简介:针对当前f-x域CEEMD分频去噪容易损失倾斜有效信号及Cadzow滤波法秩的大小难以选择等问题,研究了一种f-x域CEEMD结合Cadzow滤波的去噪方法.对原始含噪剖面应用f-x域CEEMD分频处理,利用其在保留水平同相轴方面的优势,将原始剖面分割为水平同相轴剖面和倾斜同相轴剖面.利用Cadzow滤波法通过保留较大的秩和较小的秩分别对水平和倾斜剖面做进一步降噪处理,恢复有效信号,将二者相加得到最终降噪剖面.该方法能恢复在f-x域CEEMD分频去噪中损失的倾斜同相轴,且能优化在应用Cadzow滤波法去噪时对秩大小的选择.数值计算结果证明了该方法的有效性.
简介:
简介:ThisstudyinvestigatesclassificationanddiurnalvariationsoftheprecipitationechoesoverthecentralTibetanPlateaubasedontheobservationscollectedfromaC-bandvertically-pointingfrequency-modulatedcontinuous-wave(C-FMCW)radarduringtheThirdTibetanPlateauAtmosphericScientificExperiment(TIPEX-Ⅲ)2014-IntensiveObservationPeriod(2014-IOP).Theresultsshowthat51.32%oftheverticalprofileshavevalidechoeswithreflectivity>-10dBZ,and35.06%ofthevalidechoprofilesproduceprecipitationattheground(precipitationprofiles);stratiformprecipitationwithanevidentbright-bandsignature,weakconvectiveprecipitation,andstrongconvectiveprecipitationaccountfor52.03%,42.98%,and4.99%oftheprecipitationprofiles,respectively.About59.84%oftheprecipitationoccursintheafternoontomidnight,while40.16%oftheprecipitationwithweakerintensityisobservedinthenocturnalhoursandinthemorning.Diurnalvariationofoccurrencefrequencyofprecipitationshowsamajorpeakduring2100-2200LST(localsolartime)with59.02%beingthestratiformprecipitation;thesecondarypeakappearsduring1300-1400LSTwith59.71%beingtheweakconvectiveprecipitation;thestrongconvectiveprecipitationoccursmostly(81.83%)intheafternoonandeveningwithtwopeaksover1200-1300and1700-1800LST,respectively.Startingfromapproximately1100LST,precipitationechoesdevelopwithenhancedverticalairmotion,elevatedechotop,andincreasingradarreflectivity.Intenseupwardairmotionoccursmostfrequentlyin1700-1800LSTwithasecondarypeakin1100-1400LST,whilethetopsofprecipitationechoesandintenseupwardairmotionreachtheirhighestlevelsduring1600-1800LST.Theatmosphericconditionsintheearlymorningaredisadvantageousforconvectiveinitiationanddevelopment.Aroundnoon,theconvectiveavailablepotentialenergy(CAPE)increasesmarkedly,convectiveinhibition(CIN)isgenerallysmall,andasuper-dry-adiabaticlayerispresentnear