Utilizing Spaceborne Radars to Retrieve Dry Snowfall

Abstract

A dataset consisting of one year of CloudSat Cloud Profiling Radar (CPR) near-surface radar reflectivity Z associated with dry snowfall is examined in this study. The CPR observations are converted to snowfall rates S using derived Ze?S relationships, which were created from backscatter cross sections of various nonspherical and spherical ice particle models. The CPR reflectivity histograms show that the dominant mode of global near-surface dry snowfall has extremely light reflectivity values (?3?4 dBZe), and an estimated 94% of all CPR dry snowfall observations are less than 10 dBZe. The average conditional global snowfall rate is calculated to be about 0.28 mm h?1, but is regionally highly variable as well as strongly sensitive to the ice particle model chosen. Further, ground clutter contamination is found in regions of complex terrain even when a vertical reflectivity continuity threshold is utilized. The potential of future multifrequency spaceborne radars is evaluated using proxy 35?13.6-GHz reflectivities and sensor specifications of the proposed Global Precipitation Measurement dual-frequency precipitation radar (DPR). It is estimated that because of its higher detectability threshold, only about 7%?1% of the near-surface radar reflectivity values and about 17%?4% of the total accumulation associated with global dry snowfall would be detected by a DPR-like instrument, but these results are very sensitive to the chosen ice particle model. These potential detection shortcomings can be partially mitigated by using snowfall-rate distributions derived by the CPR or other similar high-frequency active sensors.