in the spectra profiles may be used for mixed-phase cloud studies and for model validation.
Between cloud top and melting layer/cloud base, the fall velocities have amplitudes increasing with decreasing height. We consider stratiform cloud systems, for which it is possible to minimize the effect of the vertical air motion. We evaluate the gradient of the vertical velocity as a function of altitude and the evolution of this gradient with time. We find that the vertical velocity increases with an average of about 10 cm/s per km fall path, that can be explained as consequence of particle-growth due to various microphysical interaction processes. If further measurements confirm that the observed vertical velocity gradient in clouds is of general validity, it should be reproducible by clouds resolving models with explicit microphysics.