Snow Detection and Estimation from Ku- and Ka-Band Dual-Wavelength Radar

Ku- and Ka-band dual-wavelength radar data have shown promise in accurately inferring microphysical properties of snow. Studies show that the median volume diameter of snow can be reasonably derived from dual-wavelength measurements, and that the estimates are nearly independent of the snow density. The Global Precipitation Measurement (GPM) Dual-frequency Precipitation Radar (DPR) has been proposed to globally map precipitation (rain and snow) in conjunction with onboard passive microwave sensors. As the GPM extends precipitation measurements to mid- and high-latitude where snow becomes a larger fraction of the total precipitation, it is imperative to develop dual-wavelength radar algorithms for estimates of snowfall rate and parameters of the size distribution. Toward this end, we will focus on several issues in this study that are associated with radar algorithms for snow retrieval. To apply a radar algorithm for estimating snowfall rate, an important question is whether proper identification can be made of the precipitation. Use of radar reflectivity at one frequency alone is not sufficient to distinguish snow from rain echoes because light rain exhibits a similar range of reflectivities as snow. However, the differential frequency ratio (DFR) between Ku and Ka bands, provides useful information that might be used to distinguish snow and rain. For this study we will perform radar simulations of snow at Ku and Ka bands and consider in particular, whether solving the dual-wavelength equations under the assumption of rain leads either to no solution or physically unrealistic results. The measurements made by the JPL dual-frequency airborne Ku and Ka band radar during the Wakasa Bay AMSR-E validation campaign over the Sea of Japan on board a NASA P-2 aircraft, will be used as test data to assess the method. In addition, the accuracy of the dual-wavelength radar technique based on DFR and radar reflectivity will be examined for snow retrieval in terms of snow density that is assumed either fixed or variable with particle size. Other issues, such as scattering properties of non-spherical snow particles and their impact on the retrieval as well as snow attenuation, particularly at Ka-band, will also be addressed in our study.