Statistical Characteristics Of Sea Clutter And Identification Of Sea Clutter With CINRAD

Radar echo from sea clutter is an unpredictable event which often contaminates theprecipitation measurement and other radar products in coastal areas. It is thus necessaryto identify and remove these clutter echoes. The purpose of the paper is to developSCDA (Sea Clutter Detection Algorithm) to fit CINRAD. According to sea clutterobserved by SA/SB Doppler radars in Fuzhou, Wenzhou, Xiamen, Zhoushan andZhanjiang, it is divided into two kinds by the time when it appears. The first kind of seaclutter with strong strength mainly appears before and after a precipitation processduring the period except typhoon. The second kind of sea clutter whose strength isgenerally weak and uniform only appears close to sea areas of the radar position duringa typhoon. The reflectivity, radial velocity, spectrum width of velocity, and their spatialvariations for sea clutter and different precipitation are examined to determine themembership functions based sea clutter detection algorithm. For the first kind of seaclutter, considering rarely overlap between sea clutter and precipitation echo, thealgorithm is based on fuzzy logic and echo pieces using SCIT. Firstly, thepre-processing criteria which can be used to determine when sea clutter is likely to takeplace so that sea clutter detection algorithm can be applied. Secondly, echoes arecombined into pieces using SCIT, if the area of an echo piece is more than100km2, theecho piece would be given an attribute value. The echoes would be calculated with analgorithm based on fuzzy logical if the attribute value of echo piece meets suitableconditions and the area of echo piece is less than100km2. Inspired by LIU’s two-stepground clutter detection algorithm, the author improved Liu’s approach to realize thedynamic threshold. For the second kind of sea clutter, the algorithm is based on fuzzylogic and the improved two-step approach. Moreover, it studies the sea clutteridentification algorithm in the ultra-low-elevation angles. The conclusions are asfollows:(1) Through the preprocessing of the reflectivity value and the radial velocity valuethreshold， it can remove the obvious precipitation echo directly and can more effectively analyze the characteristics.(2) The six characteristics of TDBZ、SPIN、GDBZ、FOP、MDVEand MDSWcan effectivelydistinguish between the first kind of sea clutter and precipitation. The fourcharacteristics of TDBZ、SPIN、TVDBZand MDVEcan effectively distinguish between thesecond kind of sea clutter and the precipitation of typhoon.(3) The two methods can effectively identify the two kinds of sea clutterrespectively. In the first kind of sea clutter, the method has good result and can identifythe most of the echo, but the remote stratiform echo has some wrong identification. Inthe first kind of sea clutter, the method brings a lower wrong identification rates.(4) Using SCIT, it obvious reduces the misjudgment of the precipitation echo; usingthe improved two-step approach, it can further improve the identification effect.(5) The radar can reasonably reduce the scanning angle to observe a moresystematic precipitation; also, it is helpful for the accuracy of the precipitation products.The statistical analysis and the analyses results show the statistical characteristics(TDBZ、SPIN、TVDBZand MDVE) of sea clutter are unchanged in the low elevation angle; the mostof sea clutter in the ultra-low-elevation angles are detection effectively by the sea cutterdetection algorithm in the routine scan way.