Application of matched statistical filters for EarthCARE cloud doppler products

This paper presents a method for filtering the random noise that affects spaceborne Doppler measurements of atmospheric velocities. The proposed method hinges on adaptive low-pass filters that apply to the measured pulse-pair correlation function. The parameters of the filters are found by optimizing the statistics of the velocity residue of the filter. The method is illustrated by simulations of the cloud-profiling radar of the future Earth Cloud, Aerosol and Radiation Explorer (EarthCARE) mission of the European Space Agency and the Japanese Space Exploration Agency. These simulations, which do not include strong convection, show the higher performance of the filters when compared with the traditional increase of the along-track integration length. The results obtained with the filters show that velocity accuracies of 0.48, 0.42, and 0.39 m · s^-1 are achievable at PRF = {6.1, 7, 7.5} kHz, respectively, while preserving the initial 500-m sampling of the measured EarthCARE data. These results also show the potential benefits of avoiding excessive alongtrack integration, for postprocessing tasks such as dealiasing or the retrieval of the vertical distribution of the atmospheric velocity (e.g., longer than 5 km for cases consistent with the climatologies represented in this data set).