8.交通地理与经济发展(Transportation)

8.TransportationGeographyandDevelopment交通地理与发展8.1Transportation:conditionfortrade交通地理的重要意义：贸易的条件IfpricedifferentialbetweenregionsiandjisUi-Vj,thetradecanberealizedonlyifTijUi-Vj.Cheapertransportationcostsovertimestrengthenedregionalspecialization&trade8.2Transportcosts交通成本构成Twocomponents:terminalcost:fixed(FC)line-haulcosts:t*dTC=FC+t*dImplicationsinindustriallocationFirmstendtolocateintheraw-materialsiteormarketsitetosaveloading/unloadingcosts.Transshipmentcostmakesthebreak-of-bulklocationwheretransshipmentbetweendifferenttransportationmodesadvantageous.ImpactonindustryIIIThesteelindustryandthebreak-of-bulklocation8.3Specifiedmarketsbyvariousmodes不同运输模式的专有市场Competitionstrategy:rateumbrellaTherailsystempracticedittoattractlong-haulbusinessTheriseoftruckingbusiness“predatorypricing”TransportdevelopmentmodelTaaffe,MorrillandGould,1963(GeographicalReview53,503-529)Scatteredports;Penetrationlines;Feederlines;Interconnectionsdeveloped;Hierarchalstructure(differentiatednodesandlinkages).FourErasintheU.S.Localera:priorto1820Trans-Appalachianera:1820-1850Eraofraildominance:1850-1900Eraofcompetition:1900-presentJohnAdam’sModelofUrbanGrowthWalking-horsecarera(1800-1890)Electricstreetcarera(1890-1920)Recreationalautomobile(1925-1950)Freewayera(1950-1970)Multicentriccity(1970-present)Airtransport:Hub-and-spokenetworkThegravitymodelTheprimitiveform:Theimprovedform:indicatesthedegreeofdistancefriction2ijjiijdPPkTijjiijdPPkTVaryingvalues:gettingsmallerovertimelessfrictionaleffectofdistanceforlargercitiesvaluesaredifferentamongdifferenttransportationmodeswell-developedareaswithadvancedtransportationnetworkshavesmallervalueshigher-valuegoodsarelesssensitivetotransportationcosts,thussmallervaluesSpatialvariationofvalueThepotentialmodelTheprimitiveversion:aggregateinteractionsbetweenpointiandallotherpoints(j=1,2,3,…,n).Therevisedversion:accessibilityofpointitoallothersnjijjiidPPV1njijjidPV1TheHuff(Probability)ModelProbabilityofresidentsfromdistrictishoppingatstorejPortionofresidentscommutingfromitojkikkijjijdSdS//Pr8.4NetworkComponent1:NodeCity,stationorportInGISrepresentation,0-dimension(point)“gateway”or“hub”交通网络分析NetworkComponent2:LinkWaterway,highway,railway,airwayInGISrepresentation,1-dimension(line)Trunkline(arterial,corridor)vs.feederlineTwoformsofmatrixO-DmatrixTij:Rowsindexedbyorigins(Oi)Columnsindexedbydestinations(Dj)Binaryconnectivitymatrix:Tij=1ifthereisalinkbetweeni&j,Tij=0otherwiseWidelyusedingraphtheoryBasicTerms:Vertex&edge(node&link)Planar(edgesintersectonlyatvertices)vs.nonplanar(e.g.,airlineorpipelinenetworks)PathDiameter:numberofedgesneededtoconnectthetwomostremotenodesDegree:#directlinksfromanodetoothernodesTree:Minimally-connectednetworkEachvertex(node)isconnectedtothenetwork.Theremovalofasingleedge(link)willdividethenetworkintotwodisconnectedpartsemin=v-1OnlyonepathbetweenverticesForest:disconnectedtreese=v-k(k:#trees)Maximally-connectednetworkEachnodehaseitheradirectlinktotheothernode,orbyaddingonelinkthelinkwouldintersectwithexistingones.nonplanarnetwork:emax=V(V-1)/2planarnetwork:emax=3(V-2)forV=3Wheretheadditionofanynodecreates3additionallinkswithoutintersectinganyexistinglinks.Thecaseofmaximallyconnectednetwork:v=4,ande=3*(v-2)=6Thecaseofmaximallyconnectednetwork:class:v=5,ande=3*(v-2)=9PlanarnetworksCyclomaticnumber(measureofcircuits):μ=e–v+pWherepis#nonconnectedsubgraphs(p=1iffullyconnected)Circuitsmeasurethegapbetweeneandv,i.e.,howmany“extra”(morethanminimal)edgespresentinthenetworkMinμ=0ifitisatree(p=1,e=v-1)Maxμ=0.5v(v-1)–v+1fornonplanarnetworkMaxμ=2v-5forplanarnetworkConnectivitymeasures:nonplanarAlphaindex:α=(e-v+1)/[0.5v(v-1)-(v-1)]Actual#circuits/maximum#circuitsinaconnectednetworkBetaindex:β=e/vAverage#edgespervertex(i.e.,degree)Gammaindexγ=e/[0.5v(v-1)]Actual#edges/maximum#edgesConnectivitymeasures:planarAlphaindex:α=(e-v+1)/[3(v-2)-(v-1)]Actual#circuits/maximum#circuitsinaconnectednetworkBetaindex:β=e/vAverage#edgespervertex(i.e.,degree)Gammaindexγ=e/[3(v-2)]Actual#edges/maximum#edgesTypesofplanarnetworksbasedonthegammaindexSpinalnetwork:1/3≤γ1/2Gridnetwork:1/2≤γ2/3Deltanetwork:2/3≤γ1Connectivityofnode:theproblemoftheleaderTopologicaldistance(0/1)matrix:GThemeaningofmatrixtoapower:G2Gij2:numberofpathslinkingiandjby2edgesThemeaningofmatrixtoapower:GkGijk:numberofpathslinkingiandjbykedgesT-matrix:T=G+G2+…Gd(d:diameter)RowsumindicatesnodeconnectivityScalarsinT-matrixbyGarrisonT=sG+s2G2+…sdGd(s1)MeasuresofDistanceEuclideandistanceGeodeticdistancethroughagreatcircleoftheearthManhattandistanceNetworkdistance:shortest-paththroughareal-worldroadnetwork2/122122112])()[(yyxxd)]cos(*cos*cossin*cos[sin*12acdbdbard||||212112yyxxdComputingNetworkDistanceandTimelabelsettingalgorithmbyDijkstra(1959)valued-graph(orL-matrix)methodNODEDISTANCEcommandinArcInfocomputesdistancesbetweennodesonthenetworklabelsettingalgorithmL-matrixmethodStreetcentralitymeasuresClosenessCCh