USF Libraries
USF Digital Collections

Integrated planning and design of intermodal systems


Material Information

Integrated planning and design of intermodal systems interim task report
At head of title:
Guideway transit and intermodalism function and effectiveness
Physical Description:
1 online resource (116 p.) : ill., col. maps. ;
Boyle, Daniel K
Newsome, Tracy H
Liu, Rongfang
United States -- Federal Transit Administration
University of South Florida -- Center for Urban Transportation Research
University of South Florida, Center for Urban Transportation Research
Place of Publication:
Publication Date:


Subjects / Keywords:
Local transit -- Planning -- United States   ( lcsh )
Local transit -- Ridership -- United States   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Includes bibliographical references (p. 111-116).
Statement of Responsibility:
Center for Urban Transportation Research, College of Engineering, University of South Florida ; project managers, Daniel Boyle, Tracy Newsome ; project statff, Rongfang Liu ... et al..
General Note:
Title from e-book cover (viewed Aug. 25, 2011).
General Note:
"July 1995."
General Note:
"... under the sponsorship of the Department of Transportation, Federal Transit Administration. ..."

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029186474
oclc - 747836164
usfldc doi - C01-00034
usfldc handle - c1.34
System ID:

This item is only available as the following downloads:

Full Text


Guideway Transit and lntermodalism: Function and Effectiveness Draft Interim Task Report \ Integrated Planning and \ \ of lntermodal Systems\ \ \ \ \ I I I I CUIR Center for Urban Transportation Research College of Engineering ot South Florida 4202 E FowlerAve nVsit AdmiNsttatiOn in the intmst of l n fOtmatiOn exchange The U S Govemmant assumes no Jjsbility for the contents or use fllereolandd06snOfeOOOrseanyvendorsorproductsmentloned i nthoreport \\ \ \ \ \ \ \ \ '\ \ \\ \ \ \ \ ', \ \ \ \ \ ,_L_ ___ \ \ \ \




Preface Over the past few decades more than a dozen U.S cities have implemented new guideway pub lic transit systems and virtually every major urban area has or is considering i ncreasing public transportation I nfrastructure Investments, frequently includ i ng the cons i deration of gui deway t ransH investments. The country's dramatic suburban i zation and socio-economic changes have p l aced new challenges on pub lic transportation. Various guideway investments a r e among the solutions that local communities have considered to meet the changing transportation needs of their communH ies. The result has been growing guideway transit ridersh i p and an increase in the importance of guideway in the overall transportation system. Gui deway transH investments are perce i ved as the public transit investment that provides an excellent opportuney to compete with auto trave l influence l and use, motivate public and business financia l support and address air qualey and environmental goal s This report does not advoca t e guideway solutions or discourage carefu l consideration of non g uideway trans p ortation investments, but prov i des a knowledge base to support those involved in gui deway planning and impl ementat ion. With the devel opment o f numerous systems over the pas t few years. a grea t deal of experience and knowledge has been gai ned about all as p ects of using gu i deway investments to meet transp o rtation and other l oca l goals. Much of this kn o wledge r es i des with l ocal plann ing agency staffs and is of great value to other urban areas if the most relevant info rmati o n can be c aptu r ed and commun i cated to the eve r grow i ng and c hanging group of professionals that are involved in guideway project p l anning and decision making. This report is one of severa l that are being p r oduced as part of a study funded by the Federal Transit Administration on intermodalism and guideway effectiveness Th i s m ulti year effort is bein g conducted by the Lehman Center for Transportation Research at F l orida I n t ernat i ona l Universijy and the Center for Urban Transportation Research at t he University of South Flori da. The b road l y-defined research project, a response to a U S congressional authoriza t ion, focuses on the exam inati on of fact o rs that influence the effectiveness and efficiency of guideway trans i t systems and i n t ermodal passenger transportation The w o r k program is driven by eight primary research tasks. eac h o f which i s being add ressed 3


through a variety of research methodologies. The overall objedlve Is to assemble existing and new information and interpret and communicate that information in a manner that supports the planning and decision making efforts of public tra nsportat io n planners. Know1edge gained in this project will provide useful information for the many communities and transportation professionals that are planning or considering guideway transit as a key component in their transportation system. In addition, many of the issues and much of the i nformation will have application for all public transportation planning. The products of this research effort in 199 5 include technical reports, case studies, and data books. 4


Ta b le of Contents UstofTables .............. ...... ......................... 9 Ust of Figures . . . . . . . . . . . . . . . . . . . . 9 Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 Section 1 . . . . . . . . . . . . . . . . . . . . . . . . . 13 Task Objectives . . . . . . . . . . . . . . . . . . . 13 Research Methodo logy . . . . . . . . . . . . . . . . . . . . . . . 15 Activities/P r ogress to Dat e . . . . . . . . . . . . . . . . . 16 Planned ActivHies . . . . . . . . . . . . . . . . . . . . 17 Section 2 . . . . . . . . . . . . . . . . . . . . . . 18 Techical Find ings .... . . .......... . . . ...... ....... ... 18 Subta s k 1 ........... .... . ..... .... . . ..... ... ....... 18 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 18 Theoretical Constructs . . . . . . . . . . . . . . . . . . . . . 18 lntennodal Connections in Practice ................. .. ....... ...... ... 38 Subtask 2: Station Location Criteria and Efficiency . ...... ..... .... . .... 52 Introduction ... .................. . .................... ...... .... 52 C onst raints t o Optimality .... ............................ .......... 5 2 South Florida Station Locati ons ... .. ......... . . . .......... ... 57 Subtask 3 ..... ...... ..................... ... ..... ... ... . 77 Intro ductio n .... .......... . . ....... .... ..... .... ................. 77 Fare Policies and Structures ........ .... . ...... .... ... ............. ... 78 Location of Fixed Guideway in Relation to the Existing Bus Network .... ....... 80 Passenger Convenience ...... .. .. .. .. .. ............ .... . .... ... . 81 Extent of Bus Network Restructing to Serve Fixed Guideway ... . ..... ........ 83 Automobile Access to Guideway Tra nsit .. .................. ... ........ 91 Tran sit Agency Organization and inter-Agen cy Cooperation .... ......... . 93 Local Policies ................... .... ........ .......... ......... 9 5 5




Subtask 4: Facilitating Convenient Transfers . . . . . . . . . . . 96 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 96 Transfer Pricing . . . . . . . . . . . . . . . . . . . . . . 97 lntrasystem verses In tersystem Transfer Polley . . . . . . . . . . 98 Schedule Coordination . . . . . . . . . . . . . . . . . . . . . 99 The Role of Monthly Passes . . . . . . . . . . . . . . 101 The Regional Pass Concept .. .. .. .. .. . .. .. .. .. . .. .. .. .. 102 Theoretica l Assessment of Fare Policy . . . . . . . . . . . . . 103 Summary . . . . . . . . . . . . . . . . . . . . . 1 OS References . . . . . . . . . . . . . . . . . . . . . 111 7




List of Tables Table 1 Selected Connectivity Measures for the Hypothetical Networks ... ..... 25 Table 2 Hypothetlcallntermodal Networ1< Iota Values for Alternative Node Weighing Schemes ......... . . .... . . 29 L ist of Figures Figure 1 Hypothetical Regional Bus Networ1< . . . .... . .......... 21 Figure 2 Hypothetical Regional Bus Networ1< ........ . ............ ... 22 Figure 3 Hypotheticallntermodal Networ1< ................ ... .. . .... 23 Figure 4 Par1<-N-Ride Lot at Santa Clarita Metrolink Station ......... ..... 45 Figure 5 Lack of Park-N-Ride Space leads to On-Street Parking at Blue Une Station . . . . . . . . . . . . . 46 Figure 6 Bus Informat ion in LA. Union Sta t ion .... ... .. ............ 49 F i gure 7 Bus Information in LA Union Station .......... .. ... ...... 50 Figure 8 Population Density for Tri -R ail Service Area ............... .... 59 Figure 9 Population Density for Metrorail Service Area ............ ... ..... 63 Figure 10 Tract Level Median Incomes for Tri-Rail Service Area ....... ........ 67 Figure 11 Tract Leve l Med i an Incomes for Metrorail Service Area ........ .... 69 Figure 12 Zero Car Households in Tri-Rail Serv i ce Area ........ ....... ..... 71 Figure 13 Zero Car Households in Metrora i l Serv ic e Area ... ............ .... 73 Figure 1 4 One Car Households in Tri-Rail Serv i ce Area ..... ...... ........ 75 9 \




Foreword This report i s one in a series of eight technical reports being prepared as part of the an overall study Guideway Transit and lntermodalism: Function and Effectiveness. Each technical report addresses one of the major theme areas of research being investigated in the overall rese arch project. The Center for Urban Transportation Research has the primary responsibility for addressing the four subject areas shown in bold in the task list below. This report is indicated by the pointer -Task 1 Task2 Task3 Task4 TaskS Tas k 6 Tas k 7 TaskS Evaluation of Complementary Policies to Support lntennodal Guideway I nvestments Evolving Technology Options Preserving lntermodal Guideway System Competitiveness Integrated Planning and Design of lntennodal guideway Transit Systems Multimodal Transportation Center Design Determining Organizational and Operat i ng Strategies Application of New Techno logies in Developing l ntermodal Tr ansit Systems Assessment of Guideway Transit System Impacts These interim task reports are one component in a broadly defined research project that examines facto rs that i nfluence the success of guideway systems and i ntermodal transportation Each task addresses one or more of the crit i cal cons i de rations These interim reports are intended to communicate find i ngs to date from t h e research that has been carried out. Th ese reports are not fina l r eports nor are they the only output documents for the research effort to date. The Interim Task Reports are complemented by a series of case study reports for selected with guideway t ransit systems and with a three part series of data books that compi l e information from guideway sites 11


The reports are presented in two major sections. The first describes the objectives of the task, the research methodo l ogy and activities completed to date and currently underway. The second section provides a report on the findings to date from research in the given subject area. As work continues the remainder of this year and next year the full work program will be completed and a report prepared that communicates the cumulative findings from the research activities. 12


Task 4: Integrated Planning and Design of lntermodal Guideway Transit Systems SECTION I Task Objectives This report provides an update of activities related to Task Four through July, 1995. Task Four examines the issue of planning and designing integrated systems which include at least one fixed guideway component. Understanding how fixed guideway transi t systems can be successfully i mplemented requires an understanding of how such systems are designed and how that design, in tum, affects system success. The overall goal of Task Four i s to determine which system elements are most important for achieving a well integrated and efficient system and how those elements should be designed. Therefore, an outcome of this task will be more conclusive evidence on the re lative roles of elements like complementary fare structures, coordinated scheduling, coordinated planning activities, and even the configuration of the routes and station locations in achieving successful implementa tion of the transit system. Furthermore, this interm rep ort prov ides examples from case study systems to completed to date of the important elements of an intermodal system. The emphasis for tihis task is on the design of the system. The term system i s used here n ot merely in the generic sense that urban areas have transit systems. but rather in the sense tihat the whole is greater than the sum of the parts that the various components of the system should be viewed as portions of the entire "organism". This is a necessary for any study of an intermodal nature The impetus for much of the current interest in intermoda lism particularly for transit, was the passage of ISTEA (lntermodal Surface Transportation Efficiency Act) in 1991. As t he name implies, the idea is that i ncreased efficiency levels are possible through the use of intermod alism the essence of 13


which is to create a "seamless" transportation system. The assumption is that well designed, fully Integrated systems will be more efficient, attract more rid ers, and be more successful overall. Much of the interest In intennodalism has focused only on the intennodal facilities, with less consideration of the overall system, particularly where public transit is concerned. While this Task considers the individual facilities, and in fact points out the importance of those facilities, it also considers other elements of the overall system. It is hoped that the results of Task Four will overcome some of the perceptual obstacles to considering the entire system can and lead to full realization of the advantages of an integrated intennodal system. The goal of Task Four is to examine the system itself, including routes stations, movement through the system, and consideration of the roles of agencies involved in planning and implementin g the system. One purpose of this research is to provide a clearing house of infonnation related to specific case study urban areas that will help to answer questions related to the design of lntennodal systems. Task Four has been broken I nto four sub-tasks which, together, address the major issues in designing an in tegrated intennodal system. Though these subtasks provide some overlap, they are treated as separate sections in this report The subtask sections include: System network topology Station locat ion criteria and efficiency Coordination of multimodal planning Facilitating convenient transfers While other questions may arise as a resuH of the dynamic research process being used, this interm report provides an update of our attempt to both more thorough ly understand the problem of implement ing lntennodal systems, as well as answering the more obvious questions related to integrated intermodal system design. 14


Guideway TtMslt and lntwmodaUsm: FUnctJon and Research Methodology This Task relies heavily on the case studies of areas that have planned and constructed new fixed guideway systems or extensions to existing systems over the past 25 years. To date. six case studies have been completed in the following metropolitan areas: Atlanta Los Angeles Miami/South Florida New Or1eans Por1Jand, Oregan San Fra.ncisco Bay Area These case study metropolitan areas form the core of info rmation currently available and utilized in this research report In addition, two other metropolitan a rea case studies are nearing completion (Sacramento and San Diego) and, where possible, information from these case studies is inco rporated The case study information is also supplemented, where appropriate, by knowledge and observation of other metropolitan areas by the project team Several of the subtasks have also relied heavily on existing literature for background to the research This is particularly true for Subtask 1, which presents a more theoretical approach to developing the background requirements of fixed guidew ay intermodalism. The l iterature review is an ongoing process and will continue in Year Three of the project. Research methodologies which deviate from the general research methodology discussed here, are discussed in more detail within appropriate subtask sections of the report. This i s mostly in reference to Subtask 1, which uses a topological approach to examining intermodal transit networks in terms of their theoretical advantages and disadvantages 15


Aspects of network topology which are Importan t to this research are discussed in detail in the section for Subtask 1. Research for this Task report also utilizes GIS technology. Subtask 2 uses 1990 census and TIGER file line data in Maplnfo (GIS) to create maps for analysis of station locations from the South Florida case study area These maps are used to evaluate the relationship between fixed guideway components and several socio-economic and demographic variables in South Florida. Again, these specific variables are discussed in more detail in Subtask 2. Activities/Progress to Date Efforts for this Task have followed several lines. First and foremost are the eight case studies, since thl!$e comprise the bulk of the information used in the Task reports These case studies have provided the basic data from which the various sections of this report draw. The second research activity has been the literature review. This effort has been more significant for some of the subtasks, in particular Subtask 1. The rev iew is an ongoing aspect of the overall re search. In addition to the information made available by the case studies and literature review, this Task also represents other efforts. Subtask 1, for examp le, includes the development of a concept paper on the advantage$ and disadvantages of intermodal transit systems. The concept paper provides theoretical background to analysis of intermodal systems and a basic introduction for the findings from the case studies. Subtask 2 has also required additional efforts beyond those in relation to the specific case studies. As mentioned, this subtask includes an analysis of locations and discussion of location criteria for fixed guideway systems. Therefore, maps were created in Maplnfo to allow a visual analysis of station locations related to demographic and socio-economic data for South Florida 16


Planned Activities The bulk of the remaining work for this task is related to the completion of the remaining case study areas. Addrtional case studies will be completed in Year Three. In addrtion to these data collection activities, the literature rev iew process will continue. Regarding individual subtasks, a choice from among the following additional research activities for Task Four will be made at the beginning of Year Three: Expand the topological discussion of intermodal networks to include an analysis of one of the case study systems. This would include some hypothetical node weightings represe nti ng transfe r penalties. Continue the GIS analysis of the South Florida case study, wherein t he optimal stat ion locations on the network are devised using the location/allocation component of the Transcad GIS. Produce map overlays for at least two more systems (Atlanta and Los Angeles) to allow evaluation of transit facilities' locations relative to demographic and socio-economic variables. These represent possible research activities in addrtion to the complet io n of the remain ing case studies. Clearly, the major research task remains the collection of data via the case studies, as well as additional literature rev iew. 17


Technical Findings SECTION2 Subtask 1: System Network Topology Introduction The overall goal of Task 4 is to examine how successful intermoda l fixed guideway systems are designed and planned. A fundamental element of integrated design is the network configuration and the interactions between the different components of that network. From this starting point, Subtask 1 takes an essentially topological view of the intermodal network many aspects of which can influence the relative success of the system. This subtask report focuses on the nodes through which passengers travel for access to the entire intermodal system and, specifically, how we ll those nodes function in that regard. A further goal of this research, which comprises the first section of this interim report, is to more thoroughly understand the nature and advantages of an intermod a l approach to public transit provision, as well as the ramifications and complexities of creating such a system. This report is designed to first simplify the problem, then to consider the difficulties inherent in providing an efficient intermodal system. This subtask report begins a theoretical discussion of the advantages and complexities of intermodalism in urban introduces the diffi cult ies raised by providing intermodal connections (still wrthin the theoretical construct), then proceeds to raise questions which must be addressed by both future research and the providers of Finally, the report offers examples from the case studies to illustrate the extent to which intermodal systems are well integrated, particu larly from the standpo in t of movement through the intermodal nodes or stations. Theoretical Constructs Graph Theory and Network Topology. A heal thy" transit system should p rovide adequate (however that is measured) transportation for the transit dependent, while also offering 18


enough flexibility or savings to attract the discretionary rider, thereby increasing overall ridership and, hopefully, revenues. The question for this research becomes: of what use is an intermodal approach in attaining this goal? The provision of p ublic transport is essentially a response to societal needs and social constructs and is inherently complex, even disregarding the basic logistics of designing, maintaining, and operating a transit system. Given this, His helpful to, at least temporarily, divorce ourselves from the complex issues and attempt to simplify the issues Involved. Graph theory provides one set of tools for such simplification, even for issues related to social needs. For instance, we can examine the poss ibl e advantages of an intermodal approach to transit provision by first asking the following questions: Will intermodalism i ncrease connectivity? Will intermodalism increase accessibility? We can define "connectivity" to mean ... the degree of interlinking of nodes by arcs in a graph or network (Cole and King, 1968 p. 649)." Arcs connect nodes and can be referred to as either edges or links, in the transit sense. Accessibility is related to connectivity, but is not an interchangeable term, as it is generally used to refer to the "ease of movement between places i n an area (Cole and King, 1968, p. 645)". More specifically, His used here to refer to the ease of movement between nodes on the network. For the purposes of this discussion, we will focus first on connectivity and then discuss accessibility since that is what a transit system is designed to provide. Figure 1 shows a graph representing a regional bus network. Graphs are differentiated from networks in part by the fact that they do not Include flows over the edges or arcs (because we are suing transit as an example of this type of ne twork we will refer here to the edges as li n ks) In the long run it will be useful to consider flows, but we are simplifying as much as possible here. The "bus routes" are considered to be links (edges) and the stops, transfer points, and park-n-ride lots are considered nodes on the graph. This topolog ical representation does not include distance relationships, though we can "see" that some routes are longer than others. Again, the goal i s simplicity. The graph is representative of a ''typical" bus system, in that it is essentially radial, with routes running 19


into the central city It is also regional, in that the separate towns also have routes running (again, radially) into the regional center. The network may be efficient from the standpoint of transit provision, but offers relatively few options for the transit user. Figure 2 shows a graph representing the rail network for the same hypothetical region. Again, the edges represent the rail routes. while the nodes represent stops on the routes, including nodes specially designated as park-n-ride lots. At this point, the differentiation between regular nodes and park-n-ride nodes is only for visualization of the network and has no bearing on any topological measurement of the network. This regional rail network, including both commuter and intercity rail, offers another option for the transit user, but when considered by Itself, is a sparse network. This is typically the real-life situation, given the relative costs of providing r ai l transit versus bus transit. Once again, the rider is limited in his or her ability to move around withi n the region. The transit dependent worker would have to devise, within the constructs of the two separate transit networks, a way to move through the region. Figure 3 shows, conceptually, the advantages of considering the entire transit system, rather than the separate modal networks previously discussed. In this case, the "new" network is not the network devised by the transit rider finding his or her way from one mode to another, but rather a network that is actually planned for and operated as an entity The intuitive advantages of considering the total transit system as a single network are easily seen in this example, in that the overall network is more dense, providing more options and opportunities for the transit user. The transit rider h as the choice, for example, of riding a train that might not be as direct a route as an available bus route, but might be faster or more comfortable. The rail commuter, when the entire network is funct ioning seamlessly has more obvious opportunities to switch to bus transit and, therefore, have more accessibility to other portions of the region, once the end of the rail line is reached. The new network incorporates the role of the automobile as well as bus and rail to provide a fuller set of options for the transit user. 20


Bus Route Park-n-Ride 0 Transfer Poim 0 Figure l. H ypothetical Regional Bus Network 21


Urban Rail Intercity Rail --Rail Sta tion 0 Park-n-Ride C Figure 2. Hypothetical Regional Rail Network 22 ,' I I I I I I \ \ \ \


Urban Rail h1terciLy Rail -8u$ROute ParknRide C (traio) \ \ t> Transrer Point Q Imermodal Station C) ParknRide 0 Rail 0 (bus) Figure 3. Hypothetical Intermodal Network 23


While the intuitive advantages of the total system networ1< are easily seen graphically, they can also be quantified, meaning that we can measure increases in connectivity brought about by intennodalism and total networ1< approach to transit provision. Having set up this discussion in the context of graph theory, we again draw upon that theory to answer our earlier question: will an in tenno dal approach increase connectivity? First. it is important to consider why increasing connectivity is useful. Basically, connectivity increases bring with them more options for movement throughout the network. Increasing connectivity of the networ1< is associated with greater ease of movement between places on the networ1< and, accord ing to the earlier definition, higher levels of accessibility This is a worthwhile goal for the transit provider, both in tenns of service to the transit dependent and the attraction of the discretionary rider. Many applications of graph theoretic measures are based on Kansky's (1963) work, which provided an early and tho rough discussion of networ1< measures including those for measuring connectivity. For more in fonnation on graph theoretic measures for transportation studies see also Lowe and Moryadas (1975) and Taaffe and Gauthier (1973). More recent applications of these measures to transportation studies include a study of the emerging airline networ1< in Africa (Gaile, 1988). There are several ways to measure the connectivity of a networ1<. Two simple and commonly used measures of networ1< connectivity are Beta and Gamma. Beta is the ratio of edges to vertices (or nodes). defined as : Beta= v while Gamma is the ratio of edges in the networ1< to the total number of edges possible given the number of nodes or vertices in the networ1<. Gamma is defined as: Gamma= 3(v-2) 24


Both measures consider the ratio of the edges to the nodes, with increasing connectivity arising from a higher ratio of edges (links, in the hy pothetical example) to nodes and hence, indicated by a higher value. Table 1 shows t he Beta and Gamma values for the three hypothetical networks Both measures show an increa sing level of connectivity for the "intermodai" network. Though Gamma shows an increase between the b us network and the rail network, this is a reflection of the scarcity of stops (nodes) rather than the fullness of the network. While both measures show a higher degree of connectivity for the intermodal network this is best illustrated by considering another interpre tation of the Gamma index Gamma can also be interpreted as the "percentage" of connectivity. Therefore, while the bus network is 34 percent connected, the intermodal networ k is 40 percent connected. This simple, hypothetical example shows that increases in connectivity occur simply by evaluating the total network as opposed to separate network for each transit mode, thereby supporting the intuitive conceptual advantages discussed above. Table1 SELECTED CONNE CTIVITY MEASURES FOR THE HYPOTHETICAL NETWORKS Networl< Type Bus Network Rail Network lntermodal Networtc Beta 1 0 1 0 1.2 Gamma .34 .39 .40 SOurce: KJ. Sf'tl.(t(Nt (;( 25


This hypothetical example also raises questions about the use of connectivity as an appropriate measure. If we now reconsider the earlier questions, we can begin to more fully understand the complexities of an intermodal system. Recall that the previously posed questions were: Will intermodalism increase connectivity? Will intermodalism increase accessibility? We have seen, using a simple example, how an intermodal approach can increase connectivity. If we assume that connectivity can be used as a surrogate for accessibility, we can also see how accessibility through the networ1< can increase with a truly interrnodal approach to transit However, It is necessary to consider two aspects of this issue further First we must consider the actual measurement of connectivity as an appropriate measure for intermodal transit systems and possible alternatives. Second, we must consider whether intermodalism might bring with it less than advantageous circumstances as well. Omnectivity in Transit Systems. Regarding the first issue, we have seen that connectivity increases when the entire networ1< is considered and we have furthermore assumed an increase in accessibility arising from that connectivity Ho wever, we must acknowledge tha t for the majority of transit providers. simply Increasing connectivity may no t be desirable or possible (which is not to say that increasing accessibility is not des irable or possible). When using standard connectivity measures for networ1< analysis, connectivity increases as a result of increas ing the number of edges (links) or decreasing the number of nodes. For a network to be 100 percent connected, there would have to be a lin k connecting each node in the networ1< to each other node. While this situation would be impre ss ive from the standpoint of the transit user it is, in the real world, infeasible. Even given the money to construct links between each node on the networ1<, physical constraints rule out very high degrees of connectivity in urban transit networ1

mathematical standpoint, connectivity can also increase by reduc ing the number of nodes in relation to the edges or l i nks. lllis makes little sense when you consider that the nodes represent places that lie on the networ1<. and, in the case of networl<.s, are essentially demand points that need to be served. lllerefore, accessibility in this case can be measured more effectively in some other way than strictly in terms of connectivity. Again, the focus is on the nodes at this point as edges can only be added up to a certain point, and nodes are unlikely to be dropped from the networ1<. (remember that the goal is to improve movement or accessibility throughout the networl<. ... that is not going to happen if nodes are dropped). Nodes take on added significance, both in terms of their location on the networ1<. and, im portantly in terms of their functionality. By functionality, we need to consider the type of function the node serves (e.g., is it a bus stop or i s it a park-n-ride lot at a rail station) in addition to how well performs that function. One way to assess the functionality of the nodes still within the graph theoretic conceptual framework, is to weight the nodes i n relat ion to the total network. Weighting the nodes allows consideration of an i nd i vidual node's functionality and the effect which that ind ividual node's functiona lity, or lack thereof, has on the entire network. Weighting the nodes also allows the network analyst to assess improvements o r problems i n the system which is essential to a system intent on providing seamless transit connections. Furthermore. a node weighting method is flexible enough to use individua l system performance measures to be considered For example, lhe we i ghts can be assigned based on the amount of t ime spent traveling through the node the "ease" with which one can travel through the node, the cost associated with traveling through the node, or any other conceivable performance measure. We have now essentially moved from try ing to show that i ntermodalism i s "good" to trying to understand how best to i llustrate and eva lu ate a particular network in relation to how "good" it is. 27


One general approach to wei ghting the nodes in the intermodal networ1<. is through the use of the Iota index put forth by Kansky (1963) Iota is given by: Iota= M /W where: M = total mileage of the network W =observed number of vertices (nodes) weighted by their function This measure was i ntended to be used in terms of freight movement, where more "high function" nodes meant more capacity through the system However, the same theoretical construct can be used to examine how passengers move through the system particula rly when a node's functionality can include a measure of how well it functions. Clearly, iota is a very flexible measure For a given network, the numerator changes as a result of any of the following changes: links are added or deleted a different networ1<. measure is used (t ime versus d i stance, e g.), or the values for specific links change. The denominator changes because: nodes are added or deleted, the functions of nodes may change, weights are assigned for different purposes, or "improvements" are made to a node or nodes II is really this last category that the transit provider or planner will need to focus on to improve access i bil ity when work ing under other constraints 28


Table 2 shows one example of how i ota can be used to evaluate nodes on a network. Using the same hypothetical intennodal network shown earlier, several weighting schemes were used to measure the iota value. The schemes include: 1) Nodes not weighted at alleach assigne d an equal value of 1 2) Nodes weighted accord in g to function all end nodes assigned a weight of 1, with other nodes' weights being the number of connecting edges multipl ied times 2. 3) Nodes still weighted according to function we ig hts from scheme 2, plus additional value for intennodal connections (e.g park-n-ri de lots) 4) Nodes still weighted according to function intennoda l weights assigned according to the number of modes connecting at the nodes. Tabla 2 HYPOTHETICAL INTERMODAL NE1WORK IOTA VALUES FOR ALTERNATIVE NOOE WEIGHTING SCHEMES Scheme 1 Scheme2 Scheme 3 Scheme4 .70 .18 .16 .1 4 K J. Kamky, Stluc:!ure of Tran$p01!'/M Hetwot*.J.' Ntwf)lt f

The resulting iota values shown in Table 2 are simply an indication of how the weighting schemes themselves might be evaluated, not an analysis of the network or of specific weights. Note that the value of iota drops as weights increase. It is conceivable that a weighting scheme could be devised wherein the values increase with i ncreasing functionality The iota values are decreasing in this example because of increasingly sophisticated weighting schemes, with higher weights attached to higher function nodes C l early, to evaluate a network and its nodes in this fashion, the weighting schemes must be devised i n the beginning. VVhatever scheme is chosen, the matter of interpretation is largely one of common sense, though understanding degrees of (quantifoed) improvement would require additional calibration. Once a particula r weight i ng scheme i s chosen whether in relation to a specific performance measure of functionality, or i n relation to general function, changes in the iota value would then indicate improvements or lack of improvement In the system. Transfer Penanies. The concept of weighting the nodes according to thei r function and/or functionality brings up the second major set of issues that must be conside r ed. Though this theoret i cal discussion has thus far illustrated the advantages of an intermoda l approach, i t is a l so necessary to consider the disadvantage of intermoda l ism This disadvantage i s evident from the previous discussion of the importance of the nodes' functions and functionality in improving access i bi l ity over the system The i mplication of that discussion is that the nodes must function as smoothly or as seamlessly as possible to imp r ove movement through the network. Indirectly, that translates to encouraging more use of the network and l eads further to a point that i s well recognized within the literature: that there are "penalties" associated with transferring from one mode to another or even within the same mode The penaHies reflect the des i re on the part of riders not to have to transfer By transferring, the act of riding transit becomes more onerous and less l i kely to occur for discretionary riders These penal ties, however they a r e measured, are a fact of life fo r an intermodal system. Furthermore, the relative importance of the nodes on the network (versus the edges) and the many possible weighting methods for those nodes imp l ies that some penalties may be higher than others. 30


Much of the literature related to transfer penalties has been in the realm of general travel behavior models but the concepts are equally i mporta n t from the perspective of the i ntermodal transit planner. Most transit operators and planners agree that the perception of inconvenience caused by transfers and the need to transfer discourage peopl e from us ing transit modes (Stuart, 1995) There Is a consensus that time spent on transferri n g is more one r ous than in-vehicle travel time, and even more onerous than other forms of out of-vehicle time, such as walking (Brand, 1989). The theoretical r ationale for such a high value placed on transfer time is that thi s represents relatively unproductive time which is often spent in an unprotected environment and is typically of uncertain duration, even where service schedules are known in advance (P i ckrell 1989 ) All of this i s potentially even more significant when the transfer between modes is considered, which usually requires overcoming even more inertia than transferring between vehicles of the same mode From the standpoint of the previous discussion on node weighting, it can be seen that a node serves a higher function if it serves more than one mode, yet there i s likely a perception of mo r e inert i a to be overcome by transferring between modes. This i s not to say that transferring is necessarily all bad. David Phraner, Director of Transit Operations of the New York Port Authority stated, in a conversation with one of the authors (February, 1995), that riders on the local bus often voluntarily transfer to the express bus to m i nim ize travel time. In Washington D.C. some commuters ride in the direct i on away from their final dest i nat i on in the morning, then transfer to the right train just to get a seat (Pratt, 1995). I n both these instances it is clear that some other factor overcame a n y reluctance to transfer, e g. travel time or comfort. When chosen vo l untarily (and offering some advantage) transfers can provide added convenience and can improve travel efficiency I n these types of cases, the transfer is a result of the rider exercising options, the likelihood of which typically improve in an i ntermodal approach to transit operat i ons, as was shown in the previous d i scuss ion. A number of studies have attempted to determine the relative importance of the transfer to potential transit riders. Many of these studies ranked various factors affecting the travel behav ior or mode choice of the trave l er (Gustafson Curd and Golob, 1971 ; Golob. Canty, Gustafson, and Viti, 1972; Horowitz, 1994) Gustafson et at (1971) i n Horowitz (1994) 31


evaluated more than 30 factors and their rankings indicate that time savings are generally more important than fare savings. Several of the factors can be related to the transfer or waiting time. These factors, along with their rank {out of 32 factors), include: No transfer trip {ranked number 3, following "arriving when planned" and ''having a sear}. Weather protection at pick-up (ranked number 5). Less wait time (ranked number 6). Choice of pick-up time (ranked number 7) Less walk to pick-up {ranked number 1 0). Direct route {ranked numbe r 12). Easy fare paying (ranked number 13). For the intenmodal transit planner, these ranked factors indicate that the transfer is quite important to the passenger, both from the standpoint of not wanting to transfer (number 3) and from the standpoint of how easy that transfer is (the other factors listed). II should be emphasized that these rankings depend on the leve l of service experienced by those interviewed (Golob, Canty, Gustafson, and Vitt, 1972) Not all of the ranked factors are directly related to transferring, but those factors having to do with the beginning or end of the trip are also quite important to the intermodal planner, in that it is clear that these factors can affect the overall impression of ease of use of the transit system. This is even more import ant when more than one mode Is in use. By implication, if the other factors are well addressed by the in tenmodal system then that can help to overcome the fact that people simply don't like to transfer. While it is clear that improving transit ridership requires eliminating (difficult in an lntermodal system) or minimizing the transfer penalty, there is very little research directly quantify ing the impact of the t ransfer penalty on transit ridership. The difficulty of measuring the transfer penalty impact or reflecting it in model structure was encountered by well known researchers in early studies. For example MacFadden and Reid (1975) tried to forecast aggregate travel demand from disaggregate b ehav ior models. Their model included one socioeconomic variable (income) one transportation variable (cost difference), and four 32


mixed variables defined by the after-tax wage for in-vehicle time walk, inHial wait, and transfer wait t ime s The probit coefficient of transfer wait time and initial wait time were relatively significant, while the coefficient for walk time was not. Without further examination, the authors combined all walk inHial waH, and transfer wait time ("second" waH time) into one variable: out-of-vehicle time, thereby suppressing the opportunity for a promising transfer penalty study A number of mode Choice studies have considered the value of time savings and estimates of time and cost elasticHies (Pratt, 1979). Among these models, the role of comfort and convenience was always referred to as an important factor but rarely incorpor ated as transportation policy-oriented variables (Alger Hansen, and Tegner, 1975) Umited by data availability and in lieu of collecting real time waiting data, most of the models assumed that waiting time was equal to one half of the headway. Few models differentiated between walking, waning, or transferring time. When the transfer factor was included i t was often combined with excess t ime, egress/ingress time, or general out-of-vehicle time (Kavak and Demetsky, 1975). There are few sources of detailed data or studies concerning the transfer time alone, particularly not for intermodal transfers. Those transfer studies which have been conducted can be loosely grouped into two categories: econometric models and psychological scaling studies. While H is beyond the scope of this report to discuss these types of models in detail, some of the major findings of those studies are worth reporting. The role of wait ing time and transit comfort first appeared in research conducted in Stockholm (Alger, Hansen, and Tegner, 1975). Their model included variables to measure travel comfort and convenience with the most important comfort variables being "number of transfers" and "chance of getting a seaf'. Since the transfer times were not explicitly recorded, the coefficient for number of transfers was used as a surrogate for the value of transfer inconvenience. By comparing d ifferent assumptions about the arrival patterns of transH riders at transit stations, the authors devised upper and lower limits on the value of waning time occurring during transfers. It was assumed that the range between these limits is a function of the distribution of the transit frequenc ies on the transit network under consideration (Alger, Hansen, and Tegner, 1975). In othe r words, a highly developed transit system wi t h higher service f requencies should reveal little variation in waiting time 33


values, whereas a system with low frequencies will have a larger range between the upper and lower limits. In addition, it was found that, among all of the inter and intra modal transfers, the bus to bus transfer was valued higher than the average transfer value. The implication of this study is exactly as we theorized eartier in this report. That by offering more options, including higher service frequencies, the rider imposes less of a penalty on the transfer and on transit usage in general. In fact, a 1993 survey of riders of the Metrolink system in Los Angeles showed that 64 percent of those surveyed say that their use of Metrolink wo uld increase with in creased service frequencies (Facts Consolidated, 1993). Granted, for the purposes of increasing ridership, it would be helpful to know whether increased service frequency would attract new ri ders but there is an assumption that such increases would occur Moreover, the riders state that the trip would be easier with higher service frequencies. Though this is not explicitly stated. the same ease of use arising from higher frequency could also lead to easier and less time consumi ng transfers, thus reducing the tra nsfer penalty and, in the long run, increa s ing ride r ship. Another transfer penalty study conducted in Taipei (Han 1987) focused on route choice rather than mode choice, since bus transi t is the most importan t public transportation mode and accounts for 40 percent of the total daily passenger trips in the metropolitan area This study collected real time data on walking wailing, and in-veh i cle time for 327 transit riders. Again, the wait time for transferring was not differentiated from the overall wait time, though the inconvenience of a transfer was taken into account. Han (1987) found that the disutility of one bus tra nsfer was perceived to be equivalent to 30 minutes in -vehicle time and that the value of wait time is three times the value of in-bus travel time. This suggests that transit planners may have consistently underestimated the transfer penalties to bus riders, and an optimal network should be more connected than previously thought (Ha n, 1987). This study focused on buses and the derived penalties are not easily repl i cable i n the North American context due to i ncomparable transit densities but it is clear that transfe r penalties do exist. It is up to the transit planner to attempt to reduce thos e penalties, particularly in an intermodal system where the potential for more options is there and should not be 34


Other studies have used psychological scaling techniques to attempt to derive estimates of both penalties and value of time. Horowitz (1978) app lied psychologica l scaling to transfers in the Chicago area. Using Chicago residents In a laboratory setting, the respondents were asked to rate the time spent on a series of common urban trips. Among these trips were several trips that varied according to duration, number of transfers, transfer time, waiting time, need to wait, and seat availability Respondents were asked to rate a series of "comparative" trips against a single "stand ard" trip for which a numerical value had previously been established. Worse trips rated higher on the scale. In this fashion, transit trips which are identical i n all but one element can be compared and the subjective value associated with that element can be isolated. Results from the study indicated that, all other things being equal respondents were indifferent between equal duration trips for automobiles and buses (Horowitz 1978) indicat ing that minutes of travel time are valued essentially th e same, regardless of mode Furthermore, it was found that the first five minutes of a wait are valued higher than either the second or the third frve minutes of wait time. The author concluded that the requirement for a transfer represents a subjective value equivalent to 13 minutes of riding for a 45 minute trip and 8.4 minutes of riding on a 30 minute trip. Further calculations indicated that the "pena lized" t rave l time more than doubles when a transfer is required While this seems very high, there are some important elements to be considered from these findings. First it is clear that the simple need to transfer is seen as more onerous than the amount of time takes to transfer. Whi l e the need to transfer is a necessary evil of an intermodal transit system, it is even more apparent that this perception needs to be overcome and dealt with as well as possible (1981) conducted a follow-up study to determine if a time-independent transfer penalty could be derived from an evaluation of actual t rips Findings from this study reinforce the ear1ier findings in t hat it was concluded that any transfer, r egard less of duration greatly r educes satisfaction with bus transit travel. Furthermore, long transfer t imes are associated with relatively l ow satisfaction and, finally, the time independent transfer penalty is large relative to penalties associated with transfer time, the exact degree of that penalty is not clear. 35


Studies which have attempted to more specifically determine transfer penalties for home based work trips have varied on their results, partially as a result of variation in methodology. As compiled by Charles River Associates (1989), a number of cities' models have incorporated the transferring wait time including Chicago, Dallas/Fort Worth Detroit, Honolulu, Houston, Portland, Seattle, the Twin Cities, and Washington, D.C. Only the Honolulu model included the independent transfer penalty in the mode equations. Though it is generally agreed that transfer times are weighted higher than In-vehicle time, the ratio of these weights varies greatly depending on whether a separate transfer penalty was used. As concluded by Brand (1989), studies that omit transfer penalties can overestimate the value of out-of-vehicle time, particularly when the independent transfer time penalty is omitted. While many models have attempted to derive the independent transfer penalty the lack of consistency and consensus in terms of value range might be another reason that these values are not widely applied. In terms of intermodal transfers, early studies (Alger et al, 1975) indicate that the monetary value and time equivalency of the in termodal transfers varied even more than the previously discussed independent transfer penalties. Since most of these values were derived from the Stockholm study, there is no way to ascertain the validity of these ranges in other settings. One of the difficulties in measuring the transfer penalty lies in the dispute over how many components are associated with the transfer Some argue that the two components include the act of changing from one vehicle to another, which is separate from the other component. waiting time (Brand, 1989; Horowitz 1981) Horowitz (1994) calculated the transfer penalty in dependently from the waiting time and found that, depe ndin g on the conditions encountened, the penalty ranges from 0 to 32 minutes. Others argue that the transfer penalty is reflected by the time it takes to complete the transfer (Stopher, 1995). When the transfe r time approaches zero, the transfer penalty diminishes (Aigers et al, 1975) The previously discussed findings indicate that there is a strong penalty associated simply with transferring supporting the first argument. The magnitude of that pena lty probably varies according to variables such as overall trip time, comfort. convenience, etc. While th ere are strong arguments for both sides, the dispute remains unsett led since no 36


zero transfer penalty situation could be observed. From the standpoint of an intermodal system, the argument that there are actually separate components of the transfer penalty suggests that there are ways to lessen the penalty. This brings us back to the original question, tho ugh now with more complex issues introduced. Will intermodalism help in terms of transit systems' ability to attract and better serve customers? From the previous discussions of both the theo retical advantages of intermodalism and the more practical disadvantages, there is clearly the potential for improvement under an intermodal, truly system-wide (or region-wide) approach to transij lntermodalism provides more opportunijies and a richer, denser network for the transij user Travelers are more likely to take advantage of a system with more opportunities, whether those opportunijies come in the form of more frequent service or more routes, access points, or modes. However, travelers are also discouraged by the need to transfer, whether that transfer occurs between modes o r between vehicles of the same mode. This "discouragement" means that the nodes represent not just demand points on the network but also obstacles to free movement over the netw ork. In an intermodal system, i t is difficult to eliminate the nodes or transfer points (we can broaden the concept of transfer point to inc lude any nodes of ingress or egress to the system, since each involves waiting time and fare provis i on, etc.). Given this, it is essential that the nodes operate as efficiently as possible. Simp ly put, the nodes (stat ions, stops, etc.) must be designed and operated in such a way as to provide necessary connections within the system while also restricting, as little as possible, freedom of movement through the system There is an inherent penalty associated with transferring and that penalty must be reduced as much as p ossi ble Transit p lanners in U.S. Metropolijan areas with guideway systems are faced with this task (though it is not necessarily couched in those terms). This has been approached more successfully in some areas than others, as will be discussed below. What types of elements must be considered in terms of movement through the nodes? Design standards for stations and stops should include cons ideration for the comfort of the patron (this is discussed elsewhere). Also, the connections between modes and vehicles should be as easy as possib le Thi s means attention must be paid to reduction of wait time 37


through more frequent service (if possible), reduction of walking distance between intennoda l ingress and egress, ease of fare payment or transfer use betwee n modes, reliable and oonvenient information availability for other modes' oonnections, and the oost of transfers. lntermodal Connections In Practice While the first portion of this report established the problem in a theoretical sense this section focuses on specific examp les of how systems are implementing intennodalism from the standpoint of creating a "seamless" system. This seaion focuses on the nodes or on the ease of transfer throug h the nodes of the intermodal system. It attempts to answer the questions : if the nodes are so important, then how are they being planned and operated in reality, and what are the difficuhies in moving people through the system? This section of the report offers examples from the case study urban systems that focus on three aspects of the intermoda l system and easy transfers through the nodes: the degree of oonnectedness of the network, the importance of providing automobile access as one form of i ntermod al oonnection, and the importance of providing service and transfer in formation to passengers. It should be noted that another important aspect of the ease of movement through the nodes is in terms of how the transfers and fare payments are made but, sense this is the focus of another subtask, this will be mentioned here only briefty. Each of the above characteristics represents just one element in the "functionality" of nodes on the intennodal network lntermodal Connections. One of the first aspects that must be considered for the intennodal network is the actual number and form of oonnections available to the passenger. The previous discussion about the intermoda l network structure and the impo rtance of the nodes on the network makes some very important points. First, if the intermodal network is to indeed offer advantages, there must be sufficient connections between the modes to allow a denser network (in the same way that the hypothetical network discussed earlier p r ovided higher degrees of connectivity). In the terms used earlier this means that there must be some "higher order" nodes on the network, which offer h igher functional levels. Second, the conneaions between the modes at these higher 38


function nodes must be as fluid as possible. In other words, the functionality of the these higher order nodes must also be high to attract and maintain usage. This section of the report looks briefly at the "higher function" nodes i n tenns of the connections offered in the case study networks. The degree of intennodal connectedness (note that this is different from the connectivity discussed in the theoretical section of this report) varies from system to system This variance is due to a variety of factors, but clearly is related to the overall size of the urban area and its network as well as the number of different operators and options available The types of possible intennodal connections range from rail to rail connections. requiring high capital investment, to pedestrian access to transit, which can be quite simple, but should also be well thought out. This section focuses on rail to rail and rail to "other tran s it" connections. Rail to automobile connection s are discussed in a separate section because of the special significance of those connections. This section looks simply at the available connections, with some attention to the physical "ease" of those connections. The case study i ntennoda l networks offer a range of levels of intennodal connections. Most of the intennodal systems have many connections available, but also have at least one focal point for the intermodal network. Sometimes that focal point o r n ode is a planned node or development and sometimes it is not. For example, the previously mentioned Los Angeles Union Station is designed as the focal intennodal node for the overall syst em. On the other hand. the Market Street Corridor in San Francisco is an area where major intennodal connections occur but is not a planned intennodal node development (CUTR, 1995e). The Los Angeles Union Station offers direct access to all Me tro li nk routes Amtrak i ntercity trains and interc ity buses, other local bus transit, and the Metro Red Line subway (which provides indirect access to the Metro Blue Line light rail route to Long Beach) (CUTR, 1995b) As such, Union Station i s planned as an i ntennodal node with special attention to the ease of access between the modes Union Station als o creates a focal point for pedestrian access to nearby areas. When all intermodal connections at Unio n Station are complete, a bus mall (connected to the dedicated El Monte Busway) will allow easy access to rail platfonns in an attractive setting. During construction these connect ions are more difficult as passengers coming off of some of the Metro link trains must walk to 39


Funeflolfand the end of the platform to gain access to the interior of Union Station. The Metro Red Line station is located on a lower level at the other older end of the station, which means that those passengers transferring between the bus mall and the Red Une will have a longer walk than those transferring to or from the Metrolink commuter rail lines However access to the Red Une is easy clean, and attractive. The Atlanta MARTA rail system also has a high function intermodal node at the Five Points station (CUTR, 1995a) H ere the two main MARTA rail lines converge at a common transfer point the shorter Bankhead-King Memoria l line, as well as 25 MARTA bus lines Connections between the rail l i nes are underground, with separate platform levels connected b y stairs, escalators, and elevators. A concourse level has a transit information center and a store for buying passes and to k ens (CUTR, 1995a) The station also has a special passageway connecting to Underground Atlanta for shopping and restaurant access The Five Points station, then, represents a high function node three rail lines, 25 bus lines, and pedestrian access as well. In Southeast Florida, there is one connection poi nt between the T ri -Rail commuter ra i l system and the Metrorail urban heavy rai l system This single station serves as the transfer point for commuters moving between the two rai l systems Approximately 1,100 people transfer from Metrorailto Tri railthrough this station during an average weekday (CU TR. 19951). The Metrorail station is e l evated above the street leve l Tri-Rai l stat i on with access p rovided by stairs escalators and elevators. Fare pol i cies allow Tri-Railticket hol ders a free transfer to Metro rail (fa r e policy and fare technology are discussed in more detail in another section of this task report) and, during peak c o ngestion times. turnst i les or gates are opened by attendants to allow unconstrained movement between trains Thi s points out an interesting adaptation Many of the case study systems use barrier free fare technology, allowing easy movement between trains or modes However in this instance where the techn o logy use d is not barrier free, a simp l e ad j ustment has been made wh i ch allows more freedom of movement when i t is most important -dur ing rush times. Othe r important intermoda l nodes on the S o utheast Fl o rida networ1< include the two stations connecting Metrora i l to the Metromover people mover sys tem operat i ng i n downtown Miami 40


(CUTR, 19951). These two stations provide an interesting contrast in terms of the design of intermodal nodes. The Government Center station is a bi-level station, with the Metrorail station buiH above the Metromover station and stairs and escalators connecting each to a mezzanine. Here again, turnstiles are in place to control access and fare payment (though transfers are free from Metrorail to Metromover). The Brickell station is no t a bi-level station The Metromove r station is l ocated to the southeast of the Metrorail station, requiring the transferring passenger to walk down from the Metrorail station and across to the Metromover. This configuration is less convenient and, thus far, transfer activity at the station is less tihan at the Government Center station (CUTR, 19951). In tenns of the major intermodal nodes already discussed for other case study systems, all incorporate bi-level configurations except for Los Angeles Union Station. There, the Metro Red Line subway is located at a lower level, which is slightly removed from the rest of the slation, while the other train platforms are located along a concourse While Union Station passengers might benefit from a more centralized configuration, that is also a historical station which must incorporate many of the original design constraints. The major intermodal nodes for the Los Angeles, Atlanta, and, to a lesser extent, Miami networks represent planned attempts to develop major high functio n intermodal nodes. The San Francisco network offers a major intermodal focal point that is not a planned development (CUTR 199 5e). The Market Street Corridor, spanning approximately two miles from the Ferry Building to the Civic Center BART/Muni station offers an array of intermodal connections including four bi level BART/Muni stations, ferry boats linking San Franc isco to Marin Alameda, and Napa counties, 26 Muni motor and trolley bus routes operating on the surface of the corridor, several express buses termina ting in the corridor, as well as Amtrak buses running to the Ferry building, and t hree cable car lines term i nating at Market Street. Surface streetcars will operate on Market street beginning in late 1995 (CUTR, 1995e) While the connections between these various modes are not as integrated as some of the other intermoda l facilities al r eady discussed, they represent ... a diversity of transit modes operations and transfer opportunities probably unequalled (sic) in North America (CUTR, 1g95e, p. 85) 41


These examples of major in termodal nodes on the case study networks do not tell the whole intennoda l story. While much effort can be expended planning and designing those major nodes there are also other connection points which serve as higher o(!:ler nodes over the intermodal network though they are not such extreme focal points. From the standpoint of the earlier theoretical discussion, these also serve as highe r order nodes which would be weighted differently from other nodes because they serve an intermoda l function. For example, while Los Angeles' Union Station is the highest order node, there are a total of 64 rail stations on the intermodal network (CUTR, 1995b). Of these. 37 have parking facilnies (discussed in the next section) and 55 have connections to "other transit", i.e bus lines. Furthermore, the Metro Red Line subway connects to the Metro Blue Line light rail system which will, eventually connect to the Green Line, scheduled to run between Norwalk and El Segundo. All of these connection points become more important over the intermodal network because they serve higher functions as transfer points and, as such need to offer higher functionality as well. In South Florida, besides the stations already mentioned which allow rail to rail transfers, there are many which have other intermodal connections. Of the 15 Tri-Rail stations, 5 are shared with Amtrak lntercijy trains, 14 inc lude parking lots, and 12 have bus feeder routes. The remaining three stations serve regular fixed route bus lines (CUTR, 19951). Furthermore, 17 of the 21 Metrorail stations have parking and each station is served by at least one bus route. Those stations that offer rail connections, auto connections, and bus connections serve higher functions on the intermodal network. In San Francis co rail to rail intermodal connections include 5 stations connecting BART to Muni light rai l, and one connection between Caltrain commuter rail and the Santa Clara County light rail system in San Jose (CUTR, 1995e). However, of the 107 rail stations in the San Francisco Bay area, 95 have direct connections to "other transit" typically bus transn. Finally, as was previously mentioned, BART does have a station loca ted near to the Ferry Building, which allows easy access to ferry operations. In fact, in this system, the most difficult intermodallin kage is probably that between the Caltrain commuter rail lines and other rail operations in San Francisco. The location of the Caltrain terminal (discussed in another section of this Task report) is removed from the center of the intermodal activity and is somewhat more difficult to access, especially for those unfamiliar with the system or those carrying loads etc. 42


Another case study system represents an evolv ing system which has yet to provide many connections between the modes. New Orleans can be considered an example of intermodallsm as "a work in progress (CUTR, 1995c). Current connections between modes are minimal. The two streetcar lines in place do not connect at all, while the Riverfron1line offers one bus transfer point and the St. Charles line has seven bus transfer points. The Riverfron1 bus transfer point Is physically removed from the rail stop and requires several blocks' walk. The New Orleans Union Passenger Terminal, which serves Amtrak and Greyhound is not currently connected to the streetcar system, but is about four blocks away. However, the planned Canal Street Corridor Project will provide linkages between the streetcar lines and a possible light rail link to the airport. The previous discussion extends the concept of some nodes o n the network being higher order or higher function nodes. Those serving more modes will, of necessity, become more cen1ral to the network. In terms of the earlier theoretical discussion, these nodes might be weighted more heavily (depending on the chosen weighting scheme). On the other hand, the degree to which t hese nodes provide easy connections between modes is also paramount to the efficient implementation of an intermodal network. The following two sections address just a couple of the elements of prov iding easy connections, though the following section expands the intermodal connection discussion to focus specifically on t he role of the automobile in a successful intermodal network Automobile Access/Parking. One element of the successful integra ted intermodal system is how well that system addresses the automobile as a component of the overall system. For example, the Los Angeles transit network includes Metrolink commuter rail, inter-urban service offered by Amtrak, the Blue Line light rail line to Long Beach t he Red Line heavy rail subway system in downtown Los Angeles and numerous bus operators. Even with all of these examples of transit o pportuntlies it is essential that the intermodai system not ignore the automobile, particularly in an automobile oriented soc iety (a society which is epitomized by the Los Angeles region) Our earlier hypothetical regional t ransit network in cluded park-n-ride lots as "higher function nodes", in dicat ing that they serve an important purpose within the overall network. That importance is demonstrated in the Los Angeles intermodal network by the fact that 77 percent of the Metrolink commuter rail patrons drive 43


alone to the Metro llnk stations (this does not in clude kiss-n-ride patrons) (Facts Consolidated, 1993). Los Angeles serves as the prototypical automobile-oriented city, but other U.S. cities that have grown rapidly during the automobile era face the same restraints Impo sed by the reliance on cars. In Miami, for example, 48 percent of Tri-Rail commuter rail patrons gain access to stations by car, while 26 percent of the Metrorail users do so (CUTR, 1995f). In San Francisco, a 1989 rider survey found that 45 percent of Caltrain commuter l ine riders drove a lone to the train station. while BART data indicate a figure of 38 percent for riders from the East Bay (CUTR, 1995e) As expected, Los Angeles residents seem most attached to the automobile as a source of access to transit, but clearly, this phenomenon i s not limited to Los Angeles. Generally, if inadequate provision is made for t hose using their car for commuter connections, they are less likely to ride Within the Los Angeles system some rail lines that have been more amenable to car drivers than others. This arises, in part, from a on the part of transit providers of the importance of park-n-ride lots. The Southern California Regional Rail Authority (SCRRA), respons ible for the Metrolink commuter rail system, requires each station coming on line to provide a minimum num ber of parking spaces if possible. For example the Santa Clarita station on the Santa Clarita Metrolink line has made very good provisions for parl<-n-ride patrons, though, interestingly, Santa Clarita has also been relatively successful at getting patrons to ride the bus to the Metro link station (Figure 4). The Santa Clarita station has parking faciiHies and is also able to expand its parking capacity in times of heavy station use. Specifically, an agreement with a neighboring speedway allowed the speedway parking lot to be used during the transportation emergency created by the January, 1994 earthquake. Parking is free (as it is for the Miam i Tri-Rail commuter t rains) and the lots are clean, attractive (for parking lots), well lighte d and promote a feeling of security 44


Figure 4 Park-n-Ride Lot as Santa Clarita Metrolink Station At the other end of the spectrum is the Metro Blue Une light rail runn ing between downtown Los Angeles and downtown Long Beach (CUTR, 1995b). While this line is well-designed and has been relatively successful, it has had to play catch-up in the realm of providing amenities to park-n-ride commuters Blue Line plans did not include park-n-ride lots at several of the stations situated between L.A. and Long Beach (those stations running through south central L.A. e g.) Furthermore planners in Long Beach apparently deliberately sought to downplay the role that park n ride commuters would play in the success of the Blue Une. Due to fears that the loca l street network could not hand l e additional traffic, there was an attempt to discourage transit users from driving into Long Beach to catch the Blue Une. Therefore relatively few park-n-ride spaces were made available. However as shown in Figure 5, commuters began to park along the sides of roads adjacent to stations 45


Figure 5 Lack of Park-n-Ride Space Leads to On-Street Parking at Blue Line Station A similar situation exists in Atlanta, where a MARTA station without parking facilities is located in a suburban center where parking is available along streets and in other off-street lots (CUTR t995a). In the Atlanta case, sufficient parking exists outside the purview of the station, whereas in Long Beach, as a result of the unexpected (or unsought) demand for parking, plans are currently underway to build more park-n-rlde spaces, including a covered parking garage to accommodate commuters in Long Beach In the Long Beach case, an attempt to avert one anticipated problem (increased traffic) led to problems which now must be dealt with retroactively. A similar a ros e in Miami with the Tri-Ra il commuter rail system (CUTR 19951). In that case original p lans for parking were inadequate due, in part, to the fact that Tri-Rail originated from attempts at congestion Inadequate parking at several of the Tri-Rail stations resulted in the Metr o Dade Authority recently opening an additional parking garage (the 2000 space garage is already operating close to capacity) As these two examples show for those urban areas intent on implementing an intermodal system, i s necessary to integrate the role of the automobile in the planning stages i n order to help make the overall system successful. Furthermore planners should consider the consequences (both financially and politically) of hav i ng to go back and p ro vide service once the network is already in place 46


In some instances, this connection between automobile interface and successful intermodal systems is l ess clear, due eijher to the uniqueness of the urban situation or the specific l ocation of the stations within the urban area For example, Portland's MAX light rai l system curren tly serves 30 stations, only 5 of which have parking available (CUTR, 1995d) Parking is planned for several other stations but will not be offered at all in the two innermost fare zones of the MAX system There a r e several reasons why, i n this instance the system can have a relatively high level of success First Portland has the advantage (from the standpoint of the transit promoter) of a public mindset that is more or less favorable toward public transit, in part due to an environmenta ll y conscious populace and also an urban history that includes early implementation of planning and land use con s t raints Secondly, those zones where no parking will be offered are the zones i n which it is u nl ike l y that a commuter wou l d be driving lo gain access to the system anyway These zones are i n the downtown area and in an a r ea whe r e access by car and parking prov ision would be difficuH due to the bu itt e n vironment. A similar snuation is seen in the San Francisco region, where parking is ava i lab l e at outlying systems' stations, but no parking is offered at the Mun i light rail stops (CUTR 1 995e ). The Muni system runs i n down town Sa n Fr anc i sco and it Is unl ikely that the average ride r needs parki n g at the stops themselves. However, parking facilities at Bart and Caltrain stations are seen as more important with Bart offering parking at 21 of its 34 stations and Caltrain offering parking at 23 out of the 34 stations (CUTR, 1995e) In At l anta 24 out of the 33 MARTA rail station have parking ava il able. Of the 9 stat i ons without parking, 7 are located in downtown Atlanta and the Midtown stat i on offers on l y 10 spaces (CUTR, 1995a). Information Dissemination. While providing an adequa t e interface between automobile and t r ansit is one way to improve in termoda l connections, it i s a l so necessary to provide sufficient i nformation to the transit user as to his or her intermodal transn options Potential transit users can be disco u raged by the realny or even the perception of inadequate i nformation about service schedules and connections Even seasoned transit users can expe ri ence frustrat i on when try i ng to swnch between modes. Simply put the intermoda l transit operators should make the connections between the modes as easily understood 47


Gultlwiay Transit and Function as possible. There should be no mystery about making intermodal connections, nor should there be any perception of mystery. The method of information dissemination can be technologically advanced, as in computerized, real-time displays of connections, or can be simple, as in the provision of printed hard-copy transfer and routing schedules. For example BART stations serving more than one li ne have electronic signs indicating which train is entering the station next Los Angeles' Union Station has an electronic board providing information about arriving and departing trains for both Amtrak and the Metrolink commuter rail system. A similar electronic board is offered for bus passengers In Portland's transit mall, but only provides information about bus connections (CUTR, 1995d). This form of information is not the most technologically advanced, but does offer a centrally located, easily understood method of dissemination. On the other hand, the Union Station board provides no informat ion on any of the other modes connecting at the Station. That Is left to the staffers at information windows which or to racks of printed bus schedules for those who want to transfer to one of the many bus routes running out of Union Station The information windows for the various operators in Union Station are not located in d irect proximity to one another. While this is not much of a problem for the average traveler, it can pose difficulties for the unseasoned traveler or those dealing with heavy baggage or who find walking and standing difficult. In terms of the bus schedule racks, these are a perfectly acceptable way to offer informat ion, but, during one visit in the middle of the day to Union Station these racks were almost empty (Figures 6 and 7). In this case, while the informatio n dissemination method is useful in theory, the realized p otential of that method suffers in practice. L os Angeles' Union Station also offers examples of the most basic type of information: signage. Signage is especially important in large, intermodal connection points like Union Station because of the multitude of modes and routes operating out of a single station. Union Station's signage ranges in usefulness though there are no significant problems. For the passenger departing fr om Metrolink or Amtrak trains and wanting to find the Metro Red Line subway or ticket informat ion for another carrier, for example, the signs are sufficient for travelers familiar to the system, but coul d be better for the unseasoned t raveler. For commuters this should not present a problem unless some potential users 48


Figure 6 Bus lnfonnation In LA. Station 49


Figure 7 Bus lnfonnation In L.A. Station are scared off by the perception of difficulty. In terms of the Metro Red Line subway, information is readily available once the passenger has descended to the lower level (which requires some vigi la nce to find on the part of the unfamiliar traveler) Once in the Red Line area, signage is clear and ticket machines are easily a ccessible Furthermore. maps of the route and the stations themselves are particularly useful. Electronic signs provide an indication of the arrival of the next train. This i s interest i ng, but the schedules posted i n the station provide sufficient and clear i nformation. At individual statio ns, the Red Line includes s i gns with "blowup" maps of the area surrounding the station This i s very useful to the passenger who is unfamiliar with the area (tourists for example) The passenger can get a feel for the di r ection to ta ke out of the station before they are actually on the street Union Station offers an example of a high profile, at1ractive intermodal node on a large and still emerging network where the abi l ity of the transit user to easily switch from one m o de to another might be hampered by the way in which information is or is not disseminated. 50


Other examples show how even the simple approaches to information dissemination can be helpful when applied appropriately In the M i ami area, for example schedules for the Tri-Rail commuter rail system, as well as bus connection schedules are offered at the Douglas Road Metrorall heavy rail station a major intermodal transfer point (CU T R 19951). In that way, it is poss i b l e for the commuter rail passenger to check connections before leav ing the station, or conversely to peruse the schedule in relative comfort whi le ri ding Th i s eliminates the stress of arriving at an intermodal t ransfer point and having to decipher schedules while trains or buses are actually arriving and departing. I n Atlanta, a ll MARTA ra il stations offer transit information (CUTR, 1995a) in the form of area maps, train schedules, and connect i ng bus lists. Furthermore the Five Points MARTA station which serves as the focal point of the i ntermoda l network, offers a transit information center on its concourse level wh i ch is manned and offers all of the above forms of information as well as bus schedules for all r outes and sales of passes and tokens i n bu l k Th i s is possible, in part, because the transit network is a l most totally under the contro l of a single agency (trains at one MARTA station connect to Cobb County Transit buses ) therefore a ll owing cen t ralized i nformation dissemination. Contrast this w i th the need to dea l with several different age n cies ( l ocated at different windows i n the terminal) at los Angeles Un i on Station, for example and it is easy to see how the larger more dece n tral ized systems have a g r eate r d i fficulty in some respects. Whi l e a large system such as that found in los Angeles or San Francisco offer, on the one hand the p otential for a denser network they also bring with them i nter-agency re l ated complex i ties that run the gamut from planning an d ope r at i ons to simply providing i nformation These examples high li ght an i mportant aspect of designing i ntermoda l trans i t networks: if the passenger perceives that it is difficult to figure out transfers then the penalty" associated wit h that transfer is like l y t o be h i gher. T his discussion has focused on the ava il ability of connections between modes for the case study systems, as well as some aspects of how easy that connection i s This is not a comprehensive l i sting of the elements that make up a convenien t transfer, but i nstead rep r esents some of the most important aspects of devis i ng an integra t ed inte r modal system particular l y with respect t o the 5 1


previous theoretica l discussion of the adVantages and disadVantages of intermodal systems. Subtask 2: Station Location Criteria and Efficiency Introduction The previous examples have focused on the ease with which transfers are made in terms of provision of automobile access, information, and the availability of access to other modes. Another element of the overall network that must be considered is the efficiency of station locations. The previous discussion focused on the functions of the nodes and some elements of the functionality of the nodes Those nodes, or stations, are transfe r points between the modes and routes of the same mode and they are demand points for access to the network Therefore, the location of the nodes on the network and in relation to the urban area through which the network runs is vital to a successful intermodal network. Constraints to optimality Station locat ions, and the affects of those locations, are significant in an intermoda l system for two main reasons. The first, and most obvious, reason is that they serve as access points to the system and must be located appropriately for access to major origin and destination points In the urban area. Secondly, stations and entire routes are often located for reasons other than efficiency or optimality. Particularly in buil t up urban areas or in cases where existing or historic lines are used, the stations are located according to factors other than optimality. Political divisiveness, financial constraints, environmental obstacles, lack of adequate land, inertia or neighborhood opposition can derail (!) plans for individual optimal station locations. Given these constraints, "optimality" becomes an inte rest ing concept. On one hand optimality can be defined strictly in terms of stations' locations relative to demand points on the network (wherein they would be located so as to minimize distance traveled for access 52


by the most users). On the other hand, the loc:ation selected for an essentially public facility such as a transit sta tion or route must also consider equity issues. Trans it must serve the transit dependent as well as the discretionary rider, implying that a station's location relative to specific transit dependent demand points could be used to determine optimality. This is particularly important in cities such as Miami or Los Angeles, where equity issues have been raised concerning the amount of public spending on rail versus bus In these cases, rail transit is seen as catering to a wealthier and smaller population, while bus service serves a large r population, but is provided with less money and fewer service increases. In theory, some combination of these considerations should be considered optimal. However in practical terms, and given the previously mentioned constraints faced by transit planners, "optimality" may instead be defined according to finding the location that best serves the purpose while minimizing related problems thereby becoming a "satisfactory" location. In terms of major intermodal stations as high function nodes on the network, the situation in San Francisco offers an example of the importance of good locat ion versus a satisfactory l ocatio n As was mentioned previously, the focal point of the intermodal system in San Francisco is the Market Street Corridor (CUTR, 1995e). This does not represent a single node on the network, but a series of closely tied nodes which, together, provide many intermodal opportunities. Missing from this intermodal focal point, however, is the Caltrain commuter rail system. The CaHrain terminal is located 3/4 mile away from the combined BART/Muni station at Powell Street. Bus service provides access to the Caltrain terminal, but the intermodal connections are not good, particularly for passengers who are in a hurry or having to carry luggage. Several solutions to this problem have been proposed including a surface extension of the Muni light rail line to the Caltrain terminal which could begin operation as early as 1997. An alternative proposal is to relocate the Caltrain terminal, clearly an expensive undertaking Given constraints imposed by the surrounding urban environment, a move to the Embarcadero BART/Muni station (approximately 2 1/2 miles from the existing Caltrain terminal), will require underground trackage and an underground station. While either of these alternatives will require heavy investment the potential for improvement of the transit network is great since the current l ocation is detrimental to the overall intermodal network. 53


Severa l examples from the case study urban areas demonstrate further how outside influences affect station locations and, In some cases affect the efficiency of the station on the intermodal networl<. The Metro Blue Line light rail runs between downtown Los Angeles and downtown Long Beach. An official with Long Beach Transit has described the station location and plann ing process for the B lue Line as a series of give and take discussions among a variety of interested parties Included in discussions at various times in the planning process were homeowners, business leaders and representatives of Long Beach an d the Los Ange les Metropo l itan Transit Authority, which operates the Blue Line. The inclusion of these groups was in recogn i tion of the political aspects of providing rail transit. Inclusion in the planning process of these groups with their different interests resulted in at least o ne station site being abandoned. It was originally intended that a Blue Line station be l ocated at the historic Cerritos station bui l ding, a station on the o l d Red Car line between Long Beach and Los Angeles early in this century. However, homeowners in an adjoining wealthy neighborhood, fearing noise and other negative impacts, raised opposition whi ch resulted in the demolition of Cerritos station and construction of the Wardlow station further south along the Blue Line. Since this increased ridership and a lack of l and for expanded parking at Wardlow stat i on has raised d emand for a new station located ironically, near the original Cerritos station si t e. However concerns that the "new" (old?) site may not be environmentally feas i ble because of oil wells and potential contamination This single station l ocation choice provides an example where historica l precedence (inert i a), citizen opposition, land avai l ability prob l ems and environmental concerns have affected the decision. Another i nstance where cit i zen opposition affected a station l ocation is in the Metrorail system in Miami As men t ioned in a previous section the Brickell station on Miami s Metrorail line is an intermodal transfer point to the Metromover people mover (CUTR 19951). The original plan was for the Metrom o ver to run along SW 1Oth Street, but property owners and the City of Miami opposed that alignment, resulting in the alignment along SW 11th Stree t. further to the south. The result is an interrnodal transfer sta t ion with an inconven i ent transfer. The o riginal intent was to provide a bil eve l stati o n with easy 54


connections between the light rail and the people mover, but the actual configuration requires a transferring passenger to descend from the Metrora il station cross to the Metromover station and then climb to the station. In this instance, changing the station location created a less efficient intermodal transfer point. Surveys have shown that this station has low transfer activity when compared with the Government Center transfe r station (CUTR 19951), though the effect of the station relocation and resul t ing re configuration is not definitive. While citizen or homeowner opposition can force site changes, other groups might encourage a specific stte for other political or financial reasons. I n Santa Clarita, a town on the Metrolink commuter route running from the high desert into Los Angeles, one local developer applied pressure to locate a station in a suboptimal location. While no t located close to existing demand points, the intent on the part of the developer was to encourage development on land already owned by the developer. This is obviously not without precedent, since many trolley lines built during the late 19th and early 20th century were built at the behest of real estate speculators, who benefited from the increased acc ess I n this case, the station was not located on the develope(s land even though the proposal included land for the station. From the standpoint of the total network, there may actually be no specific method or set of criteria for locat ing individual transit facilities. For example, the Southam California Regional Rail Authority (SCRRA), which operates the Metrol i nk commuter rail system i n the Los Angeles reg ion, does not have specific guidelines by which station locations are selected Rather, municipalities along the routes (the routes reflecting existing lines to a great extent) must ask for and be willing and able to fund stations. Once an agreement i s reached on whether a station will be placed in a municipality, the actual station location decision depends on a variety of localized factors The SCRRA offers a set of guidelines for design ing the stations, but the location decision i s up to the municipal planners with input from SCRRA. While the specific station location situation in the previous example requires agreements between the transit agency and the municipality, it is often the case that no specific method 55


for site location is used. F or older systems, in particular, it is difficult to reconstruct how actual site decisions were made For example, the Peninsula Joint Powers Agency, which provides service between San Francisco and San Jose, uses lines and stations that date from the 1880s (CUTR, 1995e). Station locations were chosen to develop new town sites or they followed new, emerging settlement nodes. For newer, heavy rail systems (MARTA, e.g.), proximity to major arterials and local highways have been important selection criteria for station locations. However, even in newer systems. historical precedence can be important (for example, the attempt to use the Red Car station in Long Beach, discussed above) Many systems rely on existing lines to begin service, such as the Metrolink system in Los Angeles, which made extensive use of the ability to acquire existing l ine and right-of way from freight companies. Along the same lines, a proposed five line light rail commuter system in Tampa, Florida i s under d iscussi on only because a relatively low cost approach using existing trackage, Is possible. In that case, the existing lines run through high demand areas, though construction of new lines would allow better alignment through high density areas. An important point to consider is that once station and route decisions are made, they continue to affect locations for years to come. This has been the case in the past and it w i ll l ikely continue to be the case. There are two main reasons for this "inertia" First the cost of developing and constructing new lines is prohibitive, as in the Tampa situation or, for that matter, the Metrolink service in Los Angeles, where the timely opportunity to acquire trackage and right-of-way from freight operators allowed the project to go forward. Secondly, development patterns emerge in part from transportation improvements, including rail. This is certainly evident in the Los Angeles region, where many bedroom communities flourished during the early part of this centuJY as a result of the Pacific Electric Red Car rail service to Los Angeles (Crump, 1978). The Red Car lines affected the urban development pattern of the Los Angeles region. Given the chicken and egg argument concerning the ties between development and transportation, it is likely that most station locat i on decisions reinforce the existing development patterns if they affect them at all and, furthermore, that current location decision which utilize existing lines or stations, will in fact be serving major demand points barring large-sca le changes in land use. 56


South Florida Station Locations. Having discussed the many external influences on location decisions in guideway this section of the report examines the South Florida intermodal network from the standpoint of station locations and geographic accessibility to the network. The discussion centers on a simple analysis of the efficiency of the locations, but also raises equity issues The South Florida system covers Palm Beach, Broward, and Dade Counties and includes the Tri-Rail commuter rail, Metrorail, and Metromover people mover, as well as fixed route and feeder buses. Shaw (1991) has examined accessibility to the Tri-Rail system in Broward County. That analysis used buffers in a GIS to determine the leve l of accessibility at both origin and destination ends by automobile and feeder buses. That study concluded that for the Broward County stations, automobile access (which is most important at the home end of the trip) was sufficient in that most of the portions of the study area were within 20 minutes of a station. This was deemed sufficient because earlier surveys had shown that 80 percent of Tri-Rail riders traveled 20 minutes or less to their station (Shaw, 1991; Shaw, 1989). Bus service on the other hand, provided less accessibility on the work end of the trip. By assuming a 25 minute maximal bus travel time from Tri-Rail, it was found that slightly less than 35 percent of Broward County employees had access by direct bus route, to Tri-Rail (Shaw 1991 ). By considering both fiXed route and feeder bus service, w ith the same time constraint approximately 39 percent of Broward County employees had access to Tri-Rail under the existing feeder bus and route structure (Shaw, 1991 ) The study points out that feeder bus service is more popular with Tri-Rai l users because of shorter wait times, hence easier transfer between the modes. The study also proposed an alternate feeder system that would increase the number of employees "covered" by 20 percent. The above Tri-Rail study completed raises some interesting issues. As Shaw (1991) points out, the preponderance of riders using automobiles for access to Tri-Rail at t he home end of the trip supports the necessity to provide more parking access. This is in keeping with the earlier discussion of automobile access to intermodal systems. However, the study also points out that accessibility for those without automobiles is not good. If most of the 57


populace of Broward County is within a 20 minute automobile ride to Tri-Rail, what about those who do not have access to automobiles? This Is particularly important In that the bus service evaluation was not as conclusive in terms of easy accessibility . Shaw (1991) suggests identifying areas with low auto ownership levels and providing trans", but it was already shown that trans" does not provide as much access i bility even at the work end of the trip. To gain a better understanding of the location of Transit facilities in South Florida a series of maps was created in the Maplnfo GIS. These maps were created us i ng TIGER line files and 1990 census data. The data were mapped at the census tract level and overlays were created in the GIS for the following variables: Populat i on Density Median I ncome Households Wrth No Vehicles Households Wrth One Vehicle The demograph i c and socio-economic overlays were compared with overlays for the major fixed guideway systems in South Florida to evaluate station and route locations. Maps were created for both the Tri-Rail commuter rail line and the Metrorail system Because of the broad geographic extent of the Tri-Rail line separate maps were created for the Metrorailline. Note also that the Metromover people mover line in downtown Miami is not specifically Included This line runs through the downtown area and a demographic analysis of the station locations would not prove useful. In the future, an analys i s of surrounding land use and employment pockets would prove more useful. Figure 8 shows the Tri-Rail system in re l ation to populat ion density over the three county region of Broward Palm Beach, and Dade counties Popu l ation dens"y is an important indicator of potential trans" ridership (CUTR 1994) However, the nature of commuter rail, in that it connects outlying, suburban areas to more centrally located urban areas i s that stations have a higher likelihood of being located in relatively low-density areas Furthermore, the high cost and other locational constraints (particularly land availability) 58


Til-Rai l T fi-R ail stops ........ Metrorail 0 Intercounty Bus Transfer Stations Popu l ation Density (Persons per square mile) 7700 to 31400 (133) ssoo to noo (116) 3700 to 5500 (131) IJ 2000 to 3700 ( 1 30) 0 0 to 2000 (132) Figure 8 Population Dens i ty for Tri-Rail Service Area 59


This is a blank page


previously discussed can also discourage station locations in high density areas. This is the case for Tri-Rail, which was buiH using existing trackage In lower density areas. Of the fifteen Tri-Ra il stations shown in Figure 8, only two are located in census tracts which fall into the highest density category (J700 to 31,4000 persons per square mile). An additional four stations are located in or very near tracts which fall into the second highest density category (5500 to 7700 persons per square mile) The remaining nine Tri-Rail stations are located in tracts with relatively low population density. It is Important to note that this three county region has relatively low population density overall with Dade County having the highest density (J ,538 persons per square mile is the average tract level density), and Palm Beach County having the lowest density (1 ,989 persons per square mile is the average tract level density) (CUTR, 1994). For comparison Figure 9 shows population density surrounding Metrorail stations in Dade County. Metrorail serves a much more urbanized area than does Tri-Rail. Of the twenty one Metrorail stations all but seven are located i n (or imme diately adjacent to) tracts of the highest density category. Four of the remaining seven are located in tracts falling into the second highest density category. Th e lo west density stations are those serving t he area south of downtown Miami, an area with lowe r populat i o n densities overall. From the standpoint of strict population density (and overall demand) Metrorail stations appear to be better located with respect to p opu lat ion densities than are Tri-Rai l stations. On the other hand commuter rail lines such as Tri-Rail are, by their nature more likely to run through lower density, suburban areas. Also as mentioned, other constraints both fiscal and environmental, can inhibit locat ions based solely on population densities. One question for Tri-Rail then becomes whether locating more stations in higher density areas would increase ridership? Much is made about the demograph ics and higher socio economic status of the average commuter rail rider in U .S. metropolitan areas, and these station locations might best serve the riding populace. This is particularly true given that the commuter rail lines are mostly funneling passengers between the suburbs and downtown locat ions. The socio-economic status of downtown workers is generally higher than that of typical transit riders. This situation gives rise, in part, to the ongoing debate about the equity of provi d ing funding for rail transit serving the wealthier discretionary rider 61


versus funding for bus service for the transit dependent. The demographics of the commuter rider could be significantly d i fferent if stations are located strictly according to density. While other criteria should also be considered, density does provide the proximity to attract riders. However, proximity is a relative condition. As was stated eartier in relation to Broward County, most of the population of Broward County is within a 20 minute ride of a station (Shaw, 1991). This might provide proximity for the person who drives to a stat i on (though that's not an insignificant drive in but does not mean much for the rider who has no car). Clearty, there are other variables that can be used to evaluate station locations, particularly when equity issues and service for the discretionary rider versus the transit dependent rider are cons idered. Figures 10 and 11 show the median i ncomes of areas surrounding both Tri-Rail and Metroraillines. For Metrorai l stations, the pattern i s basically the reverse of the density pattem. The majority of the stations which were located in higher density tracts are also located in or adjacent to lower and lower-middle income tracts. For the most part, the stations located in lower density areas to the south of downtown M i ami are serving higher income areas. In this instance Metrorai l is serving (at least from the geographic standpoint, if not from the ridership standpoint) both groups of riders. Those stations located in tracts high densities and lower incomes offer efficiency from the standpoint of offering service to large numbers of people and are also offering service to people who are more likely to be dependent. In the southern port ion of the Metrorailline the densities are lower and the incomes are higher Here. it is less likely that the dependent are being served (at least from the home end of the trip), but densities a r e still high enough theoretically, to warrant service For the Tri-Rail stations, the situation is less clear cut As shown in Figure 10, the majority of Tri-Rai l stations are located in or adjacent to tracts With median incomes above $20,000. From the standpoint of service for the transit dependent, number of cars in the household should a lso be eva l uated Figures 12 and 13 show the percentage of households, by tract, with no cars for Tri Rail and Metrorail respectively. Of the fifteen Tri-Rail s t ations seven 62


Figure 9. T ri-Rail Stops Metrorail 0 Metrorail Stops Population Density (P....,.,. per sqUOte -> 7700 to 31400 (133) 5.500 to 7700 (116) 3700 to 5500 (131) 0 2000 to 3700 (130) 0 0 to 2000 (132) Popu lation Density for Metrorail Service Area 63 .


This is a blank page


are located in or adjacent to tracts with 15 to 40 percent of the households being cartess. Furthermore, as shown in Figure 14, a majority of Tri-Rail stations are located in tracts with a high percentage of one car households However, without knowing the household size and age characteristics of these tracts it would be premature to assume that these represented transit dependent households Either way, Tri-Ra i l stations are at least at the broad geographic scale of census tracts located well with respect to zero car households and fair1y well with respect to low income households. Given that the majority of Tri-Rail riders drive from their homes to Tri-Rail stations, it appears that the geographic relationsh i p between stations and tracts with a high percentage of zero car households is better than ridership from these households would indicate. For Metrorail, the situation is much the same. As shown in Figure 13, thirteen of the twenty-one stations are located in or adjacent to tracts with relatively high percentages of zero car households. Eleven are in or adjacent to tracts whose percentage of zero car households range from 40 to 100 percent (the top two categories used). This is in keeping with the ear1ier discuss i on of Metrorail stations in relation to density and median household income, wherein most of the stations are located in higher density tracts with lower incomes. These characteristics are l ogically related to more zero car households. As also expected, those stations located further south along the route, in the lower density h i gher income areas also are l ocated in tracts with lower percentages of zero car households However, even in those areas, no Metrorail stations are located in tracts with the lowest category of zero car households (0 to 5 percent) though one station is l ocated adjacent to such a tract. Strictly from the standpoint of geographic proximity at the census tract level Metrorail stations are well located with r espect to potential transit dependent househo l ds, as measured by percentage of zero car households in the tract The degree to which these stations are actually accessible to the transit dependent in these areas depends, in part, on how well other transit modes connect to these stations. This is discussed i n more detail e l sewhere and is a point which should be considered further in future research This section of the task report serves as an introduction to basic issues related to station location. In many i nstances, stations locations are the result of compromises brought about by outside influences, such as land availability land costs, existing land use and trackage, 65


neighborhood opposition, and political or economic concerns. Examples of these influences have been g i ven i n relation to the case studies, including the South Florida case study. Evaluation of the South Florida station l ocations with respect to several i mportant variables indicates that, at the census tract level, stations are located well, both in terms of proximity to population concentrations and potential transit dependent riders. Future research shou l d look more closely at the interrelationships between these variables and station locations and should be conducted at a smaller geographic sca l e Furthermore, a more detailed study of accessibility via other transit modes would prove useful. 66


Tri-Rail Tri-Rail Stops Metrorail 0 Inte rc ounty Bus Tra n sfer Stations Median _Income SOOOOto 15100 0 (62) 32000 to 50000 (199) 20000 to 32000 (257) 0 10000 to 20000 (99) 0 Oto 10000 (25) F i gure 10. Tract Level Median Incomes for Tr i-Rail Serv ice Area 67


This is a blank page


Figure 11. TrJ..Rail Tri-Rail Stops -Metrorail 0 Metrorait Stops 50000 to 151000 (62) 32000 to 50000 (199) Ill 20000 to 32000 (25 7) 0 10000 to 20000 {99} 0 0 to 10000 (25} Tract Level Median Incomes for Metrorail Service Area 69


This is a blank page


Tri-Rail T ri-Rail Stops -Metro rai l 0 Intercounty Bus T ransfe r Stations Percent of Households with ZeroVetas 60to 100 (9) 40 to 60 (37) 15 to 4 0 (135) 1m 5to 15 (247) 0 Oto 5 (214) Figure 12. Ze r o Car Households in Tri Rail Service Area 71


This is a blank page


) '' . . ', 1 . .. "., -_ .. Figure 13. Tri-Rail Tri-Rail Stops Metrorail 0 Metrorail Stops Zero Vehicles 60to 100 (9) 40 to 60 (37) 15 to 40 (135) 0 5to 15 (247) 0 Olo 5 (214) Zero Car Households in the Metrorail Service Area 73


This is a blank page


Tri-Rail Tri-Rail Stops -Metro rail 0 Intercounty Bus Transfer Stations Percent of Households with One Vehicles 48 to 100 (138) 42 to 48 (124) 38 to 42 (99) [) 30 to 38 (144) 0 0 to 30 (137) Figure 14. One Car Households in Tri -Rail Service Area 75


This is a blank page


Subtask 3: Coordination of Multimodal Planning Introduction Fixed guideway systems are not planned and built in a vacuum. A metropolitan area's decision to invest in fixed guideway is generally driven by a desire to improve the current transportation system and provide a more attractive alternative to the single-occupant automobile. Yet cities for which fixed guideway is a reasonable aHemative generally have an extensive bus network in place, and may have other fixed guideway modes as well. At the network level, there should be an explicit recognition of the need for coordination of planning for the two (or more) transit modes. One potential problem w ith transit planning is that it tends to be organized in modal fashion, and there is not always good communication between rail and bus planners. The p lanning for a new fixed guideway facility can be so involved with a myriad of details that connections with the existing bus system receive less attention than they should. The problem to be addressed in this part icular subtask is the potential reliance on narrow modal approaches to planning and operation of an interrnodal transit network. The concept of interrnodallsm will be applied on three levels recognizing that these levels are closely interconnecte d: fixed guideway to fixed guideway (e.g., commuter rail to heavy ra il ); fixed guideway to other transit (generally the existing bus network) ; fixed guideway to automobile (e.g., park and ride lots). The purposes of this research are to reveal f acto rs that affect intermodal connectivity for new/extended fixed guideway systems, and to assess their ultimate relevance in terms of how the overall transit network functions after the guideway system is operational. Factors may include the timing of intermodal considerations in system planning the nature of the existing bus network and the extent to which it was restructured to serve the new fixed guideway, transit agency organization, and political influences in the p lann ing process. 77


GulcMway Tranalt Mrd lnfllmlo

Two relevant points regarding fare policy deserve elaboration here. The first is that coord i nation of fare pol i cies does not mean that all agencies will have identical policies. A clear example of this can be -n in any metropolitan area w ith commuter rail service. All commuter rail lines in the States now have a zonal fare structure, whereas many bus systems charge a flat fare This difference is a function of the variations i n trip lengths on commuter rail as opposed to loca l bus, and is perfectly logical. It may be useful fo r trans i t agencies to coordinate fa r e policies on issues such as age (or height) qual i fications for reduced or free fares, but even then there may be historica l reasons for differences among agencies. The second point addresses the role of fare policy in the perception of transit as an integrated whole w ithi n the metropo l itan area. The concept of "seamless" transportation has gained in interest and popularity in recent years, particularly since the passage of ISTEA. In terms of fares a seamless" system is one in which a user pays once in order to have access to all modes operated by a specific agency, and i n s ome cases to other modes as well. A Metrolink month l y commuter rail pass for example, permits free transfers to the Metro Red line at Uni o n Station in downtown L o s Angeles. In Port l and Tri Met's light rail ticket is good on all bus l ines as well as long as the passenger boards wit hin one hour and 45 minutes of purchase MARTA in Atl anta is genera ll y regarded as an excellent examp l e of seamless transporta t ion Not only are transfers between rail and bus free but several stations are co nfigured to buses to enter the fare contro l area and pick up and discharge passengers No transfer or farecard needs to be shown at these locations thus making i t as convenient as possible for passengers This is somet i mes known as the Theme Park policy: the passenger pays one admission and is then welcome to ride any transit mode. MOTA in Dade County (South F l orida) t akes a different appr o a c h r equ i ring a 25 cent charge for transfers between Metrorail and Metrobus Th i s approach which is the more traditional policy among mutt i modal systems i n the United States does not make a clear identification of the bus and rail systems as one i ntegrated networl<. MOTA personnel however argue that their system is exactly like Atlanta's : the same pri ce is charged f o r a 79


bus/rail transfer as for a bus/bus transfer. A transfer charge should not be interpreted as discouraging multimodal usage when is standard for the bus system, brings in a small but steady source of revenue, and helps to control fare evasion and transfer misuse A major demonstration project is underway in the San Francisco Bay Area to establish a universal ticket good on all systems in the region. The Translink project is bringing agencies on line one at a time, and so far appears to be successful. With the rapid pace of developments i n farecard technology, it is likel y that Translink is the wave of the future in metropolitan areas served by more than one transit mode o r agency. Location of Fixed Guideway in Relation to the Existing Bus Network Fixed guideway systems are sometimes located opportunistically a long existing freight right-of-way. The San D iego Trolley is the best-known example of this phenomenon. Tri Rail in South Florida, the first new commuter rail line in the United States in over 20 years at the time of its start-up in 1989 operates on existing CSX freight tracks that have been upgraded to accommodate commuter rail. A similar arrangement is in effect for Metrolink commuter rail in Los Angeles. Subway l ines such as BART in the central business districts of the Bay Area and the Red Line in Los Angeles are not constrained by use of existing l ines The impact of fixed guideway location on multimodal planning is in part, related to the industrial land use prevalent in proximity to freight rights-of-way A fixed guideway system operating in such an area is more dependent than usual on intermodal connections to bring passengers to and from stations, because industrial land use is generally not supportive of transit. There are exceptions to this rule, however. The south portion of MOTA's Met rorailline follows freight r ail right-of-way that parallels the South Dixie Highway (U. S 1} a major transportation corridor in southem Dade County. Thus, the south line occupies a centra l location in the region, is readily accessible and operates where residents are accustomed to travel. 80


Passenger Convenience Subtask 4 of this task report takes up the issue of convenience from a policy perspective as it relates to in termodal transfers. In this section, site planning considerations that can maximize the integration of separate modes are addressed. The rule in older fixed guide way systems was that the bus stopped on the street near a staircase up or down to the rail station. Special provisions to ease connections were the exception, atthough major intermodal centers such as commuter and intercity rail stations and ferry terminals would have a variety of modes under one roof or at one site The typical rail/bus connection was on the street and functioned as a regular bus stop, only with considerably more activity. Among the case study metropolitan areas New Orleans is an example of an older fixed guideway system with no special attention given to intermodal linkages. Each of the streetcar lines (St. Charles Avenue and Riverfront) operates as a separate entity, with no direct linkage between the two lines There are seven bus transfer locations along the St. Charles Avenue line and one bus transfer l ocation a l ong the Riverfront line These transfer locations represent very informal modal linkages, with no conventional platforms, stations or parking facilities provided at the sites where the two modes meet. RTA's plans for New Orleans are in ten ded to promote inte rmoda lism by i ncreasing and improv ing connections among the streetcar lines, including the proposed ne w Canal Street line. and between streetcars and buses. Newer fixed guideway systems have given greater attention to in termo da l connectio ns, espedally with feeder bus routes. The Lenox station o n MARTA's system is one example of a seamless connection within the fare control area itselt, a passenger never has to leave the station to trans fer to a bus. Off-street bus bays have been established at key intermodal stations such as the Gateway station in Portland and the Douglas Road station in Miami. The Banfield light-rai l line Is at grade at Gateway, and t he bus bays are at the outer periphery of the rail platforms At Douglas Road Metrorait is elevated but passengers descending a short flight of stairs can gain ready access to each of the seven 81


bus routes serving the station immediately outside the fare control area Many BART stations have an intermodal int erface sim ilar to the Douglas Road station in Miami, while Sacramento's light rail stations often feature an interface similar to Tri-Met's Gateway. These off-street bus bays generally have a canopy to sheHer passengers in bad weather Miami and Portland both have examples of less desirable intermodal connections. At Tri Met's Coliseum station, for example, bus routes operate on adjacent city streets and transfers between modes involve use of sidewalks and crossing of roadways with vehicular traffic. Several Metrorail stations do not feature off-street bus bays, but in stead use regular on-street bus stops for intermodal transfers. This suggests that the driving force behind good intermodal site design is not the trans" agency's or local planning department's awareness of the desirabil"v of fostering intermoda l connections, since good and bad examples exist within a single system. The street network at a particular station location and, perhaps most importantly, the availability and affordability of land to construct an off street facil"v appear to be the determining factors. The land affordability issue explains why new off-street facilities are not constructed in central business districts, where the cost of land tends to be prohibitive. Off-street bus bays are generally more convenient in terms of bringing the two modes as close together as possible and thus minimizing walk distance fo r transferring passengers. There are certain advantages to on-street bus stops, however. In many locations, newsstands and stores catering to waiting passengers have sprung up at major intermodal sites. These stores serve a dual function of making the wait for a bus more pleasant and providing sheHer from inclement weather. Commercial activ"v on the street attracts other pedestrians and can create a safer environment than an off-street faci l ity at a station in the midst of a large parking lot. With good directional signage and passenger amenities, on street bus stops can rival off-street intermodal facilities in terms of passenger convenience. Connections among fixed guideway systems are also important. The case studies provide examples of good and bad designs, in terms of passenger convenience and ease of understan ding and use. At Union Station in Los Angeles, the connection between Metrolink commuter rail trains and the Red Line is direct. On the other hand, in San Francisco the 82


CaiTrain commuter rail tenninalls 1.5 miles east of downtown San Francisco, w ith no direct connection to BART, Muni Metro or cable cars. I n Dade County, there are two stations that serve both Metrorail and Metromover the downtown people-mover system. Government Center features a direct and logical connection. The Metrora il station (center platform) is located two levels above the Metromover station with side platforms, and access between the two is via escalators/staircases and an i ntermediate mezzanine level ( where passengers transferring from Metrorail to Metromover select the appropriate staircase for t he inner or the outer l oop). Brickell Station is designed differently, with the new Metromover station approximately one block to the south of the Metrorail station. Transferring passengers must first descend almost to street level, and then go back up to the desired platfonn. The Brickell Metromove r station was opened in 1994 as part of the Met romover south leg extension, and thus is not as familiar as the Government Center facility. Nevertheless, t he overwhelming majority of transfers continue to take place at Government Center, suggesting that the less direct connection at Brickell may be discouraging transfer activity. In summary passenger convenience is an important factor in encouraging interm odal transit usage. MARTA 's Lenox and Tri-Met's Gateway stations are excellent examples of designs that maximize transfer convenience. San Francisco's isolated CaiTrain station is an example of a location that discourage s intermodal use. Transit and planning agencies are well aware of the importance of proximity and passenger amenities at intermodal centers but are limited by site facto rs and the unavailability or unaffordability of land. With good planning however on-street facilities can be attractive and can serve to encourage in termoda l use Extent of Bus Network Restructuring to Serve Fixed Guideway As noted in the introduction to this subtask, fixed guideway systems are not planned and built in a vacuum, but are optimally designed as a component of an overall public transportation network. Generally, there is an ex istin g and reasonably extensive bus 83


network, and there may a lso be o t her fixed guideway modes. The bus network may be radial or grid in natu re, but is structured to serve the same central locations as served by the new fixed guideway systems The extent to which the bus and fixed guideway modes c o mplement each other within a give n metropolitan area has been suggested as a major factor contributing to the success of fiXed guideway systems At the extremes, there are two ways to plan for the bus network to serve a new fixed guideway service. The first is to reorient all bus routes as feeders to the fixed guideway route; the second is to do nothing to the bus network, assuming that fixed guideway will find its own market. Both options are unrealistic The first treats the two modes as interchangeable and thus ignores the d i fferent travel needs met by bus and fixe d guideway sys t ems, while the second ignores reality. The options are presented in their starkest form to illuminate the dilemmas faced by transit agencies implementing new fixed guideway systems The goal is t o take maximum advantage of each mode's functional strengths while ensuring a who l e integrated publ i c transportati o n system In theory, this is relatively straightforward, but in practice agencies have struggled with avoiding duplicative services while maintain ing mobility at all levels within a metropo l itan area. The bus network typically serves shorter trips while a fixed guideway service serves intermediate (heavy night rai l ) or long (commuter rail) trips. Fixed guideway networks tend to be regional in nature (although not necessari l y at first), wh il e bus networks are local. Along with g r eater average travel distances, fixed guideway systems operate at higher speeds because the excl us i ve guideway right-of-way a ll ows fiXed guideway vehicles to avoid traffic congestion. The most l ogica l alternat ive for a metropolitan area is to desi gn a fixed guideway system to serve intermediate and long trips while allow ing the bus network to focus on shorter trips This alternative allows each mode to do what it does best. The problem arises with the exceptions to this general ru le. Express bus routes serve regional markets and Los Angeles particularly has some very long bus r outes that carry as many passengers as fixed guideway systems elsewhere Some forms of fixed guideway such as Dade County's downtown people mover, excel at short trips through 84


congested central business districts. Light rail systems do not necessarily avoid traffic congestion along joint rights-of-way. Historical travel patterns and loca l condHions can and do complicate dedsions regarding bus networ1< restructuring to serve new fixed guideway systems. Grid systems that provide coverage at reasonable headways to all or most neighborhoods wHhin a transit agency's service area may require fewer route adjustments than radial systems oriented, like the new fixed guideway system, to the central business district. Changes to the bus networ1< have been implemented In conjunction with virtually all new fixed guideway systems The only exception to this statement, due to its specialized tourist/convention orientation, is the Riverfront streetcar line in New Orleans. The case studies to date suggest that the extent of the bus networ1< reorganization is related to a number of factors, including : type of fixed guideway system; type of bus network (radial or grid); demand within a given corridor; locat ion within the metropolitan area. Type of Fixed Guideway System. As a general rule, restructuring of the bus networ1< is more extensive for heavy rail systems than for light rail systems The exclusive right-of-way and higher operating speeds for heavy rail permit greater travel time savings for heavy rail whereas light -rail systems operating on surface streets in central business districts are subject to traffic delay Travel time savings is a useful measure for transit agencies to use when making decisions on rerouting or truncat ing speafic routes. I f a trip can be made more quickly by taking a feeder bus and the fixed guideway system than by the bus routing prior to the opening of the fixed guideway, then that route is an obvious candidate fo r rerouting. Calculations such as these must take into account the t ime required to transfer and the wait t ime for the second leg of the trip In the case of rail systems, passengers may be willing to sacrifice 85


a WIN minutes of travel time in order to enjoy the enhanced amenities of rail in comparison to bus Lightrail systems that operate In mixed traffic wi ll usually supplant existing bus routes within the same corridor. As discussed below, this is not always the case for heavy rail systems Commuter ra i l lines a r e regional in nature, and tend not to replace existing bus routes. Stations along the lines become i mportant as intermodal points. Feeder bus service and pari<. and ride lots are both needed at stations Downtown commuter rail stations often rely on other fixed guideway systems to serve as downtown distributors, as in Los Angeles. Commuter rai l stations are increasingly as potential major intennodal centers as the transit networl<. expands The generally dense downtown bus network also carries commuter rail passengers to their final destinations. San Francisco s CalTrain station does not connect with other fixed guideway transit and CaiTrain passengers must rely on buses jit n eys, trains or walking to reach their downtown destination. Tri-Rail, the commuter rail line in South Ron da, presents an interest ing example of how a new fixed guideway mode can fit in to the ex i st i ng transit networl<.. Tri-Rail began operat i on as a temporary traffic mitigation measure during the reconstruction of 1-95 in Dade Broward and Palm Beach Counties Tri-Rail uses existing CSX freight tracks located to the west of 1 -95, away from the downtown areas. The location of th i s r a i l line meant that prov i ding access in the forms of feeder bus l ines a n d parl<.-and-ride lots was v i tal to ensuring tha t intending passe n gers could get whe r e they needed to go. The loca l transit agenc i es d i d not want to dis ru pt their bus networl<.s to serve a tempo r ary commuter rail l ine, and i nstead established dedicated feeder routes Incidentally Orange County in Ca l ifornia took a similar approach in provid ing feeder service to the new Metrolink commuter rail l ine I n Dade County a new Metrora i l station was constructed at the intersection of Metrorai l and Tri-Ra il. Met r orail was then able to provide access for commute r ra i l passengers to downtown Miami a ma jor destination point. Through a comm i tment to intermodal connections, the local transit and planning agencies were ab l e to overcome t h e uniq u e obstac les of the temporary (at the outset) natu r e of commuter r ail serv i ce an d the lack of 86


a downtown terminal. Me trorail continues to serve as the primary feeder/distributor for Tri Rail in Dade County; PalmTran in Palm Beach County is planning a major reconfiguration of transit system that will rely on Tri-Rail as the north-south spine of the new network. In summary, light rail often displaces existing bus routes within a corridor, although if it operates in its own right-of-way may also induce restructuring {as in Portland). Heavy rail sys tems are more likely to cause a restructuring of the bus network, due to greater travel time savings. Commuter rail requires bus service at stations in residential neighborhoods and suburban communities, and generally relies on the existing transit network to distribute its riders throughout dow ntown Type of Bus Network. The nature of the existing bus network also contributes to the need to restructure routes to serve and avoid duplication of a new fixed guideway route. Bus networks can be categorized as grid or radial systems. In a grid system. routes travel in each direction at regularly -spaced intervals and thus provide a high degree of service area coverage. In a radial system, routes radiate outward f rom a central hub {usually downtown) like spokes on a wheel, leaving wide swaths of o utlying areas out of reach of the bus network. Most new fixed guideway systems are radial in nature and oriented toward downtown. Thus, radial routes in close t o the fixed guideway system are likely to be restructured, whereas many routes in a grid system can readily serve as feeder routes perhaps with minor modifications to serve the station locations. The example of MOTA's attempt to reorient its bus network is relevant to this discussion. Following the opening of Metrorail, a significant restructuring of the bus system was implemented. The plan, known as Network '86, inc reased the number of grid routes and restructured several rou tes as feeders to Metrorail. Its broader goal was to rationalize and in many ins tances decrease, bus service throughout the county. As noted below many route changes were withdrawn or revised in response to community concerns, but the MOTA planners clearly saw the benefits of strengthening grid routes to work in conjunction with Metrorail as an integrated system. 87


Demand within a Given Corridor. In addition to converting certain routes to feed the fiXed guideway stations, transit agencies often discontinue routes paralleling the fixed guidewey route Portland and Atlanta took this action, and Dade County discontinued express routes when the south line of Metrorail opened. This type of action is logical to avoid duplication of service, but two examples from the San Francisco Bay Area involving BART and AC Transit reveal that the issues can be less obvious. The first example Involves two of AC Transit's busiest routes Route 72 along San Pablo north through Berkeley and Route 82 along East 14th street through East Oakland. These routes operate in corridors that closely parallel BARrs service to Richmond and Fremont. Both routes are sufficiently patronized to have 24 hour service, and buses on both are full for much of the day. The routes divert to serve BART at selected stations, and so funct ion partly as feeder service but their through ridersh i p within the corridor is significant. Interestingly, there appears to be considerable turnover in ridership along the routes, suggesting that the buses are serving primarily a local market. BART is carrying the l onger haul passengers bound for San Francisco or downtown Oakland while the AC Transit routes are primarily meeting the short-haul demand for transit. A person looking at a map might conclude that the bus routes duplicate BART service and should be discontinued, but the transit planners correctly recognize that the buses and BART serve different travel markets The second example is the cross-bay express bus service operated by AC Transit. BART is the dominant public transportation mode for cross-bay travel between Oakland and San Francisco but AC Transit continues to run express r outes via the San Francisco-Oakland Bay Bridge These express routes are the direct successors of rail lines operated as part of the Key System until 1958, and serve areas not directly served by BART AC Transit is pleased with the patronage on the express routes but their continued operation also seems to be closely tied to AC Transit's view of i tself as a regional transit agency. BARrs opinion Is that the express bus routes duplicate Its rail service, and that feeder routes from neighborhoods served by the express buses to BART stat ions nearby would be a much more efficient way to provide transit service. The issue has neve r been resolved, although budget problems may force AC Transit to address the issue from an efficiency standpoint. 88


Guideway Transit end Function llfld The examples highlight the need for careful judgment in making the decision to discontinue bus routes when or after fiXed guideway systems are introduced. While some dupl i cative services may be traceable to i nteragency turf conflicts, there are also cases I n which apparently duplicative routes are serving different markets and are both needed to provide adequate transit service. Location within the Metropolitan Area. Adjustments to the exist ing bus network i n response to a new fixed guideway system are more likely to take place in outlying port ions of the service area. While minor adjustments may be needed in or near central business districts, there is usually a well-developed grid network in the city center that can adequately serve to distribute fixed guideway passengers to their downtown destinations. Major bus route restructurings are more appropriate in outlying areas, where fixed guideway offers greater travel time savings and where the density of the bus network is lower Bus networks can often take on a hybrid form when restructured in order to accommodate a new fixed gu i deway system. As seen in the north line Metrorail example in South Florida a grid system might be implemented in densely deve loped sections of an urbanized area. In outlying, suburban sections, especially where the distance between stations on the fixed guideway network is g reater, a radial bus mini-network centered on fixed guideway stations may develop. BART MARTA and TriMet's Banfield line have influenced the development of radial bus mini -networks in suburban areas, with rail stations as hubs. In the San Diego metropolitan area, the Chula Vista and National City transit systems have oriented their bus networks to serve the San Diego Trolley and El Cajon to the east has developed into an I mportant intermoda l center. An integrated bus-rail transit system was established as a goal of Tri Met from the earl i est days of planning for a fixed guideway system. Even before light rail became a serious discussion item for Portland's transit future T r iMet r ecognized the importance of connectivity to its bus network. The concept of radial routes extending outward from a key downtown mall link sub-area transfer centers in suburban and outlying Portland ne i ghborhoods where local routes c l uste r ed and crosstown routes linking these peripheral transit centers is a clear application of i ntermoda l ism 89


In Atlanta, an integrated bus-rail transit network was an early MARTA policy goal. The rail system design incorporated plans to rest ructure the bus network, and stations were designed to facilitate that Interface t h ro ugh street extensions, bus bays and special bus lanes. Not every fixed guideway system needs to become a hub i n the subulban bus network Tri Met restructured its bus network to provide feeder services at selected rail stations. Stations that are in primarily residential areas or that are difficutt to access via major streets are not good candidates for hubs, atthough they may be served by a nearby bus line. When BART was constructed, some local communities changed the zoning in the areas of the BART station to restrict commercial development and avoid disturbing the existing residential character of the neighborhood. On the other hand, stations located at major intersections where there is already some retail or commercial development are prime locations as focal p o ints for bus network restructuring. I t should be noted that the nature of bus networks are changing, even in the absence of fixed guideway systems, in respons e to population and employment shifts. As suburbanization trends have continued and Edge Cities are increasingly part of the metropolttan landscape, transit agencies have wrestled with the best way to provide service to population, employment, and retail centers in suburban areas. A modified radial network with major activity centers such as regional malls as hubs is one potential solution to the challenge of serving suburban loca tions with public transportation. Fixed guideway systems can contribute to t his service pattem by providing natural activity hubs at t heir outlying stations. Joint development projects, d iscussed in other task reports, enhance the level of activity at stations and create stronger transit centers and the opportunity for more frequent feeder bus service. Feeder bus routes extend the market area for a fixed guideway system by providing access to the fixed guideway system for those beyond walking distance. Thus, a reorganization of the bus network to prov ide service to fixed guideway stations and to discontinue duplicative routes can make a strong contribution to fixed guideway's ability to attract riders. In areas with a strong grid system, major reroutings are generally not necessary. 90


Restructurings are more common in outlying suburban areas, where bus service is often reoriented in a modified radial pattern with fixed guideway stations as hubs. This approach is congruent with current attempts to refocus bus networks in suburban areas to serve major activity centers such as regional malls. The guideway station becomes another type of activity center to be served. Automobile Access to Guideway A major rationale behind many new fixed guideway systems is to reduce traffic congestion by attracting commuters and other travelers out of their cars. Older American fixed guideway systems were constructed whe n high urban densities were the norm and b efore the advent of the age of the automobile Now that the automobile is a fixed star in the constellation of travel alternatives, new guideway systems must lake account of the need to provide access for automobiles. This has generally taken the form of park-and-ride lots at guideway stations. There are sound reasons for prov iding park-and-ride access for automobiles at fixed guideway stations As is the case with feeder bus service park-and-ride lots extend the market area for fixed guideway beyond walking distance of the station. In fact, park-and ride lots can be more effective than feeder bus in extending the market area within easy reach of a station, because its effects are not l imite d to a specific route or routes but are omnidirectional within the constraints of the roadway network. Also, park-and-ride lots function as intercepts for the automobile, and do not require a complete change of mode for the overwhelming majority of commuter who currently use automob iles; the automobi le is still available on the journey home to pick up children at day care or to run errands. Automobile d rivers and passengers are also the major source of new transit riders. Automobile access is particu larly important in suburban locations. Commuter rail lines includ ing Metrolink in los Angeles, CaiTrain in San Francisco, and Tri-Rail in South Florida, depend heavily on the automobile as an access mode to their stations. BART, SCCTA, MARTA and MOTA provide parking facilities at most of their outlying station locations 91


Inner city and downtown stations generally have fewer parking spaces available. The Blue Une in Los Angeles has little parking at most of its stations, pfimarily due to their inner-city locati on In Portland, more positive community attitudes toward transit and an extensive transit network appear to have obviated the need for extens ive parking along the Banfield l ight rail line Most new fixed guideway systems, however, have planned for ready access to their lines for the automobile Tri-Rail's original status as a temporary agency meant that it was constrained from spending any money to acquire l and for parking As commuter ra i l ridersh i p has increased in South Fl orida many parking lots squeezed into available land at T ri-Rail stat i ons overflowed with commuter automobiles. The Florida Department of Transportation (FOOT) stepped in with plans to construct major park-and-ride lots at several locations along the right-of way, and these lots have driven decisions on locating new stations Inte r est i ngly, while Tri-Rail station locations are placed prominently along 1 -95, the agency has noted that most of its park-and-fide patrons arrive on l ocal streets from adjacent neighborhoods. Tri Rail believes that once a driver gets on 1-95, he/she will not get off to park at a rail statio n and change modes. The automobile i s expected to continue to be a ma jor access mode for fixed .guideway systems, part i cularly in outlying or suburban areas Since many of the planned guideway extensions (BART and MARTA, for example) will take place in growing suburban areas, the automobile is likely to become even more important i n providing access to fixed guideway systems. In unusual circumstances such as in Portland, park and-ride lots may not be needed. Generally however easy access to fixed guideway by automobile will be a necessary componen t i n increasing the appeal and competit i veness of fixed guideway systems 92


Transit Agency Organization and Inter-agency Cooperation The ability to integrate the plann ing and operation of multiple trans i t modes can be enhanced by the organization of the transit agency and by the number of agencies within a metropolitan area with responsibi l ity for operation. The presence of a strong regional planning or transportat i on agency can contribute significantly to the effectiveness of intermodal cooperation. The simplest organizational situation is when one agency is responsible for all or virtually a ll public transportation within a given metropolitan area. Among the case study cities Atlanta is a prime example of this situation, with MARTA operating the heavy rail system and the preponderance of bus service in the region MARTA planned for intermodal connections well in advance of the construction of the rail system and supported the intermodal concept with both a seamless fare policy that encourages transfers between modes and station design that facilitates these transfers Location of responsibility for all modes within a single agency does not always ensure cooperation. In some of the older fixed guideway cities transit agencies a r e organized on a moda l basis and there is a long history o f l ack of cooperation While several agencies with newer guideway systems are also organized modally, the relative newness of the fixed gu i deway component has prevented a bureaucratic hardening of responsibilities so l ely by mode. In Los Angeles, the agencies responsible for buses (Southern California Rapid T r ans i t District) and new rai l construction ( Los Angeles County Transportation Comm i ssion) recently merged to form the Los Angeles County Metropolitan Transportation Authority. As might be expected, a merger of this magnitude did not happen without difficulty, but the idea of joining the two agencies together is very much in keep i ng with the i ncreasingly i ntermoda l nature of public transportat ion i n Southern Ca li fornia. Conflicts among transit agencies in the San Francisco Bay Area were cited above i n the case of AC Trans i t's trans-bay express buses competing BART heavy rail service The 93


Guideway TIMSitMd Function Bay Area is one of the most fragmented on paper in terms of four major fiXed guideway operators (BART, Muni the Peninsula Joint Powers Board, and the Santa Clara Count y Transportation Agency) a l one The MPO for the Bay Area, the Metropolitan T ransportation Commission or MTC, is a strong and forward-looking regional agency that helps to coordinate and unify public transportat i on in the metropolitan region MTC is sponsoring Translink an i nnovative program with the ultimate goal of providing a single fare medium good on any transn service in the Bay Area. MTC a lso publishes a regional trans i t guide that prov ides information on how to use all transit services. San Diego i s another metropolitan area with a number of transit providers I n the San Diego region. however the Metropolitan Transit Development Board is a unifying factor i n transn plann i ng The San Diego Association of Governments (SanDAG), the region's MPO, a l so works closely with transit agencies to ensure coordination and cooperation. The examples of San F rancisco and San Diego demonstrate that an intennodal sp i m can be fostered in regions with mu l tiple providers if there is a strong force for regi onalism South Florida provides an interesting example in wh i ch the regional force was the F l orida Department of Transportation (FOOT) When Tri-Rail began operation in three count i es in 1969, its service area spanned three MPOs, two FOOT d i strict offices and two r egional planning councils. FOOT purchased the CSX right-of-way and was a major source of funding to get the commuter rail service up and running. It brokered agreements and provi ded partia l funding for feeder bus service, and pai d for the construction of a new Metrorail station at the i ntersection of Metrorail and Tri-Rail to ensure access to downtown Miami for Tri-Rail passengers. The South Florida region has not gener ally been noted as an example of regional cooperation but the success of Tri-Rai l is justifiably viewed with pride as an example of what can be done when the various agencies in the region work together in common cause State Departments of Transportation have also been key supporters of fixed guideway and intermodal connections in most other metropoman areas where case studies were done MARTA personnel cited the role of the Georgia DOT in support of intermodal po li cies 94


MPOs, regional planning agencies and state DOTs can ensure coordination among multiple transrt agencies within a metropolitan area and promote intermodal cooperation. Local Policies Local policies (and local politics) can help or hinder transit agencies in their pursuit of systemwide integration and rationalization. As discussed earlier, the political fallout from MDT A's attempt to restructure its bus service in part to feed Metrorail, prevented the plan from going through in its entirety. Transportation dE!Qsions, especially regarding major components of the loca l transportation network such as fiXed guideway, are nearly always influenced to some degree by local politics. It is in teresting to note, however, that policies more conducive to transit and particularly to i ntermodalism are being adopted or considered in many metropolitan areas This may be traceable to the Federal ISTEA legislation, but it may also be a result of lessons learned from the experiences of the past several years Other sections of this report go into much greater depth regarding land use, congestion management, and specific transportation policies that support intermodalism One interesting finding of the case studies is that future plans for transit are based on an intermoda l approach involving a current or future rail station. As noted in the Los Angeles case study, Union Station is already an i ntermodal center, but future transit developments will accentuate this role. New Orleans' streetcar lines do not connect with each other and have only casual interfaces with RTA's bus system New Orleans is planning a new Canal Street line that will make connections with both existing lines and also serve the New Orleans Union Passenger Terminal, anticipated to become an intermodal focal p o int. Of all the case study metropolitan areas South Florida is the most illuminating in its plans for intermodal connections based at a rail station. Downtown Miami no longer has a rail station; Amtrak service terminates in the northwest section of Dade County, near the Metroraii/Tri-Rail station. Tri-Ra il's terminal is near Miami International Airport well west of downtown in an indus trial area with o nly dedicated feeder bus connections. The Miami lntermodal Center plan calls for a new station approximately one mile closer to the airport 95


along the CSX right-of-way. The station will serve Tri-Rail and Amtrak, both of which will be extended to the new tenninal, and will have a people-mover to connect passengers directly to Miami International Airport. The proposed Metrorail East-West line will also use this station, and will provide connections to the Miam i Seaport and its booming cruise business. Metrobus, intercity bus, taxi service and car rental agencies would also be available at the Miami lntennodal Center. Thus, near1y all fonns of local, regional, national and international transportation modes will be concentrated in or easily reachable from a single location. The choice of l ocation immediately adjacent to the airport recognizes the possibility of landside access modes functioning as distributors for air passengers, just as heavy rail lines are distributors in central business districts for commuter rail passengers, and feeder buses distribute heavy rail passengers ISTEA proponents have been eager to see the intennodal concept applied at a major airport. the Miami lntennodal Center concept, the South Florida region and transportation community has taken an important step toward innovative intennoda l connectivity. Subtask 4: Facilitating Convenient Transfers Introduction The success of an intennodal transit system rests to a large degree on the extent to which the modes are connected. A seam less" system designed to give the passenger the perception that he/she is rid ing on a single system no matter which mode is being used is the intennodal ideal. While many factors enter into the design of such a system, fare policy and schedule coordination can make a significant contribution to the rider';s perception that the transit network is an integ rated whole. This sub-task examines fare policies at multi-modal transit systems to iden tify and assess the different approaches taken to transfer pric ing The role of monthly fares as a payment method is also assessed. The objective Is to ident ify different pricing strategies a n d their 96


Guldewg TntiWt and lntermodlll&m: Futw:flon and er.ctfwneM underlying rationales. Service-related policies that encourage intermodal use, such as schedule coordination between modes, will a lso be examined. The purposes of this research are to identify the various approaches taken to fare policy, to highlight current trends in the area of intermodal fares. to understand how different policies can encourage or discourage intermodal usage, and to assess the extent to which service policies, incl uding schedule coordination. can affect passenger perceptions of system unity. Factors to be addressed include transfer pricing, intrasystem versus intersystem transfers. schedule coordination, the role of monthly passes, the regional transn pass concept, and the practical impacts of theoretical assessments of fare policy. This subtask report is organized into sections focused on each of these relevant factors. Transfer Pricing An important component of fare policy involves the treatment of transfers wnhin or between modes. A recen t survey of 17 transn systems.' incl uding five with in the case study metropolitan areas, discovered a range of transfer options, from free transfers to no transfers. i.e . full payment of the appropriate fare each time a passenger boards a transit vehicle. The average transfer charge for the 17 systems was 15 cents. From an intermo dal perspective, free transfers are obviously more effective in encouraging transfers and intermodal usage. Systems that charge for transfers often express concern wnh transfer abuse. A paper transfer is worth the equivalent of a full fare, although there are generally time limnations for ns use Requiring at least a nom inal charge for transfers discourages riders from obtaining transfers for resale, in the view of those concerned with transfer abuse. In increasingly constrained fiscal times transfer fees also provide a small but steady sou rce of agency revenue. Center for Urban Transportation Research, Transit Fare Pol icies and Strategies. Tampa: prepared for Metro-Dade Transit Agency by CUTR, September 1994. p 19. 97


Systems that require payment of a full fare each time a passenger boards a vehicle claim that a no-transfer system eliminates the need to print, distribute, track and collect transfers and so reduces administrative headaches Bus operators avoid transfer-related conflicts with passengers in a no-transfer system. Some agencies have considered l owering their base fares and eliminating transfers as a means of simplifying fare co llect io n and increasing revenue. Free transfers are an essential component of a seamless transit network. This concept is exemplified by MAR TA, which has pursued a transfer policy treating bus and rail as part of a single unified system. Tri-Met has a similar policy, enacted through its proof-of payment fare system. A validated fare ticket is good on any bus or ligh t rail vehicle for one hour and 45 minutes after validation Tri-Met also has a fareless zone covering 100 blocks of the Portland central business district. MOTA i n South Florida and RTA in New Orleans have transfer charges for transfers within or between modes that they operate The cost is 25 cents at MOTA and 10 cents at RTA. MOTA cla ims that a transfer charge i s not necessarily a barrier to intermodal connections since the agency treats a bus/rail transfer in exactly the same fashion as it treats a bus/bus transfer. New York City is an example of a system that has free bus to bus transfers but char ges an add i tional full fare for bus/rail connections. There are sound arguments for charging a transfer fee, but from the rider's perspective cost is a factor in discouraging intermodal connections. lntrasystem versus Intersystem Transfer Polley The development of new fixed guideway systems has changed the general rule regarding intrasystem versus intersystem transfers. In older guideway cities, systems have historically been more likely to offer free transfers to other components of their service and to charge for transfers to service operated by othe r agencies To help ensure the success of new fiXed guideway systems, particularly commuter rail start-ups, local transit agencies 98


have been willing to accept fare media from the new guideway systems without additional charges The Coaster commuter rail line in the San Diego metropolrtan area and Tri-Rail in South Florida exempl i fy this trend. A Coaster rail ticket is accepted by San Diego Trolley and San Diego TransH as full payment. Tri-Rall reached interiocal agreements with the three transit agencies within Hs service area to permH commuter rail riders to board local transrt veh i cles at or near rail stations without charge. MOTA has gone beyond this agreement to allow Tri Rail passengers wHh any type of ticket or pass (other than a one-way ticket, which must be surrendered upon transfer) to board any MOTA vehicle free of charge anywhere in the system. In the reverse direction, passengers transferring from a loca l transit vehicle to Tri Ra i l or the Coaste r have the local cash fare deducted from the commuter rail fare. Among transH agencies in the Los Angeles region, Santa Clarita has moved farthest i n encouraging intermodal connections wHh Metrolink. T ransit agency representatives in the L os Angeles area suggest that H is unl ikely that a completely coordinated transH fare policy will develop, because there are too many providers, with responsibility to too many jurisdictions, to expect that a single fare structure could be implemented. Most major trans i t providers in the reg i on howeve r do have fare agreements designed to encourage intermodal connections. Schedule Coordination The timing of transit services at intermodal centers can dramatically affect the ease of transferring between modes. Many bus networks rely on a timed-transfer system in which buses are scheduled to meet downtown or at a major activity or transfer center. Buses are held until all buses arrive ensuring that transfers can be made without delay The application of this concept across modes can be more difficult especially s i nce a fixed guideway system is likely to create new focal points for transfers at Hs stations and so have many timed tra n sfers to meet. 99

PAGE 100

Heavy or light rail systems operate with sufficient frequency during most of the day that timed transfers with other modes are not a major issue. Coordination tends to take place more often during the lat e evening hours when transit service is less frequent. Commuter rail schedule coordination with other modes can be troublesome, particularly in outlying areas lacking operational frequencies typical of grid service. Tri-Rail in South Florida is an Interesting example of schedule coordination policies. The Tri-RaiVMetrorail connection is the major intermodal connection in the region, and is vita l because it allows Tri-Rail passengers access to downtown Miami. Metrorail frequencies are sufficient to ensure reasonable connections with hourly Tri-Rail service. However neithe r system holds its trains if the other is delayed. This occasionally causes serious problems if Metrorall suffers a power outage or some other delay, since transferring passengers then must wait one hour for the next Tri-Rail train At most othe r stations, T riRail has dedicated feeder bus service. As noted in a previous report, this s i tuation arose owing to the temporary origins of Tri-Rail. Local transit agencies were unwilling to disrupt their existing bus networks to serve a rail l ine that might operate for only two or three years In addition, Broward County Trans i t and PalmTran re l y on timed transfers at central l ocations T o fit bus schedules with Tri-Ra i l service and timed transfers e l sewhere would have been a challenge and it would have been more difficult to hold the buses if Tri-Rail were late. TriRail has had problems with schedule reliability, because the signal system i s old and n ot appropriate for commuter ra il operations and because the CSX right-of-way consists of a single track with sidings. Thu s a dedi cated feeder bus network was imposed over existing routes. The schedu l es of the feeder buses are driven by train arrival times and they will be held if the train i s late Most othe r systems in the case studies have at l east a minimal l evel of schedule coord i nation. The benefits to passengers are that it provides them with convenient connections and with a sense of confidence that t h e bus or train w ill be there when they transfer Thus, schedule coordination supports i ntermodal usage. 100

PAGE 101

The Role of Monthly Passes Nearly all transrt systems offer monthly passes as a fare payment option The passes are typically priced at the cost of between 32 and 40 single fares> Aside from the volume discount offered for regular transit users, passes provide a convenient way of pay ing for transit service and eliminate the need to carry exact change. Monthly passes also p l ay a role in encouraging intermodal usage. For systems with transfer fees. monthly passes elim i nate the need to pay an additional transfer charge. It is hypothesiZed that monthly passes are more appealing to riders who regularly transfer, because the savings and convenience are even greater. For intersystem transfers, monthly passes from one system are often accepted on othe r systems. this is perhaps most common for commuter rail passes A Metrolink pass in the los Angeles area, for example, is accepted by several local transit systems The liberal transfer provisions accorded to Tri Rail riders by MOTA results in the Tri-Rail monthly pass functioning effectively as a regional transit pass. MOTA is one of the few systems in the case studies that charges for parl
PAGE 102

' The Regional Pau Concept The Metropolitan Transportation Commission in the San Francisco Bay Area is spearheading an innova t ive project to develop a universal transit pass good on all of the transit systems in the Bay Area. The Translink project, as it is known, was in troduced in 1994 on BART trains and express buses and on County Connection buses in Contra Costa County and Is being gradually extended to other transit systems in the nine county region Translink uses smart card tedlnology that has the potential to foster major changes in transit fare payment media over the coming years. The Metropolitan Transportation Commission has as its goal an in tegrated transit network with a single fare card usable throughout the region by early I n the 21st century. Short of a regional pass using smart card tedlnology, other developments in the Bay Area are also intended to make a single system's transit pass more valuable to pass holders. The BART Plus program provides a BART pass that can be used as a flash pass on local bus service and as a stored value pass on BART. Participating bus systems i nclude Muni, AC Transit, SamTrans in San Mateo County, the County Connection in Contra Costa County, and SCCTA. BART Plus passes are valid for only half a month, but two can be purchased at the same time The concept of a regional pass for South Florida was examined in a recent study .3 The study concluded that the Tri-Rail monthly pass was, for all intents and purposes, the equivalent of a regional pass. The only places where the Tri-Rail pass is not accepted is on local bus routes in Broward and Palm Beach Counties that do not pass by Tri-Rail stations. As noted ear1ier, the approach taken to accepting the Tri-Rail pass is an example of intermodal cooperation among several agencies. Center for Urban Tran.sponation Research, Southeast Florida Regional Transit Fare Coordination Study Tampa: prepared for CoTran/Broward County Transii/Metro-Dade Transit Agencyrrri-Rail by CUTR. December 1994. 102

PAGE 103

The Translink project in the Bay Area is likely to be only the first attempt to use fare media technology as a means to integ rate transit systems within a region into a seamless network. As smart card technology continues to evolve in areas outside of transit, the application of this technology to improve intermodal connections should occur at an accelerated pace. Theoretical Assessments of Fare Policy A recent theoretical contribution to fare policy decisions has been the "deep discount" concept Proposed by Richard Oram, this Idea relies on market segmentation and differing sensitivities to fare level s to encourage transit ridership while maximizing transit revenue While not directly related to intermodal issues. the deep discount approach has impl icat ions for various pricing dec isions. Oram began by segmenting transit riders i nto three groups. The first Is very frequent riders, those who use transit intensively for work and other purposes. Next is the somewhat frequent riders, who may only use transit for the journey to work or for other limited and well-defined purposes. The final group is the infrequent riders, those who use the system only occasionally. Each of these groups has different sensitivities to fare levels, and different fare media are appropriate for each group. The very frequent riders are likely to be most interested in a monthly pass. Somewhat frequent riders are sensitive to fares, while infrequent riders are relatively insensi tive due to only occasional use. A monthly pass targeted at very frequent riders will have a higher price tag tha n the standard equivalent of 32 to 40 monthly fares. Very frequent riders can take 80 trips per month on the transit system, and a lower price results in revenue foregone fr om this group. Thus, deep discount theory advocates a relatively high price for the monthly pass in spite of conventional w isdom. The price should put it beyond the reach of the middle group of riders. at perhaps 50 or even 60 trips per mon t h Note that frequent riders will still get a 103

PAGE 104

discount from what they would pay in individual fares. Continued rewards for frequent riders at a higher level is akin to recent changes in frequent flier programs by some airlines. The next group addressed by deep discount theory is the infrequent users. Since they are only occasional users, the theory argues that they are not sensitive to price; they are likely to use transit only when they must. The base fare should be set at a relatively high level, to maximize revenue from the infrequent user group If monthly pass prices are high and the base fare is high, then how can the middle group of riders, those who use regularly but only for one purpose, be accommodated? For these riders deep discount theory recommends prepaid fare media, tokens or tickets, sold in bulk (e.g., ten-packs) at a discounted price of at least 25 percent below the base fare This pricing provides a (deep) discount for the somewhat frequent riders, but requires advance purchase. The theory is that when the fare media have been purchased and are available, somewhat frequent riders will be more likely to use transit for other travel purposes. Deep discount programs are typically at times of fare increases, and provide a way of shielding riders from the effects of a fare increase If they choose to shift to prepaid media. Presumably, those riders who are sensitive to fare levels will want to do so, while those who are relatively insensitive will cont i nue to pay the base fare on a trip-by-trip basis Some have argued that low-income riders are not l ikely to take advantage of prepaid media because of the large one-time purchase price and so are forced to pay the high base fare Results of deep discount programs have been mi xed. There are cases in which ridership and revenue have both risen, but there are also cases In which revenue increases have occurred in conjunction ridersh i p decreases, the typ i cal result of a fare increase. A deep discount scheme could be applied with provisions that the prepaid fare med i a be accepted on multiple modes. Thus, deep discount theory is neutral with regard to interrnodal integration In practice, however, reliance on the monthly pass as a means for interrnoda l users to avoid transfer charges would mean that the increased pass price could 104

PAGE 105

discourage lntermodal use. A marl
PAGE 106

The following represent the major findings of th i s task report. These findings arise both from the theoretical perspective presented in subtask one and the empi rical evi dence offered i n all of the subtasks. Use of graph theoretic techniques offer one method for evaluating the advantages of intermoda l transit systems, as well as a method for evaluat ing i mprovements within the overall system, particularly those related to the nodes or stations on the i ntermodal network It was shown using the theoretical constructs why nodes have added significance in an intermodal network Since, practically speaking, it is not possible to continually add links to the lntermodal netwo r k, the nodes offer the potent ial for both improvements to the system or, conversely, fo r impeded flows through the system. Even with the added significance of the nodes, It Is necessary and usefu l to consider the entire network While intermodalism offers an advantage from the standpoint of more connections through the region and increased accessibility, i t also encourages more transferring between modes, thereby r aising the potential for transfer penalties to make travel through the network more d i fficult thereby poss i bly reducing ridership From this perspective intermoda l ism is a double edged sword. The potentia l for increased connections and access is accompanied by the potential fo r increased penal ties associated with transit use. This further supports the need to car efully cons ider how those transfers are made and to attempt to reduce the i mpedances through the network nodes as much as possible. There are many issues which must be considered i n relation to movement through the network and specifically, the nodes or stat ions. One important element of the in t ermodal netwo r k is simp l y to provide adequate physical connections between the modes. Most of the case study systems have a few n odes upon which the i ntermodal network is focused In addition. it was f ound that provision of parkand ride and bus connections at rail stat ions throughout the network will encourage 106

PAGE 107

ridership. For most of the case study systems, automobile access was essential, as the majority of rail riders use automobiles at the home end of the trip. Another aspect that must be considered is the ease with which passengers can assimilaet infonnation about the intermodal network. Passengers, particularly those new to transn use, need to be able to easily understand scheduling and movement through the intermodal connections. The methods used for information dissemination vary from system to system and from station to station. Most systems provide adequate information to the intermodal passenger, but some examples showed that that is not always the case. Station locations represent another major element of the intermodal network. Transit facilttles are major public facilities which must be located well with respect to b oth the transit dependent population and the discretionary rider. However, there are many outside influences which can cause the actual station locations to deviate from the optimal. Foremost among the examples offered were influences related to the use of existing trackage (usually a fiscal consideration), public or citizen opposition to proposed alignments and station locations, and environmental and land use constraints to station locations. Even wnh locational constraints systems may find satisfactory station locations with respect to efficiency and equtty concerns. South Florida offers an example of one system that has faced both locational constraints and equity concerns in relation to the use and locat ion of rail facilities. A simple evaluation of sta tion locations with respect to demographic and socio-economic variables shows that these locational constraints don't ne<:essarily have to re sutt in a suboptimal system, though it should be recognized that different types of rail location decisions will have different locational requirements and solutions. Fare policies and structures that avoid or minimize monetary transfer pena lties are most effective in encouraging intermo dal usage. 107

PAGE 108

Guideway TtaMitand lntermodallsm: Fund/on and E/feclhwt111 Locat ion of a new fixed guideway system affects the importance of feeder bus services and park-and-ride lots. F i xed guideway utilizing freight rights-of-way tend to have less supportive land uses surrounding stations, whereas newly constructed fixed guideway can have station locations selected at sites that are already activity centers The convenience of interrnodal connections is vital to encouraging use. Distances should be minimized, off-street bus bays utilized, and (including clear signage) provided. Design elements of interrnodal centers can enhance passengers' experience of transferring. Bus network restructuring is often necessary to maximize interrnodal usage and allow each mode to carry best function. Grid bus systems generally need only minor adjustments in station areas, but radial systems require more extensive changes Feeder bus service is more importan t at outlying stations, where the fixed gui deway portion of the total trip is larger and travel time savings are greater. Man y areas new guideway systems have evolved a bus network that is a grid in the centra l business district and the inner city, and a modified radial system with activity centers (including guideway stations) as nodes in suburban outly ing neighborhoods All stations do not have to become major nodes, and apparently duplicative bus service can actually serve a different market from fixed guideway. Planners' knowledge of the area and understanding of the transportation network come into play ; there is no fixed set of rules that apply in all settings. Automobile access to guideway systems is critical, since automobile drivers and passe ngers are the major untapped market for fixed guideway P ark -and-r ide lots are especially important in outlying suburban areas. Park-and-ride lots and feeder bus service both extend the market for fixed guideway transit beyond wa lk ing distance from a station. Because automobile access is omnidirectional {with in the constraints of the roadway network) park-and-ride lots have a greater effect i n t his regard. 108

PAGE 109

lntennodal integration is easier to achieve in metropoman areas where a single transn agency operates the major modes. However, operational fragmentation does not necessarily mean that intennodalism cannot work. In cases such as San Francisco and South Florida, regional or statewide agencies such as MPOs and DOTs can play a vital unifying role in encouraging lntennodalism lntennodalism is increasing ly recognized in Mure plans as a sound policy. Railway stations are major focal points in intennodal planning. Airports may also play an increasingly significant part in integrating all transportation modes The best transfer pricing policy is to provide free transfers between modes This encourages lntennodal use and gives the sense that once the passenger i s aboard a transit vehicle, he/she has access to the entire system. Interagency transfer agreements have been traditionally more difficuH to reach than intra-agency policy. Recent new foced guideway systems have changed this trend, as existing local agencies have ta ke n an intennodal approach encouraging use of all modes by offering free transfers. At least a minimal leve l of schedule coordination is provided by transit agencies in the case study areas. Schedule coordination is most i mportant at times and places where service is relatively infrequent. Monthly passes provide a volume discount and increased convenience for transit riders In systems with a transfer charge, passes allow transferring passengers to avoid paying the charge, and thus are more attractive to i ntermodal riders The Metropolitan Transportat i on Commission in the San Francisco Bay Area has embarked upon a regional pass project tha t could be a forerunner of fare policies in major metropolitan areas in the next century. The goal of the Translink project is to create a universal fare medium accepted by all transit agencies within the Bay Area. 109

PAGE 110

The most recent theoretical development in the fate area is the deep discount concept. This involves a high monthly pass price, a high base fare, and deeply discounted prepaid media (tokens or tickets) available in bulk. Deep discount proponents claim that it can increase both ridership and revenue, aHhough results have been mixed. Deep discount should be neutral with regard to intermodal integration, although in practice its effect on pass prices may discourage intermodal use to some extent. 110

PAGE 111

References Amtrak. 1994 Amtrak California Timetab l e, Effective October 30, 1994. Bernick M. and R. Cervero. 1994 Transit-Based Development in the United States. passenger Transoort, Jan., 1994 Catellus Development Corporation. 1994 Highway and Transit Legislation: DRAFT report City of Los Angeles Planning Department. 1993. Land Usecrcansoortalion City of Los Angeles P l anning Department. 1993 Land Use/ Transportation Policy: A Guide to Transit Oriented Development. Cole, J P and C.A.M King. 1968. Quantitatlye Teoiaues and Theories in Geo graphy. London: John Wiley and Sons Ltd. Crump, S. 1978. Henrv Hyntingtpo and the Pacific E l ectrjc Glen d ale, CA: Trans Anglo Books. CUTR. 1994 Commuting Allematjyes in the Unijed Stales. Center for Urban Transportat i on Research Tampa. CUTR 1995a. Guideway Transit and lntermodalism: Function and Effectiveness. Case Study, Atlanta. Draft Center for Urban Transportat i on Research Tampa. CUTR. 1995b. Guideway Trans i t and lntermodalism: Function and Effectiveness. Case Study, Los Angeles Draft. Center for Urban Transportation Research Tampa CUTR. 1995c. Guideway Transit and lntermodalism : Function and Effectiveness. Case Study New Orl eans. Draft. Center for Urban Transportation Research Tampa. 111

PAGE 112

CUTR. 1995d Guid eway Transit and lntermodal i sm: Function and E ffectiveness. Case Study, Portland. Draft. Cente r for Urban Transportation Research, Tampa CUTR. 1995e. Gui deway Trans i t and lntermoda li sm: Function and Effectiveness. Case Study San Francisco Draft. Center for Urban Transportation Research T ampa CUTR 1995f. Gu i deway Transit and lntermodal ism: Functio n and Effective n ess Case Study, South Florida Draft. Center for Urban Transportation Research, Tampa. CUTR 1994 Tri-County Commute r Rail Authority Transit Development Plan 1995 1999 F in al Report Cente r for Urban Transportation Research, Tampa Dav i s M 1 990 City of Quartz New Yo r k : Vintage Books Dougherty J. 1994 Transit Moves as L.A Quakes Passenger Iransoort, January 24, 1994, p 1 Ehrenkrantz and Eckstut Arch i tects. 1991.A i ameda D i strict P l an. Env i ronmenta l Protect i on Agen c y Office of Air Quality Pla n n i ng and Standards 1991 Ozone and Ca r bon Monoxide Areas Designated Non-attainment. Facts Consol i dated. 1993. Metro l ink Riders Survey Report prepared for Commute r T ransportation Services Inc. G i lson James R. and F Michael Francis 1993 P l anning fo r Joint Deve l opme n t i n Los Angeles Urban L a n d June 1993, pp. 30 32. Hawthom, Ga ry. 1991. Tran s portation Prov i s i ons i n the Clean Ai r Act Amendments of 1990. I TE Jouma April 1991 pp. 17-24 112

PAGE 113

Kansky K.J. 1963. Structure of Transportation Networks: Relationships Between Network Geometry and Regional Characteristics. U n iversity of Chicago, Research Paper No. 84. Long Beach Transit. 1992 Transfer S u rvey; Internal Memorandum. May 28, 1992. Los Angeles County Transportation Commission. 1993 Metro Rail Transit System Planning Guide : An Overview of the Rail Transit Devel opment Process. Los Angeles County Metropolitan Transportation Authority 1993. A Look at the Los Angeles County Metropo li tan Transportation Authority. Los Angeles County Metropol it an Transportation Authority No Date Blue Line Features (Brochure) Los Angeles County Metropolitan Transportation Authority. 1993 Congestion management Program for Los Angeles Cou nty L os Angeles Cou nty Metropo l itan Transportation Authority 1994. Ejscal Year 19941995 Budget. Los Angeles County Transportation Commission. 1993. Inner CitY Transit Needs Assessment Study. Los Angeles County Transportation Commission. 1991. LACTC Recommended Integr a ted Transportation Plan. L os Angeles County Met ropoli tan T ransportation Authority 1994 Metro Green Line Northern Extension : Final Supplemental E n vironmental Impact Report Los Angeles County Metropo l itan Transportation Authority 1994. M T A Long Range Transportation P l an Reassessment. Executive Summary 113

PAGE 114

Los Angeles County Metropolijan Transportation Authority. 1993. Short Bange Transit Plan and Transportation Improvement Program: Fiscal Years 1994-1997. Los Angeles County Metropolitan Transportation Authority 1994. Short Bange Transit P lan and Traosoortatjon Program : Ejscal Years 1995-1998. Los Angeles County Metropo l itan Transportation Authority 1992. Thirty Year Integrated Transportation Plan. Los Angeles County Metropolitan Transportation Authority. 1994. Union Station Gateway Project Report on Headguarter:s Eacjljtv and Gateway lnteanoda l Transit Center. Metro Magazine. 1990 Build ing a Bail Line Means Making Passengers Happy. MEmo Magazjne, May/June 1990, pp. 38 -39. Metro Magazine. 1993. Day-care Centers Planned at Rail Station METRO Magazine, May/June, 1993. Metro Magazine. 1993. L.A. Commuter Bail Opens: 400 miles due. METRO Magazine. January/February, 1993 p. 8. Metro Magazine 1990. Los Angeles Opens 22m ile Light Rail Line METRO Magazjoe. September/October, 1990, pp.108 -111. Metro Magazine. 1993. L.A. Subway Opens $1.4 Billion Segment. METRO Magazine, March/April, 1993 Pacific/West Communications Group, Inc. 1994 1994 Metrolink on-board Bider Survey, Executive Summary of Key Fi ndings (Topl ine Report) Passenger Transport. 1994. Los Angeles Breaks Ground for 6 3 mile Red Line Extens i on Passenger Iransoort. March 21, 1994, p. 8. 114

PAGE 115

Passenger Transport. 1991. L.A. Outlines Prop. C Spending. passenger Transport. January7, 1991, p.4 Passenger Transport. 1991. Los Angeles Marks Six-Month Anniversary of Metro Blue Line Passenger Transoort, January 28, 1991, p.3. Passenger Transport. 1994. Los Angeles MTA to Cut, Costs, Staff PassengerTransoort. April4, 1994 p. 2. Passenger Transport. 1994. S. California's Metrolink Responds as an Earthquake Veteran. Passenger Transport February 14, 1994, p. 3. Schaeffer, K.H. and E. Schar. 1975 Access for All: Transportation and Urban Growth. Baltimore: Penguin Books. Southern California Rapid Transft District. 1990. Short Range Transit Plan: Guideway Plan, Fiscal Years 1991-1995. Southern California Rapid Transit District 1993. Short Range Transit Plan: Guideway Plan, Fiscal Years 1993 1997. Southern California Rapid Transit District. 1993. Short Range Transit Plan: Facilities Plan, Fiscal Years 1993-1997. Southern California Regional Rail Authority. 1993. Metrolink Statjon Design Manyal Southern California Regional Rail Authority. 1994. Metroljnk Perfonnance Summaries, July Southern California Regional Rail Authority. 1994. Preljrojnarv Bydget. Ejscal Years 1994 115

PAGE 116

Southern California Regional Rail Autho rity. 1994. Regjonal System Plan Update. Southern California Regional Rail Authority. 1991. Southern California Commuter Rail 1991 Regional System Plan. Sullivan, J.P. 1992. Security on L.A. Metro: No Barriers, Few Fare Evaders Railway Age, November 1992, pp. 78 -79. Taaffe, E.J. and Howard L. Gauthier. 1973. Geograohy ofTraosportation. Englewood Cliffs, N.J.: Prentice-Hall, Inc. University of Southern California. 1990 Air Pollution Study Shows Lung Damage in Los Angeles Youths. L.A. U.S.C., March 20 1990. Press Release. U.S. Department of Commerce, Economics and Statistics Administration Bureau of the Census 1991. State and Metropolitan Area Data Book. U.S. Department of Transportation, Federal Transit Administration 1993. Review of the Transportation Planning Process in t he Southern California Metropolitan Area. 116

xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ia 4500
controlfield tag 001 029186474
005 20110825131553.0
006 m o d
007 cr cn|||||||||
008 110825s1995 fluab ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a C01-00034
b .I58 1995
0 245
Integrated planning and design of intermodal systems
h [electronic resource] :
interim task report /
Center for Urban Transportation Research, College of Engineering, University of South Florida ; project managers, Daniel Boyle, Tracy Newsome ; project statff, Rongfang Liu ... [et al.].
1 246
i At head of title:
Guideway transit and intermodalism :
function and effectiveness
{Tampa] :
University of South Florida, Center for Urban Transportation Research,
1 online resource (116 p.) :
ill., col. maps.
Title from e-book cover (viewed Aug. 25, 2011).
"July 1995."
"... under the sponsorship of the Department of Transportation, Federal Transit Administration. ..."
Includes bibliographical references (p. 111-116).
Local transit
z United States
x Planning.
Local transit
United States.
Boyle, Daniel K.
Newsome, Tracy H.
Liu, Rongfang
United States.
Federal Transit Administration.
University of South Florida.
Center for Urban Transportation Research.
Print version:
t Integrated planning and design of intermodal systems.
d {Tampa] : University of South Florida, Center for Urban Transportation Research, 1995
w (OCoLC)269707794
Center for Urban Transportation Research Publications [USF].
4 856