珠江三角洲机场的竞争(1)

Ref:T02155AAIIRRPPOORRTTCCOOMMPPEETTIITTIIOONNIINNTTHHEEPPEEAARRLLRRIIVVEERRDDEELLTTAABBeecckkyyPP..YY..LLOOOODDeeppaarrttmmeennttooffGGeeooggrraapphhyy,,TThheeUUnniivveerrssiittyyooffHHoonnggKKoonnggaannddHH..WW..HHOOaannddSS..CC..WWOONNGGDDeeppaarrttmmeennttooffCCiivviillEEnnggiinneeeerriinngg,,TThheeUUnniivveerrssiittyyooffHHoonnggKKoonngg4.1-1AIRPORTCOMPETITIONINTHEPEARLRIVERDELTABeckyP.Y.LOODepartmentofGeography,TheUniversityofHongKong,PokfulamRoad,HongKongSAR,ChinaandH.W.HOandS.C.WONGDepartmentofCivilEngineering,TheUniversityofHongKong,PokfulamRoad,HongKongSAR,China1.INTRODUCTIONIn1979,therewereonlytwointernationalairports(atHongKongandGuangzhou)intheHongKongandPearlRiverDeltaregion,hereaftertheHK-PRDregion.Afterabouttwodecades,therewerefourinternationalairports(anadditionaltwoatShenzhenandZhuhai)inthisregionof48,0000sq.km(Loo,1999).Isthereanyneedforairportco-ordination?Before1997,theHongKongInternationalAirportoperatedandthrivedundertheBritishcolonialrule.Ontheonehand,itfacedvirtuallynocompetitionfromairportsacrosstheborder.Ontheotherhand,fewpeopleandfreightfromthatareaactuallyuseditsairportfacilities.After1997,thesituationgraduallychangedunderthe“onecountry,twosystems”policy.TheHongKongInternationalAirporthastofacecompetitionfromthreeotherlargeinternationalairportsintheregion.ItiseasierforHongKongpeopletouseairportfacilitiesintheMainlandandviceversa.Mostimportantly,thedemandforairtravelinalargepartoftheregionhasgreatlyexpandedbecauseofrapidincomegrowthandindustrialisation.Undersuchcircumstances,howwouldthegeographicaldistributionofairpassengerschange?Ifoneconsidersairportstobefacilitiesprovidingservices(ratherthanjustintermediatestops),thedelimitationofmarketareasofcompetitivetransportfacilitiesbecomessimilartotheclassicallocationproblemsofvonThunen’sisolatedstates,Christaller’sCentralPlaceTheory,andHotelling’stravellingsalesmanorretaillocations.Theprovisionofairportservicesrepresentspartofthemovementspaceoftheairpassengers(Borgstrom,1974),andairportservicesgenerateutilitytoairpassengers.Whenairportsareviewedasservicepoints,airpassengerturnoversareconceptualisedastheresultsofaggregatepersonaldecisionstousethoseairportfacilities.Onthedemandside,thedecisiontochooseaparticularairportisnolongerpurelythe(shortest)Euclideandistancebetweenthepassenger’shome(ortheoriginofthefreight)andthedepartureairport,orthedistancebetweenthefinaldestinationandthearrivalairport.Onthesupplyside,airportchoiceisaffectedbyfacilityattributesincludingasetofprimarylevel-of-service(LOS)factors,whichrelatetothepriceandfrequencyofflights,andasetofsecondaryLOSfactorsaboutthequalityofairportsandairlines.Inparticular,secondaryLOSfactorstendtodominateinamulti-airportcontextwhendifferencesinprimaryLOSfactorsanddistancesbetweenairportsarenotgreat(Bradley,1998;RubinandFagan,1976).Inamulti-airportregion,thechoiceofanairportfacilityisnolongerperipheraltoanindividual’sdecisiontotravel;andservicesatdifferentairportsare“differentiatedproducts”.Consequently,4.1-2thedelimitationofthemarketareasofairportsbecomesmuchmorecomplicatedthanasimplecaseofnearest-centreassignment(BryanandO’Kelly,1999).InthepioneerstudyofNdohetal.(1990),revealedpreferencedatahavebeenusedtoexamineairtravellers’choicesofdepartureairportsfromamongtheCentralEnglandairports,thatis,EastMidlands,Birmingham,Manchester,andLiverpoolAirports.Similarly,Bradley(1998)findsthattheairtravellers’choicefromamongthecompetitivedepartureairportsinEuropeisatleastaffectedbytwelveLOSfactors(Bradley,1998,p.158).Inlightofthesefindings,thereisaneedtore-calibrateairportattributes(otherthansimplyanairport’sphysicallocationinrelationtodemand)toreflectrealairportchoices.TheHK-PRDregion,withfourinternationalairports,representsanidealmulti-airportregionfortestingthehypothesisthatairportattributesareimportantinunderstandingthegeographicaldistributionofairpassengers.Traditionalairtrafficforecastingtechniquesbasedonnationalforecastsandlocalmarketsharesarefarfromsatisfactory.Whenthesetechniquesareapplied,“itisunusualforairportserviceareastooverlapandevenmoreunusualforanoverlaptobeconsideredintheforecastingprocess”(RubinandFagan,1976,p.1).Suchforecastsaremadeasinasingle-airportregionwithawell-definedandfixedmarketarea.Inreality,however,theoverwhelminglysimplifiedassumptionofasingle-airportregionisofteninapplicable.Thecontinuousequilibriummodeloffersanalternativetounderstandairtrafficdistributioninamulti-airportregion,whereairports’marketareasareneitherwelldefinednorimmutable.TaguchiandIri(1982)wereamongstthefirstwhodevelopedapromisingnumericalproceduretosolvetheproblemofcontinuoustransportationsystems,forageneralcityconfiguration,byusingthefiniteelementmethod(FEM)(ZienkiewiczandTaylor,1989)tosolvethreecontinuumnetworkproblems,namely:themaximumflowproblem,theshortestrouteproblem,andtheminimum-costflowproblem.Foruserequilibriumproblems,adual-basedformulationwasgivenbySasakietal.(1990),inwhichtheuserequilibriumprobleminthecontinuoussystemwassolvedbyminimisinganobjectivefunction,thatwassubjecttoasetofconstraints.T