A Study On Operational Application Of Doppler Radar Data

Firstly, This study reviews past attempts to mitigate ground clutter contamination of radar data resulting from anomalous signal propagation, and presents a new algorithm for radar data quality control. The new automated procedure has been developed that makes use of the three-dimensional reflectivity structure. In particular, the vertical extent of radar echoes, their spatial variability, and vertical gradient of intensity are evaluated. Then, the paper aims at Doppler velocity dealiasing. A operational algorithm provided by WSR-88D is introduced. Because of its defaults, an additional process is applied to the first gates.Secondly, three storm automatic identification algorithms by Doppler radar are discussed. The first one provided by WSR—88D Build 7.0 (B7SI) tests the intensity and continuity of the objective echoes to build three-dimensional storms. When storms are merging, splitting, or clustered closely, errors may occur in B7SI. The second algorithm (B9SI) is part of the Build 9.0 Radar Products Generator of the WSR-88D system. It uses multiple thresholds of reflectivity, designs the technique of cell nucleus extraction, and processes the close storms. So B9SI is capable of identifying embedded cells in multi-cellular storms. But, B9SI can't give information on storm convection strength, because texture and gradient of reflectivity are not calculated and radial velocity data are not used. Then, the third algorithm (CSI) is addressed detailedly. By using fuzzy logic technique, CSI processes radar base data and the output of B9SI, in which the levels of the seven reflectivity thresholds are lowered, to obtain storm convection index. For the CSI algorithm, a set of features is combined to describe the convective characteristics of storm, and each feature is given a weight. These features include texture and gradient of reflectivity, VIL, and standard deviation of radial velocity. Then, the likelihood values that the features match the objective storm are calculated by the linear membership functions, which are from the feature field histograms of the historical data. Finally, the convection index is the weighted average of all the likelihood values.Thirdly, the basic technique, functions and key parameters of the Regional CINRAD Mosaic System is documented in this paper. And it is introduced that the system was applied to the severe convection weather and typhoon detecting and warning in Guangdong province recently. Thehorizontal wind field in the meso-scale weather system can be derived by applying the TREC technique to mosaic products.Finally, a fuzzy logic radar echo classifying scheme for CIN-98SA is designed, and it consists of four data fusion algorithms including the AP Detection Algorithm(APDA), the Precipitation Detection Algorithm(PDA), the Insect Clear Air Detection Algorithm(ICADA) and the Sea Clutter Detection Algorithm(SCDA). For each algorithm, a set of features is combined to describe the characteristics of the objective echo type, and every feature is given a weight. These features include mean, median, standard deviation, texture in reflectivity, radial velocity or spectrum field. Then, the likelihood values that the features match the objective echo type are calculated by the linear membership functions, which are from the feature field histograms of the history data. By comparing the chosen threshold and the weighted average of the likelihood values, whether the observed echoes are the objective type is decided.