Dual-wavelength Doppler spectra radar techniques for rain and dynamics retrievals with the ARM Ka-W radars

Description

Against common expectations, millimetre wavelength cloud radars are well capable of quantifying precipitation via their attenuation signals. For zenith observations, the vertical gradient of Ka-band reflectivity in stratiform precipitation can be used to retrieve rain-rate profiles. However, in presence of vertical variability of hydrometeor profiles, the vertical gradient of reflectivity is not driven by attenuation only, leading to inaccurate retrievals. To avoid such a limitation, the cloud radar reflectivity can be compared with non-attenuated Rayleigh reflectivity measured by a collocated centimetre wavelength radar. However, for improving the retrieval accuracy, the Mie effects - i.e., the effects related to the wavelength dependence of the radar backscattering by rain drops - must be known as well. Furthermore, because of the wide difference in the transmitting frequency, it is very challenging to match the beam widths of a millimetre and centimetre wavelength radar, thus introducing further uncertainty to the problem. Another possibility is to use the recently developed dual-wavelength Doppler Spectra Ratio (DSR) technique for two collocated millimetre radars at Ka- and W-band. This technique is capable of disentangling the Mie and attenuation effects and opens the way to various methods for the characterization of rainfall. For example, the turbulence is known to broaden the measured Doppler spectra, and hence to alter the shape of the DSR. The effectiveness of the DSR technique has been demonstrated under low turbulence conditions where the DSR reduces to a universal shape. For higher turbulence, a deconvolution technique of the DSR can be used to retrieve the amount of turbulence. Furthermore, after the deconvolution of the turbulence, the full bin drop size distribution can be derived, without the need of using parametrized DSD. Likewise, this methodology can also provide an estimate of the vertical wind by recognizing the features known to correspond to a specific diameter, and hence to a specific fall velocity, in a fashion similar to the W-band spectral method, but with the possibility of extending the range of applicability to lighter rain scenarios. Finally, a variational approach can be used to consistently retrieve all these parameters. The recent advances in this field will be presented.