Effects of along-track integration on Doppler velocity bias with a spaceborne cloud-profiling radar

Doppler operation of spaceborne cloud radar is useful for characterization of cloud convective motions, estimation of sedimentation velocity in cirrus clouds, and quantification of drizzle fluxes in stratocumulus. EarthCARE (Earth Cloud, Aerosol, and Radiation Explorer) CPR (Cloud-Profiling Radar) is the first spaceborne cloud radar with Doppler operation, and will contribute to cloud, aerosol and climate studies through measurement of radar reflectivity and vertical velocity of clouds globally. For spaceborne radar, if distribution of cloud and precipitation within radar beam is not uniform (NUBF; Non-Uniform Beam Filling), observed Doppler velocity has Bias, and sometimes aliasing is caused. For EarthCARE CPR, pulse-pair operation will be adopted for measurement of Doppler velocity. Along-track integration of 500m-10km is calculated in order to reduce random error. In this study, effects of along-track integration on NUBF-induced bias are evaluated, and availability of larger along-track integration for correction of aliasing is shown.

Doppler velocity bias is simulated for various length of along-track integration under simple scenes of radar reflectivity and vertical velocity. Initially, scene that only radar reflectivity is changed but velocity is uniform is assumed. Positive (upward) bias appears when large reflectivity area exists at front of radar beam. When length of along-track integration is larger, NUBF-induced bias gets smaller but position of the bias moves ahead. NUBF-induced bias is proportional to gradient of radar reflectivity, as previous studies showed. Next, scene that only vertical velocity is changed but radar reflectivity is uniform is assumed. When vertical velocity changes from zero to downward, apparent upward velocity appears, which original cloud does not have. When vertical velocity changes from downward to zero, much more downward velocity than that of original cloud appears. This apparent velocity is not symmetric between increasing and decreasing areas of vertical velocity. Such apparent velocity is reduced when length of along-track integration is larger. Finally, assuming the existence of precipitation area between non-precipitation clouds, scene that both radar reflectivity and vertical velocity are changed is considered. Apparent velocity is also non-symmetric between front and rear sides of precipitation area. Sometimes aliasing occurs only at rear side of precipitation. When length of along-track integration is larger, aliasing is reduced. This indicates that longer along-track integration can be used as a reference of aliasing correction. Furthermore, simulation of Doppler velocity bias based on realistic cloud dataset obtained by 95-GHz ground-based radar (SPIDER) will be discussed.