ISSN 1055- 1425
April 2009
This work was performed as part of the California PATH Program of the
University of California, in cooperation with the State of California Business,
Transportation, and Housing Agency, Department of Transportation, and the
United States Department of Transportation, Federal Highway Administration.
The contents of this report reflect the views of the authors who are responsible
for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of the State of California. This
report does not constitute a standard, specification, or regulation.
Final Report for 6406
CALIFORNIA PATH PROGRAM
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
A Combined Quantitative and Qualitative
Approach to Planning for Improved Intermodal
Connectivity at California Airports
UCB- ITS- PRR- 2009- 27
California PATH Research Report
Xiao- Yun Lu, Geoffrey D. Gosling, Avi Ceder,
Steven Tung, Kristin Tso, Steven Shladover,
Jing Xiong, Sangwon Yoon
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
A Combined Quantitative and Qualitative
Approach to Planning for Improved
Intermodal Connectivity at California Airports
Final Report
Task Order: 6406
Project Team:
Xiao- Yun Lu, Geoffrey D. Gosling, Avi Ceder, Steven Tung, Kristin Tso,
Steven Shladover, Jing Xiong, and Sangwon Yoon
- ii -
Key Words
Intermodal connectivity, airport ground access, air passenger mode choice, transportation
provider behavior, project evaluation, system performance measures, connectivity performance
measures, policy recommendations
Abstract
This report has been prepared as the final deliverable for a research project developing a
combined quantitative and qualitative approach to planning for improved intermodal
connectivity at California airports. The quantitative approach involves the development of an
Intermodal Airport Ground Access Planning Tool ( IAPT) that combines transportation system
performance measurement, an air passenger mode choice model, and a model of transportation
provider behavior, and is designed to interface with a traffic network analysis model. The
qualitative approach is used to enhance the quantitative analysis to account for factors that are
difficult to quantify and to provide recommended policy and planning guidelines.
This report documents the progress on the project over the past three years. It describes
the following main tasks:
• Identification of opportunities for improved intermodal connectivity at California
airports
• Research into techniques for modeling air passenger mode choice and
development of a mode choice model for use in subsequent analysis
• Development of techniques for modeling transportation provider behavior
• Performance measurement definitions and calculations addressing both system
performance and connectivity performance
• Design and development of a prototype version of an Intermodal Airport Ground
Access Planning Tool
• Use of the prototype IAPT to evaluate selected projects at three Bay Area airports
• Development of policy recommendations and guidelines for project evaluation
- iii -
IAPT provides a standard way for quantitative project evaluation at an airport level. A user
friendly graphical interface makes it easy for a user to define projects for a given airport, input
data, select model parameters, choose performance parameters for comparison, run the analysis
process, and view the output in different ways. Requirements for further development of the
IAPT are discussed and recommendations for future study of airport ground access planning
issues are presented.
- iv -
Acknowledgements
This work was performed as part of the California PATH Program of the University of
California, in cooperation with the State of California Business, Transportation, and Housing
Agency, Department of Transportation ( Caltrans); and the United States Department of
Transportation, Federal Highway Administration. The contents of this report reflect the views of
the authors who are responsible for the facts and the accuracy of the data presented herein. The
contents do not necessarily reflect the official views or policies of the State of California. This
report does not constitute a standard, specification, or regulation.
The guidance and support from Colette Armao, Terry Barrie and Debbie Nozuka, of the
Caltrans Division of Aeronautics and Dan Lovegren, the Project Manager at the Caltrans
Division of Research and Innovation, are gratefully acknowledged.
- v -
Contents
Page
Key Words and Abstract ............................................................................................................ ii
Acknowledgements .................................................................................................................... iv
Table of Contents ........................................................................................................................ v
List of Figures and Tables ....................................................................................................... viii
List of Acronyms ........................................................................................................................ xi
Executive Summary ................................................................................................................. xiv
Chapter 1 Introduction .............................................................................................................. 1
1.1 Scope of this Report ................................................................................................... 2
1.2 Role of Modeling in Quantitative Analysis ............................................................... 3
1.3 Dynamic Interactions in Airport Ground Access Activities ...................................... 5
1.4 Capabilities and Limitations of Modeling ................................................................. 6
1.5 Structure of this Document ........................................................................................ 8
Chapter 2 Literature Review on Intermodal Airport Ground Access ............................... 10
2.1 Airport Ground Access Planning ............................................................................. 10
2.2 Intermodal Transportation Planning Principles ....................................................... 11
2.3 Quantitative and Qualitative Approaches in Airport Planning ................................ 14
2.4 Policy and Institutional Issues ................................................................................. 17
2.5 Mode Choice Modeling and Analysis ..................................................................... 18
2.6 Airport Ground Access Travel Information ............................................................ 19
2.7 The Government Accountability Office Study ........................................................ 20
Chapter 3 Opportunities for Improving Intermodal Connectivity at California
Airports ................................................................................................................... 25
3.1 California Airport Ground Access Needs ................................................................ 25
3.2 Potential Strategies to Enhance Intermodal Connectivity ....................................... 30
3.3 Intermodal Opportunities at Selected California Airports ....................................... 33
3.4 Further Analysis of Potential Intermodal Opportunities ......................................... 42
3.5 Intermodal Air Cargo Considerations ...................................................................... 58
3.6 Institutional Issues ................................................................................................... 67
- vi -
Page
Chapter 4 Design and Development of the Intermodal Airport Ground Access
Planning Tool ......................................................................................................... 68
4.1 Overview of the IAPT ............................................................................................. 68
4.2 Functionality of Main Components ......................................................................... 69
4.3 Software Structure and Data Flow ........................................................................... 74
4.4 Graphical User Interface and Functionality ............................................................. 78
Chapter 5 Passenger Mode Choice Modeling ....................................................................... 97
5.1 Air Passenger Mode Choice Model Development .................................................. 98
5.2 Literature Review on Air Passenger Mode Choice ............................................... 108
5.3 Data Preparation for Modeling .............................................................................. 117
5.4 Model Development and Calibration ..................................................................... 125
5.5 Model Validation ................................................................................................... 134
Chapter 6 Modeling Transportation Provider Behavior ................................................... 136
6.1 Operational Considerations ................................................................................... 138
6.2 Literature Review .................................................................................................. 151
6.3 Interaction Between Passenger and Transportation Provider Decisions ............... 159
6.4 Simplified System Representation and Justification of Assumptions ................... 165
6.5 Transportation Provider Costs ............................................................................... 169
6.6 Model Implementation Issues and Current Status ................................................. 174
Chapter 7 Intermodal Airport Ground Access Systems Performance Measurement .... 177
7.1 Planning Considerations ........................................................................................ 178
7.2 Developing Measures of Airport Connectivity ...................................................... 180
7.3 Detecting Weaknesses in Intermodal Airport Access Trip Chains ....................... 182
7.4 Measuring Intermodal Airport Ground Access Connectivity ................................ 183
7.5 Performance Measurement Definitions and Implementation in the IAPT ............ 188
7.6 Example of MOP Calculation ................................................................................ 191
Chapter 8 Guidelines for Project Evaluation Using the IAPT .......................................... 195
8.1 Project Definition ................................................................................................... 195
8.2 Project Evaluation .................................................................................................. 196
8.3 Institutional Considerations ................................................................................... 198
- vii -
Page
Chapter 9 Potential Bay Area Case Study Analysis ........................................................... 204
9.1 Definition of the Analysis Scenarios ..................................................................... 205
9.2 Representation of the Case Studies Using the IAPT ............................................. 208
Chapter 10 Policy Recommendations .................................................................................. 212
10.1 Potential Opportunities for Improved Intermodal Connectivity ........................... 213
10.2 Institutional Aspects ............................................................................................. 213
10.3 Project Funding ..................................................................................................... 215
10.4 Technical Support ................................................................................................. 216
Chapter 11 Concluding Remarks and Recommendations for Future Development ...... 218
11.1 Potential Follow- on Research ................................................................................ 218
11.2 Future Development of the IAPT .......................................................................... 223
References ............................................................................................................................... 225
Appendices
A. Details of Recent Air Passenger Model Choice Models ................................................. A- 1
B. Implementation of Nash Game Modeling of Transportation Providers ...................... B- 1
C. Measures of Performance and Evaluation Analysis ...................................................... C- 1
D. IAPT Data Table Specifications and Data Preparation ................................................ D- 1
E. Technical Aspects of the IAPT ......................................................................................... E- 1
F. Sample Data Files Used for the Development of the IAPT ............................................. F- 1
- viii -
List of Figures and Tables
Figure Page
1- 1 Feedback Between Transportation Provider and Airport Traveler Behavior ................... 6
3- 1 Comparison of Ground Transportation Options .............................................................. 28
3- 2 Coliseum Station Rendering ............................................................................................ 44
3- 3 Oakland Airport Station Rendering ................................................................................. 45
3- 4 Alignment of Oakland Airport Connector Preferred Alternative ................................... 46
3- 5 AirBART Roundtrip Route ............................................................................................. 47
3- 6 SJC Automated People Mover Preferred Alignment ...................................................... 48
3- 7 Airport Flyer ( VTA Line 10) Route ................................................................................ 49
3- 8 Proposed Ferry Service Routes to South San Francisco ................................................. 50
3- 9 Potential Shuttle Bus Route Between Oyster Point Ferry Terminal and SFO ................ 51
3- 10 Marin Airporter Larkspur Landing Terminal ................................................................. 53
3- 11 Caltrain System Map ....................................................................................................... 54
3- 12 Potential Santa Clara County Off- Airport Terminal Location ....................................... 57
3- 13 Bay Area Air Cargo Forecast .......................................................................................... 63
3- 14 Air Pollution Caused by Goods Movement Activities ................................................... 64
4- 1 Conceptual Modeling of the Intermodal Airport Ground Access System ...................... 70
4- 2 Functional Structure of the Analysis............................................................................... 72
4- 3 Overall Picture of Software Structure ............................................................................. 75
4- 4 Flow- chart of the Mode Choice Model Module Data Flow ........................................... 75
4- 5 Hierarchical Structure of GUI: Each Tab Represents a Major Functional Component . 79
4- 6 Initial IAPT Screen ......................................................................................................... 79
4- 7 MOP Definition, Editing and Selection Screen .............................................................. 81
4- 8 Airport Definition and Relevant Parameter Input Screen ............................................... 83
4- 9 Mode Choice Model Coefficient Input Screen ............................................................... 84
4- 10 Typical Project Definition Screen ................................................................................... 86
4- 11 Data Entry Screen – Regional Data ................................................................................ 88
4- 12 Data Entry Screen – Clicking Any “ Open” Button Allows User to Input the File
from the Windows File Server .................................................................................. 88
4- 13 Data Entry Screen – Modal Fare and Cost Data ............................................................. 89
- ix -
Figure Page
4- 14 Data Entry Screen – Transportation Provider Service Data ........................................... 90
4- 15 Selection of MOP List for Viewing and Output ............................................................. 91
4- 16 Representative Model Run Definition Screen ................................................................ 92
4- 17 Selection of Different Output for Viewing ..................................................................... 93
4- 18 Viewing Output Measures by Mode ............................................................................... 94
4- 19 Viewing Output Measure by Mode ( cont.) – Right Side of Window ............................. 94
4- 20 Viewing Output by Project ............................................................................................. 95
4- 21 Viewing MOP by Project: Lower Panel is a Continuation of the Upper Panel .............. 95
5- 1 Multinomial and Nested Choice Models ...................................................................... 105
5- 2 Market Segmentation for the Bay Area Ferry Study .................................................... 115
5- 3 Classification Tree for Commuter Mode Choice in Athens ......................................... 117
5- 4 Planned Mode Choice Model Structure ........................................................................ 126
6- 1 Interactions of System Components in Transportation Provider Decisions ................. 137
6- 2 Transportation Provider Modeling ................................................................................ 165
6- 3 Path Choice is Equivalent to Primary ( Representative) and Secondary
Mode Choice .......................................................................................................... 167
6- 4 Sample Extrapolation of Hourly Rate ........................................................................... 174
6- 5 Data Aggregation and Dissemination at Each Recursive Step in the
Generalized Nash Game Optimization Process ...................................................... 176
7- 1 Overview of Three Interrelated Study Components ..................................................... 189
7- 2 Measures and Relationships Among Service Input, Service Output and
Service Consumption .............................................................................................. 191
- x -
Table Page
3- 1 Most Significant Ground Access System Deficiencies at Major California Airports .... 27
3- 2 Transit Percent Market Share by Mode .......................................................................... 29
3- 3 Operating and Patronage Assumptions for the OAC Preferred Alternative ................... 47
3- 4 Operating Assumptions for the SJC Automated People Mover ..................................... 49
3- 5 Operating Assumptions for the Proposed South San Francisco Ferry Route ................. 52
3- 6 Operating Assumptions for the Proposed South Peninsula Off- Airport Terminal ......... 55
3- 7 Operating Assumptions for the Proposed Santa Clara County Off- Airport Terminal .... 57
3- 8 Integrated Air Freight Carriers ........................................................................................ 60
5- 1 Estimated Mode Choice Model Coefficients ................................................................ 133
7- 1 Known Transit MOPs in the US – Especially for Buses .............................................. 184
7- 2 List of MOPs Accompanied by their Typical Criteria Range and Data Required ........ 187
9- 1 Input Parameters for Case Study Baseline and Sensitivity Analysis ............................ 207
- xi -
List of Acronyms
ACRP Airport Cooperative Research Program
ACTT Access travel time
AGT Automated guideway transit
AM Ante meridiem ( before noon)
APM Automated people mover
ASC Alternative- specific constant
AVM Automated vehicle monitoring
BART Bay Area Rapid Transit
BRT Bus rapid transit
CAAA Clear Air Act Amendments
Caltrans California Department of Transportation
CMAQ Congestion Mitigation and Air Quality ( program)
CPC Connectivity production cost
CSV Comma- separated values
CTPS Central Transportation Planning Staff ( Boston)
FAA Federal Aviation Administration
FEIR Final Environmental Impact Report
FHWA Federal Highway Administration
FTA Federal Transit Administration
GAO Government Accountability Office
GIS Geographic information system
GUI Graphical user interface
HOV High occupancy vehicle
IAPT Intermodal Airport Ground Access Planning Tool
IIA Independence from irrelevant alternatives
ISTEA Intermodal Surface Transportation Efficiency Act
ITLSS Intermodal Terminals Location Simulation System
ITS Intelligent transportation systems
IVTT In- vehicle travel time
LAWA Los Angeles World Airports
- xii -
List of Acronyms ( cont.)
LAX Los Angeles International Airport
Massport Massachusetts Port Authority
MBTA Massachusetts Bay Transportation Authority
MNL Multinomial logit ( model)
MOCP Measure of connectivity performance
MOP Measure of performance
MOSP Measure of system performance
MPO Metropolitan planning organization
MTS Metropolitan Transit System ( San Diego)
MTC Metropolitan Transportation Commission ( San Francisco Bay Area)
MTD Metropolitan Transit District ( Santa Barbara)
OAC Oakland Airport Connector
OAK Oakland International Airport
ODBC Open Database Connectivity
NCIT National Center for Intermodal Transportation
NL Nested logit ( model)
NOx Nitrogen oxides
OD Origin- destination
PATH Partners for Advanced Transit and Highways
PDX Portland International Airport
PFC Passenger facility charge
PM Particulate matter
PM Post meridiem ( after noon)
PT Public transportation
PVH Passengers per vehicle- hour
PVM Passengers per vehicle- mile
RB Resident business ( trip type)
RCM Recursive partitioning methodology
RFP Request for proposals
- xiii -
List of Acronyms ( cont.)
RP Resident personal ( trip type)
RP Revealed preference
RP Revenue per passenger
RTP Regional Transportation Plan
SamTrans San Mateo County Transit District
SAN San Diego International Airport
SEM Structural equation modeling
SERAS South East and East of England Regional Air Service ( study)
SFO San Francisco International Airport
SJC Mineta San José International Airport
SOV Single occupancy vehicle
SP Stated preference
SPSS Statistical Package for the Social Sciences ( computer software)
STP Surface Transportation Program
TAZ Transportation analysis zone
TCRP Transit Cooperative Research Program
TIFIA Transportation Infrastructure Finance and Innovation Act
UCSB University of California Santa Barbara
U. K. United Kingdom
U. S. United States
VB Visitor business ( trip type)
VHT Vehicle- hours of travel
VMT Vehicle- miles of travel
VOC Volatile organic compounds
VP Visitor personal ( trip type)
VTA Valley Transportation Authority ( Santa Clara)
WT Waiting time
WTA Water Transit Authority ( San Francisco Bay Area)
- xiv -
Executive Summary
This report describes the work conducted over the past three years for a research project
for the California Department of Transportation ( Caltrans) that has been exploring planning
techniques to support improved intermodal connectivity at California airports. The research has
included a literature review, identification of opportunities to improve intermodal connectivity at
a wide range of California airports, extensive system modeling and analysis, design and
development of an Intermodal Airport Ground Access Planning Tool ( IAPT) to evaluate
potential airport ground access improvements, definition of potential projects to enhance
intermodal connectivity at the three major airports in the San Francisco Bay Area that could
serve as case studies to demonstrate the use of the IAPT, and the development of policy
recommendations and project evaluation guidelines. The IAPT development included the design
of the overall software structure and data flow, definition and implementation of the functionality
of each sub- module, design of the supporting databases, and development of a Graphical User
Interface ( GUI). The IAPT was developed using Microsoft Visual Studio. Net. It is designed to
evaluate proposed projects at the airport level as part of intermodal ground access planning for
improved connectivity and system performance. Each chapter of the report is summarized in the
following sections.
( 1) Introduction: The first chapter describes the overall objectives of the project and the scope of
this report. It examines the role of modeling in quantitative analysis of strategies to improve
intermodal connectivity at airports and discusses the dynamic interactions that arise in airport
ground access activities, as well as the capabilities and limitations of such modeling.
The scope of intermodal airport ground access planning for a given airport needs to take
into account the following considerations: air passenger mode choice behavior, transportation
provider behavior, regional transportation network traffic, and the respective roles of the airport
authority and local government agencies. The dynamic interactions between these components
of the intermodal airport access system may be unidirectional or bidirectional. Prediction of air
passenger mode use must be based on an understanding of the behavioral characteristics of each
component and interactions between them. However, since the overall system is very
- xv -
complicated, simplifications are necessary. The main assumptions related to the analytical
approach and the simplification of the overall system are:
( a) Evaluation of the comparative performance of project alternatives is usually the most
interesting consideration for decision makers. System performance can be described using
performance measures such as travel time, vehicle- miles of travel ( VMT) or vehicle- hours of
travel ( VHT), emissions, etc., as well as some composite performance measures quantifying the
connectivity within the transportation system. The values of these performance measures for any
particular project alternative are determined by the interactions between the air passenger access
mode choice, transportation provider behavior and the transportation network traffic.
( b) Competition between modes is assumed to occur collectively rather than between
individual providers within a mode; i. e. all the providers in the same mode are considered to
compete as one with their counterparts in other modes. Although this joint behavior does not
typically occur in practice, the assumption is reasonable in the sense that the behavior of
transportation providers for a given mode can be considered as a collective behavior averaged
over all the providers within the mode. This simplification is consistent with a mode choice
model that predicts the use of different modes rather than specific providers, and also
significantly reduces the complexity of modeling transportation provider behavior.
( c) The modeling of airport access can be distinguished from general transit system
modeling in that origin and destination patterns and vehicle routing can be greatly simplified for
planning purposes and there is no network optimization problem to be faced. An access/ egress
path can be defined that links each primary airport access mode with any auxiliary modes that
provide access to or egress from the primary mode, potentially at both the origin and destination
ends of the trip. Examples of such paths and primary modes are: ( i) single mode trips such as
taxi, shuttle van, pickup/ drop- off by private vehicle, or self- driving with airport parking, and ( ii)
combined mode trips such as rail transit with access by private vehicle parked near the station
and a shuttle bus link from the nearest rail station to the airport, or self- driving to an off- airport
parking lot with a shuttle van link to the airport, for each of which the primary mode is obvious.
The choice of the auxiliary mode( s) may not need to be explicitly considered in modeling air
passenger mode choice, although of course their travel times and costs should be considered
together with those of the primary mode. If only the primary mode choice is considered, there is
a one- to- one correspondence between mode choice and access/ egress path choice. The main
- xvi -
advantage of this simplification is to avoid the need for modeling the selection of the auxiliary
modes throughout the regional transportation network.
( d) The relationship between airport ground access activities and decisions by the airport
authority and local government agencies is unidirectional, and is reflected through the
regulations and policies regarding access to the airport terminal by the transportation providers
and the fees charged for that access. These factors affect passengers indirectly through prices
and waiting times experienced using different modes.
( 2) Literature review: Chapter 2 provides a review of the relevant literature on intermodal
ground access to airports, including the broad range of issues in airport ground access planning,
general principles of intermodal transportation planning, modeling and analysis of airport
traveler access/ egress mode choice, and the role of ground access travel information. The
chapter also discusses the findings of a recent study by the United States Government
Accountability Office on potential strategies that would redefine the Federal role in developing
airport intermodal transportation capabilities.
Intermodal airport ground access planning should follow accepted principles of
intermodal transportation generally, i. e. convenient connectivity, flexible choices of different
modes, coordination between transportation providers, and cooperation and collaboration among
transportation providers and governmental agencies at all levels to ensure seamless service. Thus
the literature review examined the overall picture: intermodal transportation planning principles,
quantitative and qualitative approaches for planning, including airport access planning, passenger
mode choice modeling, transportation provider behavior modeling, planning tool development,
and addressing California airport ground access needs from a strategic planning perspective.
( 3) Opportunities for improving intermodal connectivity at California airports: The third chapter
examines a range of opportunities to improve intermodal connectivity at airport in California.
Many of the these opportunities leverage existing investments in improved public transportation,
particularly rail services, by improving the connections between the airports and nearby rail
services or the regional bus route network. These range from extending light rail lines to
airports, as currently planned for Sacramento International Airport, to improved bus service to
- xvii -
airports or improved links between airports and nearby rail stations, such as the people- mover
connections planned for both Oakland International Airport and San José International Airport.
Many opportunities exist at different levels for improving intermodal connectivity at
California airports: a strategic level, a regional level and an airport operational level. The
chapter reviews the findings of previous work at a strategic level that addressed the needs for
improving the connectivity at California airports for both passenger and commodity movement
( Landrum and Brown, 2001). The chapter then identifies and discusses a range of potential
projects that could improve intermodal connectivity at a large proportion of California
commercial service airports.
( 4) IAPT design and implementation: Chapter 4 describes the design and implementation of a
computer modeling tool, termed the Intermodal Airport Ground Access Modeling Tool ( IAPT),
that has been developed as part of the current research in order to support the analysis of projects
to improve airport intermodal connectivity.
The IAPT has four major components – regional transportation network traffic data for
the different access modes serving the airport, an air passenger mode choice model, which
generates predictions of air passenger mode use across the available modes and thus vehicle trips
for the given airport; a transportation provider behavior model, which predicts the changes in the
service characteristics of the available modes in response to changing traffic levels, and a
performance measurement module, which calculates system performance and connectivity
performance measures. Iteration between the mode choice and provider models leads to the
prediction of vehicle trips and related system performance measures. The performance
evaluation module forms the final step of the IAPT analysis process and generates measures of
the change in system performance and connectivity offered by different alternative projects.
Those components and the underlying database structure are linked with a GUI, which allows
users to define airport characteristics and projects, select alternatives, enter and update relevant
data, run the analysis, and view the outcome and performance measures for comparison in
decision making. The GUI effectively hides the complexities of the modeling and data flow
processes and provides the user with a friendly and standardized planning environment for the
evaluation and comparison of multiple projects in decision- making.
- xviii -
( 5) Passenger mode choice modeling: The following chapter discusses the passenger mode
choice modeling approach adopted in the IAPT in more detail, including a review of the relevant
literature and the development and estimation of a preliminary airport access mode choice model
for Oakland International Airport.
The mode choice model component of the IAPT predicts how air passengers would
choose a mode for their airport access trip based on their air party characteristics and the service
levels offered by the available access modes. A discrete multinomial logit mode choice model
has been adopted in the initial implementation of the IAPT for modeling air passenger mode
selection. The perceived attractiveness of each mode is reflected in a utility function that depends
on several parameters. The essence of the mode choice modeling approach is to calculate the
probability of each air party in a representative sample of air parties choosing each of the
available modes. These choices are then factored up to the total number of ground access trips,
based on an assumed or known total number of passenger trips to and from the airport. This
aggregate demand is obtained from airport traffic statistics or demand forecasts while the mode
choice model parameters are determined from air passenger survey data for the given airport.
( 6) Transportation provider behavior modeling: Chapter 6 discusses the approach to modeling
the behavior of transportation providers serving the airport ground access system within the
analytical framework of the IAPT.
Ideally, such a model should represent the competitive behavior of transportation
providers within and between modes, but in the current implementation of the IAPT the model
only focuses on the collective competition between modes. The most common way of thinking
about provider behavior is to focus on the elasticities of demand that can be observed empirically
by the providers, considering the ridership changes that result from changing service variables,
such as increasing or decreasing the fare or changing the operating frequencies. This approach
can predict the outcomes of unilateral actions by individual modes, but it is more difficult to
predict the outcomes caused by near- simultaneous actions of multiple providers. A game theory
approach, on the other hand, can be applied to capture the dynamic effects of the interactions
among decisions by multiple transportation providers. A few researchers have begun to attack
the problem using this approach, in which a passenger mode choice model is tightly coupled with
- xix -
a provider behavior model. The chapter examines how such an approach can be applied within
the framework of the IAPT and presents a preliminary framework for such an analysis.
( 7) Performance measures: Performance measures are critical for the evaluation of projects for
decision- making. Traditional methods focus on system performance measures such as VHT,
VMT, revenue, travel time, and emissions. Recently, the transportation community has begun to
quantify the connectivity in addition to those system performance parameters. It has been
suggested that both system performance measurement and connectivity performance
measurement be used for project evaluation. Chapter 7 describes the approach to measuring both
system and connectivity performance within the IAPT. The connectivity performance measures
depend on the following factors:
• Average and variance of walking time ( to a service point)
• Average and variance of waiting time ( for scheduled/ non- scheduled services)
• Average and variance of travel time ( on a given mode and path)
• Average and variance of scheduled headway
• Number of transfers required.
Although connectivity measures can be defined based on these factors, the challenge of how to
combine those factors in a single performance measure that can effectively reflect the different
considerations, particularly for airport ground access planning, is not yet well understood and
needs further study in the future.
( 8) Guidelines for using IAPT in airport ground access planning practice: Chapter 8 presents
guidelines for using the IAPT to analyze airport ground access projects. The overall framework
of the IAPT has been developed to provide generic analysis capabilities that could in principle be
used for any airport ground access planning study as long as the corresponding data and models
are available in the required formats. To use the IAPT for project evaluation it is first necessary
to prepare the required data, depending on whether the project involves a new service at an
airport for which data has already been assembled, or a new airport or region. What is needed to
run the IAPT are:
• A passenger mode choice model for the airport( s) in question
- xx -
• Data on the trip and party characteristics of a sample of air parties ( often the same
sample used to develop the mode choice model)
• Regional highway network travel times and distances
• Regional transit system service data
• Other transportation provider service data, including service locations, fares,
schedules, and travel times.
These form the essential information that needs to be provided when analyzing a new region or
airport, or updated to the same time period in cases in which analysis is being repeated to reflect
more recent values for some of the above data ( for example, more recent air passenger survey
data has become available). Once this has been done, the project evaluation process using the
IAPT follows a standard sequence of steps, comprising airport selection, project definition,
performance measure definition and selection, data input, performing the analyses, and
displaying the results in various ways. The IAPT provides the flexibility to allow the planner to
focus on the quantitative comparison of different alternative projects for the selected airport
without becoming unduly involved in managing the underlying analytical processes.
( 9) Potential Bay Area case studies: Chapter 9 describes five potential projects to improve
intermodal connectivity at the three major Bay Area airports that were defined for future analysis
using the prototype IAPT in order to demonstrate the use of the performance measures identified
in the research and incorporated in the IAPT. The five projects are:
• The Oakland Airport Connector automated people- mover between the Coliseum
Bay Area Rapid Transit ( BART) and Amtrak stations and Oakland International
Airport
• A planned automated people mover at San José International Airport connecting
the airport to nearby stations of the Santa Clara Valley Transportation Authority
light rail system and the Caltrain commuter rail line serving communities between
Santa Clara County and San Francisco
• A proposed ferry service linking a terminal serving San Francisco International
Airport with downtown San Francisco and the East Bay
- xxi -
• An off- airport terminal located in the South Peninsula serving Oakland
International Airport and potentially San Francisco International and San José
International Airports
• An off- airport terminal located to the south of downtown San José providing
service between Santa Clara County and both Oakland International Airport and
San Francisco International Airport.
Due to the resource constraints of the current phase of the research, detailed analysis of these
case studies was deferred to future work.
( 10) Policy recommendations: a Chapter 10 presents a number of recommendations that have
been developed to help guide future efforts to improve airport ground access planning in
California, including both passenger and freight movement, based on the research undertaken in
the project.. These address potential opportunities for improving intermodal connectivity at
California airports, institutional aspects involved in pursuing these opportunities, the need for
additional guidance material and coordination to facilitate funding airport intermodal
connectivity projects, and requirements for technical support to airports and regional
transportation planning agencies to assist them in analyzing the potential ridership and economic
feasibility of proposed projects to improve airport intermodal connectivity.
( 11) Concluding remarks and recommendations for future research: The final chapter presents
some concluding remarks from the research undertaken during the project and presents
recommendations for future research. Several aspects of the research undertaken in the course of
this project require further work and the development of IAPT is only at a preliminary stage.
The software itself needs further development. The transportation behavior modeling needs to
be refined and the convergence of the Generalized Nash Game process needs further study. In
addition, airport employee ground access behavior should be studied so that it can be modeled
and incorporated in the IAPT. The IAPT is currently structured for planning ground access
projects for a specific airport. In practice, passengers choose among available airports in a
region and this choice in influenced by the airport ground access system at each airport. This
airport choice behavior affects both airport passenger traffic levels and the associated ground
access activities. Future enhancement of the IAPT needs to take airport choice into account to
- xxii -
allow the IAPT to be used in regional- level airport ground access planning. Air freight has been
a steadily expanding segment of air transportation and is critical for the U. S. economy in a
globalized market. Further research into the characteristics of the air freight market is needed in
order to determine how to model the factors affecting ground movement of air freight to and
from airports. Including air cargo truck trips in the IAPT would provide important analytical
capabilities for Caltrans.
Chapter 1. Introduction
This research report documents the work done under the project titled A Combined
Quantitative and Qualitative Approach to Planning for Improved Intermodal Connectivity at
California Airports. The project was sponsored by California Department of Transportation and
undertaken by the Partners for Advanced Transit and Highways ( PATH) under task order
TO5406- 6406 during the period 10/ 01/ 2004 to 10/ 31/ 2007.
The objective of the project is to use a combined qualitative and quantitative approach to
analyze the effectiveness of alternative strategies for improving intermodal connectivity at
airports. The qualitative approach involved a case study analysis of a selection of representative
airports to identify and evaluate the potential effectiveness of alternative projects to improve the
connectivity between the airports and the rest of the intermodal transportation system. It is
envisaged that this would be supplemented as part of future work by a more detailed quantitative
analysis of selected case study airports utilizing a mathematical model, termed the Intermodal
Airport Ground Access Planning Tool ( IAPT), which has been developed in this research as the
main product. The IAPT has been designed to provide an analytical environment that integrates
existing data sources and transportation network analysis software with improved models of air
passenger travel choice behavior in order to evaluate the costs and benefits of proposed projects
to improve intermodal connectivity for airport ground access. The goal of developing the IATP
is to ensure a consistent approach to analyzing alternative projects and simplify the complicated
modeling and computational aspects by providing decision makers and planners with a user-friendly
interface to a standard set of analysis modules. Planning guidelines have been
developed on how to use the IAPT for implementation- related project evaluation. Based on the
results of the qualitative case study analysis, policy recommendations have been developed and
reviewed with Caltrans and other stakeholders.
The motivation to improve intermodal connectivity at airports results from growing
pressures to reduce the volume of highway traffic generated by airport access and egress trips
and to facilitate the ability of airport travelers to use high- occupancy modes. Continuing growth
in air travel and air freight is generating increasing volumes of surface traffic traveling to and
from airports, particularly major airports. This traffic arises primarily from air passenger trips,
but airport employees and air cargo movement also contribute significant volumes of traffic at
- 2 -
large airports. These vehicle trips contribute to congestion on the regional highway network and
the local street system in the vicinity of the airport, as well as adversely impact air quality
through increased vehicle emissions. The goal of improved intermodal connectivity is to
encourage greater use of high- occupancy transportation modes for airport trips, particularly rail
modes that do not involve use of the highway system ( other than for access and egress trips to
the rail stations) and in many cases use electrical power, thereby potentially reducing emissions
in the area served by the airport. Improving the connectivity to rail modes leverages the public
investments that have been made in these modes, and to the extent that these modes are operated
below capacity ( as is commonly the case) makes use of excess capacity that would otherwise
remain unused.
The IAPT has been developed as an analytical tool to support airport- level planning of
ground access projects that can enhance intermodal connectivity. Its focus is on the efficient
evaluation of a wide range of project alternatives at a specific airport, rather than strategic
planning at a regional or statewide level, such as that undertaken in the earlier Caltrans Ground
Access to Airport Study ( Landrum & Brown, 2001).
1.1 Scope of this Report
This report documents all the deliverables of the project, including the structure of the
IAPT and the technical details of the various components of the tool. It includes the following
elements:
• An extensive literature review on intermodal transportation planning with
emphasis on the intermodal airport ground access planning
• Identification of opportunities for improving intermodal airport ground
access in California airports
• Development of a modeling framework for analyzing improvements in
airport intermodal connectivity, including the design and development of a
prototype version of the Intermodal Airport Ground Access Planning Tool
( IAPT)
• Systematic consideration of intermodal airport ground access systems
performance measurement, including the definition of measures of systems
performance and connectivity performance
- 3 -
• Development of guidelines for project evaluation using IAPT
• Policy recommendations for improving intermodal connectivity at
California airports based on the findings of the research
• Recommendations for further development of the IAPT and technical
support for intermodal airport ground access planning
• Appendices containing technical details of the mathematical modeling and
data specification and preparation for the IAPT
Each of these elements is described in more detail in the following sections.
Although airport access and egress traffic is generated by air passengers, airport
employees, and air cargo activities, as well as airport support functions and other ancillary
activities that occur on the airport, the version of the IAPT described in this report and the focus
of the research has been on air passenger trips. It is anticipated that future enhancements to the
IAPT would be desirable to address airport employee trips, air passenger airport choice and air
cargo truck trips.
1.2 Role of Modeling in Quantitative Analysis
The objective of quantitative analysis in assessing proposed improvements in airport
ground access systems, and enhancements to intermodal connectivity in particular, is to provide
a basis for estimating the likely usage of proposed facilities or services, the resulting revenues
and costs involved in implementing the proposed improvements, the economic impacts on other
ground access services at the airport, and changes in the environmental impacts of the ground
access system. These estimates are required for planning the details of the proposed
improvements, assessing their feasibility, and developing the necessary environmental impact
documentation that will be required in many cases before a project can proceed. They are also
likely to be of considerable interest to both the airport operator and other ground transportation
providers serving the airport due to the anticipated effect on the economics and operation of the
airport and other ground transportation services.
These assessments are inherently quantitative and will generally require some form of
mathematical modeling. The circumstances at each airport are sufficiently distinct that the
experience at one airport is not readily transferable to another without extensive adjustments to
- 4 -
account for the different situations. Since it is typically not obvious how to determine a priori
what are appropriate adjustments, this is usually addressed by developing a mathematical model
of the system and using this model to predict the effect of changes to the system. Such models
also have the advantage that they can be designed to readily generate a large amount of situation-specific
data that is required to perform related analyses, such as estimating changes in highway
traffic conditions and vehicular emissions for the purpose of air quality analysis.
The central component of these analytical activities is the modeling of airport traveler
mode choice behavior. The ability to predict the changes in the use of the different components
of the airport ground access system in response to any given change in the system obviously
depends on the ability to predict how those traveler choices will change. However, as discussed
in the following section, it is also necessary to be able to model the resulting decision process of
the various transportation providers as they also respond to changes in the system. The nature
and extent of these choices and decisions are not usually self- evident, and an important purpose
of developing formal models of how the system will respond to any given change is to help
decision makers to better understand these complex and interactive factors.
It is therefore important that the modeling activities are not viewed ( or used) as a “ black
box” that produces numerical results in a way that the decision- makers do not or cannot
understand. A situation in which decisions are being made on the basis of the results of a model
that nobody can really explain why it gave the values that it did is not only unsatisfactory for the
decision- makers, since they do not know how much they should trust the results, but prevents
any validity checking of the model itself. This is critically important in any complex situation
such as an airport ground access system, where any analysis is very dependent on a large number
of assumptions that are often deeply buried within the models. It is therefore essential to be able
to understand how changes in the assumptions affect the results. If the results are largely
insensitive to a particular assumption, then decision- makers do not need to worry too much if
that assumption turns out to be incorrect. However, if the results of the analysis turn out to be
highly sensitive to a particular assumption, then those using these results need to satisfy
themselves that the assumption is reasonable and to understand how changes in the assumption
would affect the results.
- 5 -
1.3 Dynamic Interactions in Airport Ground Access Activities
The airport ground access system consists of a large number of different service
providers in competition with each other ( directly or indirectly) to meet the ground access needs
of airport travelers. In turn, those travelers select their ground access travel mode on the basis of
the characteristics of the alternative services available. However, for many of these services,
their characteristics are affected by their utilization. Service frequencies can be increased with
more riders. Fares can be reduced if higher average load factors can be achieved. Shared- ride
door- to- door services involve less circuity picking up passengers in areas of higher trip end
density. Conversely, the more operators that are attempting to serve the same market, the less
traffic each will have and the harder it will be to achieve economies of density. Similarly the
more airport travelers who decide to drive a private vehicle to the airport, the more congested the
approach roads and terminal curb- front will become.
Therefore introducing a new or improved service will not only change the use of the
other ground access services, but will result in changes in their service characteristics. Some of
these changes will occur naturally due to the change in utilization while others will represent
decisions by the operators to respond to the changed situation. Thus in order to properly assess
the effect of a change in any service, such as an improvement in intermodal connectivity, it is
necessary to account for these dynamic feedback effects and resulting decisions by the other
operators. This requires not just a way to model how airport travelers choose their access mode
in the light of a given set of service characteristics, but also how the transportation providers will
modify their service characteristics in the light of changes in airport traveler mode choices.
For the purposes of the IAPT, the critical transportation provider behaviors that need to
be modeled are decisions regarding changes in service attributes that affect the modeling of air
passenger mode choice. This is represented in the diagram shown in Figure 1- 1 on the following
page.
The approach being taken to modeling the feedback process shown in Figure 1- 1 forms
the central focus of this report. Subsequent chapters discuss the overall modeling framework of
the IAPT, the details of the mode choice model, and the approach proposed for modeling
transportation provider behavior.
- 6 -
Transportation
Provider
Behavior
Mode Choice
Model
Fares
Frequencies
Routes/ Service Area
Traffic by Mode/ Service
Figure 1- 1: Feedback Between Transportation Provider and Airport Traveler Behavior
1.4 Capabilities and Limitations of Modeling
In spite of the essential role of formal modeling in the quantitative assessment of
proposed improvements to airport ground access systems, or indeed any transportation system, it
is important to also appreciate the capabilities and limitations of particular modeling approaches.
In general, the more disaggregate the modeling approach, the more detailed the results can be.
For example, predicting airport traveler mode choice decisions at the level of trips from
individual analysis zones allows the analysis to consider resulting changes in highway traffic at
the level of individual links of the regional highway network. In fact, since airport traveler mode
choice decisions are influenced by individual air party or airport employee characteristics as well
as the service characteristics of the different ground access modes, which necessarily differ for
different trip end locations in the region, any meaningful analysis needs to be undertaken at the
level of individual travel parties using a fairly disaggregate zone system.
The other level of detail that is germane to the results of airport ground access analysis is
the extent to which the different ground transportation providers and services are explicitly
identified in the analysis. For example, does the mode choice analysis distinguish between the
different off- airport parking lots, or even between on- airport and off- airport parking? The level
of aggregation at which the different transportation services are identified affects the type of
question that the analysis can address, as well as how the modal service levels are expressed.
While it may not matter from the perspective of the ridership on an improved intermodal
connection which parking lot is used by those air parties that drive to the airport and park, it most
certainly matters to the parking lot operators.
- 7 -
Therefore the complexity and structure of the mode choice model needs to reflect the
questions that the analysis is designed to address. Since these questions may not be fully known
at the time the model development is commenced, there is an understandable ( and justifiable)
tendency to develop mode choice models that are as detailed as the underlying data can support.
However, this brings up an important constraint on the modeling process. Model development
requires data on which they can be estimated. In the case of air passenger mode choice models,
this includes the results of air passenger surveys that identify the ground access modes used by
the travelers. If the survey questions do not identify the ground access choices at a sufficient
level of detail ( for example failing to ask which parking lot was used), it will be much more
difficult to develop a mode choice model that can predict those choices at the level of specific
services or facilities. The development of the IAPT as part the current research has been based
on an air passenger survey undertaken by the San Francisco Bay Area Metropolitan
Transportation Commission ( MTC) in 2001 and 2002.
Another consideration that arises with airport ground access mode choice models is how
to represent new services or modes that do not currently exist at the airport in question. It is
obviously not be possible to include these services or modes in mode choice models that are
estimated directly from existing data for that airport. Where similar services exist at the airport,
it may be possible to modify the model after it has been estimated to incorporate the new service
based on the representation of the existing services in the model. However, where a proposed
mode does not exist at all at the airport in question, determining how to modify the model to
incorporate the new mode is much more challenging. This issue is discussed further later in this
report.
A different type of limitation that can arise in airport ground access analysis results from
the level of temporal resolution of the model. A model that is estimated on the basis of travel
conditions on an average day of the year will be unlikely to do a very good job of predicting the
difference in travel patterns between those at 5 pm on a Friday afternoon and those at 10 am on a
Sunday morning, or between a given weekday in March and the same day in August. An
analysis framework that is required to generate results that distinguish between different times of
day and days of the week, or seasonal effects, will be significantly more complex and costly to
develop than one that simply predicts the average use of different modes throughout the year.
- 8 -
1.5 Structure of this Document
The remainder of this document consists of ten more chapters and six appendices.
Chapter 2 documents the literature review undertaken as part of the research, while Chapter 3
presents some of the opportunities for improving intermodal airport access in California.
Chapter 4 describes the design and implementation of the IAPT, including all the functional
components, the software structure and data flow, and the graphical user interface. The
following two chapters present the development the two key analysis components of the tool.
Chapter 5 describes the air passenger mode choice model development for the IAPT, while
Chapter 6 addresses the transportation provider behavior modeling component. These chapters
describe the process being followed to develop the model components, review the relevant
literature on modeling approaches, and present the results of the model development work.
However, the details of the mathematical modeling have been put in the appendices. Chapter 7
discusses a number of issues involved in measuring airport intermodal connectivity, including
how system performance measures and intermodal connectivity performance measures have been
addressed in more general public transit systems, the development of appropriate measures of
airport intermodal connectivity, and ways to identify weaknesses in intermodal connectivity and
capacity constraints in airport ground transportation systems. The mathematical definitions of
the proposed measures of performance and their calculation formulae have been put in an
appendix.
Chapter 8 provides guidelines for using the IAPT for intermodal airport ground access
project evaluation, using a tutorial approach. Chapter 9 discusses how five potential Bay Area
intermodal airport access projects identified in Chapter 3 could be evaluated using the IAPT,
including project definition, data preparation and selection of alternative analysis scenarios.
Chapter 10 presents a number of policy recommendations for enhancing intermodal connectivity
at California airports that were developed in the course of the research. These recommendations
should be reviewed and refined in the light of the results of the analysis of the Bay Area case
study projects that has been deferred to a later phase of the research. Finally, Chapter 11 offers
concluding remarks and recommendations for the future development of the IAPT.
Appendix A documents the technical details of a number of representative airport access
mode choice models that have been developed in recent years. Appendix B provides the detailed
mathematical derivation of the transportation provider behavior modeling described in Chapter 6.
- 9 -
Appendix C documents the details of the performance measure calculations discussed in
Chapter 7. Appendix D documents the structure of the data tables that form the basis of the initial
IAPT implementation, while Appendix E provides additional details of the technical aspects of
the IAPT. Appendix F presents a number of sample data files used in the development and
testing of the IAPT.
- 10 -
Chapter 2. Literature Review on Intermodal Airport Ground
Access
This chapter presents a review of recent literature relevant to intermodal airport ground
access planning. The objective of this project is to improve intermodal connectivity in airport
ground access using a combined qualitative and quantitative approach. This involves a number
of different aspects of airport ground access planning. On the qualitative side, it involves
institutional issues, political relationships and coordination between planners in different types of
organizations, as well as the role of air passenger information systems in air traveler mode choice
behavior and understanding and accounting for the relevant decision making behavior of ground
transportation providers. On the quantitative side, it involves how to measure the performance of
an intermodal transportation system so that those performance measures can be used to guide
decision- making at different planning levels. This in turn requires the ability to model the air
passenger mode choice and transportation provider decision making behavior, their interactions
and the resulting effects on the number of vehicle trips generated by the airport, as well as the
impact of these trips on traffic conditions on the street and highway network and air pollution.
The review has examined recent literature addressing these aspects of airport ground access
planning as well as some principles, viewpoints, analysis methods, and recommendations from
the literature on general intermodal transportation that are relevant to the particular case of
airport ground access travel. The review also gives particular attention to the distinction between
qualitative and quantitative analysis approaches, and includes the application to airport ground
access issues of relevant modeling and analysis methods from general urban intermodal
transportation.
2.1 Airport Ground Access Planning
There is an extensive literature on the many different aspects of airport ground access
planning, particularly the planning and design of specific airport ground access facilities. This
section summarizes some of the more recent key documents and studies that are particularly
relevant to intermodal aspects of airport ground access planning. All of these documents and
reports contain extensive bibliographies, from which the interested reader can obtain more
detailed information.
- 11 -
In 1994 the Federal Aviation Administration ( FAA) sponsored two workshops on ground
access to airports that were organized by the Institute of Transportation Studies at the University
of California at Berkeley. These examined the role of off- airport terminals and institutional and
funding issues in developing improved airport ground access services and systems ( Gosling,
1994). Subsequently, a contract was let by the FAA in association with the Federal Highway
Administration ( FHWA) to develop a planning guide for intermodal access to airports ( Shapiro,
et al., 1996). The importance of viewing the airport ground access system as an intermodal
interface and the role of such airport ground access systems as rail links and off- airport terminals
was further developed in a paper by Gosling ( 1997).
The growing interest in improving public transportation access to large airports, and in
particular proposals to develop very expensive rail links at an increasing number of airports,
began to become of concern to the FAA and other Federal transportation agencies. Together
with the Federal Transit Administration ( FTA) and the Federal Highway Administration, in 1998
the FAA requested the Transit Cooperative Research Program to undertake a comprehensive
study of strategies for improving public transport access to large airports ( Leigh Fisher
Associates, 2000, 2002).
At about the same time, as part of the growing interest in developing intermodal strategies
to address airport ground access, the Texas Department of Transportation sponsored an extensive
study on the topic that undertook a comprehensive review of the literature, identified best practices
and developed case studies, and performed an assessment of alternative strategies ( Mahmassani et
al., 2000, 2001, 2002a, 2002b). While this study was primarily interested in methodology, another
comprehensive study in California ( Landrum & Brown, 2001) assembled information on the
ground access conditions and needs at a large number of airports in the state, and examined the
roles and responsibilities of different agencies. The findings of the latter study are discussed in
more detail in the following chapter.
2.2 Intermodal Transportation Planning Principles
Since airport ground transportation can be considered as a particular subset of the more
general intermodal ground transportation system, airport ground access planning should be
guided by generally accepted principles for intermodal transportation planning. The National
- 12 -
Center for Intermodal Transportation ( NCIT) has proposed the following four principles for the
development of the intermodal transportation system ( NCIT, 2001):
Connection: All modes should be well connected with one another to accomplish the
convenient, expeditious, and efficient movement of commodities and people. Connecting points
should be conveniently located and connections timed to facilitate movements from one mode to
another.
Choices: The intermodal network should offer choices, allowing its users to select the
mode that can most efficiently satisfy their transportation needs.
Coordination: The transportation infrastructure should be planned, designed, and built
in a way that brings the modal networks sufficiently close together so that connections can be
made relatively effortlessly. In addition, transportation providers must coordinate their
schedules to reduce dwell time between intermodal movements.
Cooperation: There should be cooperation and collaboration among transportation
providers and governmental agencies at the federal, state, and local levels to ensure that the
needs of the users for seamless service are realized.
One definition of good intermodal connectivity is as follows: Advanced and attractive
systems that operate reliably, and relatively rapidly, and form part of the passenger and freight
door- to door chain with smooth and synchronized transfers.
However, in order to apply these principles effectively, it is necessary to understand the
similarities and differences between planning for intermodal urban transportation in general and
airport ground access in particular.
Similarities:
• As an example of intermodal transportation, the principles guiding airport ground
access planning are similar to those for intermodal transportation in general;
• Basic requirements for facilities and service follow those identified by Homburger et al.
( 1996): adequacy to handle expected demand, compatibility with existing master plans,
environmental compatibility, acceptability to decision makers and the public, and
financial feasibility;
- 13 -
• Planning processes in both cases involve institutional issues, political relationships,
identification of needs for enhanced facilities and services, and development of
recommended changes to policy guidelines;
• Many of the factors influencing passenger mode choice decisions are similar;
• Public and private transportation providers operate in a similar way in serving airport
ground access and general urban transportation trips.
Differences:
• Compared to general urban travel, airport ground access travel typically involves many
more distinct modes and services;
• Airport ground access and egress trips by air passengers involve considerations not
typically addressed in general urban travel, such as the need to carry luggage, round
trips involving travel duration of many days, and a significant proportion of trips by
visitors to the region;
• Many airport employees have shift patterns involving travel outside the usual commute
times and the regular work week;
• Travel purposes for airport ground access trips are limited to airport related activities,
which are not as diversified as those involved in general urban transportation;
• The total demand for airport ground access travel can be estimated from the air
passenger traffic level at the airport and airport employee counts;
• Airport authorities typically maintain information on available airport ground access
services at their airport( s), providing a current and consistent source of information;
• Airport access and egress trips involve travel to or from a single location, which
reduces the complexity of travel patterns compared to general urban intermodal
transportation;
• Transportation providers are generally subject to airport regulation which makes their
behavior more predictable and can provide a source of statistics on operational traffic
- 14 -
and activity levels, which is not always the case for more general urban intermodal
transportation;
• Surveys of air passenger and airport employee travel patterns can be performed
relatively easily at the airport, since these trips involve a common location.
These differences make some aspects of the modeling and analysis of airport ground access more
challenging than that for general urban intermodal transportation while making other aspects
easier.
2.3 Quantitative and Qualitative Approaches in Airport Planning
To develop a combined quantitative and qualitative approach for intermodal airport
ground access planning, it is necessary to look at quantitative, qualitative and combined
approaches used for planning in previous work. This requires a way to distinguish between
quantitative and qualitative approaches. For the purpose of this discussion, a quantitative
approach is considered to be one that involves modeling for analysis regardless of how simple or
how complicated the model may be. Most previous work used either a qualitative or a
quantitative approach. Few studies tried to combine them for intermodal transportation planning.
Cunningham and Gerlach ( 1998) discuss the use of decision support systems for airport
ground access planning using both quantitative and qualitative approaches. The approach used in
this work included: ( 1) a literature review to obtain background information concerning the airport
ground access problem and analysis of various proposed ground access solutions; ( 2) telephone
interviews with airport and regional transportation officials to clarify issues and identify key
transportation officials familiar with airport ground access planning; and ( 3) focus group meetings
with airport ground transportation managers, local metropolitan planning organization ( MPO) staff
directly involved in airport ground transportation planning, and relevant staff from local transit
authorities at a selected number of case study locations. Participants were encouraged to provide
their opinions as well as factual information regarding the planning process and the extent to which
decision makers relied on quantitative models and qualitative information to reach a decision.
However, this study does not discuss how the quantitative approach was conducted in the
locations examined. Instead, the main part of the study discusses some practical problems
encountered in ( or controversial attitudes towards) the use of quantitative analysis in intermodal
- 15 -
airport ground access planning. The main findings are the following: ( 1) on the one hand, decision
makers need a decision support system to provide numerical results as references for decision
making; ( 2) on the other hand, using quantitative modeling for strategic decision support is very
difficult. This difficulty arises because ( a) modelers are not confident about the accuracy of their
models and transportation officials believe that the information supplied is flawed by a number of
defects that minimize the value for the decision maker, which in turn leads to the situation that
decision makers lose confidence in the quantitative method; and ( b) modeling is generally believed
to be very costly and difficult – human behavior is not sufficiently understood to accurately predict
how travelers make individual transportation decisions. To avoid these difficulties, the authors
propose the following solution: ( i) improve quantitative modeling such that the model can actually
reflect passenger mode choice behavior; ( ii) use a combined quantitative and qualitative approach
for decision making, where the qualitative approach involves the use of such techniques as
community or airport user focus groups to identify attitudes toward airport ground access issues
and likely use of proposed new services, the use of expert opinion to supplement analytical
modeling, comparative analysis with airport ground access systems in other regions, and
consideration of the potential implications of longer- term visions for land use development in the
areas around the airport or the evolution of the regional transportation system.
They point out that planning, designing and building a transportation system involves
multiple constituencies, as well as multiple decision variables and criteria. This suggests that
decision- making at different levels of government needs to consider the interests of the different
constituencies at the local, regional, and state level.
Cunningham and Gerlach suggest that decision makers, particularly higher level officials,
tend to rely on a qualitative analysis based on a subjective assessment that draws on their
background, beliefs, and experience. They might never make use of model results or cost- benefit
numbers generated from models, in part because such models do not generate the type of
information that they need. In practice, decision makers often base their strategic planning on their
“ vision” of how they think the transportation system should evolve based on their intuition and
experience.
Reliance on a vision of how an organization, community or region should evolve is also a
widely used tool for decision making in the business and political community. Quantitative
information is then often developed to support that vision. In this case, the main information used
- 16 -
by the decision makers in practice is derived from the vision with only minor information from
quantitative analysis. In consequence, decision makers often choose to rely on models that are
consistent with their visions. Where the results of analysis are in conflict with their vision,
decision makers often choose to base their decisions on their vision rather than the analysis. This
can often arise from a large gap between the quality of the analysis tools and how the decision
makers perceive those tools.
The authors also identify a number of concerns and limitations with existing approaches to
quantitative analysis:
( a) Small amounts of data often only allow models to provide a general
representation of complex phenomena, such as the use of average daily traffic
levels;
( b) Models are often too sensitive to key inputs and too easily manipulated;
( c) Models sometimes do not predict what is really happening;
( d) There is often difficulty modeling the effect of new modes or services using
models that have been calibrated on data for the existing pattern of services.
The authors propose a number of ways to remedy these modeling limitations in order to
win the confidence of decision makers:
( a) Involve the transportation agency decision makers in the modeling process by
organizing a committee to oversee the design and use of transportation models;
( b) Ensure that the modelers make clear to decision makers the following aspects:
• What assumptions have been made
• What data are to be used and why
• The methodology to be used, which should be documented in a form that decision
makers can understand;
- 17 -
( c) Encourage the decision makers to use the results of model analysis as prudently
and conservatively as possible.
Ceder ( 2004) discusses the major elements and challenges surrounding the introduction
of new or improved public transportation ( PT) systems or services. The choice between public
and private transport is an individual decision that is influenced by government and community
decisions. These decisions often send mixed signals to the public transport passengers and
potential users while failing to recognize system- wide considerations and integration
implications. This paper attempts to summarize the current state of PT practice and to cover
issues affecting the use of PT including the willingness of users to pay for improved service,
assessment and projection of economic viability, the effectiveness of new initiatives mostly in
Europe and North America, and strategies to achieve multi- modal service integration.
Ceder discusses the use of a qualitative analysis approach to address factors which affect
the quality of service offered by the intermodal system but are difficult to quantify. These
factors could include the introduction of the following intelligent transportation systems ( ITS)
technologies and other measures to improve user comfort and convenience:
• Automatic vehicle monitoring ( AVM)
• Signal priority for public transportation vehicles
• Traveler information systems
• Stability of perception of service
• Ticketing integration
• Improved terminal, interchange and park and ride facilities
• Coordination between different modes to reduce total travel time
• Increased passenger comfort
• Introduction of different modes to increase system capability.
2.4 Policy and Institutional Issues
Lacombe ( 1994) suggests that inadequate ground access facilities may limit airport
capacity. This paper examines the requirements in the Clear Air Act Amendments ( CAAA) and
- 18 -
Intermodal Surface Transportation Efficiency Act of 1991 ( ISTEA) as they affect airport ground
access planning. The paper examines the effect of institutional constraints and funding limitations
that hinder intermodal approaches to improving airport ground access, and points out the necessity
and opportunity for cooperation between airport authorities and urban transportation planners.
Yevdokimov ( 2000) examines the use of cost benefit analysis and related techniques to
analyze the effect of transportation investment on economic growth. Microeconomic and
macroeconomic simulations are used to support the benefit measurement.
2.5 Mode Choice Modeling and Analysis
To date the quantitative approach to air passenger ground access model choice analysis
has almost exclusively used a form of logit model ( multinomial logit or nested logit). Past model
development efforts have been summarized in a recent review of the relevant literature by
Gosling et al. ( 2003). A number of subsequent studies and alternative approaches are discussed
below.
Tam and Lam ( 2005) studied the mode choice pattern for ground access travel to Hong
Kong International Airport using a survey of air passengers. Their results show that due to very
low car ownership and relatively short travel distances to and from the airport, access is mainly by
public transport such as bus and train or light- rail. Passengers are divided according to arrival,
departure and transit/ transfer. The authors conclude that business travelers are less concerned with
the cost of service than with travel time and convenience. This paper is primarily focused on the
design of the survey and explanation of the results rather than their use for model development.
Arentze and Timmermans ( 2005) discuss the application of formal decision rules, such as
parametric action decision trees, to explain travelers’ mode choices. According to the authors,
using discrete choice models ( such as the logit model) could limit the sensitivity of the model to
travel time and travel cost. This paper uses a hybrid model to reduce such defects. It is claimed
that the hybrid model can reproduce realistic price elasticities of travel demand. The authors assert
that decision trees have the advantage of being consistent, exclusive, and complete compared to
other methods for formal representation of decision making, such as belief networks, association
rules and production systems. However, this paper does not specifically address airport ground
access travel.
- 19 -
Two recent papers by Outwater et al. ( 2003, 2004) describe a market segmentation
modeling approach to predicting the effect on mode choice of introducing a new mode, in this
particular case the introduction of ferry service in markets not currently served by ferries. Two
types of models were considered: multinomial logit and nested logit. The authors found that the
later did not give any improvement and thus based their analysis on the former. Stated- preference
survey data were used to calibrate the model. The calibrated model was then used to analyze three
future year alternatives and to test sensitivities to pricing, service changes and alternative modes.
According the author, previous mode choice modeling work has tended to emphasize the following
factors: trip purpose, geographical location, and travel time. However, the focus of the current
paper was to extend this mode choice modeling approach to reflect the effect of passengers’
attitudes toward improvement in ferry service and apply this to the forecasting ferry ridership in the
San Francisco Bay Area. Six attitudinal factors were identified: desire to help the environment,
desire for time saving, need for flexibility, sensitivity to travel stress, insensitivity to transport cost,
and sensitivity to personal travel experience. Three of these were used to partition the potential
ferry- riding market into eight segments and develop demand estimates for each segment.
Lo, Yip and Wan ( 2004) incorporated the competitive behavior of transit services in an
intermodal planning model using a nested logit approach. However, the competitive behavior
between transit providers was considered in a static manner rather than a dynamic interaction
between the transportation provider decisions and the passenger travel choices. The effect of the
transportation provider behavior on passenger mode choice was reflected through the relationship
between fare changes and ridership. Using their model, the authors studied the effect of fare
changes on overall network congestion. A case study of travel between Hong Kong International
Airport and the Downtown Area was used to illustrate the method.
2.6 Airport Ground Access Travel Information
A key aspect of air passenger choice of travel mode for airport trips is the information
available to them about travel options. It is self- evident that travelers will not use transportation
options that they are not aware of, but an equally important consideration is whether they can
readily obtain the necessary information to decide whether to use a particular service. In the
absence of accurate information, their perceptions of travel times or costs may be sufficiently
- 20 -
biased to cause them to reject options that in fact might work very well for them. In spite of the
importance of this issue, it has received relatively little attention in the literature.
In the early 1990s the California Department of Transportation ( Caltrans) funded a
research project to examine how advanced technology might be used to improve information
available to air passengers to help their airport ground access decisions ( Du & Gosling, 1994).
Subsequently, Caltrans funded a demonstration project at several airports in the state in which
automated ground transportation information kiosks were installed in the airport terminals.
These kiosks used a touch- screen display to provide information on alternative travel options and
contained a database for all the airports in the demonstration program. Thus air passengers
waiting for their flight at one airport could obtain information about ground transportation
options at their destination airport. As part of the demonstration program, a series of surveys
were conducted of air traveler and airport user information needs and the effectiveness of the
kiosks at meeting those needs ( Gosling & Lau, 1995). The survey results found that kiosk users
generally found the information provided by the kiosks helpful and that they liked being able to
obtain information about their destination airport in advance to arriving there.
A similar survey was undertaken a few years later at George Bush Intercontinental
Airport in Houston ( Burdette & Hickman, 2001). The latter survey only addressed the needs of
departing air passengers and included information related to the flight ( such as gate information
and flight delays) as well as ground access information. It focused on traditional highway travel
information issues, such as traffic delays and road conditions, rather than the type of information
needed to make an informed access mode choice.
More recently, Lo and Szeto ( 2004) studied how to model traveler response to advanced
travel information systems using both static and dynamic paradigms. Although not directly
applied to air passenger travel decisions, their approach may offer some insights as to how to
better understand the role of travel information systems in airport ground access travel decisions.
2.7 The Government Accountability Office Study
The United States ( U. S.) Government Accountability Office ( GAO) published the results
of a major study on potential strategies that would redefine the Federal role in developing airport
intermodal transportation capabilities ( GAO, 2005). This report explored the possibility of
integrating passenger air transportation with intercity passenger rail transportation in the U. S.,
- 21 -
based on the analogous experience in Europe. The “ intermodal” transportation that is
emphasized here is not the local transit access to and from the airport that our project is
addressing, but rather the possibility of Amtrak intercity rail linkages for air travelers. In the
course of the study, however, this report provided useful background information about both
current and planned local transit intermodal linkages to airports in the U. S.
Major airports in Europe ( Frankfurt, Paris, Brussels, Amsterdam) are increasingly well
integrated with the European high- speed intercity rail network, with rail stations built adjacent to
or beneath the airport terminals. This has made it possible for airlines to offer code- share
arrangements with the railroads for passengers traveling to and from smaller nearby cities, and
has led to the reduction of short- haul flights at these airports. National governments have
encouraged these trends by providing financing for the construction of the new rail lines and
stations at airports.
The GAO report notes that the European experience is not readily transferable to the U. S.
for a variety of reasons:
• The Amtrak passenger rail network is not nearly as well developed nor
heavily used as its European counterparts. It does not provide the breadth
or frequency of service to make it an attractive alternative for passengers
or a code- share partner for airlines ( which would require a service
frequency of at least one train per hour).
• The trip ends for travelers to and from U. S. airports are not nearly as
focused on the urban core locations that could be served effectively by rail
as in Europe.
• U. S. airports are disinclined to encourage new access modes that could
lead to a reduction in on- airport parking, which is an important revenue
source for them.
• Space and cost constraints make it difficult to build large new facilities at
major airports in the U. S.
• Cars remain more convenient and economical for airport access than other
modes in the U. S., in contrast to the situation in Europe.
- 22 -
The report suggests a couple of potential policy alternatives to the federal government:
( 1) providing more flexibility and alternative funding concepts to enable
state and local agencies to take a more system- wide approach to
providing intermodal access to airports, without any more direct
federal role;
( 2) increasing the federal role in planning and funding to proactively
promote integration of air transportation with intercity rail and bus
services. This latter strategy was dismissed because of its expected
high costs relative to its benefits, especially based on expected low
levels of demand in most places.
The report includes much useful background information on the current state of ground
access to airports in the U. S. and the federal programs that could fund airport access projects.
This was based on a survey of 72 airports ( including the 68 largest ones, all large and medium
hubs, accounting for 90% of U. S. enplanements in calendar year 2003) and case studies of
16 airports ( including Los Angeles International ( LAX), San Francisco International ( SFO),
Oakland International ( OAK) and Mineta San José International ( SJC)). These case studies each
include a table summarizing the local officials’ assessments of the primary benefits and barriers
to intermodal access facilities at their airports, up to one page of text describing their existing
intermodal access facilities and identifying the key local stakeholder organizations and their
roles, and a one- page schematic diagram showing the locations of the access points to the
intermodal facilities relative to the airport terminal and parking lots.
Of the 72 airports that were surveyed:
• 64 had access by local buses
• 27 had access by local rail transit ( all but one of which also had local bus
access)
o 13 of these could be accessed by automated people movers or
walking
o 22 of these could be accessed by shuttle buses
- 23 -
• 19 were connected to nationwide intercity bus or rail services
o 13 were connected to Amtrak ( only Newark had a direct people
mover)
o 12 were connected to intercity bus services.
The California airports that were identified as having local rail access included Burbank,
LAX, OAK, SJC, and SFO. Among the airports that do not currently have local rail transit
access, there are plans for adding rail transit access at ten: Cincinnati, Denver, Houston
Intercontinental and Hobby, Jacksonville, Memphis, Phoenix, Salt Lake City, Seattle- Tacoma,
and Tampa.
The Newark Airport example was particularly interesting because of the direct access to
Amtrak’s highest- density Northeast Corridor services. This led to the creation of some code
sharing arrangements with Continental Airlines, some reduction of short- haul flights to and from
Philadelphia, and significant usage of the Amtrak station at the airport by travelers from
Philadelphia and Washington DC. The costs of the people movers used for airport connections
were cited for Newark’s low- speed, low- capacity, short- distance link ($ 357 million) and JFK’s
faster, higher- capacity and somewhat longer link ($ 1326 million).
Both federal and state/ local funding sources that have been used to pay for intermodal
access projects are identified in the report:
Federal
• FTA New Starts program for major fixed- guideway systems [ competition
at national level to get on the approved list of New Starts]
• FHWA Surface Transportation Program ( STP) and Congestion Mitigation
and Air Quality ( CMAQ) Program [ competition at state and local levels to
get allocations from these formula grant programs]
• FAA Airport Improvement Program, for projects at airports with
commercial air service and at least 10,000 annual enplanements
• Specific Congressional earmark projects
- 24 -
• Transportation Infrastructure Finance and Innovation Act ( TIFIA) credit
assistance for development of revenue- producing facilities that will be
able to repay the TIFIA loans
State and local
• Allocations from Highway Trust Fund
• Local tax revenues, including regional sales taxes allocated for
transportation improvements
• Revenues from toll facilities ( Port Authority of New York and New
Jersey)
• Local transportation improvement districts making special assessments
• State credit assistance programs analogous to TIFIA
• Passenger Facility Charges ( PFCs) for projects on and owned by the
airport, subject to FAA approval
• General airport revenues
• General airport revenue bonds ( only for on- airport facilities)
- 25 -
Chapter 3. Opportunities for Improved Intermodal Connectivity at
California Airports
Research in the first year of this project identified opportunities for improving intermodal
connectivity at California airports and performed a preliminary analysis of a sample of
representative projects at selected airports. The results of this analysis were documented in the
working paper Opportunities for Improved Intermodal Connectivity at California Airports ( Lu,
Gosling & Xiong, 2005). This section summarizes the findings of the case studies presented in
the working paper. These examined a range of strategies to improve intermodal connectivity at
airports, including the provision of direct rail service to the airport, the creation of improved
links to nearby rail stations, and the development of express bus services to off- airport terminals
or regional intermodal terminals. In order to better understand the factors affecting the
feasibility and likely contribution of these alternative strategies, a series of more detailed case
studies of potential opportunities for enhancing intermodal connectivity at airports in the San
Francisco Bay Area was undertaken. More detailed analysis of the Bay Area case studies is
presented in Chapter 9. This analysis examines the likely ridership levels and economic
feasibility of the different strategies, and provides a quantitative basis for considering the effect
of airport traffic levels and other factors that are likely to influence the viability of potential
projects.
3.1 California Airport Ground Access Needs
In 2001 the California Department of Transportation ( Caltrans) released the findings of
an extensive study of airport ground access issues in the state that had been undertaken by a
consulting team led by Landrum & Brown ( 2001). The study report consisted of an Executive
Summary and three working papers. A major focus of the work by the Landrum and Brown
Team was to identify airport ground access needs and specific problems at a wide range of
California airports. This information was then used to develop recommended policies and
guidelines to address these problems and needs. These policies and guidelines provide a
strategic or high- level approach but do not get into the details of how this could be accomplished
in practice. For example, they suggest that improved coordination between airport authorities
and ground transportation agencies is needed but do not address how this coordination could be
- 26 -
facilitated. The report recommended that the next step for improving the various planning,
programming and implementation processes for California airport ground access is for Caltrans
to develop a specific improvement plan.
3.1.1 Airport- Identified Issues and Problems
In order to identify the ground access issues and problems at each airport, the study
undertook a survey of airport managers throughout the state. The predominant issues identified
by the survey respondents are summarized as follows:
• Large and medium sized commercial airports are primarily concerned with
major regional mobility issues.
• Small and non- hub commercial airports tend to have more localized problems
associated with roadway geometry and immediate terminal area requirements
of curbside and parking.
• Issues and needs at general aviation and business airports are also more
localized in nature, and are generally related to parking, roadway geometry
and roadway conditions.
• Cargo airports are often served by an infrastructure of local roads that are
inadequately constructed to meet the truck traffic demands generated by the
airports.
3.1.2 Performance Based Needs
The study asked each airport to identify the most significant ground access system
deficiencies that they faced and to assess how severe the inadequacy was currently and was
expected to be in the future. The specific deficiencies identified by each airport were grouped into
the following categories:
􀂾 adequacy of alternative modes
􀂾 auto access
􀂾 curbside
􀂾 goods movement
􀂾 airport parking.
- 27 -
The study findings presented only those airport needs for which the existing or future
conditions were deemed to be moderately to severely inadequate. The specific deficiencies at each
airport were listed in the report and then summarized for each airport in terms of the occurrence of
deficiencies in each of the five aspects of the airport ground access system described above.
Table 3- 1 presents these summary findings for the large and medium hub commercial
service airports in California, since it is these airports where intermodal connectivity issues are
likely to be most relevant.
Table 3- 1: Most Significant Ground Access System Deficiencies at Major California
Airports
Adequacy of
alternatives
Auto
access
Terminal/
curbside
Airport
parking
Goods
movement
Airport
2000 2020 2000 2020 2000 2020 2000 2020 2000 2020
Burbank ( a) 􀁑 ( a)
John Wayne ( a) 􀁑 􀁑 􀁑 􀁑 􀁑 ( a)
Los Angeles 􀁑 ( a) 􀁑 􀁑 􀁑 􀁑 􀁑 􀁑 ( a)
Oakland 􀁑 ( a) 􀁑 􀁑 􀁑 􀁑 􀁑 ( a)
Ontario ( a) 􀁑 􀁑 􀁑 􀁑 􀁑 ( a)
Sacramento ( a) 􀁑 􀁑 ( a)
San Diego ( a) 􀁑 􀁑 􀁑 ( a)
San Francisco ( a) 􀁑 􀁑 􀁑 􀁑 􀁑 ( a)
San Jose ( a) 􀁑 􀁑 􀁑 ( a)
Source: California Ground Access to Airports Study, Working Paper Two: Issues and Problems ( Landrum
& Brown, 2001), Table 2- 4.
Note: ( a) Deficiencies not assessed for 2020.
- 28 -
3.1.3 Availability and Use of Public Transportation Modes
Working Paper Two also discussed the tradeoff between the cost of travel and the level of
service offered by different modes, as illustrated in Figure 3- 1 and percent market share of
different public transportation modes at major California airports shown in Table 3- 2.
Appropriately reflecting the tradeoff between cost and level of service provided to the users by
different modes is obviously critical to accurately modeling of air passenger mode choice and
transportation provider behavior in the current project. The data shown in Table 3- 2 demonstrate
the relatively low market share that has historically been achieved by conventional transit
services at California airports, although it should be noted that since the table was prepared the
Bay Area Rapid Transit ( BART) system has been extended to San Francisco International
Airport and use of that service has significantly increased. In addition, use of BART for trips to
and from Oakland International Airport has also increased since the data shown in Table 3- 2 was
assembled by the Transit Cooperative Research Program study from which the data was
obtained.
Source: California Ground Access to Airports Study, Working Paper Two: Issues and
Problems ( Landrum & Brown, 2001). Original source: Transit Cooperative
Research Program, Report 62, Improving Public Transportation Access to
Large Airports, Transportation Research Board, 2000.
Figure 3- 1: Comparison of Ground Transportation Options
- 29 -
Table 3- 2: Transit Percent Market Share by Mode
Source: California Ground Access to Airports Study, Working Paper Two: Issues and
Problems ( Landrum & Brown, 2001)
3.1.4 Recommendations
The Landrum and Brown study also made recommendations as how to improve the
planning, programming and implementation of airport ground access at a strategic level. In
particular, recommended criteria were provided for the selection of ground access projects at
different types of airport: commercial ( large, medium and small), general aviation, cargo and
military. They can be summarized as the following five points:
􀂾 Choose cost effective projects;
􀂾 Maintain or improve passenger/ cargo ground accessibility to airports including
road quality and signage and minimize delays at curbside, including providing
adequate curbside space;
􀂾 Maintain or improve passenger accessibility to local, regional, intra- state, or
international air service;
􀂾 Mitigate neighborhood, local, and regional highway traffic by maximizing the
use of the transit network to decrease vehicle miles of travel and reduce traffic;
􀂾 Promote safety.
These criteria can be considered as addressing four different concerns: cost effectiveness,
accessibility, environmental impacts, and safety. The recommendations suggested that the project
- 30 -
selection should be performance- based and that the performance should be quantifiable, although
they did not provide any specific guidance on how to achieve this.
The recommendations of the Landrum & Brown study can be viewed as based on the
principle that, since each airport is neither isolated from the larger concerns of society nor from the
surface transportation network, it should be viewed as an integral part of the overall transportation
system. Thus decisions by each party involved in airport ground access issues affect others
directly or indirectly. This requires the decision making in the planning process to consider the
problem as a whole and will involve different levels of government. Thus there needs to be
effective coordination among decision makers in government agencies and other organizations to
address airport ground access planning issues at different levels: airport, regional, state, and
federal. There should also be coordination between decision makers involved in planning the
aviation system and those involved in planning the ground transportation system.
3.1.1 Need for Planning Guidelines
According to the Landrum & Brown study, many airport managers are frustrated by the
lack of guidance from local, regional, state and federal agencies to help them implement ground
access projects. Development of planning guidelines based on a combined qualitative and
quantitative approach is one of the objectives of the current project. The quantitative approach is
reflected in the development of an Intermodal Airport Ground Access Planning Tool ( IAPT) that
provides the capability to undertake systematic modeling of airport traveler and transportation
provider behavior in order to support airport ground access planning and project implementation.
The IAPT is designed to generate measures of system performance, which can be used to guide
decision- making by planning agencies or decision makers at different levels of government.
3.2 Potential Strategies to Enhance Intermodal Connectivity
The working paper prepared during the first year of the study identified three principal
strategies to improve intermodal connectivity at airports:
• Direct rail service to the airport
• Improved links to nearby rail stations
• Express bus service to off- airport terminals or regional intermodal
terminals.
- 31 -
Although direct rail service to an airport station has been proposed or implemented at an
increasing number of large airports worldwide, it is typically a very expensive solution. Except
in rare cases where an existing rail line runs within close proximity to an airport terminal, the
engineering required to bring a rail line into a station in the airport terminal complex requires
substantial capital investment. In the case of a dedicated airport line, the operating costs of
maintaining an adequate train frequency must also be considered. While such an approach may
be justified at the very largest airports, in general this is not an appropriate strategy for most
airports.
Improving links to nearby rail stations is generally a much less expensive strategy and
more appropriate for smaller airports. These links may take the form of a dedicated shuttle bus
service or an automated people- mover. The later may provide a higher level of service to the
user, and eliminates the vehicle trips associated with a shuttle bus service, but is generally more
expensive to construct and operate. The attractiveness of such links will depend on the
frequency of service of both the link itself and the rail service to which it connects, as well as the
fares charged for the use of the link and by the rail service. While there is no need to operate the
link at a higher frequency than the rail service that it serves, it is important for less frequent rail
services that the connecting link schedule be coordinated with the rail service schedule, so that
the users do not incur a long wait twice.
The provision of express bus services to off- airport terminals located some distance from
the airport provides another strategy to reduce the volume of vehicle trips to and from the airport.
Such off- airport terminals typically provide parking at lower rates than at the airport, as well as
waiting facilities for bus passengers or those waiting to pick up bus passengers. Larger facilities
may also provide ancillary services, such as a newsstand or food and beverage concessions, and
some have provided airline ticketing or check- in. While the ability to check baggage at a remote
location has often been proposed as a feature of off- airport terminals, it is unclear whether this is
a significant factor in the attraction of such a facility and justifies the logistical complexities
involved. The principal advantages of an off- airport terminal to the users are the reduction in the
driving time and distance compared to driving to and from the airport, particularly for passengers
being dropped off or picked up, as well as any saving in parking costs or taxi fares for those
using taxi to get to or from the off- airport terminal, compared to taking a taxi all the way to or
from the airport. Locating an off- airport terminal at a major transit hub also allows airport
- 32 -
travelers to use transit to get to and from the terminal, which is likely to provide better service
than taking transit all the way to or from the airport. Similarly, providing express bus links
between the airport and regional intermodal terminals, such as central rail stations or transit hubs,
can allow airport travelers to utilize the better rail or transit service at those locations to travel to
and from their ultimate trip end, while increasing the ease of travel between the airport and those
facilities.
3.2.1 Examples of Existing Services
Services representing each of the foregoing strategies currently have been implemented at
various California airports.
The extension of the Bay Area Rapid Transit ( BART) system to San Francisco
International Airport ( SFO) that opened in June 2003 provides direct rail service to the second
largest airport in the state. The BART system provides an extensive and frequent region- wide
network with 43 stations serving Alameda, Contra Costa, San Francisco, and northern San Mateo
counties. In addition, the Millbrae BART station provides an interchange with the Caltrain rail
line that serves the Bayshore corridor of eastern San Mateo County and northern Santa Clara
County.
There is also direct rail service at Burbank/ Bob Hope Airport, where the Burbank Airport
Station is located adjacent to the airport within an easy walk of the airport terminal. Even though
it is a very short walk between the train station and the airport terminal, there is shuttle bus
service between the two locations with a direct- line telephone at the train station that airport
travelers can use to call for a shuttle. The station is served by both Metrolink and Amtrak trains
that provide service between Los Angeles Union Station and communities in the San Fernando
Valley and along the coast in Ventura and Santa Barbara counties. However, trains are relatively
infrequent outside of weekday commute hours ( Metrolink is primarily a commuter rail service),
with fewer trains serving points north of Moorpark in the San Fernando Valley.
Several California airports have dedicated shuttle bus service to nearby stations. At Los
Angeles International Airport ( LAX) there is a shuttle bus operated by Los Angeles World
Airports to the nearby Green Line Metro station. In the Bay Area, the AirBART bus operated by
the Port of Oakland connects Oakland International Airport ( OAK) and the Coliseum BART
station as well as the Oakland Coliseum Amtrak station that serves the Capitol Corridor route
between San Jose and Sacramento. However, AirBART bus does not serve the Amtrak station
- 33 -
directly. An Amtrak passenger has to walk to or from the AirBART bus stop at the BART
Coliseum station. At San José International Airport ( SJC) the Route 10 Airport Flyer bus
operated by the Santa Clara Valley Transportation Authority ( VTA) connects the airport
terminals with the Metro/ Airport station on the Alum Rock- Santa Teresa Light Rail line and the
Santa Clara station on the Caltrain line that that serves communities in the U. S. 101 corridor
between Santa Clara County and San Francisco. The Port of Oakland and BART are currently
pursuing a joint project to construct an automated people mover to link OAK to the Coliseum
BART and Amtrak stations ( U. S. Federal Transit Administration, 2002) and San José
International Airport is pursuing an automated people mover link between the airport and the
VTA light rail station, with a possible future extension to the Caltrain station ( Lea+ Elliott, 1999).
Two California airports currently have express bus service to off- airport terminals. The
Los Angeles World Airports ( LAWA) operates the Van Nuys FlyAway service between LAX
and an off- airport terminal adjacent to the Van Nuys airport in the San Fernando Valley. This
terminal provides long- term parking and waiting facilities. LAWA has recently modernized the
terminal building and provided additional parking in an adjacent structure. In the past, a number
of airlines maintained ticket offices at the terminal, although there was no provision for baggage
check- in. In March 2006 LAWA opened a second FlyAway service from the Patsaouras Transit
Plaza adjacent to Union Station in downtown Los Angeles. In the Bay Area, Marin Airporter
operates a scheduled bus service between SFO and two off- airport terminals in Marin County, at
Larkspur and Ignacio near Novato ( North Hamilton Parkway). Both terminals provide long- term
parking and waiting facilities.
Scheduled airport bus service is also available to regional transit centers at a number of
airports. Marin Airporter buses to and from the Hamilton terminal stop at the Central San Rafael
Transit Center, as do Sonoma Airport Express buses serving both SFO and OAK. In Southern
California, Airport Bus of Anaheim provides scheduled bus service between the Anaheim Bus
Terminal and LAX and John Wayne Orange County Airport.
3.3 Intermodal Opportunities at Selected California Airports
The working paper identified a number of opportunities to improve intermodal
connectivity at thirteen California airports, including some that had been previously identified in
the Ground Access to Airports Study performed for the California Department of Transportation
- 34 -
( Landrum & Brown, 2001), and also presents a preliminary qualitative assessment of their
feasibility. In those cases where the intermodal opportunities have already been subject to more
detailed quantitative analysis as part of other studies, the results of this analysis are discussed in
the working paper.
3.3.1 Southern California
Burbank/ Bob Hope Airport is currently served by Metrolink and Amtrak trains, although
these are relatively infrequent. However, the Red Line of the Los Angeles Metro terminates at
North Hollywood station, about 4 miles to the southeast of the airport. An extension of the
system beyond North Hollywood to Van Nuys, Reseda and Canoga Park in the San Fernando
Valley using buses on a dedicated guideway, termed the Orange Line, opened in October 2005.
The Red Line provides frequent service to downtown Los Angeles seven days a week and
connections to other Metro lines that provide service to large parts of the Los Angeles basin.
Since October 31, 2005 Burbank Bus, the local transit system for the City of Burbank, has
operated its NoHo- Emprire route between the North Hollywood station and the area immediately
to the east of Bob Hope Airport. The closest bus stop to the airport is on North Hollywood Way
and Thornton Avenue at the entrance to the airport, a short walk from the terminal building.
However, the bus route makes a one- way loop in the area to the east of the airport, and travelers
from the North Hollywood station to the airport have a somewhat longer ride than travelers from
the airport to the station. The service operates weekdays with departures from North Hollywood
station between 6 am and 10 am and from 2: 45 pm to 7: 23 pm.
At present, the majority of Burbank air passengers come from the San Fernando Valley to
the west of the airport or communities in the San Gabriel Valley to the east of the airport. An
improved link to the North Hollywood station would enhance service to communities between
North Hollywood and downtown Los Angeles served by the Red Line as well as communities in
the San Fernando Valley served by the Orange Line. Travelers to Burbank Airport from
communities in the San Gabriel Valley would need to take the Gold Line into downtown Los
Angeles to connect to the Red Line in order to use Metro Rail. Since there are fairly direct
freeway links between the San Gabriel Valley and Burbank Airport, it can be expected that
relatively few airport travelers from the San Gabriel Valley would find this an attractive way to
reach the airport. However, the Red and Orange Lines would serve a significant share of the
Burbank Airport market. A transit advocacy group in the San Fernando valley, The Transit
- 35 -
Coalition ( www. thetransitcoalition. us), has proposed extending the Orange Line north from its
current terminus at North Hollywood station to Burbank Airport along Vineland Avenue, using
arterial streets rather than a dedicated guideway.
John Wayne Orange County Airport currently has no dedicated link to any regional rail
system. The Orange County Transportation Authority has plans to implement a bus rapid transit
( BRT) route between the Irvine Station to the east of the airport and the City of Brea to the north
( http:// www. octa. net/ brt. aspx). It is anticipated that service would commence by 2010 and the
planned route includes the airport and the Santa Ana Depot transportation center that would
provide access to Metrolink and Amtrak trains serving communities between downtown Los
Angeles and San Diego, as well as the Metrolink Inland Empire- Orange County line serving
communities in Riverside County and connections to other Metrolink and Los Angeles Metro
lines that provide service to large parts of the Los Angeles basin. An alternative, and more
direct, connection would be provided by establishing a shuttle bus link between the airport and
the Tustin Metrolink and Amtrak station about 4 miles to the northeast of the airport. However,
relatively few air travelers using John Wayne Airport have trip ends outside Orange County due
to the more extensive air service available at Los Angeles International Airport to the northwest
and Ontario International Airport to the north. It is therefore unlikely that improved intermodal
connections at John Wayne Airport would attract significant numbers of air passengers with trip
ends outside Orange County. While the communities served by the Metrolink Orange County
Line account for about 60 percent of the Orange County residents using John Wayne Airport, for
many of these trips the time involved in accessing the nearest station, riding the train, and then
riding a bus to the airport would be significantly longer than driving to the airport. In particular,
most trip origins in Irvine, which account for about 12 percent of the total, are closer to the
airport than to the Irvine station. Therefore it is likely that the percent of air passengers who
would use such a service would be quite small. However, it may attract a number of airport
employees who are more likely to be familiar with the train schedules since they make the trip on
a regular basis.
Long Beach Airport currently has no dedicated link to any regional rail system. However
the Blue Line of the Los Angeles Metro Rail system runs about a mile and a half to the west of
the airport and connects downtown Long Beach with downtown Los Angeles. A bus link to the
Willow station on the Blue Line would provide access to communities between Long Beach and
- 36 -
downtown Los Angeles, as well as connections to other Metro lines that provide service to large
parts of the Los Angeles basin. The airport has recently experienced a significant growth in
traffic as a result of the introduction of air service by jetBlue Airways and other airlines serving
the airport. In consequence, it is likely that the airport is now drawing air passengers from a
wider area in the Southern California region. This suggests that an improved connection to the
regional rail system might attract some of these air passengers. Also, since the air service at the
airport is primarily targeting low- fare travelers, it is likely that many of those air passengers
would be attracted to an improved transit connection. At present local bus service between the
airport and stations on the Blue Line is relatively infrequent, particularly at weekends, and rather
circuitous. A shuttle bus link to the Blue Line Willow station would take about 10 minutes in
each direction, so it would be possible to provide service every 30 minutes with only one vehicle
per shift. A less expensive way to provide equivalent service would be to modify the route of the
Long Beach Transit Route 102 bus, which currently provides half- hourly service on weekdays
with stops at the Willow station and on Spring Street on the southern boundary of the airport, but
does not serve the terminal, to include the airport terminal in the route and add evening and
weekend service. This might attract sufficient additional riders to be attractive to the transit
operator without any subsidy from the airport.
The California Ground Access to Airports Study identified four potential intermodal
connectivity projects at Los Angeles International Airport ( LAX): expansion of the current Van
Nuys FlyAway bus terminal in the San Fernando Valley; development of new FlyAway
terminals elsewhere in the region; an extension of the Metro Green Line to the Airport; and an
airport people- mover link to the Green Line. The expansion of the Van Nuys FlyAway bus
terminal was initiated by LAWA and completed in summer 2005 and the FlyAway service from
Union Station commenced in March 2006. The Metro Green Line currently extends past LAX to
a terminus in Redondo Beach, with a station ( Aviation/ LAX) adjacent to the airport and served
by a free shuttle bus connection operated by LAWA. The recent LAX master plan update
envisages a major reconfiguration of the airport terminal area, with an automated people- mover
link to an intermodal facility located at the Aviation/ LAX station. Therefore additional FlyAway
terminals at other locations in the region would appear to be the only intermodal connectivity
project identified in the study that remains to be addressed. In 2001 LAWA commissioned a
market analysis of a number of potential sites for new FlyAway facilities in the region ( Leigh
- 37 -
Fisher Associates, 2001). The analysis examined alternative sites in four corridors, as well as the
feasibility of a terminal at Union Station in downtown Los Angeles, and developed estimates of
average daily ridership from each site for the peak month ( August) in 2000, 2005 and 2010. The
sites were then compared using a scoring system and the preferred site identified in each
corridor.
Ontario International Airport currently has no dedicated link to any regional rail system.
However the Metrolink San Bernardino Line runs about 2 miles to the north of the airport, while
the Metrolink Riverside County Line runs about one mile to the south of the airport. A shuttle
bus link serving the Rancho Cucamonga station on the San Bernardino Line and the East Ontario
station on the Riverside Line would provide access to Inland Empire communities served by both
lines, as well as connections to other Metrolink and Los Angeles Metro lines that provide service
to large parts of the Los Angeles basin. In 2004 Ontario International Airport