Reliable remote detection of cloud liquid water content (LWC) and droplet sizes is important both for understanding cloud microphysics and meteorological processes and for providing improved diagnostics of icing potential for the aviation community. One particularly promising technique makes use of the ratio of reflectivities measured by co-located millimeter radars operating at different wavelengths, such as Ka and W-band or X and Ka-band radar pairs. This dual-wavelength ratio (DWR) is used to obtain signal attenuation and thereby infer the liquid water density along the radar beams. Unfortunately, useful implementation of the DWR technique has been hindered by its sensitivity to reflectivity measurement error, Mie scattering, and artifacts due to mismatched radar beam sizes and radar locations. The retrieval of reasonable LWC values has usually required significant averaging, which seriously degrades the resolution of the retrievals.

The present paper elucidates the mechanisms responsible for contaminating the DWR retrievals, describes appropriate censoring and smoothing techniques for mitigating this contamination, and presents methods for boosting the resolution of the smoothed values to nearly the resolution of the raw reflectivity measurements. A fuzzy-logic approach is used to assign and propagate confidence weights throughout the processing. Results from simulations are used to illustrate the artifacts generated by beam mismatches, and data from the Mount Washington Icing Sensor Project (MWISP) and the Alliance Icing Research Study (AIRS) field programs are used to demonstrate and validate the performance of the enhanced DWR retrieval technique.

This research has been performed at the National Center for Atmospheric Research under the auspices of the Federal Aviation Administration's Aviation Weather Research Program.