【英文标准名称】:StandardGuideforStatisticalEvaluationofAtmosphericDispersionModelPerformance
【原文标准名称】:大气散射模型性能的统计评价的标准指南
【标准号】:ASTMD6589-2005
【标准状态】:现行
【国别】:
【发布日期】:2005
【实施或试行日期】:
【发布单位】:美国材料与试验协会(US-ASTM)
【起草单位】:D22.11
【标准类型】:(Guide)
【标准水平】:()
【中文主题词】:大气的;分散;分散模型;评定;模型;性能;统计
【英文主题词】:airquality;model;modelperformance;qualitativeevaluation;quantitativeevaluation;statistical
【摘要】:Guidanceisprovidedondesigningmodelevaluationperformanceproceduresandonthedifficultiesthatariseinstatisticalevaluationofmodelperformancecausedbythestochasticnatureofdispersionintheatmosphere.Itisrecognizedthereareexamplesintheliteraturewhere,knowinglyorunknowingly,modelswereevaluatedontheirabilitytodescribesomethingwhichtheywereneverintendedtocharacterize.Thisguideisattemptingtoheightenawareness,andthereby,toreducethenumberofx201C;unknowingx201D;comparisons.Agoalofthisguideistostimulatedevelopmentandtestingofevaluationproceduresthataccommodatetheeffectsofnaturalvariability.Atechniqueisillustratedtoprovideinformationfromwhichsubsequentevaluationandstandardizationcanbederived.1.1Thisguideprovidestechniquesthatareusefulforthecomparisonofmodeledairconcentrationswithobservedfielddata.Suchcomparisonsprovideameansforassessingamodel'sperformance,forexample,biasandprecisionoruncertainty,relativetoothercandidatemodels.Methodologiesforsuchcomparisonsareyetevolving;hence,modificationswilloccurinthestatisticaltestsandproceduresanddataanalysisasworkprogressesinthisarea.Untiltheinterestedpartiesagreeuponstandardtestingprotocols,differencesinapproachwilloccur.Thisguidedescribesaframework,orphilosophicalcontext,withinwhichonedetermineswhetheramodel'sperformanceissignificantlydifferentfromothercandidatemodels.Itissuggestedthatthefirststepshouldbetodeterminewhichmodel'sestimatesareclosestonaveragetotheobservations,andthesecondstepwouldthentestwhetherthedifferencesseenintheperformanceoftheothermodelsaresignificantlydifferentfromthemodelchoseninthefirststep.Anexampleprocedureisprovidedintoillustrateanexistingapproachforaparticularevaluationgoal.Thisexampleisnotintendedtoinhibitalternativeapproachesortechniquesthatwillproduceequivalentorsuperiorresults.AsdiscussedinSection,statisticalevaluationofmodelperformanceisviewedaspartofalargerprocessthatcollectivelyisreferredtoasmodelevaluation.1.2Thisguidehasbeendesignedwithflexibilitytoallowexpansiontoaddressvariouscharacterizationsofatmosphericdispersion,whichmightinvolvedoseorconcentrationfluctuations,toallowdevelopmentofapplication-specificevaluationschemes,andtoallowuseofvariousstatisticalcomparisonmetrics.Noassumptionsaremaderegardingthemannerinwhichthemodelscharacterizethedispersion.1.3Thefocusofthisguideisonendresults,thatis,theaccuracyofmodelpredictionsandthediscernmentofwhetherdifferencesseenbetweenmodelsaresignificant,ratherthanoperationaldetailssuchastheeaseofmodelimplementationorthetimerequiredformodelcalculationstobeperformed.1.4Thisguideoffersanorganizedcollectionofinformationoraseriesofoptionsanddoesnotrecommendaspecificcourseofaction.Thisguidecannotreplaceeducationorexperienceandshouldbeusedinconjunctionwithprofessionaljudgment.Notallaspectsofthisguidemaybeapplicableinallcircumstances.Thisguideisnotintendedtorepresentorreplacethestandardofcarebywhichtheadequacyofagivenprofessionalservicemustbejudged,norshoulditbeappliedwithoutconsiderationofaproject'smanyuniqueaspects.Theword"Standard"inthetitleofthisguidemeansonlythatthedocumenthasbeenapprovedthroughtheASTMconsensusprocess.Thisstandarddoesnotpurporttoaddressallofthesafetyconcerns,ifany,associatedwithitsuse.Itistheresponsibilityoftheuserofthisstandardtoestablishappropriatesafetyandhealthpracticesandtodeterminetheapplicabilityofregulatorylimitationspr......
【中国标准分类号】:A47
【国际标准分类号】:13_040_20
【页数】:17P.;A4
【正文语种】: