Statistical Modelling of Severe Wind Gust

《Statistical Modelling of Severe Wind Gust》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、StatisticalModellingofSevereWindGustX.G.LinRiskModellingProject,MineralsandGeohazardsDivision,GeoscienceAustralia,Canberra,AustraliaE-mail:xunguo.lin@ga.gov.auAbstract:Themodellingofseverewindgustisafundamentalpartofanywindhazardassessment.Statistica

2、lmodellingapproachesareusuallyusedtodescribetheprobabilityofoccurrencesofextremewinds.Theclassicalextremevaluetheorybasedonthegeneralizedextremevalue(GEV)distributionandthemorerecentlydevelopedpeaksoverthreshold(POT)andtheuseofgeneralizedParetodistri

3、bution(GPD)areinvestigated.Practicalissuessuchasthethresholdselectionandmodelvalidationarealsodiscussed.SeverewindgustmodelsweredevelopedforanumberoflocationsinPerth.Estimatesof50and100yearreturnperiodeventsarecomparedwithestimatespublishedbytheStand

4、ardsAustraliaforPerth(RegionA),whicharebasedonasinglesite.Forcoastsites,thepredictedwindgustspeedsaresimilartothosepublishedbytheStandards,whiletheinlandsiteshavesmallerwindgustspeeds.Keywords:Windgustmodelling;Extremevaluetheory;GeneralizedExtremeVa

5、luedistribution;GeneralizedParetodistribution;Peaksoverthreshold;Returnperiods1.INTRODUCTIONintroductionofextremevaluetheorycanbefoundinEmbrechtsetal.(2001)andColse(2001).ExtremewindisoneofthemajornaturalhazardsexperiencedinPerth.Theseextremewindsare

6、generallyproducedbycoldfrontsandnotfrom2.1GeneralizedExtremeValueDistributioncyclones.TopredicttheextremewindwithagivenreturnperiodstatisticalanalysisofwindTheclassicalextremevaluetheoryisbasedonrecordswasused.threeasymptoticextremevaluedistributions

7、identifiedbyFisherandTippett(1928).TheExtremevaluetheoryisastatisticaltechniqueforGeneralizedExtremeValue(GEV)distributiondescribingtheunusual(extreme)ratherthantheintroducedbyJenkinson(1955)combinestheusualevents.Thisworkquantifiesthehazardthreedist

8、ributionsintoasinglemathematicalcomponentsinariskassessmentofthePerthformwiththecumulativedistributionfunctionmetropolitanareatowindsthatresultindamage(CDF):tobuildingsandinfrastructures.Thesearethe1/ξextremewindsofinterest.x−µ−1−ξH(x;ξ,σ,µ)=e