地铁乘车需求的影响因素外文文献翻译.docx

地铁乘车需求的影响因素外文文献翻译.docx

《地铁乘车需求的影响因素外文文献翻译.docx》由会员分享，可在线阅读，更多相关《地铁乘车需求的影响因素外文文献翻译.docx（6页珍藏版）》请在冰豆网上搜索。

地铁乘车需求的影响因素外文文献翻译

地铁乘车需求的影响因素外文翻译2019

英文

DoesdemandforsubwayridershipinManhattandependontherainfallevents?

ShirinNajafabadi，AliHamidi

Abstract

TheNortheastUnitedStates,particularlyNewYorkStatehasexperiencedanincreaseinextremedailyprecipitationduringthepast50years.Recenteventssuchas Hurricane IreneandSuperstormSandy,haverevealedvulnerabilitytotheintenseprecipitationwithinthetransportationsector.InthescaleofNewYorkCity,wheretransitsystemisthemostdominantmodeoftransportationanddailymobilityofmillionsofpassengersdependsonit,anydisruptioninthetransitservicewouldresultingridlocksandmassivedelays.Toassesstheimpactsofrainfallonthesubwayridership,wemergedhighresolutionradarrainfallandsubwayridershipdatatoconductadetailedanalysisforeachofthe116subwaystationsattheboroughofManhattan.Theanalysisiscarriedoutonbothhourlyanddailyresolutionlevel,whereaspatial-temporalBayesianmulti-levelregressionmodelisusedtocapturetheunderlyingdependencybetweentheparameters.Theestimationresultsareobtainedthrough MarkovChainMonteCarlo samplingmethod.Theresultsfordailyanalysisindicatethatduringweekdays,transitridershipinthestationslocatedincommercialzonesarelesssensitivetotherainfallcomparedtotheonesinresidentialzones.

Keywords：

Bayesianmulti-levelregressionmodel，Subwayridership，MCMCsampling，Radarrainfall

Largecitiesaroundtheworldrelyonpublictransportationinfrastructuretomaintainagoodlevelofserviceandtoincreasemobilityandeconomicproductivity.Inyear2015,morethan10.7billiontransittripswerereportedintheUnitedStates（Matthew Dickens,2016）.AccordingtotheAmericanPublicTransportationAssociation（APTA）every$1investedinpublictransportationgeneratesapproximately$4ineconomicreturnsthroughincreasedemploymentrate,businesssales,andenhancedpropertyvalues.Thefullreturnoninvestmentfortransportationsystemscanonlybeachievedwhencitiesoptimizeandplanthemaintenanceandgrowthoftheirpublictransportationsystemsthroughbetterrealizationofthedemandlevelfortransportationalongwiththeidentificationofotherinfluentialparameters.Transportationsystemperformancedependsonthegeometryofthenetwork,aswellasotherexternalfactorssuchasaccidents,operationaltearandwear,disruptionsondependentsystems（e.g.thepowergridforsubways）,effectiveautomobileregulation,weather,etc.（DeGrangeetal.,2012）.TheobjectiveoftheworkpresentedhereistoprogressfurthertheunderstandingofimpactsofrainfalleventsonthesubwayridershiplevelinManhattan,NewYork.

Duringthepast50years,NewYorkStatehasexperiencedanincreaseinextremedailyprecipitation（Hortonetal.,2011）alsoithasbeenheavilyaffectedbyunusualweatherpatternssuchas Hurricane IreneandSuperstormSandy.InNewYorkCity,withover4Milliondailycommutetrips（MossandQing,2012）,theevidenceintheaftermathofextremeweatherpatternshaverevealedvulnerabilitiestointenseprecipitationwithinthetransportationnetwork,yieldingenormouseconomiclossesfortheCity（Brian Tumulty,2012）.Abetteranalysistoquantifytheeffectsofvariousweatherconditionsonthetransportationnetworkwouldresultinwell-informedandefficientpolicy-makingdecisions.Literatureinrecognitionofweatherinfluenceontransportationandmobilitycanbeclassifiedintotwogroups.Thefirstgroupoftheliteraturemeasurestheinfluenceofweatherontheperformanceoftransportsystems,whilethesecondgroupstudiesitsbehavioralimpactsoncommunities.Studiesontheinfluenceofweatherovertheperformanceoftransportationnetworksbrushoverlargespectrumoftopicsincludingtrafficflowandroadcapacity（e.g. Kyteetal.,2001; Mashrosetal.,2014; Mazeetal.,2006）,infrastructureperformance（e.g. KoetseandRietveld,2009）,trafficsafety（e.g. AndreescuandFrost,1998; KoetseandRietveld,2009）andchangesinthequalityofservice（e.g. Coolsetal.,2010; KhattakandDePalma,1997）.Thesestudieshaveshedsomelightonthemostimportantaspectsoftransportationsystemsandserviceaffectedbyweather,whicharevaluableinformationforcityadministrators.Ingeneral,traveldemand,trafficsafety,andthetrafficflowarethethreedominantfactorsimpactedbyadverseweather（Mazeetal.,2006; Mashrosetal.,2014）.AccordingtoastudyconductedinManchestercityby JaroszweskiandMcNamara（2014）,duringrainfallstherateofroadaccidentsincreasesby50%.Meanwhile,throughamorecomprehensivestudycarriedoutbyHofmannandO'Mahonyonvariousperformancemeasuresoftransitsystems,itisrevealedthatbadweatherconditionsdegradethelevelofserviceoftransportationsystems（HofmannandO'Mahony,2005）.

Behavioralimpactsofadverseweatherconditionsarestudiedthroughmeasuringitsimpactsonpublictransportdemand（e.g. Guoetal.,2003; Singhaletal.,2014; Zhouetal.,2017）,modalshift（e.g. KhattakandDePalma,1997; KoetseandRietveld,2009）,androuteanddestinationchoice（e.g. Guoetal.,2003）.Astudyshowstheadverseweatherconditionschangesmodechoice,routechoice,anddeparturetimeofautomobilecommutersinBrussels（KhattakandDePalma,1997）.Dependingonthetrippurpose,thechangeinthemobilitybehaviorduringdifferentweatherconditionssuchasrain,snow,temperatureandfogwouldvary（Coolsetal.,2010）.

Ourstudyfocusesontheinfluenceofrainfallconditionsonsubwayridership.Despiteitsimportance,thedependencybetweenweatherconditionsandtransitridershiphasseldombeeninvestigated,especiallyatafinerresolutionusingspatiallydistributedrainfalldata.Existingstudieshavelimitedtheirattentiontomeasuringtheeffectsofweatherconditionsontheridershipdemandoftransitsystemonaggregatelevel（daily）（e.g., Aranaetal.,2013; Changon,1996; Guoetal.,2003; Kashfietal.,2013; StoverandMccormack,2012）.Thesestudieseithermeasurethedirectandindirectimpactsofextremeweatherevents（e.g.,increaseintraveltime,waitingtime,operationaldelaysinthesystem）ontransitusers（Guoetal.,2003; HofmannandO'Mahony,2005）. Aranaetal.,2013,analyzedchangesinthenumberofdiscretionarytrips（measuredviadailybusridershipovertheweekends）inresponsetoweatherconditions,wheretheanalysisdemonstratesdecreaseintransitridershipinrainyandwindydaysandincreaseduringhotdays.Trippurposeplaysacrucialroleinridershipdemandunderadverseweathercondition;travelersaremorelikelytopostponeleisuretripswhilemandatoryandworktripsarelesslikelytobedeferred（Coolsetal.,2010; KhattakandDePalma,1997; Mazeetal.,2006）. KhattakandDePalma（1997） foundthattransitridershipmayincreaseduringextremeweatherconditionsduetoshiftsfromothermodesoftransportationsuchasautomobiles,walking,andbiking. Table1 presentsmoredetailsonsomeotherstudiesfocusedonassessingtheimpactsofweatherinstabilitiesontransitridership.Amongstthesestudies,itisworthnotingthat StoverandMccormack（2012） foundrainfalltobethemostinfluentialweatherfactor（amongwind,temperature,rainandsnow）leadingtoridershipdecrease. Kashfietal.（2013） conductedadailyanalysisandfoundastrongnegativerelationshipbetweenrainfallandridershipdemand.

Allthesestudiespresentimportantqualitativeandquantitativeinformationontheimpactofweatheronridership,buttheanalysisweremostlyconductedatalargespatialscale-e.g.citylevelaggregatescale-whichisnotnecessarilyaproperscaleforpredictingtransitdemand.Itispreferablefortransitplannerstoanalyzetransitdemandanalysisatthetransitstoplevel,asitisthespatialscalebywhichtransitusersusethesystem（Dilletal.,2013）.Usingspatialpointscalelevelisalsousefultoconnecttransitdemandwithlandusecharacteristicsofthestations.Fewstudieswereconductedatadetailedspatialortemporalresolution.Forexample, Singhaletal.（2014）,utilizedhourlyridershipdatatoevaluatethesystem-levelridershipinManhattan.However,theydidnotincludespatiallydetailedweatherdataintheiranalyses.Similarly, Zhouetal.（2017） employeddisaggregatelevelsmart-cardinformationtoinvestigatetheimpactofweatherontransitridershipinShenzhen,China,reportingthatrainfallhasanegativeimpactonridershipduringoff-peakhoursandduringweekends.Majorityofthestudiespresentedin Table1 appliedstepwiselinearregressionsmodelsorsimilarstatisticalmodelstoestimateridershipindifferentweatherconditions（e.g. Aranaetal.,2013; Guoetal.,2003; Sabiretal.,2008; Singhaletal.,2014）.Furthermore,mostexistingstudiesusedsurvey-baseddatawithrelativelysmallsamplesizeasanexample. Coolsetal.（2010）conductedasurveyovertwoyearswith586respondentsfocusedondailyaggregateridershipwithaggregatedweatherconditionsdatacollectedfromfewweatherstationsinthestudyareaofinterest.Forinstance Singhaletal.（2014）,utilizedinformationcollectedfromoneweatherstationintheiranalysis.While Zhouetal.,2017 useddisaggregatelevelsmart-cardinformationandtheirdatasourcewaslimitedtoonemonth.

Afullyfunctionalmodelofweatherpatternsovertransitridershiprequiresanunderstandingofdetailedspatialtemporaldemandvariabilitytopredictfutureridershipdemandinvariousweatherconditions.Thecurrentstudyisastepinthisdirection.ThemaingoalofthispaperistofindouthowsubwayridershipinManhattanischangingduringrainfalleventsandisitbeingimpactedbythetimingofserviceandspatial