Meteorological Radar Data Quality Control Studay And Application

All kinds of radar data quality issues have very serious effects on radar data application,especially on radar data assimilation and quantitative precipitation estimation, so studyingradar data quality problems (non-meteorological echo, especially non-meteorological echocaused by the radar hardware limitations or failure, and velocity aliasing) observed fromoperational weather radar (China New Generation of Weather Radar CINRAD, WSR-88D,WSR-88D dual polarization radar) has important scientific meaning and practicalsignificances. Firstly, analyzes and processes the EMI (Electromagnetic Interference) echoobserved from the CINRAD operation, and then develop "Test Pattern" identificationalgrithom, remove pollution cause by them for both reflectivity and radial velocity data.Analyze aliased radial velocity from hurricane (typhoon) wind, develop fitting vortex modelto aliased velocity method, and generate background vortex wind field, dealiase observedvelocity automatically. Finally, analyze biological echo from polarmetric radar observation,develop birds echo identification method to remove the radial velocity containments causedby birds echo. These works improve the quality of weather radar based data, and provide astrong foundation for the quality of the radar product and its application. The main contentsare summarized as follows:（1）Collect the EMI echo observed from operation CINRAD, aimed at resolving theEMI abnormal echo problem, focus on understanding the graphic display characteristics ofEMI echo, analyze the characteristics of the doppler (CINRAD/SA) radar EMI echo, whichknows as isolation and the antenna-like distribution, combine with current filter methods,simulate the logic of identify EMI echo from human eye, propose an automatic method tofilter out EMI echo, and good results achieved in the experiments by using this method.（2）“Test Pattern” caused by test signal or radar hardware failures in CINRAD SA andSB radar operational observations are investigated. In order to distinguish the “Test Pattern”from other types of radar echo such as, precipitation echo, clear air echo, all other kinds ofclutters, five feature functions including reflectivity effective echo rate Rz, RF (RangeFolding) value rate Rrf, effective echo mismatching rate Rvz, reflectivity gate to gate changealong radial RNaz，reflectivity gate to gate change along azimuth RNrz, are proposed. Based onfuzzy logical method,“Test Pattern” identification algorithm has been developed. Thestatistical results form all kinds of radar echoes indicated the performance of the algorithm.Individual case analysis of “Test Pattern” with heavy precipitation echo inside was showed. （3）A least-squares method is developed to estimate the maximum tangential velocityVMand its radial distance RMfrom the hurricane vortex center by fitting a parametric vortexmodel directly to raw aliased velocities scanned from a hurricane. In this method, aliasing-caused zigzag-discontinuities are formulated into the cost-function via an unconventionalapproach to ensure the cost-function to be smooth and concave around the global minimum.Simulated radar velocity observations are used to examine the cost-function geometry aroundthe global minimum in the space of control parameters (VM, RM).（4）This AR vortex analysis is refined in this paper to estimate VMand RMas functionsof height, so a suitable reference radial-velocity field can be produced on each tilt of radarscan for the reference check in the first step of dealiasing. This upgrades the previouslydeveloped VAD-based dealiasing method adaptively for applications to radar radial velocitiesscanned from hurricanes. The robustness and improved performance of the upgraded methodare exemplified by the results from extensive tests with severely aliased radial velocitiesscanned by WSR-88D radars from hurricanes（602volumn scans） and CINRAD/SA fromtyphoons.（5）Based on the fuzzy logic algorithms, processing the collected birds echo andinsects echo from operation KICT polarimetric radar in the fall of2012respectively bystatistical analysis, then developing birds echo identification algorithm. This technology canbe achieved through the following two steps:(1) using existed polarimetric WSR-88D radarecho classification algorithm to simply divide radar echo into three categories: precipitation,ground clutter and biological echoes.(2) Using birds echoes and insects echo analysis,extracting the polarization characteristic parameters to distinguish them, then examining theecho points that has been determined as biological echo, to identify birds echo then mark itand remove it.