分布式降水估算的模型与方法分析-model and method analysis of distributed precipitation estimation.docx

下载文档 降价啦

13
0
约4.74万字
约 67页
2018-06-04 发布于上海
举报
版权申诉
保障服务

分布式降水估算的模型与方法分析-model and method analysis of distributed precipitation estimation.docx

1、本文档共67页，可阅读全部内容。
2、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

分布式降水估算的模型与方法分析-model and method analysis of distributed precipitation estimation

AbstractPrecipitationisnotonlyanimportantfactortoaffectresource,environmentand disaster,butalsoisanimportantparameterinhydrologyandintheresearchofwater resources。Butsinceaperiodoftime，theprecipitationdataobtainisonlyforsomedispersepoint，and thatcan notgettingmorematerial rainfall of area。Therefore，In this paper,aestimationmodelisbuiltwhichutilizesthespecialinterpolationtechnologyand the MODIS datum, andget thedistribute rainfall.Studyonthespecialinterpolationmethodofhydrologicalmodelhaveagreat practicalvalues,andtheArtificialneuralnetworkhasverycapabilityindealingwith hispaper,themodelingmethodofartificial neural networksstudying onvalleydistributedrainfallmodelisadopted;Anditsperformanceiscomparedwith othertraditional algorithms such asThiessen polygon,Inversesquaredistance, etc.Inthispaper,theparametersofcloudandatmospherewhichareinvolvedwiththe physicalmechanicsofrainfallareanalyzed,andobtainedwiththecloudproductdataof MODISsensor,which,togetherwithdataobtainedbysurfacemeasurementofrainfall, composetheoriginalsample,andbuildtherainfallestimationmodelwhichutilizestheBP network.AimingatsomeweaknessoftheBPNeuralNetworkintheactualapplications,such astheastringencyslow,easilyfallingintolocalsolutionsconcerningbiggersearchspace, anothertheNeuralNetworkconnectionweightshavingnowayindefinition,Optimized theconnectionweightsbymeansofgeneticalgorithmwhichhasthebestglobalsearch ability,builttheoptimizedBPNeuralNetworkestimationrainfallmodel,andtestedwith theoriginalsample.theresultindicatesthatthenetworkmodeloptimizedbythegenetic algorithmisimprovedbothintheprecisionandstability,reflectstheactualrainfall instancebasically,atthesametime,itrevealsthatitbearspracticalsignificancetoapply thehighspectralresolutionandrichdataproductsofMODIStosatelliteestimationof precipitation.Theprecipitationdatainthispaperisobtainedfromobservationdataofrainfallintheautomaticobservationstations,HubeiProvince,in11,24,2006.TheMODISsatellitedata product,adoptedinthisthesis,isreceivedandprocessedbyt