EPA 600/R-14/278 | September 2014 | https://www.epa.gov/research
United States
Environmental Protection
Agency
v>EPA
Synthesis Paper on Sustainable
Transportation:
Prepared for the Sustainable and Healthy
Communities Research Program - Theme 4
Office of Research and Development

-------
1
SYNTHESIS PAPER ON SUSTAINABLE TRANSPORTATION
Prepared for the Sustainable and Healthy Communities Research Program - Theme 4
Authors:
Paper Coordinator: Nick Flanders
Contributors: Nick Flanders, Jeff Yang, Rebecca Dodder, Gregg Furie, Rich Baldauf, Laura Bachle, Andrew
Bostrom, Laura Berry, Claudia Walters, Jane Bare, Tim Barzyk, Randy Bruins, Ellen Cooter, Francesca
DiCosmo, Tarsha Eason, Tom Fontaine, Laura Jackson, Nathan Schumaker, Jim Weaver
Task Lead: Rich Baldauf

-------
2
TABLE OF CONTENTS
1	INTRODUCTION	6
1.1	Purpose of this Document	6
1.2	Community Priorities	7
1.2.1	Community Planning Process and Decision Making Framework	7
1.2.2	Overview of Community Transportation Priorities	8
1.2.3	Sustainability Priorities	8
1.3	Decision Agents that May Use this Document	9
1.4	Structure of the Document	9
1.5	Research Methods	10
1.6	Limitations	11
2	U.S. TRANSPORTATION BACKGROUND: HISTORY AND TRENDS	11
2.1	The Building of the Motorized Transportation System in the United States	11
2.2	The Movement of U.S. Urban Populations in the Era of Motorized Transportation	12
2.3	Present Conditions and Predicted Directions of U.S. Transportation and Land Use	14
2.4	Current Ideas that Break with the Past	15
3	SUSTAINABILITY ASSESSMENT	16
3.1	Integrated Tools, Resources, and Indicators	16
3.1.1	Limitations	20
3.1.2	Specific Models, Tools, and Other Resources from the U.S. Environmental Protection
Agency (EPA)	21
3.1.3	Specific Models, Tools, and Other Resources (Non-EPA)	25
3.2	Drivers of Transportation Behavior	28
3.2.1	Correlations and Elasticities Regarding the Built Environment and Travel Behavior	31
3.2.2	Parking	32
3.2.3	Neighborhood-Scale Travel versus Regional Travel	32
3.2.4	Transportation System Capacity as a Driver of Travel Behavior	34
3.2.5	Induced Demand	35
3.2.6	Peak Spreading	35
3.2.7	High Occupancy Toll (HOT) Lanes	36
3.2.8	Trip Aggregation	36

-------
3
3.2.9	Crowding on Transit Vehicles	37
3.2.10	Economic, Social, and Psychological Motivations for Travel Behavior	37
3.2.11	Self-Selection	39
3.2.12	Traveler Characteristics	39
3.2.13	Travel to School	40
3.2.14	Travel by the Elderly	41
3.2.15	Transportation-Behavior Modeling Techniques	42
3.2.16	Data Availability for Modeling Travel Behavior	45
3.2.17	Specific Resources and Tools for Determining Travel Behavior Outcomes	45
3.3	Air Quality Issues and Related Tools, Resources, and Indicators	46
3.3.1 Specific Resources and Tools for Determining Air Quality Outcomes	50
3.4	Energy Use and Climate Change Issues and Related Tools, Resources, and Indicators	52
3.4.1	Transportation Demand	54
3.4.2	Travel Mode Choice and Public Transit	55
3.4.3	Fuel Economy and GHG Standards	55
3.4.4	Operational Considerations Affecting Energy Use and Emissions	56
3.4.5	Alternative and Renewable Fuels	57
3.4.6	Vehicle Electrification and Related Infrastructure	58
3.4.7	Greenhouse Gas Footprint	59
3.5	Water Issues and Related Tools, Resources, and Indicators	60
3.5.1	Runoff from Impervious Surfaces	60
3.5.1.1	Mitigation Strategies	61
3.5.1.2	Assessing the Effectiveness of Mitigation Strategies	62
3.5.2	Groundwater Contamination from Leaked Fuel	62
3.5.3	Relationship between Transportation Infrastructure and Water Infrastructure	62
3.5.3.1	Water Planning and Adaptation Methods: Implications for Transportation Choices ... 63
3.5.3.1.1	Current Practice	63
3.5.3.1.2	Urban Form Transformation: Scenario Planning	66
3.5.3.1.2.1	Urban Population and Land Use Projection	67
3.5.3.1.2.2	Water Planning and Engineering	68
3.5.3.1.3	Urban Form Transformation: Smart Growth and Water Systems Optimization	68
3.5.3.2	Known Effects of Transportation Decisions on Water Infrastructure Outcomes	69

-------
4
3.5.3.2.1	City Development Examples	70
3.5.3.2.1.1	Rapid Urban Sprawl	70
3.5.3.2.1.2	Leap-Frog Development and Small/Community Water Systems	71
3.5.3.2.1.3	High-density Urban Centers	72
3.5.3.2.2	Smart Growth	73
3.5.4 Transportation Fuel Choice and Impacts on Water Resources	73
3.6	Human Weil-Being Issues and Related Tools, Resources, and Indicators	75
3.6.1	Physical Health Outcomes	75
3.6.1.1	Direct Impacts	75
3.6.1.1.1	Air Pollution	75
3.6.1.1.2	Noise	76
3.6.1.1.3	Injuries	76
3.6.1.2	Indirect Impacts	78
3.6.1.2.1	Health Behaviors	78
3.6.1.2.2	Social Determinants of Health	79
3.6.1.2.3	Air Pollution and Climate Change	80
3.6.1.3	Tools	81
3.6.2	Social Interaction Outcomes	84
3.6.3	Equity Outcomes	85
3.7	Economic Issues and Related Tools, Resources, and Indicators	87
3.7.1	Accessibility	91
3.7.2	Announcement Effect	91
3.7.3	Non-Accessibility Benefits from Transportation Projects	92
3.7.4	Agglomeration Effects	92
3.7.5	Regional Adjustment Model	93
3.7.6	Models of Transportation-Infrastructure Land Use and Other Land Uses	94
3.7.7	Market Imperfections	94
3.7.8	Import Substitution	95
3.7.9	Cost-Effective Use of Government Budgets	95
3.7.10	Equity in Transportation Funding Mechanisms	96
3.7.11	Economic Analysis Software for Transportation Planning	97

-------
5
4	EXAMPLE APPLICATION OF SUSTAIN ABILITY PRINCIPLES AND ASSESSMENT TOOLS: COMPLETE
STREETS	97
5	IMPORTANT INFORMATION GAPS	106

-------
6
1 INTRODUCTION
1.1 Purpose of this Document
This document synthesizes current assessment tools and resources, best practices, and research needs
related to sustainable transportation development and implementation. This document does not reflect
an exhaustive literature review, but instead highlights relevant work that will aid the EPA in its efforts to
establish sustainability research. The vision statement developed by the Environmental Protection
Agency (EPA) Office of Research and Development (ORD) for the Sustainable and Healthy Communities
Research Program (SHCRP) reads:
"The Sustainable and Healthy Communities Research Program (SHC) will inform and empower decision-
makers in communities, as well as in federal, state, and tribal programs, to effectively and equitably
weigh and integrate human health, socioeconomic, environmental, and ecological factors into their
decisions in a way that fosters community sustainability."
When state and federal government agencies set policies meant to promote sustainable development
and sustainable transportation systems, local communities and their individual members often
implement and adhere to these policies, both in law and in spirit. The SHCRP prepared this document to
provide a summary of existing tools, resources, and indicators that may be used in the creation and
implementation of sustainability plans, with special attention given to those tools, resources, and
indicators that relate to sustainable transportation and are rooted in a systems perspective of
interrelated variables. Ideally, communities will be able to use these tools, resources, and indicators to
assess their current local conditions, forecast the likely future consequences of various different policy
alternatives, and monitor whether existing transportation and/or sustainability policies are meeting
their objectives. Many of the tools and resources presented here will allow communities to better
identify practical, effective, and equitable ways to meet both present and future needs of the natural
environment, the economy, and human society, both within their own local context and within the
context of the country and world at large.
This document also identifies new tools being developed and research needed to further advance
sustainable transportation development. Eventually, local communities will also be able to take
advantage of a new tool being developed under the SHCRP and to which this paper contributes, called
the Total Resources, Impacts, and Outcomes (TRIO) framework. This dynamic framework for analyzing
sustainability impacts will be based upon many different causal relationships and feedback loops that
are established as existing between various measures of a community's past, present, and future
situations. This document attempts to highlight many of these causal relationships and feedback loops
related to transportation. Also identified in this document are correlations that may be indicative of
causal relationships, as well as theoretical relationships that have not yet been researched enough to be
either proven or disproven. The primary goal of this document is to help researchers and practitioners
to identify existing resources for implementing sustainable transportation planning, as well as to identify
knowledge gaps, and the research needed to fill those gaps, in order to enhance and improve the

-------
7
planning process and advance the ultimate goal of sustainably designed transportation networks that
enable and support community sustainability as a whole.
1.2 Community Priorities
The EPA recognizes that communities want to become more sustainable and are moving in this
direction. Amongst a variety of other outreach activities, EPA ORD held seven Community Listening
Sessions in 2011 to gain insight into the issues that communities are facing as they implement
sustainable practices to help inform the design of the SHCRP. Although the challenges faced and
approaches to addressing these issues were vastly different, many issues emerged as common themes
during these conversations. The Community Listening Sessions for the Sustainable and Healthy
Community Research Program document (EPA 675/R-11/001 2011) describes each common theme in
depth, identifying collective issues as well as specific challenges to communities related to each theme.
1.2.1 Community Planning Process and Decision Making Framework
The community planning process is comprised of multiple steps. In general, sustainability planning
comprises five stages: information gathering, interpretation, collaborative design and policy
formulation, implementation, and monitoring and evaluation (for example, see the Green Communities
Toolkit at http://www.epa.gov/greenkit/index.htm). Communities generate plans and projects to
address local challenges and prepare for a sustainable future. In implementing these plans and projects,
communities develop policies, such as executive directives, to implement sustainable practices. In
addition, they utilize approaches (e.g., Complete Streets) and scientific and other tools developed in-
house, by EPA or other organizations. In addition, communities' everyday decisions greatly affect their
long-term sustainability, by way of putting in place designs that predetermine local activity or long-term
infrastructure costs, and so need to be evaluated in the light of lasting implications.
The Listening Session communities were further analyzed to determine the "state of sustainability
practice" of communities with regard to decision making and the science they use. The purpose of the
analysis was to gain a better understanding of how communities, in general, make their decisions—
"community lens"—to identify the science that communities are using in their decision making and their
needs for scientific tools. In other words, the analysis was performed to understand the "state of
sustainability practices" that communities are implementing. Overall, the analysis organized the
Listening Session information into a framework of major and secondary decisions, and their priorities,
currently being addressed by communities toward becoming sustainable: Community Decision Sectors
and Cross-Cutting Issues.
Community Decision Sectors represent the fundamental areas for decision making, and often impact
how they spend their funding. Cross-Cutting Issues are broad sustainability issues, while important to
communities, are usually not the first and foremost issue in their decision making. They are often
considered in the context of the Community Decision Sectors and traverse multiple sectors. Social
acceptance of sustainability principles, for example, affects the adoption of green building practices, use
of public transit, and consumption of local foods and products. The development of performance
indicators will facilitate the planning and assessment of sustainability actions and results across many

-------
8
topics. The identification of Cross-Cutting Issues is important to understand which factors influence or
are influenced by decision-making sectors and to what extent.
1.2.2	Overview of Community Transportation Priorities
Although individual transportation preferences center on private automobile transport, many
communities would like to provide citizens with alternative modes of public transit to reduce vehicular
emissions and traffic congestion. Transportation in rural communities is used to connect residents with
job centers and services, often located in nearby urban areas, yet most rural communities have a very
limited public transportation infrastructure. Although cities recognize the value of public transportation
from a sustainability perspective, rural communities often place a lower priority on it because of cost,
limited ridership, the complexity of setting up regional partnerships with neighboring counties/cities,
and the car-centric preferences of most community members.
1.2.3	Sustainability Priorities
Increase Public Transportation. Access to public transit in rural areas is limited. Communities must
accept that personal vehicles will be part of the transportation infrastructure, while simultaneously
trying to improve and promote public transit. Increasing public transit options within urban areas
improves convenience and might influence additional citizens to utilize more sustainable options.
Provide Multi-Modal Transportation Options. Communities have explored multi-modal transportation
options (e.g., walk, bike, transit) that are inclusive of people of all demographics, thereby reducing
dependency on cars. Planning for "complete streets" that support walking, biking, and riding buses as
well as driving can promote sustainable development. Many communities believe that a public rail
system would be a useful method to improve access to workplaces while reducing individual automobile
use.
Adopt New Technologies. Technology improvements continuously occur in areas relevant to
transportation planning. These technologies can include improved vehicles (motorized and non-
motorized), signalization, monitoring, and infrastructure. For example, the technology to create
permeable surfaces for roads is available, and demonstration projects exist, but the practice has not
The Community Decision Sectors are:
The Cross-Cutting Issues are:
1.	Land Use
2.	Buildings and Infrastructure
3.	Transportation
4.	Local Food Systems
5.	Water Management
6.	Waste Management and Site
1.	Economics
2.	Social Issues
3.	Environmental Justice
4.	Climate Change
5.	Energy
6.	Human Health
7.	Environment
8.	Performance Measures and Metrics
Revitalization

-------
9
been widely adopted in the United States. People do not necessarily adopt new practices on their own
initiative, and they need to understand that their decisions should address multiple problems (e.g., a
new road should improve transportation and water permeability). Ordinances, permits, and model
design approaches could facilitate a transition to sustainable thinking and decisions regarding
transportation options.
Encourage Acceptance. Public transportation services often are used only by those who have no other
options. Socio-economic class plays a role, as some communities (urban and rural) find that low-income
residents are unwilling to use public transportation if they have other options, because of the economic
status associated with owning a car. Middle-income residents are generally more open to choosing
public transit over other modes. Some cities have developed incentives to encourage commuters to
leave their vehicles at home.
1.3	Decision Agents that May Use this Document
Because substantial stakeholder involvement is an important element of nearly any community
planning, almost anyone with a stake in their community's future may derive some use from the tools,
resources, and other knowledge presented here. A critical use of this kind of scientifically-validated
information is to overcome the assumptions and inertia that can perpetuate old patterns and problems.
In addition to local, tribal, state, and federal government agencies, decision agents and stakeholders in
transportation planning include private land developers, the end-users of the transportation system,
and various external stakeholders, such as nonprofit organizations or other residents and businesses
located near the transportation system. Transportation sustainability issues that tend to be important
to local government agencies include assuring a high quality of life for their constituents, achieving
social justice, and ensuring that their plans for the future are financially feasible. Meanwhile, private
developers are understandably concerned foremost with the economic implications of transportation
projects and sustainability policies. As for the end-users of the transportation system, personal interests
and conveniences tend to take priority, such as the affordability of housing and commercial real estate,
high-quality and timely access to destinations and services, and ensuring that the transportation system
helps produce natural and built environments that are comfortable and healthy to live in. Many
stakeholders, who may nonetheless be very directly affected in other ways by a given proposal under
consideration, tend to be the most motivated when they perceive the possibility of exposure to some
significant risk or danger. At times, various interests may come into conflict with one another, creating
the necessity to agree on what specific issues are at stake and how they should be prioritized
(Wallbaum, Krank, and Teloh 2011). It is in this identification and refinement of issues and prioritization
of objectives that the tools, resources, and other information in this document may be especially useful.
1.4	Structure of the Document
This document presents key issues for consideration in sustainable transportation planning and
development, especially at the level of community-scale decision-making.

-------
10
First, background is provided on the history of surface transportation in the United States, as well as
current trends in transportation. This discussion of transportation trends is provided to convey the
issues created by those patterns.
The main body of the document discusses various tools, resources, and known causal relationships and
correlations that may help communities evaluate the sustainability of their current transportation
systems and predict what they must do in order to optimize the system's sustainability in the future.
The relationships identified in this synthesis paper support the development of the Total Resources,
Impacts, and Outcomes (TRIO) framework, a tool being developed by the EPA to help communities make
more sustainable decisions through illuminating the comprehensive implications of decision options. In
addition to identifying known tools, resources, indicators, and relationships, this paper also identifies
issues related to sustainable transportation that represent important knowledge gaps and ought to be
studied more in the future. The following are the major topics addressed in the main body of the paper:
•	A compilation of existing tools, resources, and methods for integrating the various aspects of
sustainable transportation into a single analysis, as opposed to addressing each sustainability
goal separately. This includes a discussion of inherent obstacles to such integrated sustainability
analyses.
•	A discussion of factors that determine how people will use the transportation system and of
ways in which those factors might be analyzed. Whatever decisions a community makes about
the future of its transportation system, a large proportion of the sustainability outcomes of
those decisions will depend upon the ways in which they do or do not motivate individual
travelers to change their travel behavior.
•	Transportation impacts on air quality.
•	Transportation impacts on energy use and climate change.
•	Transportation impacts on water quality and quantity, including water that serves human needs
and water that serves other ecological functions.
•	Transportation impacts on human health and well-being. This includes both physical well-being
and psychological and social well-being.
•	Transportation impacts on economic prosperity, equity, and sustainability, including impacts on
government budgets.
Near the end of this synthesis, various sustainability components discussed throughout the paper are
considered in the context of the specific issue of "Complete Streets" design practices, with frequent
reference to the specific assessment tools identified throughout the paper and many causal
relationships illustrated through system-dynamics diagrams. "Complete Streets" was chosen as the
focus of this illustrative discussion because it is a transportation topic with a strong connection to a wide
variety of community-sustainability issues.
1.5 Research Methods
After a general outline was created of the topics deemed most pertinent to this synthesis paper, the
paper was created by compiling sections written by various subject-matter experts within the EPA. Each

-------
11
author was provided with the same set of general guidelines regarding what types of information to look
for on their particular research topic, so that all sections would address the same basic themes.
The authors of this synthesis paper gave particular attention to two tasks. The first task was identifying
known cause-and-effect relationships between communities' transportation-related decisions and
sustainability outcomes, as well as possible cause-and-effect relationships that warrant additional
research. The second task was to identify tools and resources that might be used by communities or
other relevant decision-makers to assess positive and negative outcomes from various transportation-
related decisions that they are considering making that may affect community sustainability.
Due to limitations of time and personnel, this is not an exhaustive review of the available literature
related to sustainable transportation. However this document provides resources and best practices for
implementing TRIO and research needed to further the SHCRP.
1.6 Limitations
Because this research synthesis does not represent an exhaustive review of the literature or resources
publicly available, not all aspects of the relationship between transportation and sustainability are
addressed. Where possible, the synthesis focuses on relevant variables that are within the capacity of
community decision-makers to affect. In addition, this synthesis includes only a handful of the many
tools and resources that could be useful for measuring and/or advancing sustainable transportation.
The aspects of sustainable transportation that are included in this synthesis paper reflect those research
areas that are pertinent to the mission of EPA (widely held in common with other federal, state, and
local agencies) and/or reflect the latest research, based on the knowledge of the authors.
2 U.S. TRANSPORTATION BACKGROUND: HISTORY AND TRENDS
Before getting into the question of how United States communities' transportation systems can be
made more sustainable, a clear idea of how those systems have developed and changed over the years,
what factors caused them to change, and how those changes in the transportation sector have affected
other characteristics of this country's communities is important. In addition, understanding what
present-day trends are molding the future of American transportation, American society, and their
measures of sustainability helps determine whether individual communities want to encourage or
discourage these trends.
2.1 The Building of the Motorized Transportation System in the United States
Motorized passenger transportation in the United States began in the late 1800s and early 1900s, as a
series of transitions were made from horse-drawn railcars to steam-engine transit to cable cars to
electric streetcars to heavy-rail subway lines to bus transit to travel by personal automobile on paved,
all-weather roads, with each of these transitions being motivated by the desire for greater speed and
independence (Sinha 2003). Of all the public transit modes that predated the popularization of private
automobiles, electric streetcars were one of the most influential. In the year 1920, there were 40,000
miles of electric rail lines in the United States and the mean number of transit trips per person per year

-------
12
was 250, which is approximately eight times contemporary transit usage (Sinha 2003). Even before the
affordability of personal automobiles vastly accelerated the process, electric rail transit was helping to
bring about the rapid expansion and decentralization of urban areas in the United States (Levinson 2004,
Sinha 2003).
The rise of private automobiles as the dominant mode of U.S. transportation was facilitated not just by
the production of motor vehicles that the majority of working adults can afford, but also by the creation
of a highly sophisticated system of highways. The first thirty years of the 20th century saw a series of
uncoordinated intercity roads give way to a national system of paved and numbered roadways that
were equipped with uniform signage. Some major milestones in this process were the Federal-Aid Road
Act of 1916, which established 50-50 federal-state cost sharing for highway construction, and the
Federal-Aid Road Act of 1921, which defined primary roadways, secondary roadways, and urban road
systems that were eligible for government aid. During the 1930s, limited-access highways started being
built within individual metropolitan areas, leading to the 1940 opening of the Pennsylvania Turnpike, the
first intercity limited-access highway in the country. In 1944, the federal government defined the
National System of Interstate and Defense Highways, the financing of which was established as a 90-10
federal-state split by the Federal-Aid Act of 1956. These federal actions resulted in the modern
interstate highway system, which in turn served as inspiration for the construction of many more
highways in the post-World War II era (Levinson 2004).
During the postwar highway-building era, transportation engineers concerned themselves primarily with
reducing traffic congestion and maintaining design standards, as opposed to mitigating impacts for the
various dimensions of what is now called sustainability (Levinson 2004, Mercier 2009). However, in
1962, the federal government began requiring that urban areas have ongoing comprehensive and
cooperative transportation planning processes in order to prevent conflicts between past, present, and
future transportation projects and between transportation projects and other features of the urban
environment. Then, in 1969, the National Environmental Policy Act (NEPA) started requiring that
assessments be made of the impacts of proposed infrastructure projects on the environment and how
those impacts could be mitigated. In 1970, the Clean Air Act was passed, which set transportation
emissions targets and ambient air quality standards. As of 1973, federal highway money can be spent
on public transit improvements. More significant changes to the thinking behind transportation policy
would come in the 1990s, when additional federal funding would be provided for improving air quality,
mitigating traffic congestion, and improving transportation options for pedestrians and cyclists (Levinson
2004). Also, the importance that federal law places on involving the general public in transportation
planning decisions was enhanced by both NEPA (1969) and the 1991 Intermodal Surface Transportation
Efficiency Act (ISTEA) (Lemp et al. 2008). One caveat to all of these laws and policy changes from the
last half century, though, is that as state and federal requirements regarding how transportation
projects are planned, designed, and constructed have gotten more complex, the time and money
needed to complete any given project have noticeably increased (Levinson 2004).
2.2 The Movement of U.S. Urban Populations in the Era of Motorized Transportation

-------
13
In the early 1900s, many urban theorists saw decentralization as an effective means of addressing a
variety of social and health problems that urban areas of that time were experiencing. The
decentralization that these theorists advocated has now come into being, thanks in no small part to the
modern transportation system, and is now faulted by many for helping to create numerous present-day
sustainability issues, such as urban sprawl and reliance on personal motor vehicles (Tomalty 2009). Both
the populations and the spatial extents of metropolitan areas have grown especially rapidly since the
end of World War II, correlating with the expansion of the freeway transportation system in the United
States, as well as increases in average wealth and government programs expanding the population able
to afford a single-family home (Chi and Stone 2005, Sinha 2003, Sukkoo 2007). At the same time, sprawl
development that takes place beyond the urban fringe is also partially the result of people seeking out
attractive natural scenery and cheaper land prices (Carruthers and Vias 2005). Still, highway
interchanges have emerged as significant attractors of economic development (Levinson 2004),
contributing to declines in the percentage of a metropolitan area's population that still has ties to the
central city (Sinha 2003). Manufacturing jobs have largely moved out of cities' central business districts,
which are now mostly dominated by service-sector industries (Chang 2007).
One way of measuring the simultaneous growth and decentralization of urban populations is the density
gradient, comparing population density with distance from the central business district. Over the last
120 years, the slope of the density gradient of U.S. metropolitan areas has consistently become less and
less pronounced, indicating less and less difference between the population density of the city center
and each successive ring of suburbs, as shown in Figure 1 (Sukkoo 2007). Meanwhile, there exists
evidence that average urban population densities actually increased slightly during the first half of the
1900s, before falling rapidly during the second half of the century. In 1940, the average population
density of U.S. metropolitan areas was 347 people per square mile, which grew to 589 people per
square mile by 1960 and then dropped to 288 people per square mile by 1990. Prior to 1960,
metropolitan populations were both growing and moving out from the central cities, but increases in
metropolitan land area did not always keep up with these changes. After 1960, the addition of new land
to metropolitan areas greatly outpaced population increases (Sukkoo 2007). In 1970, the U.S. census
confirmed that for the first time in history, a majority of the population of U.S. metropolitan areas was
actually living outside of those metropolitan areas' central cities, a far cry from the pre-World War II
model of dense concentric rings of development around an urban core (Sinha 2003). Throughout all of
these changes in population patterns that have accompanied the era of motorized transportation in the
United States, not only have existing cities become less concentrated, but many new major cities have
come into being that, because of the time in history when their basic structures were established,
adopted far less dense development patterns than what most of the older cities have come to possess
(Sukkoo 2007). Finally, the trends that have been described here for the United States as a whole
generally also hold true for each of the various regions within the country (Sukkoo 2007). More
information on trends in U.S. land use may be found in the concurrent SHCRP Theme 4 synthesis paper
on land use.

-------
14
o
o
o
o
o
o
o
oc
1950
1890
2000
o
o
o
-------
15
most U.S. cities that public transit is largely regarded as something that is only used by people for whom
driving a car is not a viable option (Sinha 2003). At the same time, the historically low average densities
of U.S. urban areas have both made it more expensive than it used to be to provide basic infrastructure
to a given number of users and increased the number of miles that transit vehicles must travel in order
to accommodate the same number of customers - to the point where no transit system in the country
can operate without large government subsidies (Sinha 2003). Further deepening the disparity between
private and public transportation, the consequences of peak-hour automobile travel on U.S. highways
are generally not reflected in the price of taking part in it (price of fuel, tolls, etc.) (Levinson 2004).
Based on these and other trends, a number of projections can be made regarding surface transportation
and land use in the United States. First of all, social and economic problems are likely to arise from the
combination of great travel distances caused by a low-density development pattern and high fuel prices
caused by dwindling fossil fuel reserves (Mercier 2009). Second, if present trends continue, the demand
for roads will continue to grow faster than the population. However, as the percentage of Americans
who own automobiles reaches saturation level and stops increasing, the rate of growth in the demand
for roads is likely to be mitigated (Levinson 2004). Projections suggest that most U.S. population growth
will continue to be located in urban areas, which will continue to become more dispersed and develop
multiple centers of activity, even though their original central cities will continue to play an important
role (Levinson 2004). Most policies and plans intended to affect urban land use and transportation
systems in the United States will continue to face the challenge of metropolitan areas being divided into
a very large number of largely uncoordinated municipal governments (Levinson 2004, Sinha 2003).
2.4 Current Ideas that Break with the Past
One particular idea that has gained traction in recent years is that of promoting sustainability through
the creation of an urban environment where all modes of transportation can compete on even terms, as
opposed to automobile travel being underpriced while other means of travel are not feasible across
wide areas (Sinha 2003).
Another contemporary idea is that of changing the land use patterns in a given area for the express
purpose of changing how people choose to use the transportation system. Many of the strategies
spawned by this idea, including policies that promote denser development and less separation of land
uses, are included under the umbrella term of "smart growth," especially when implemented at a city-
wide or regional scale; more on the topic of smart growth may be found in the concurrent SHCRP
synthesis paper on sustainable land use practices. Meanwhile, some people advance the ideas of
neotraditional development, wherein urban forms are created that resemble those which existed prior
to when automobiles became the dominant transportation mode, while others advance the idea of
Transit-Oriented Development (TOD), wherein compact, easily-walkable, mixed-land-use neighborhoods
are built around transit stops, bringing customers to the transit lines rather than the other way around
(Sinha 2003).
Finally, when road projects are brought under consideration, the concept of context-sensitive design,
agreeing to the increase in project costs that comes from considering factors other than the safety and
-------
16
capacity of a given part of the road system, has become increasingly popular. Ideally, the design process
would be collaborative, involve substantial stakeholder input, and produce solutions that consider the
ecological, social, economic, and aesthetic characteristics of the area around a transportation facility at
the same time as satisfying the more basic objectives of safety, system efficiency, and cost-
effectiveness. Examples of context-sensitive design practices may include building highways along
routes where their negative impacts are likely to be the least, building below grade or underground,
incorporating other land uses into a facility's right-of-way, or even choosing in a particular case to satisfy
transportation demand through an expansion of the public transit system and walking and cycling
facilities instead of an expansion of the highway system or other road network (Levinson 2004, U.S.
Department of Transportation).
3 SUSTAINABILITY ASSESSMENT
3.1 Integrated Tools, Resources, and Indicators
For some time now, there has been a consensus that the three major (interrelated) elements of
sustainability are environment, economy, and society (or human well-being). However, these elements
can be refined and expanded to include a variety of more particular factors of relevance, including
natural and human ecology, political concerns, technological limitations, regulatory frameworks,
resource conservation, human health, and demographic, socioeconomic, geographic, and
intergenerational equity (Koo, Ariaratnam, and Kavazanjian 2009, Mercier 2009, Ramani et al. 2011). In
order to effectively assess how well a given community's transportation decisions are supporting the
various elements of sustainability and create plans in response to that assessment, establishing a
hierarchy of goals, objectives, subobjectives, and performance measures can be helpful. The
overarching goal of sustainability is divided into a series of goals that are specific to the various elements
of sustainability, which are then further divided into a series of increasingly more specific objectives,
based upon social theories of which human actions have a meaningful effect on the objectives and goals
above them. Ultimately, these objectives reach a level of specificity where attainment is believed to be
met when certain quantifiable performance measures reach specific target values. Then, armed with a
means of assessing sustainability in concrete terms, corresponding policy instruments can be created to
help achieve the various performance measures (Black, Paez, and Suthanaya 2002, Ramani et al. 2011).
Before performance measures are used to make any actual policy decisions, they may also be used to
facilitate communication between the various actors and stakeholders in the transportation planning
process (Ramani et al. 2011). Even though the best time to utilize sustainability performance measures
is at the beginning of the transportation planning process (e.g., to set goals), they can also be used to
track the achievement of sustainability-related objectives at most other project stages, including the
selection of building materials, design, construction, operations, maintenance, and after the
decommissioning of infrastructure (Koo, Ariaratnam, and Kavazanjian 2009). Major categories of
decisions that performance measures can be utilized to support include future system capacity,
predicting future levels of demand, selecting construction materials and methods, amounts of land to
use, and what future upgrades and rehabilitations require investment (Koo, Ariaratnam, and
Kavazanjian 2009). Sustainability assessment tools are strongly recommended to be incorporated into
existing planning processes, rather than operating as stand-alone, often disconnected, inputs to the
-------
17
decision-making process (Wallbaum, Krank, and Teloh 2011). Sustainability goals that a community
adopts should also be mutually reinforcing since at least some trade-offs must eventually be made
(Tomalty 2009). Typically, only one or two performance measures/indicators are recommended for
each lower-level objective and these measures and indicators should be understandable to a very broad
audience (Ramani et al. 2011).
The various indicators and performance measures that a community adopts may be distinguished from
one another by a variety of different means, serving the purpose of helping to ascertain how they can
best be used. For example, a given indicator may measure either small-scale or large-scale effects and
either temporary conditions or permanent conditions (Koo, Ariaratnam, and Kavazanjian 2009).
Performance measures may also be categorized according to whether they are ordinary quantifiable
measures of the achievement of a subobjective, a composite measure of all of the different aspects of a
given objective, or a qualitative assessment of progress towards a given goal (Black, Paez, and Suthanaya
2002). Finally, certain performance measures may be subjectively deemed to be of greater significance
than others (Koo, Ariaratnam, and Kavazanjian 2009). For example, certain elements of an
unsustainable society (such as climate change, the loss of soil, and the loss of biodiversity) are arguably
of greater impact than others (such as noise and traffic accidents), based upon the duration and extent
of their impacts (Black, Paez, and Suthanaya 2002).
Even though the ideal is to develop indicators and performance measures that relate directly to desired
sustainability outcomes, there are certain indicators that do not automatically represent sustainability
or unsustainability, but may be at the root of what causes other indicator values to be either "good" or
"bad". In systems approaches to transportation planning, which this paper has as a focus, the first
impacts of the transportation system to be modeled are generally those on population and land use
patterns, which are then used as the basis for calculating other impacts (Sastry 1973). Population
density and job density have been shown in studies to explain more of the variation in numerous
sustainability indicators than any other factor (Sinha 2003). Population and employment growth rates
have cumulative effects, a quality which enhances their influence over other indicators (Duthie et al.
2010). Other major determinants of model outputs include trip-generation rates and how people carry
out their transportation mode choices (Duthie et al. 2010). Travel patterns have an effect on other
sustainability performance measures and can easily be used as a proxy for transportation energy use
and emissions (Black, Paez, and Suthanaya 2002). Drivers of individual travelers' transportation
behavior will be discussed shortly.
When performance indicators are applied to the act of transportation and land use planning, they may
be employed not just in conventional modeling activities but also in visioning exercises. The primary
function of most traditional land use and transportation models is to extrapolate existing trends into the
future, with the process usually controlled by technical analysts. Visioning involves projecting several
different future scenarios, usually including both a so-called "ideal" scenario and a do-nothing scenario,
with the process requiring extensive stakeholder involvement. When planning professionals assemble
traditional mathematical models of transportation and land use, they input data on land use inventories,
zoning policies, existing highway networks, employment, and household distribution, typically
-------
18
aggregated into zones. Some of these models are also able to account for future changes in government
policy and in the transportation system, but may still be invalidated by various unexpected events, such
as changes in migration patterns, and by faulty assumptions about the preferred behaviors of various
actors. Meanwhile, visioning processes entail a lengthy public-participation process that first produces a
set of very general guiding principles and then uses those principles to evaluate different development
scenarios, which are often projected as GIS maps for purposes of comparison. A series of specific
strategies can be drafted for the purpose of implementing whichever scenario is deemed to be the most
preferable. Many of the computer programs used for these scenario-building and visioning processes
also incorporate a conventional travel demand model, but usually not a rigorous land-use model.
Ultimately, modeling and visioning serve two different purposes, the former the purpose of projecting
the likely outcomes of historical trends and the latter the purpose of creating an integrated vision of
what direction would be most beneficial for a community (Lemp et al. 2008).
A number of different models are capable of projecting land use and transportation effects in an
integrated fashion (Black, Paez, and Suthanaya 2002). One common type is called a gravity model,
wherein the locations of job centers and households are determined by where the transportation
system will offer the greatest accessibility to other destinations. Another common type of model is a
cellular automata model, in which a community is divided into a series of cells, each affected by their
nearest neighbors. Also of note are input-output models, which are geared towards economic impacts,
large-scale analysis, and travel costs, and discrete response simulations, which are based on predicting
the choices that individuals make. These models require very large quantities of data to work properly,
especially discrete response simulations (Lemp et al. 2008).
Table 1: Types of integrated land use/transportation models (Lemp et al. 2008)
Gravity models
Predicted locations of job centers and households are determined by
where the transportation system will offer the greatest accessibility to
other destinations.
Cellular automata models
A community is divided into a series of cells, each affected by their
nearest neighbors, in order to predict how private-sector actors will
change each cell's land use.
Input-output models
Geared towards large-scale analysis and estimating economic impacts
and travel costs in a given scenario.
Discrete response simulations
Based on predicting the choices that individuals make that affect the
relationship between transportation and land use. Require especially
large quantities of data (though all of these model types have
substantial data requirements).
In addition to transportation and land use models, other types of decision-making tools exist that are
not specific to any particular context, providing flexibility to incorporate the multitude of variables and
causal relationships that are part of sustainability. These can be used, for example, to discover indirect
effects and perform benefit-cost analyses. Two tools that fall into this category are system dynamics
models and Multi-Criteria Decision-Making (MCDM) models. System dynamics models, which are
characterized by causal loop diagrams, got their start in the early 1960s. Originally, this was a tool to
-------
19
anticipate the outcomes of decisions in an industrial setting, although it has since been adapted to many
other applications. In the context of community planning, subsystems are created within the larger
model, typically including macroeconomic factors, regional economic factors, environmental conditions,
and the state of the transportation system. One noteworthy limitation of system dynamics models is
not accounting for how particular indicators vary by geographic coordinates across the area being
studied (Black, Paez, and Suthanaya 2002). Once causal links within a system are characterized, the
mechanisms by which a single policy action has the potential to advance several different goals at the
same time can be more evident (Mercier 2009). An example of a system-dynamics diagram, created
using the freeware version of the computer program Vensim, may be seen in Figure 2, showing inputs to
the amount of automobile travel from a neighborhood's residents. (More such diagrams will appear
later in this synthesis paper, in a section applying the principles of sustainability to an analysis of
"Complete Streets" design principles for transportation corridors.)
<# of jobs w/in 30-minute
peak-period drive of
neighborhood>

<% of dwelling units that
come with parking?*



# o^personal veh.s parked
off-propeity at home end of
tips in neighborhood



neighborhood of one hr. of
laboi>
car-only-pers on-miles /
resident/day

car popularity-
w/ car option
¦
<$ spent on car trips per
resident w/ car option per
day>

<% ofpopulation
that is children?*
Passengers
per car
car-trip
length
person-miles of
car
travel/re side nt/day
VM T/reside nt/day
Figure 2: Segment of a system-dynamics diagram showing inputs to neighborhood residents' Vehicle Miles Traveled (VMT).
Red arrows represent negative causal relationships and blue arrows represent positive causal relationships. A negative
causal relationship means that an increase in value or quantity of one component leads to a decrease in value or quantity of
the other component (and a decrease in value or quantity of the first component leads to an increase in value or quantity of
the second component). A positive causal relationship means that an increase in value or quantity of one component leads
to an increase in value or quantity of the other component (and a decrease in value or quantity of the first component leads
to a decrease in value or quantity of the second component).
Meanwhile, MCDM models operate by assigning weights to a number of different variables and then
adding them together to form a single composite measure (Koo, Ariaratnam, and Kavazanjian 2009).
Some common forms of MCDMs include the Multi-Attribute Utility Theory (MAUT), the analytical
hierarchy process, and the outranking method. Under MAUT, the process starts with identifying a list of
quantifiable measures, ranking them by importance, assigning them values between zero and one that
are in keeping with their respective ranks, and identifying the best and worst possible values for each
measure. The actual values of each measure under a given scenario are transformed into a proportion
of the difference between the theoretical worst and best possible values. These proportions are
multiplied by the respective values between zero and one assigned to each measure as its weight earlier
-------
20
in the process. When the results of these calculations are added together, they produce a weighted
composite measure of all of the variables under consideration, on a scale of zero to one (Ramani et al.
2011).
There are correlations identified among transportation-related variables that are simultaneously
important for transportation, land use, and various other decision sectors (Sinha 2003):
•	Population density and job density, known together as activity density, are positively correlated
with transit boardings per person per year.
•	Meanwhile, transit demand per capita is negatively correlated with:
o the provision of roads,
o the rates of ownership and use of private vehicles,
o and the number of parking spaces per employee in central business districts.
•	Transportation energy consumed per person is:
o negatively correlated with activity density and
o positively correlated with both the demand for personal motor vehicles and the
provision of roadways.
•	The percentage of work trips that people make by public transit is negatively correlated with:
o carbon-dioxide emissions from transportation,
o transportation-accident fatalities per capita,
o the amount of investment that is made in roadways,
o and the price of gasoline.
Of all the variables described above, population and job density are the ones with the greatest impact
on the others, and changes in these densities produce especially large effects in the other indicators
when the starting densities are particularly low (Sinha 2003). Subsequent sections of this paper will
discuss these and other correlations, as well as the causal relationships that are theorized to explain
them, grouped according to which dimension of sustainability they most closely relate to.
3.1.1 Limitations
No matter what system for creating and using sustainable transportation performance measures is
implemented, certain limitations will be inherent. For example, sustainability has many elements with
many ways of assessing each of those elements, so judgments must be made regarding which elements
to include and regarding the best assessment methods for the elements. In addition, many of the
decisions involved in creating a system of sustainability performance measurements may be subjective
in nature, such as deciding how to quantify an indicator when extremely high values may be sustainable
in some contexts while extremely low values may be sustainable in other contexts (El-Diraby, Abdulhai,
and Pramod 2005). A further challenge can be when the costs of a transportation project, especially the
financial and economic costs, are much easier to calculate than the benefits (Sastry 1973). Beyond that,
many performance measures are either difficult to obtain data for or simply too politically sensitive to
be put to effective use (Ramani et al. 2011). When analyzing performance measures together, capturing
multiple dimensions of the situation in a single variable can be difficult. However, the analysis of
-------
21
multiple variables can become confused by correlations among them (Cervero and Duncan 2003). Some
limitations to the creation of effective sustainable transportation performance measurements can be
remedied through expanded research efforts. Some examples include:
•	insufficient research on the measurement of sustainability in the regular functions of a
transportation agency (Ramani et al. 2011),
•	uncertainty in the inputs of transportation and land use models (Duthie et al. 2010), and
•	insufficient research on how the built environment affects pedestrian and bicycle travel, as
opposed to motorized transportation (Cervero and Duncan 2003, Wong, Faulkner, and Buliung
2011).
Meanwhile, when performance measurements are used in visioning processes, the issue of feasibility is
often neglected in the creation of preferred scenarios, and the attendant stakeholder-involvement
process may last so long that new events in the community render the scenarios under development
obsolete. Conversely, transportation and land use models frequently have the problem of requiring very
large and detailed datasets and considerable expertise to work properly (Lemp et al. 2008). In addition,
many of the software packages that are used to analyze sustainable transportation performance
measures lack the ability to exchange data with one another (El-Diraby, Abdulhai, and Pramod 2005).
3.1.2 Specific Models, Tools, and Other Resources from the U.S. Environmental Protection Agency
(EPA)
•	The Green Communities Program helps communities to better understand sustainable
development by introducing them to a basic planning process and then assisting them to
implement that process by providing access to a vast array of tools and information.
Encouraged to look hoiistically at their current situations, planners are asked to envision their
community as it is (Step 1: Community Assessment), as it will be in the future if no action is
taken (Step 2: Trends Analysis), and as it could be if the appropriate actions were taken (Step 3:
Visioning Process). Communities are then asked to develop an action plan (Step 4) that will help
them realize their sustainable development goals. Implementation of the action plan is the final
step (Step 5). A Green Community uses and encourages modes of travel other than the
automobile; street and circulation patterns encourage pedestrian movement, and efficient
transportation systems maximize accessibility and the movement of people and goods,
providing economic benefits. Increased opportunities for the use of pedestrian and bicycling
facilities help to create a sense of community, reduce car-miles driven, and contribute to
general well-being and a good quality of life. The Green Communities planning process and
web-based toolkit encourage communities to consider various transportation issues, including
the provision of public transit, as a part of their sustainable development planning processes.
The Green Communities website, http://www.epa.gov/greenkit/index.htm, includes sample
transportation action plans and information on action plan indicators (U.S. Environmental
Protection Agency).
•	EnviroAtlas is an interactive, web-based decision-support mapping tool designed to provide
information about ecosystems and their benefits to society. EnviroAtlas also addresses point-
-------
22
and nonpoint-source pollution, landcover conversion and fragmentation, resource restoration,
and shifts in demographics and natural hazards, all of which can drive changes in the production
of and demand for these benefits. Decisions about resource use and environmental policy are
often made with an incomplete understanding of the interactions between human activities and
beneficial ecosystem processes and functions. EnviroAtlas provides users with maps, analytical
tools, and other information for interpreting the distribution of ecosystem services across the
conterminous United States, and for understanding how they can be conserved and enhanced in
the future. In the context of transportation projects, EnviroAtlas can help guide where best to
preserve, restore, or construct ecosystems for maximum public benefit. Data in EnviroAtlas may
be used to inform methods to value and pay for ecosystem services, which may also influence
restoration decisions and project locations. EnviroAtlas presents data at national and
community scales. The national component summarizes data by 12-digit hydrologic unit codes
(HUCs; approximately 90,000 in the conterminous U.S.). The community component
summarizes data by census block group for selected cities and towns, with some data (like
vegetative cover) at a finer scale. Both components also include pixel-based and other spatially
detailed maps; all of these data are available via web services and direct download. EnviroAtlas
is designed for staff from all levels of government, environmental and public health
professionals, researchers, educators, nongovernmental organizations, and anyone else with an
interest in ecosystem services and their role in supporting sustainable and healthy communities.
It does not require special software, technical expertise, or a scientific background. In beta
release through December, 2013, EnviroAtlas is accessible via
http://www.epa.gov/research/enviroatlas/index.htm. Part of EnviroAtlas is the Eco-Health
Relationship Browser, which makes information on health and well-being benefits of ecosystems
easily accessible. The Browser is already available online at
http://www.epa.gov/research/healthscience/browser/introduction.html.
•	The U.S. Environmental Protection Agency has developed a Database of Sustainability Indicators
and Indices (DOSII), as well as a corresponding framework document on the selection and use of
sustainability indicators (U.S. Environmental Protection Agency 2012c). DOSII is a searchable
inventory of peer-reviewed sustainability indicators classified into a single taxonomy system,
supporting decision making by providing candidate indicators (and indices) relevant to specific
sustainability-related interests (e.g., air, water, energy, communities, transportation, etc.). In
addition, an interactive web-based tool (e-DOSII) is being developed to extend DOSII's search
capabilities to communities and to provide information through a user-friendly web interface.
While the target audience of this work is EPA personnel, external organizations interested in
measuring some aspect of sustainability have found it useful, too. The current version of DOSII
and the related framework document are presently available, and e-DOSII is scheduled for
delivery in October 2014.
•	Not yet publicly available, Decision Analysis for a Sustainable Environment, Economy, and
Society (DASEES) is a web-based decision-support framework intended to help community
stakeholders arrive at sustainable courses of action. DASEES is organized around the five steps
of 1) establishing the context for a decision, 2) determining what objectives are meant to be
achieved, 3) listing various options for accomplishing those objectives in the given context, 4)
-------
23
evaluating the relative desirability of the listed options, and 5) implementing the best available
option. For each of these steps, DASEES provides various analytical tools, allowing stakeholders
in a given community to customize their modeling efforts to fit the unique circumstances
surrounding a particular decision. The tools in DASEES are able to combine disparate outcome
metrics into one composite measure, whose inputs are weighted both according to a given
community's priorities and according to the inverse of how much uncertainty is attached to
each measurement or calculation. Outputs may take the form of tables, charts, or GIS maps.
Along with the actual tools, DASEES provides case studies and other guidance on the use of the
framework (Stockton et al. 2011).
•	The Community-Focused Exposure and Risk Screening Tool (C-FERST) is used for assessing
cumulative public health effects in a given community from a variety of stressors, based on
location-specific data, so that problems and solutions can be prioritized. C-FERST was created
because it can be very difficult for a community to draw the necessary linkages between the
presence of a given stressor, the degree of human exposure to that stressor, and the various
health impacts resulting from that exposure, and because it is also difficult to determine what
effects various stressors are likely to produce in combination. Because C-FERST's outputs
include GIS maps of these cumulative effects, the model is useful for addressing issues of
environmental justice. In the current version of C-FERST, which is still being pilot-tested, only
chemical stressors are considered. However, future versions are expected to incorporate a wide
variety of both chemical and nonchemical stressors, consequently making C-FERST relevant to
far more than just human health outcomes. Future versions of C-FERST are also expected to
predict what the outcomes would be of various hypothetical actions that a community might
take, integrate with other decision-support tools, and possess various other enhancements. In
general, C-FERST is intended for use by community leaders and is not well-suited for on-the-fly
assessments by ordinary residents (U.S. Environmental Protection Agency 2013a).
•	The Tribal-Focused Environmental Risk and Sustainability Tool (Tribal-FERST), still being pilot-
tested and enhanced, is an Internet-based decision-support tool specifically designed for use by
tribal communities to address the unique sustainability issues that they face, especially those in
the areas of human health and ecology. Tribal-FERST will be set up to provide step-by-step
guidance for determining a priority order in which to address various problems and risks and for
assessing the results of different actions. At each of the steps in this guidance, users will be
provided with scientific information relevant to their specific situation, links to other tools that
might be of help to them, and the ability to create overlay maps of many different datasets (U.S.
Environmental Protection Agency 2013d).
•	The Tool for the Reduction and Assessment of Chemical and other environmental Impacts
(TRACI) allows the characterization of the following impact categories: ozone depletion, global
climate change, acidification, eutrophication, smog formation, human health, ecotoxicity, and
fossil fuel depletion (Bare 2012, Bare 2011, Bare et al. 2002). Midpoint-level characterizations
(e.g., ozone depletion potentials (ODPs) and global warming potentials (GWPs)) allow minimal
incorporation of value choices and assumptions while also allowing the most comprehensive
inclusion of endpoint impacts (Bare et al. 2000). When combined with local data or life-cycle
-------
24
inventory data (e.g., air, water, and soil emissions), TRACI can be used to compare the relative
impacts of two or more options, as a way of supporting sustainability decisions. For example,
scenarios could be developed for two alternative transportation-related decisions to determine
which of the options would be expected to have a less negative environmental impact in one or
all of the above impact categories.
•	The Final Ecosystem Goods and Services Classification System (FEGS-CS) is a database currently
under development by the EPA that consists of quantifiable metrics that describe the ultimate,
tangible ways in which the natural environment hypothetically affects the lives of humans and
which is meant to be used by both public- and private-sector actors. While one might deem a
given ecological feature to be generally beneficial, the exact character and magnitude of its
impacts are context-dependent. By including these contextual considerations, the metrics in
FEGS-CS may serve as inputs to an overall accounting of the positive and negative outcomes of a
given action. As a result, it may also be possible to determine what ecologically beneficial
actions would be sufficient to compensate for the effects of a given action that is ecologically
detrimental. Classifying ecosystem services in terms of their final effects on humans also has
the advantage of linking the natural and social sciences and the advantage of avoiding the
double-counting of benefits (U.S. Environmental Protection Agency 2012b, Landers and Nahlik
2013).
•	The National Ecosystem Services Classification System (NESCS) is being developed concurrently
with the FEGS-CS. NESCS will link the theoretical constructs of ecology and economics in order
to create an exhaustive accounting of all of the known pathways by which humans derive value
from ecosystems. To date, most economic analyses have been blind to those things to which a
concrete price cannot be attached and most ecological analyses have been blind to the ways in
which humans value ecosystem goods and services. NESCS will use FEGS-CS to define and code
all of the conceivable inputs to the accounting, ensuring that direct and indirect effects are not
confused with one another and that drivers whose influences are distinct from one another are
not counted together. By mapping mutually exclusive pathways, NESCS will ensure that all
effects are counted and that none of them are double-counted. The resulting algorithms may
be used to calculate the value to humans of any given natural area, contributing to decisions on
preservation, mitigation, and restoration priorities.
•	Also currently under development at the EPA is the Ecological Production Function Library (EPF-
L). This will be a searchable database describing mathematical functions that various research
efforts have shown to link the inputs and outputs of a given ecological process. The EPF-L is
meant as a resource for the developers of decision-support tools, guiding them to equations
that they may decide to use to represent individual causal relationships within their models (U.S.
Environmental Protection Agency 2012a). This resource is expected to be finished in 2014 and is
expected to be available over the Internet in 2015.
•	One commonly-cited element of sustainability is the health of wildlife populations, especially
"charismatic mega-fauna" such as songbirds, raptors, anadromous fish, large carnivores, and so
on. Wildlife populations often respond dramatically to human activities, especially when they
are impacted by multiple natural and human-derived disturbances. A key activity affecting
-------
25
wildlife populations across multiple taxa is the development of transportation networks, which
can result in habitat loss, habitat fragmentation, and a loss of landscape connectivity. Roads can
also be a source of significant direct mortality due to collisions with vehicles. Range shifts
resulting from climate change, or modifications to species' movement patterns resulting from
human disturbance or invasive species, may also increase the impacts that transportation
networks have on wildlife. Forecasting what impacts human activities might have on wildlife in
the future represents a significant research challenge. The few "off the shelf" analytic tools
available for human-wildlife impact studies are unable to simulate complex stressor
interactions. To fill this computational void, the EPA created the HexSim model. HexSim is a
general simulation framework within which plant and animal simulation models are
constructed, designed specifically to meet the type of forecasting problems mentioned above. A
wide range of plant and animal models can be developed within HexSim, and any number and
type of landscape structures or disturbance regimes can be represented. HexSim is particularly
well-suited for capturing stressor interactions and other "real world" complexities, and for
generating map-based outputs that appeal to stakeholders. HexSim simulations can range from
extremely simple (with few parameters) to highly detailed and complex (with many
parameters). These sorts of design decisions are made by users, and based on data availability,
the nature of the questions being asked, and other concerns. HexSim has been reviewed by EPA
and cleared for release, forming the basis of multiple research projects and peer-reviewed
publications (Schumaker).
3.1.3 Specific Models, Tools, and Other Resources (Non-EPA)
•	An important tool supported by the Federal Highway Administration (FHWA) is the
Infrastructure Voluntary Evaluation Sustainability Tool (INVEST). Version 1.0 of this tool was
released in October 2012. INVEST is free and available for use by the general public. The
program is designed as a scorecard that includes criteria for evaluating environmental,
economic, and social impacts of transportation projects at a variety of scales and at various
stages throughout the planning and implementation processes, including system planning,
project development, and operations and maintenance. INVEST primarily focuses on highway
projects (U.S. Department of Transportation).
•	The Leadership in Energy and Environmental Design (LEED) rating system, a product of the
United States Green Building Council (USGBC), is a comprehensive, flexible, and very widely
used tool for rating the environmental sustainability of the design, construction, and operation
of individual buildings and neighborhoods, although not geared specifically to transportation
planning (U.S. Green Building Council). Foreign counterparts to LEED include Canada's
SBTool07, which includes economic and societal indicators, and the United Kingdom's Building
Research Establishment Environmental Assessment Method (BREEAM), which, much like LEED in
the United States, primarily focuses on the natural environment and energy consumption
(Wallbaum, Krank, and Teloh 2011). The portion of LEED most relevant to sustainable
transportation is the LEED for Neighborhood Development (LEED-ND) rating system, which is
applied to geographic areas within cities, rather than individual structures. Among other things,
this rating system gives points for neighborhood design features that are likely to reduce vehicle
-------
26
miles traveled (VMT) and increase nonmotorized transportation. The following are some
transportation-related features that LEED-ND awards points for (U.S. Green Building Council,
Congress for the New Urbanism, and Natural Resources Defense Council 2012):
o Development activities that take place near public transit stops at which transit vehicles
stop sufficiently frequently,
o Developments that are near existing developed areas, such that traveling to those areas
will entail shorter trips than would otherwise be the case,
o Highly-connected, grid-like transportation corridors that form small city blocks,
o Adequate transportation facilities for both motorized and nonmotorized modes of
travel.
o Close proximities between housing units and employment locations,
o Neighborhood centers with many different kinds of destinations in close proximity,
o Short distances between building entrances and transportation corridors,
o Dense and/or compact development patterns that place more destinations within a
shorter trip distance.
o Pedestrian and bicycle facilities that are safe, appealing, and comfortable to use.
o Small quantities of land dedicated to motor vehicle parking.
o Transportation Demand Management (TDM) programs that offer financial incentives for
people to drive less.
o Public schools that are small, numerous, and located in close proximity to the homes of
a large percentage of the students who attend them.
•	The GreenLITES programs (from the New York State Department of Transportation) and the
Greenroads rating system have their primary focus on issues of civil engineering, construction,
and maintenance. These rating systems are based on the LEED rating system (Ramani et al.
2011).
•	A key issue in transportation planning is the impact that transportation and land use have on
one another and how they collectively impact the environment. The Transportation Research
Board has developed a decision support tool called Transportation for Communities: Advancing
Projects through Partnerships (TCAPP), available at www.transportationforcommunities.com.
TCAPP has a broad focus, but it in part zooms in on the linkage between transportation and land
use. It provides examples of integrating transportation and land-use decisions from Maryland,
Oregon, New York State, the Thurston region in Washington State, and the Sacramento area.
Maryland and New York are given as examples of states that promote smart growth in their
transportation decisions. Oregon uses GreenSTEP, a modeling tool that calculates measures of
such things as household travel and vehicle emissions and can be used to evaluate
environmental effects of land use patterns and transportation services at a fine level of detail
(http://www.oregon.gov/ODOT/TD/OSTI/docs/media/model.pdf). The Thurston Regional
Planning Council, in addition to its various policies promoting smart growth, uses Community Viz
to develop and compare scenarios. Community Viz and other geospatial visioning tools are
powerful and flexible planning models that can calculate environmental impacts and a variety of
other outcomes resulting from given land use and transportation scenarios. Tools such as
Community Viz can be linked to external models, including travel demand models. An array of
-------
27
other tools and methods that are more broadly applicable to transportation and ecological
decision-making are listed and briefly described in TCAPP, in its section on the Integrated
Ecological Framework, a process for accounting for ecological values early in the transportation
planning process.
•	The System for Planning and Research in Towns and Cities for Urban Sustainability (SPARTACUS)
indicator system is a four-step forecasting model that involves identifying land use,
transportation, and pricing policy elements and then testing scenarios for different
combinations of packaged policies to determine their long-term sustainability. Included within
this modeling system is MEPLAN, a land use/transportation interaction model. Also included are
the ability to analyze data at a very fine geographic scale using GIS and a decision-support tool
that allows the user to create weighted composite indicators of sustainability (Black, Paez, and
Suthanaya 2002).
•	The Integrated Transportation Gravity-based Land Use Model (ITGLUM) is a very simple land
use/transportation gravity-based model (Duthie et al. 2010).
•	Slope, Land-use, Exclusion, Urban extent, Transportation, and Hill shade (SLEUTH) is an example
of an integrated transportation and land use model that uses the cellular automata approach of
defining cells within a community and then determining how each of those cells is affected by its
immediate neighbors (Lemp et al. 2008).
•	TRANUS and the Random-Utility-Based Multiregional Input-Output (RUBMRIO) model are land
use/transportation models that use an input-output model to produce results primarily focused
on forecasting economic and trade effects. (Lemp et al. 2008).
•	UrbanSim is an open source modeling program that users can expand to meet their needs. This
model is dynamic, highly disaggregated in its scale of analysis, and accounts for the behavior of
households, private companies, real estate developers, and public officials (Lemp et al. 2008).
•	The Transportation, Economic, and Land Use Model (TELUM) is a free program from the FHWA
that makes use of a gravity-based model, wherein data is geographically aggregated into zones
(Lemp et al. 2008).
•	The California Urban Futures models (CUFI and CUFII) are discrete response simulations that are
meant to carry out GIS analyses of large metropolitan areas, with CUFI placing particular
importance on real estate profitability as a determinant of development patterns (Lemp et al.
2008).
•	The Comprehensive Econometric Micro-Simulator for Daily Activity-Travel Patterns and the Mid-
Ohio Regional Planning Commission's activity-based travel-demand model are two more
programs relevant to transportation and land use applications (Lemp et al. 2008).
•	Some examples of available GIS scenario-building/visioning tools include Community Viz
(already mentioned), California's PLACE3S, and the Charlottesville, Virginia region's CorPlan
(Lemp et al. 2008).
•	Some land-use models have the potential to serve as scenario-planning tools. The program
"What if?" provides functions for its users to assign areas of land weights according to their
relative importance, along with ratings of their suitability for development and how easy to
convert their land use is. The program UPlan allows its users to specify attractiveness criteria for
-------
28
land areas that represent the land's value and accessibility. In both of these programs, the
object of the user-inputted ratings and criteria is to allocate demand across a geographic area
(Lemp et al. 2008).
3.2 Drivers of Transportation Behavior
Although there are some sustainability outcomes that a piece of transportation infrastructure will affect
regardless of whether and how people use it, one of the most important questions in ascertaining the
sustainability-related implications of a transportation policy or investment is what impact it will have on
people's transportation-related behavior. This includes choices regarding transportation modes, trip
frequencies, trip lengths, and trip purposes. The purpose of this section is to synthesize what is
currently known about the factors that influence travelers' choices in these matters. Factors to be
considered here include characteristics of the transportation system, governmental policies related to
transportation, and drivers of behavior that a community's transportation decision-makers can only
influence indirectly, such as travelers' attitudes.
The most heavily researched topic in all of urban planning is the effect of the built environment (i.e.,
land use patterns and physical infrastructure) on transportation demand and transportation behavior
(Ewing and Cervero 2010). However, in spite of all of this research, causal linkages between the built
environment and transportation behavior have been difficult to establish and quantify, leaving
knowledge gaps (Frank et al. 2008, Scheiner and Holz-Rau 2013). One reason for this difficulty is the
highly interdependent nature of the relationship between transportation systems and land use patterns.
First, changes to the size and nature of the transportation system alter the amount of accessibility that
people have to various destinations. Then, this change in accessibility prompts the development
community to alter land use patterns in such a way as to make the most profitable use of the
transportation system, which also alters the amount of accessibility that people have to various types of
destinations. All of these changes in accessibility affect the frequency, length, duration, mode, timing,
destination, and purpose of the trips that people choose to make. Finally, these changes in travel
behavior affect the balance between transportation demand and transportation supply and the
distribution of travel demand across the transportation network, which prompts transportation planners
to make additional changes to the system (Kitamura 2009, Szeto, Jaber, and O'Mahony 2010). More on
the relationship between land use and travel behavior may be found in the concurrent SHCRP Theme 4
synthesis paper on land use.
Ewing and Cervero (2010) sort the various aspects of the built environment that are relevant to
transportation behavior into a series of broad, overlapping categories that they refer to as the "Five Ds":
Density (of population, employment, dwelling units, commercial floor space, etc.), Diversity (of land
uses), Design (of the transportation network), Destination Accessibility, and Distance to Transit. Ewing
and Cervero regard these built-environment variables as affecting travel behavior alongside the other
relevant factors of Demand Management (which includes, but is not limited to, the supply and cost of
parking) and Demographics. In the same paper, they synthesize 62 separate studies of the relationship
between the built environment and travel behavior. This synthesis includes tables of average elasticity
values from the pooled samples of the various studies analyzed (each elasticity value describing the
-------
29
strength of the relationships between various aspects of the built environment and people's travel
behavior, where a 1% change in variable A is predicted to produce an X% change in variable B), as well
as an appendix containing elasticity estimates from the original studies. If a transportation planner finds
a study in the appendix that had good methodology and which is from a geographic location sufficiently
close to the one they serve, the planner could potentially borrow an elasticity value directly from that
study for use in their own calculations. The average elasticities from the pooled samples, meanwhile,
could potentially be used in various sketch-planning situations to estimate amounts of driving, walking,
and transit use that may result from a given act of government or development project, relative to a
base scenario. An important caveat to the use of this resource, however, is that the average elasticities
contained in it have not been tested for statistical significance (Ewing and Cervero 2010, Gim 2012).
The most-studied of all of the characteristics of the built environment that affect travel behavior is
density. The reasons for this particular research focus are that density is a relatively easy variable to
measure, a relatively easy driver to influence through policy decisions, and highly correlated with many
other built-environment characteristics that tend to change travel behavior in similar ways (Gim 2012).
However, when other built-environment characteristics are controlled for, the influences of population
density and employment density on travel behavior are actually quite modest. Among the significant
built-environment characteristics that are correlated with high population and employment density in a
neighborhood are highly-mixed land uses, small city blocks, and a location near the center of a
metropolitan area, all of which are associated with fewer Vehicle Miles Traveled (VMT) and greater
mode shares for walking, biking, and public transit (Ewing and Cervero 2010). Even though the
association between density and travel behavior is based to a large degree on correlations, it may still
serve as a useful tool for perceiving important relationships. For instance, one analysis found that the
smaller an area's population density, the greater the change in travel behavior would be from a given
change in density (Sinha 2003). However, another analysis found that in European cities, which are
generally denser than U.S. cities, the effect of density and urban form on travel behavior is significantly
stronger than in U.S. cities, based upon a review of the conclusions of travel-behavior studies from each
of those two regions (Gim 2012).
Another much-studied aspect of the built environment in transportation-behavior research is land-use
diversity, which consists of both the number of different land uses within a given area and the relative
amounts of each of those uses that are present. When this aspect of the built environment is studied,
the most commonly used measures are entropy metrics, wherein a scale is established where a low
number represents a single land use within an area and a high number represents a great variety of uses
within that area. However, the less-used measure of the ratio between jobs and housing in an area (or
jobs and population) has a stronger relationship with nonmotorized transportation than entropy
measures do (when the number of jobs in an area and the number of working-age residents of that area
are similar to one another, it increases the opportunities for easy commuting without a motor vehicle)
(Ewing and Cervero 2010). Research to date has given little attention to the travel-behavior effects of
the land uses that surround a person's place of work, as opposed to those that surround their place of
residence (Frank et al. 2008).
-------
30
The effect of the design of the transportation network on travel behavior is especially difficult to
measure, because it represents the collective impact of many different types of design features. For
example, aspects of transportation network design include the size of city blocks, the number of
intersections per square mile, the proportion of intersections that are four-way intersections, the
proportion of the street network that has sidewalks, the width of the streets, the distances that
buildings are set back from the curb, the number and quality of marked pedestrian-crossing locations,
the distinction between a grid-based network and a curvilinear network, and many other features.
Consequently, various composite measures are sometimes created that combine design features that
are known to affect travel behavior in similar ways (Ewing and Cervero 2010).
One common way of measuring access to public transit within a given area is to calculate the shortest
distance that someone would have to travel along the transportation network to reach a transit stop
from each residence or job location within the area and take the average of those figures. Other
measures include transit-route density, transit-stop density, and the average distance between transit
stops. Not surprisingly, there exists a strong relationship between how close someone lives to a public
transit stop and how likely they are to use public transit, a fact which reinforces the common practice of
transit agencies of trying to run buses within a quarter mile of most of the households in their service
areas (Ewing and Cervero 2010).
It is to get the best possible use out of an area's transit accessibility that there exists the practice of
Transit-Oriented Development (TOD). In this practice, a tight cluster of mixed-use development is
created around a public transit station, such that people who either live there or work there may be
within easy walking distance of both the transit station and most of the businesses and amenities within
the development. Usually, these developments that are built to be conducive to transit use are
centered around a rail station, but they may sometimes be centered around a bus stop. Most of the
existing literature on Transit-Oriented Development is focused on the policy tools used by government
agencies in their creation, as opposed to addressing design concerns such as ascertaining the
appropriate amount of travel by each non-transit mode of transportation to accommodate (Jacobson
and Forsyth 2008).
Of all of the categories of built-environment characteristics identified by Ewing and Cervero (2010),
Destination Accessibility has the strongest influence on travel behavior. One may consider accessibility
at either a local scale (for example, the distance from a person's home to the nearest store) or a regional
scale (for example, the distance to the nearest central business district or the number of jobs or
attractions located within a given travel time of a given location). Most of all, having greater access to a
greater number of destinations is highly associated with lower VMT per capita. At the core of an urban
area, there are more destinations that can be easily reached without the use of a personal motor
vehicle, and even when a personal motor vehicle is used, the necessary trip lengths are shorter. Most
developed areas near the center of a city produce less VMT per capita than a pedestrian-oriented,
compact, mixed-use development in a remote location. VMT is also reduced (to a lesser degree) by high
intersection density and street connectivity, while a very small dampening influence is exerted on VMT
by population density and job density, with the influence of population density being stronger than the
-------
31
influence of job density. Meanwhile, the built-environment characteristics that have the strongest
(positive) relationship with travel by foot are high intersection densities, a balanced ratio between jobs
and dwelling units, and short distances between homes and stores. There is evidence of a stronger
relationship between intersection density and walking than between street connectivity and walking;
one possible explanation is that a given study area might have highly connected streets, but also very
long city blocks that inhibit walkability for people who must travel around them. The transportation
network design elements of intersection density and street connectivity also constitute the second tier
of major built-environment drivers of transit use (after Distance to Transit), enabling less circuitous
travel to and from transit stops and allowing transit providers to consider a greater number of potential
transit vehicle routes. The third tier of built-environment drivers of transit use consists of measures of
land-use mixture, possibly due to sufficiently mixed land uses enabling the running of errands on the
way to or from a transit stop.
3.2.1 Correlations and Elasticities Regarding the Built Environment and Travel Behavior
The following are some numerically-expressed relationships between the built environment and travel
behavior that various studies have quantified:
•	In a study of National Ambient Air Quality Standard (NAAQS) Nonattainment Metropolitan
Statistical Areas (NMSAs), researchers estimated that if all households in NAAQS NMSAs were
located within 0.1 miles of a transit stop, it would reduce the number of private motor vehicles
by 9% and reduce VMT by 11%. The authors estimated that this is similar to the effect that
would be produced by a 50% increase in gasoline prices (Kim and Kim 2004).
•	In a synthesis of the results of many different studies of built-environment effects on travel
behavior, Ewing and Cervero (2010) arrived at an elasticity between VMT and employment
accessibility by automobile of -0.20. A 10% increase in employment accessibility by automobile
would be associated with a 2% reduction in VMT. This was almost as strong as the combined
elasticities with VMT found to exist for density, land use diversity, and transportation network
design. In a different synthesis of built-environment-and-travel-behavior studies from 2001, the
same authors stated that doubling a neighborhood's density would cause both VMT per capita
and vehicle trips per capita to go down by 5%.
•	A 10% increase in roadway capacity, measured in lane-miles, is associated with an increase in
VMT of between 5% and 10% (Sinha 2003).
•	The amount of road length per capita in different areas explains 70% of the variability in private
motor vehicle ownership rates and 76% of the variability in private motor vehicle use rates
(Sinha 2003).
•	High population densities are associated with a relatively high ratio between a transit agency's
revenue from fare collection and that same agency's operating expenses. Population density
explains 62% of the variation in these fare recovery ratios (Sinha 2003).
•	A change in the ratio of the cost of private-mode transportation per mile traveled to the cost of
public transit use per mile traveled is associated with an opposite-direction change in the
percentage of work trips that are made by transit, a relationship which explains 32% of the
variation in the latter variable (Sinha 2003).
-------
32
•	Activity density (calculated as the sum of population density and employment density) explains
88% of the variation between different areas in how much they require new road capacity
(Sinha 2003).
•	Transit demand per person is negatively correlated with the amount of parking per employee in
a metropolitan area's Central Business District (CBD), automobile ownership, automobile use,
and road provision. The percentage of work trips made by public transit is negatively correlated
with the amount of road investment in an area and with fuel prices (Sinha 2003).
•	Job density and population density are both significantly correlated with public transit boardings
per person per year (Sinha 2003).
•	The Land Use, Transportation, Air Quality, and Health (LUTAQH) Study in King County, WA found
the following relationships (Frank et al. 2008):
o The biggest determining factor in people's travel choices is the amount of time that they
spend traveling, indicating the significance of both traffic congestion and time spent
waiting to transfer between transit vehicles,
o A 5% increase in residential density, the connectivity of the street network, land use
mixture, and the ratio of retail floor area to land area used for retail would be
associated with a 32% increase in walking,
o Each additional motor vehicle that a household owns decreases the odds of people in
that household using public transit by 42%.
o A quarter-mile increase in the average distance from a person's home to the nearest
transit stop would be associated with a 16% decrease in transit use.
o A quarter-mile increase in the average distance from a person's workplace to the
nearest transit stop would be associated with a 32% decrease in transit use.
o Each additional institution or recreational facility within a one-kilometer walk of a
person's home increases their odds of walking by 20%.
3.2.2	Parking
There exists a shortage of available data on the effect of the parking supply on people's transportation
mode choices. However, there is enough evidence to determine that the dampening effect of scarce or
expensive parking on automobile use is significant. Travelers perceive one additional minute spent
walking to or from a parking space as a greater burden than one additional minute spent driving (Frank
et al. 2008).
3.2.3	Neighborhood-Scale Travel versus Regional Travel
Individual neighborhoods and other geographic subdivisions of a large study area are bound to have
unique characteristics that influence travel behavior within them. Therefore, consideration should be
given to both local drivers of short-trip behavior and regional drivers of long-trip behavior, bearing in
mind that the regional transportation environment is influenced by the various neighborhood-scale
transportation environments that are contained within it (Black, Paez, and Suthanaya 2002). When a
person habitually uses a particular mode of transportation for long trips, the odds of them also using
that same mode of transportation for short trips increases (or vice versa). For example, people who
-------
33
possess bus passes (indicating that they are probably frequent transit users) are more likely than other
people both to use public transit for long trips and to use public transit for short trips (Kim and Ulfarsson
2008). However, substantial differences between people's favored transportation modes for trips
within their home neighborhoods and trips within a larger region (urban or otherwise) also commonly
exist. In a study from the borough of the Bronx in New York City, an increase in trips that last less than
30 minutes was associated with more automobile travel and an increase in trips that last between 30
minutes and 90 minutes was associated with more travel by public transit. The authors of the study
explained these associations by observing that, in New York City, long distance car trips are discouraged
by a combination of traffic congestion, scarce parking, and expensive tolls, while the financial cost of
using transit is unaffected by distance and the average speed of a short transit itinerary will tend to be
lower than the average speed of a long transit itinerary on account of a large percentage of a short
itinerary's duration being spent on travel to and from transit stops (by other modes) and waiting for
transit vehicles to arrive. Consequently, the authors concluded that an increase in the number of jobs in
the Bronx would increase automobile usage by the borough's residents and increase the number of
people commuting into the Bronx by way of that particular region's most-favored transportation mode
for long-distance commutes (i.e., public transit) (Berechman and Paaswell 1997).
The case of such transit-heavy locations as New York City notwithstanding, most people's favored mode
of transportation for long trips across a region is the private automobile. Because of low-density
development patterns and the speed and flexibility of private motor vehicles, altering the built
environment in such a way as to substantially reduce private automobile use on long trips would be a
very difficult undertaking (but one generally regarded as being helpful in the advancement of various
sustainability goals). However, the conversion of short trips from motorized modes to nonmotorized
modes could likely be achieved with greater ease. Among other things, the promotion of short,
nonmotorized trips could reduce traffic congestion, since a large proportion of the automobile trips that
short, nonmotorized trips would be capable of replacing are non-work trips carried out during rush hour,
such as to run errands, drop off children, or engage in recreational activities. Nonetheless, drivers of
travel behavior on short, non-work trips are currently little-studied, especially in the United States (Kim
and Ulfarsson 2008).
The largest driver of people's mode choices for short trips is the purpose for which the trip is being
undertaken. When people choose to use an automobile for a short trip in spite of the presence of the
necessary infrastructure to make that trip efficiently by a different mode, one of the most common
reasons for that choice is the need to transport heavy cargo that would be difficult to carry under their
own power. Other common reasons include the transportation of passengers, the need to save time,
the knowledge that one will require an automobile for some later trip that will start from the endpoint
of the current trip, physical infirmity, and simply using a car for the reason that it is available. According
to studies from the United States and Europe, among non-automobile modes of transportation used for
short trips (defined variously as those trips with a maximum distance of between one and two miles),
walking is more common than either public transit or cycling, which are rarely used for such trips.
People are reluctant to use transit for short trips because it would require coordinating their actions
with a transit schedule that might not provide much service during off-peak times of day and because
-------
34
time spent traveling to and from transit stops (by other modes) and waiting for transit vehicles to arrive
at stops reduce the average speed of a short transit trip more than they would the average speed of a
long transit trip. Meanwhile, even though cycling is faster than walking, it produces less time savings on
short trips than on long trips, leaving people with less reason to ignore any perceived safety concerns
with cycling, or the risk of their bicycle being stolen, or the difficulty of riding a bicycle in inclement
weather. People are more likely to ride a bicycle on a short trip if that trip is across level terrain (Kim
and Ulfarsson 2008).
As with long trips, congestion charging and a limited parking supply discourage the use of automobiles
for short trips. The use of public transit for short trips is greater in more urban areas, where there is less
distance between transit stops, where transit fares are lower, and where the overall level of service of
the transit system is greater. Nonmotorized modes of transportation are more likely to be used when
safe, comfortable paths are provided for them which are separated from the flow of automobile traffic,
there is an adequate supply of bike racks, and short door-to-door travel distances exist between a given
origin point and various destinations. The short trips on which people are most likely to walk are ones
that are for social or recreational activities (including eating out at restaurants). People are less likely to
walk on shopping trips, possibly out of a reluctance to have to carry their purchases home unassisted.
Families with children (that need an efficient mode of transportation for running errands and that take
part in more group activities) tend to drive more on short trips. The elderly are less likely than other
people to walk on short trips while people who have lived at their current address for less than one year
are more likely than other people to walk on short trips. Finally, one of the most important factors in
determining a person's mode choices for short trips (and which is reflective of other factors) is their
individual threshold distance below which they see walking as preferable and above which they see
driving as preferable (assuming that both of those modes are available to them). Not surprisingly,
frequent walkers have higher threshold distances and frequent drivers have shorter threshold distances.
Short threshold distances are also associated with perceived exertion from walking and with age (Kim
and Ulfarsson 2008). In addition to the previously discussed factors of destination accessibility,
intersection density, and traffic safety, walking is also encouraged by aesthetics (both natural and
architectural), safety from crime, low noise levels, low pollution levels, the absence of "physical
disorder," and amenities such as benches and sidewalk retail/restaurants (Neckerman et al. 2009).
3.2.4 Transportation System Capacity as a Driver of Travel Behavior
When new capacity is added to a congested transportation system (e.g., in the form of new roads or
lanes), one can assume that people's travel behavior will change in response. When a transportation
agency decides to address the issue of traffic congestion by adding capacity to the roadway system,
most of the new capacity is added to arterial roads and highways, rather than smaller collector streets.
The ability to travel at high speeds on recently decongested arterial roadways motivates people to
engage in more driving, especially for long trips. However, most trips that make use of an arterial
roadway both begin and end on lower-order roadways. As a result, when a capacity expansion reduces
congestion and increases traffic volume on arterial roads, traffic volumes also increase on roads that
have not been (and frequently do not have the room to be) expanded. Therefore, expanding an arterial
roadway may simply shift the traffic congestion burden from the arterial road to city streets in an urban
-------
35
core, within which automobile travel may consequently be discouraged for short trips, even while it is
being encouraged for long trips (Frank et al. 2008). Meanwhile, capacity is sometimes added to the
transportation system in the form of a new exclusive right-of-way for transit vehicles (either trains or
buses). When this is done, it is likely that most of the transit riders on the new right-of-way will be
people who were already using public transit prior to its opening, as no other mode has been rendered
any less convenient than before. However, if an automobile traffic lane is replaced with an exclusive
public transit right-of-way that is similar to the remaining parallel automobile lanes in terms of both
speed and level of service, some people may be motivated to switch from driving a car along that
particular corridor to riding transit along it (Kennedy 2002).
3.2.5	Induced Demand
A major point of contention in the study of travel behavior is the existence of "induced demand." In
theory, once capacity is added to a congested road system, people will go on more trips and longer trips,
make more of those trips by private automobile, and perform more of their trip making during the peak
travel periods of the day. By allowing people to act upon a latent demand for travel that had previously
been suppressed by traffic congestion, at least a portion of the initial congestion relief from a capacity
expansion is expected to be undone by increases in traffic volume (Kitamura 2009, Kristoffersson and
Engelson 2009). However, the phenomenon of induced demand is not well understood, would require
very complex theories to model, and is not incorporated into standard travel demand forecasts
(Kitamura 2009). Assuming that induced demand exists, it is not limited to the mode of personal
automobile travel. There is evidence of significant latent demand in communities for more walkable
environments; however, people consider walkability to be one of many neighborhood characteristics to
be traded off with others when choosing where to live (Frank et al. 2008).
3.2.6	Peak Spreading
Another important concept related to roadway capacity is that of "peak spreading." When traffic
congestion gets bad enough during the peak travel periods of the day, some drivers will choose to avoid
the worst of the congestion by commencing their trips either a little before or a little after their
preferred time, with the result being a daily pattern of less congested traffic over a longer period of time
instead of more congested traffic over a shorter period of time. Because, in most locations, non-peak-
period traffic volumes are well below congestion levels, peak spreading is something that transportation
planners and engineers often encourage. One of the ways in which greater peak spreading is achieved is
through congestion charging, wherein travelers are charged tolls for using major thoroughfares,
sometimes with the tolls dependent upon the time of day during which one travels. Potential changes
to travel behavior in response to a tolling scheme include traveling by a different mode, traveling along a
different route, traveling to different destinations, reducing the frequency of one's trips, and altering
one's usual departure times. A toll that remains the same at all times of day may induce people to avoid
the toll road by traveling by a more circuitous route that requires them to commence their trips at an
earlier time than they would be able to if they took the toll road. However, a greater degree of peak
spreading would likely occur if off-peak periods of the day were subject to either no tolls or reduced
tolls. In that event, people would have a direct and unambiguous financial incentive to depart on trips
at times other than their most preferred times. It is also worth noting that if a congestion charging
-------
36
scheme succeeds in promoting peak spreading, it likely also succeeds in reducing overall traffic
(Kristoffersson and Engelson 2009, Ng and Small 2012). Peak spreading may also be facilitated by
employers who grant their workers flexibility in determining their hours, so that not everyone is
necessarily commuting at the same time (U.S. Environmental Protection Agency 1998).
3.2.7	High Occupancy Toll (HOT) Lanes
If people are not willing to institute congestion charging on an entire thoroughfare, the option exists of
creating High-Occupancy Toll (HOT) lanes, which are lanes of a freeway that vehicles with multiple
occupants may use for free and that vehicles with only one occupant may use for a fee. The idea is that
such restrictions will make the HOT lanes less congested than other lanes, give people a motive to
carpool, and provide people with an opportunity to make a tradeoff between time spent traveling and
money spent on tolls. Even though only ten HOT lanes existed in the United States as of 2009 (not
counting High-Occupancy Vehicle lanes, which only vehicles with multiple occupants are allowed to use),
some insights can be gained from existing HOT-lane case studies. First of all, if vehicles with multiple
occupants are only able to use the HOT lane for free if they first submit some sort of registration, they
are less likely to do so. Multiple-occupant vehicles are also less likely to use a HOT lane if they are
merely charged a discounted toll instead of no toll at all. When public transit express buses are allowed
to use these special lanes, it increases their speed relative to parallel automobile traffic, which increases
transit ridership. When unemployment goes up, the number of people paying to use HOT lanes goes
down. Finally, gasoline prices are positively correlated with both carpooling and transit use and
negatively correlated with the number of people choosing to pay tolls to use HOT lanes (Goel and Burris
2012).
3.2.8	Trip Aggregation
Another concept used to address limited roadway capacity is that of "trip aggregation." Trip
aggregation attempts to achieve a combination of the efficiency that comes from reducing the ratio of
motor vehicles to people and the flexibility that makes personal automobiles a more attractive mode of
travel than public transit in most cases. One form of trip aggregation is carpooling; however, carpooling
is not a feasible option for the majority of trips. Another form of trip aggregation is car-sharing clubs. In
these clubs, people who already drive less often than the average person (and perhaps wish to
supplement their use of other transportation modes) pay to rent vehicles for as little as one hour and
then return the vehicles to designated locations, as opposed to paying the fixed costs of owning an
automobile; the same business model is used in bike-sharing programs. Trip aggregation can also take
the form of government-subsidized demand-responsive transit services, which at the present time
typically serve either residents of rural areas or people with physical handicaps. Due to low passenger
volumes and long travel distances, though, these services tend to result in very little trip aggregation.
However, through the use of GPS, cell phones, and a centralized system for collecting, organizing, and
distributing information, more advanced forms of demand-responsive transit could conceivably be
created to transport four to eight passengers in one vehicle from locations within one small area to
locations within another small area on relatively little notice and with greater speed and flexibility than
ordinary public transit (Tuomisto and Tainio 2005). In the U.S., trip aggregation programs (such as
vanpools) may be run privately or by public transit agencies.
-------
37
3.2.9	Crowding on Transit Vehicles
The impact of capacity constraints on transportation behavior is mostly considered in the context of
roadway capacity. However, public transit systems also have limited capacities that, if insufficient to
meet demand, may also affect people's travel behavior. The effects of crowding on public transit
vehicles is not often modeled, in large part because this is only a significant issue on a small number of
transit systems and only on certain parts of those systems or at certain times of day, and it is a
particularly difficult phenomenon to predict. A person may choose to take a different transit route
either because their preferred route is so crowded that when a transit vehicle reaches their stop there
will not necessarily be room for them to get onboard or because vehicles on their preferred route tend
to be crowded to the point of discomfort. If a passenger experiences crowding-induced discomfort on a
transit vehicle, that discomfort is greater if they are required to stand during the journey rather than
sitting. Also, the disutility that people derive from transit-vehicle crowding is taken to increase with the
length of the trip. However, longer transit trips are also the ones most likely to either start or end in a
low-density area, making them the transit trips on which crowding is least likely to be a serious problem.
Although the issue of crowding may motivate transit users to choose different routes or different
departure times, some early modeling efforts have indicated that the effect of transit-vehicle crowding
on mode choice is not significant (Zorn, Sail, and Wu 2012).
3.2.10	Economic, Socialand Psychological Motivations for Travel Behavior
While correlations may be established between particular environmental characteristics or government
policies and particular aspects of people's travel behavior, there does not yet exist an embracing theory
of how travel-related decisions are made, as such a theory would need to span many different social
sciences (Kitamura 2009). From an economic perspective, travel is treated as a derived demand, only
valued to the extent that it is necessary to perform other activities, meaning that the amount of use a
transportation system receives is dependent upon the destinations that it connects. In the most
simplistic of terms, those transportation systems, routes, and modes where there occurs a drop in the
price of travel (measured in terms of both time and money and assuming the presence of worthwhile
destinations) will see an increase in travel demand (and actual travel) and a decrease in travel supply
(unused capacity). An increase in average travel speeds produces likewise changes in transportation
demand and traffic volume, which (assuming volumes increase enough to produce some measure of
congestion) causes average travel speeds to decrease. The result of this process, in theory, is a
balancing loop that produces an equilibrium combination of traffic volume and average travel speed
(Kitamura 2009). Policy-makers, planners, and engineers then manipulate the transportation system in
such a way as to produce the most sustainable possible equilibrium points for volume and speed.
However, such efforts may result in conflicting objectives. For example, government actors may desire
both to decrease traffic congestion and to increase the proportion of travel that is carried out by modes
other than the personal automobile. Since reducing traffic congestion has the effect of increasing
automobile travel speeds relative to other modes of transportation, this action gives people a strong
reason to not travel by those other modes unless required. Therefore, reduced automobile traffic
congestion and reduced automobile use are goals that likely require tradeoffs (Frank et al. 2008).
-------
38
Another important social-science consideration in describing the drivers of travel behavior is that of
individual travelers' attitudes towards particular modes of transportation. According to Ewing and
Cervero (2010), transportation-related attitudes are still a less influential factor than the built
environment in determining people's transportation behavior, but a significant one nonetheless.
Various studies have found contradictory evidence regarding whether people alter their behavior to
match their attitudes or alter their attitudes to match their behavior (Haustein 2012, Popuri et al. 2011).
Under the Theory of Cognitive Dissonance, people make choices and then adjust their attitudes to
support the choice they have already made. This contrasts with the much-more-used-in-transportation-
research Theory of Planned Behavior, which states that people perform actions because they have the
intention to perform them. Under this theory, a person's intention to take part in a particular behavior
is determined by their positive or negative valuation of that behavior (attitude), by any social pressure
they perceive to do or not do it (subjective norm), and by the degree to which they perceive they are
able to do it (perceived behavioral control). Also sometimes considered to be a factor in these decisions
is a perceived moral obligation to choose behaviors that conform with one's personal values (personal
norm) (Haustein 2012). Aside from savings of time and money, other characteristics of a transportation
mode that may contribute to someone valuing it as either "good" or "bad" may include: excitement,
prestige, privacy, autonomy, comfort, convenience, perceived safety, reliability, and the ability to have a
stress-free trip (Haustein 2012, Popuri et al. 2011). Popuri et al. (2011) found that accounting for such
attitudinal factors as these in the modeling of public transit demand results in greater predicted
reductions in ridership following service reductions and fare increases. At the same time, if a particular
mode of transportation provides people with the opportunity to conduct other activities (such as
working on a portable computer or conversing with another person) during their travel, people who
take advantage of that opportunity may be better disposed towards that mode and more willing to use
it for longer trips; to one degree or another, this motivating factor may apply to any travel mode,
including both private automobile travel and public transit, depending upon the availability of an
undisruptive environment and any necessary supporting technologies (Gripsrud and Hjorthol 2012).
Of the specific factors influencing people's transportation mode decisions, travel-time reliability is
becoming a more popular performance measure among transportation planners. However, the
behavioral effects of travel-time reliability are not adequately researched and it is still unclear what
relative amount of explanatory power reliability has compared to average time savings. Part of the
argument for giving more consideration to reliability (and unreliability) is that average travel times mask
any extreme travel-delay events that are severe enough to make a lasting impression in people's minds.
Nonetheless, most existing metrics of unreliability are primarily functions of general traffic congestion,
rather than more random bottleneck events. A study conducted in the Chicago metropolitan area found
that if transit trains have a speed advantage over parallel automobile traffic, also having an advantage in
terms of travel-time reliability would increase the odds of people choosing travel by train over travel by
automobile. The authors of that study found the effect of travel-time reliability to be significant, but
also found it to not be the driving factor in people's mode choices. The same study also found evidence
that travelers respond more to travel-time unreliability in the zones around their origin and destination
points than to unreliability along their entire travel route and that travelers are more responsive to the
average amount of unreliability that exists over the course of a day than to the amount of unreliability
-------
39
that exists during any particular time of day (e.g., morning rush hour, midday, evening rush hour, or
nighttime) (Sweet and Chen 2011).
3.2.11	Self-Selection
The greatest source of doubt that exists regarding the magnitude of the effect of the built environment
on travel behavior is residential self-selection, the idea that a neighborhood built around a particular
kind of transportation behavior will simply attract residents who already had a predisposition towards
that kind of transportation behavior rather than motivating other people to change their transportation
behavior (Ewing and Cervero 2010). One aspect of this self-selection is the tradeoff that households
make between housing costs and transportation costs, as cheaper housing (by floor area) tends to be
located in areas with greater transportation costs (such as remote suburbs) and expensive housing tends
to be located in areas with lower transportation costs (such as highly walkable urban neighborhoods
with good public transit service) (Kitamura 2009). Even though some studies report that residential self-
selection reduces the impact of the built environment on travel behavior, nearly all studies on the
subject report that the relationship between the built environment and travel behavior remains
significant after controlling for self-selection (Ewing and Cervero 2010). One possible explanation for
this finding is the existence of latent demand for built-environment characteristics that are currently in
short supply, most especially those characteristics that are associated with more walking, more cycling,
more transit riding, and less driving. A new development (or redevelopment) with a more compact,
more mixed-use, and more walkable layout may attract people who have a predisposition towards
modes of transportation other than the personal automobile and who previously lived in a
neighborhood where the use of those modes was not feasible (Ewing and Cervero 2010, Frank et al.
2008). If this is the case, it may result in a greater lag time in between when a change is made to the
built environment and when any resulting travel-behavior changes are observed, since it takes longer for
new residents and workers to move into a neighborhood than for current residents and workers to alter
their habits (assuming that they alter their habits at all) (Frank et al. 2008).
3.2.12	Traveler Characteristics
A person's socioeconomic status and other personal characteristics have a significant influence on their
transportation behavior (through attitudes, travel requirements, and budgetary concerns). One's
socioeconomic characteristics have been estimated to have a greater influence than the built
environment on the frequency with which people make trips. Trip lengths have been estimated to be
determined more by the built environment than by socioeconomics. VMT and Vehicle Hours Traveled
(VHT) have been estimated to be functions of both the built environment and socioeconomics.
Transportation mode decisions have been estimated to be influenced by both factors, but probably
influenced to a greater extent by socioeconomic characteristics (Ewing and Cervero 2010).
If different modes of transportation cost different amounts to use, and this disparity results in different
income groups having different favored modes of transportation, an increase in the number of people in
a given income range will likely increase usage of that particular income range's favored transportation
mode; however, the mode of walking may not be as sensitive to income as automobiles and public
transit (Berechman and Paaswell 1997). At the same time, a person's mode choices may be affected by
-------
40
what kind of job they have, regardless of how much money they make at that job. If somebody's job
requires them to travel to multiple geographic locations within a single workday, the odds are greater
that they will choose to commute by automobile, since that particular mode provides the necessary
flexibility to make those extra midday trips (Berechman and Paaswell 1997). Conversely, if someone
consistently works at a single location (regardless of whether or not they drive there), they may benefit
from working in a "park once" district, wherein any trips unrelated to their job that they might wish to
make during the day can be made on foot, on account of a large variety of destinations having been
established within walking distance of either their workplace or the place where they parked (likely with
more parking being provided in centralized locations and less being provided at individual businesses)
(Metropolitan Area Planning Council 2010).
One of the most significant characteristics of an individual traveler influencing their travel behavior is
whether or not they own an automobile. Both because they have more opportunities to do so and
because they have evidenced an apparent intent to do so, people who own automobiles are more likely
than other people to travel by automobile (and less likely to travel by other modes), a relationship that
eclipses most other mode-choice factors (Berechman and Paaswell 1997, Kim and Kim 2004, Kitamura
2009). Furthermore, households with multiple motor vehicles tend to drive more than households with
only one motor vehicle, and the difference between the amount of driving done by multi-vehicle
households and the amount of driving done by single-vehicle households increases with the number of
licensed drivers per household. Related to this, the number of licensed drivers in a household is also the
greatest predictor of how many vehicles that household has, more so than household income (Kim and
Kim 2004).
3.2.13 Travel to School
Little research has been conducted on the manner in which parents decide how their children will travel
to school. In addition to deciding by what mode their children will travel to and from school, parents
must also decide whether or not their children need to be escorted to school. Assuming that a child that
is too young to drive lives within walking distance of their school and does not ride the school bus, their
parents' decision of whether or not they require an escort is largely based on perceptions of safety along
the route to school, both from traffic accidents and from criminals. Meanwhile, the simultaneous
decision of what mode of transportation should be used for traveling to school is mostly based upon the
distance to the school and whatever time constraints the family is subject to at the beginning and end of
the school day. If a child's parents have decided that it is safe for them to travel to school unescorted,
the parents' schedules do not represent a constraint on the child's ability to walk to school rather than
being driven. The desire for their children to get more exercise is not usually a major consideration of
parents deciding whether or not to have their offspring walk to school (Faulkner et al. 2010).
Pedestrian-friendly neighborhood design features, such as those that are implemented as part of Safe
Routes to School programs, have a noticeable effect on the percentage of students who walk to school.
However, the distance from a student's home to their school and the amount of time required to
traverse that distance are the greatest barriers to traveling to school without a motor vehicle. One
implication of this fact is that if a school is located in a particularly densely populated area, a
-------
41
correspondingly high percentage of its students will walk to school. If all students are taken to live a half
mile from school and all else is assumed constant, 34% of them are estimated to walk to school. If all
students are taken to live one mile from school and all else is assumed constant, 19% of them are
estimated to walk to school. Because distance and neighborhood density are such important factors,
the schools likeliest to achieve high rates of students walking to school are elementary schools, as they
have smaller student bodies and therefore smaller attendance zones. However, with each additional
year of age, a student's likelihood of traveling to school on foot increases by 0.4% and their likelihood of
traveling to school by automobile decreases by 1.4%. Also, children with siblings are more likely to walk
to school than single children, possibly due to a combination of parents' safety concerns being alleviated
by their offspring walking together and the inconvenience of driving to and from multiple schools each
day (McDonald 2008).
3.2.14 Travel by the Elderly
As people age, they eventually reach a point where it is no longer safe for them to drive an automobile.
When this happens, if they wish to retain their mobility, they must find other ways of getting around.
The most favored mode of transportation among people who are too old to drive is riding in an
automobile driven by a relative or friend, especially if the traditions of their family or community dictate
that an elderly person's family have a responsibility to assume a strong caregiving role towards them.
However, since the current trend is towards a greater percentage of elderly people's adult children
moving far away from their parents, a large proportion of the elderly must either adopt other
alternatives to the private automobile or assume a state of extremely limited mobility (Waldorf 2003).
To understand the travel behavior of the elderly, though, distinctions must be made among them. First
of all, there are captive automobile users and captive public transit users, who are constrained to a
single mode of transportation either by their physical condition or by a built environment that does not
accommodate other transportation modes. Second, there are elderly people who still have access to all
modes of transportation and are able to decide between them on the basis of personal preferences.
Finally, there are elderly people who are affluent and healthy enough that mobility is not a problem for
them, but they still live in an automobile-dominated environment (Haustein 2012). Elderly people tend
to place more value on perceived safety from crime than do younger people; although this would seem
to suggest a greater reluctance among the elderly to use non-automobile modes of transportation,
perceived danger is actually weakly correlated with mode choice, possibly explained by perceived-to-be-
dangerous trips being shifted to perceived-to-be-safer times of day or being made with company. Other
factors influencing the use of cars by the elderly include weather considerations, the availability of cars,
and the perceived quality of public transit in their area. The mode share of public transit among the
elderly is affected by their attitudes towards transit, how close they live to the center of a metropolitan
area, and whether automobile transportation is available to them. Meanwhile, bicycle use among the
elderly is affected by their attitudes towards it and weather considerations (Haustein 2012). Even
assuming that high-quality alternative modes of transportation are available to the elderly when they
are no longer able to travel by personal automobile, their making use of those modes depends upon
acquiring a good working knowledge of the available options, preferably well before they stop being
able to drive (Haustein 2012, Waldorf 2003).
-------
42
3.2.15 Transportation-Behavior Modeling Techniques
Most conventional travel behavior models consist of four sequential steps, based on the assumption
that 1) people decide to make a trip, 2) they decide what the destination of that trip will be, 3) they
decide what mode they will use for the trip, and 4) they decide what route the trip will follow. However,
a better reflection of reality might be modeling these decisions as being simultaneous, since they may
not always be made in the same order. Unfortunately, just like most other modeling techniques that
produce a closer representation of reality, modeling people as making multiple decisions at the same
time would be very complex, difficult, and time-consuming, resulting in a necessary tradeoff between
accuracy and expediency (Hasan and Dashti 2007, Newman and Bernardin Jr. 2010).
One particular limitation of conventional transportation modeling is the assumption that destination
decisions necessarily come before mode decisions, which implies that people are more likely to change
their minds regarding their mode of travel than regarding their destinations. In this way, modeling
destination decisions as being made before mode decisions is a technique best suited for use in large
metropolitan areas wherein the transit system is good enough to be competitive with automobile travel
and there is a large number of people who both have the option of traveling by either transit or
automobile and consider each of those modes to be a viable option. In areas that do not fit this
description, modeling destination decisions before mode decisions could potentially represent transit
ridership rates as experiencing larger changes in response to level-of-service changes in the transit
system than actually occur. On the other hand, if mode choices are not made until after destination
choices, different modes can be taken to be compared on the basis of actual travel times. Few studies
have so far examined the possibility of modeling mode choices before destination choices (Newman and
Bernardin Jr. 2010).
Another possible shortcoming of current travel behavior models is that the number of trips people make
("trip generation") is not taken to be affected by changes in the capacity of the transportation system,
even though increased capacity in the transportation system implies increased accessibility to various
desirable destinations, which could potentially motivate people to make more trips. Including this
consideration in a model would be very complicated and very difficult to generalize to the point where it
could be applied to more than one transportation system. Also, this theoretical link between
transportation system capacity and trip generation has not yet been proven. Therefore, most models
regard the number of trips into and out of a given zone as a function of sociodemographic
characteristics and land use patterns, and not as a function of system capacity (Kitamura 2009).
Assuming that all people make their transportation-related decisions in the same fashion, such as
everyone having the same perception of the opportunity cost of time spent traveling, is not realistic.
Therefore, transportation behavior modeling would benefit from dividing travelers into various
categories, based on combinations of socioeconomic characteristics and travel purposes. Ideally, the
travel costs perceived by members of each category would be affected by the travel behavior of each of
the other categories. This, too, is a technique that would increase the complexity of the modeling
process (Hasan and Dashti 2007).
-------
43
A key concept in the forecasting of travel behavior is travel-time budgets. Under this concept, demand
for travel is assumed to be incidental to demand for the things that one travels for; thus, there is only so
much of a given person's time that they are willing to spend traveling and the given person will adjust
their travel behavior in such a way as to avoid exceeding that budget. In models, this budgeting of time
spent on travel is sometimes combined with the budgeting of money spent on travel. One problem with
travel budgets, though, is that they are typically assumed to be constant, rather than being affected by
any of the outcomes of the models they feed into (Kitamura 2009). An illustration of the effect of modal
decisions on travel-time budgets may be seen in Figure 3 and travel-time budget surpluses may be seen
as an input to car-only person-miles traveled per person per day in Figure 2.

Ave. financial
travel budget
<% of population that is
children>


Ave.
travel-time
budget

< cycling- only-trip -miles/resi
dent/day>
Ave.
travel-time
budget surplus
Ave. time
spent on car
trips
Ave. time spent
on cycling trips
of car
trips by residents>


Ave. time
spent on
walking trips
Ave. time spent
on transit
itineraries
Ave. car trip speed/Ave.
transit itinerary speed
Ave. cycling
speed/Ave. walk speed
Ave. walking
speed of
residents

+Ave. transit itinerary
speed/Ave. walk speed
Ave. transit itinerary
speecj/Ave. cycling speed
Ave. car trip
speed/Ave. walk s
Ave. car trip
speed/Ave. cycling
speed
Figure 3: System-dynamics diagram of the effect of modal decisions on travel-time budgets. Red arrows are negative
relationships and blue arrows are positive relationships.
Modeling entire tours, wherein someone travels from a given origin point to any number of destination
points and then back to the origin point, can be more informative than modeling individual trips.
Typically, people decide what mode of transportation they will use for an entire tour prior to
commencing. For example, a person is less likely to take public transit somewhere if they know that
-------
44
they will not be coming home until after the transit service day has ended and will be unlikely to travel
home by some means other than automobile if they left home in a car that they are now responsible for
also getting back home. However, the mode of walking may be used for a few small legs within a tour
that is otherwise dominated by some other transportation mode. If a tour has one primary destination
and some number of "extra" destinations that will be stopped at some time prior to returning to the
origin point, the mode choice for the tour is unlikely to be greatly influenced by land-use patterns
around any stops but the origin point and the primary destination. As stated earlier, though, there has
so far been little research on the effect of land uses around somebody's place of work on their travel
behavior. If people are presented with high-enough-quality non-automobile mode options, and use
those options, a greater proportion of their tours will probably only stop at one destination before
returning to the origin point (Frank et al. 2008).
Most travel behavior models do not give much consideration to the time of day at which a person
departs on a trip, and most of the models that do give this consideration only narrow down departure
times to one of a series of broad time periods, usually a couple hours in duration. Greater precision
would be required in order to study variations in traffic congestion within a given peak travel period,
which is currently considered a high-priority research area. Issues of complexity and prohibitively long
computer runtimes have limited the number of large-scale models including departure-time choices
(Kristoffersson and Engelson 2009).
When modeling the effects of crowding on public transit vehicles, some models treat it in the same
manner as road congestion, as a function of demand volume and system capacity. However, on a
crowded transit vehicle, a greater disconnect can occur between actual travel time and perceived travel
time (based on the disutility of boarding a crowded vehicle) than would occur on a congested roadway.
As a result of this complication, many models that describe the behavioral effects of transit crowding do
not feed those effects back into the model as drivers. Another limitation of such models is the difficulty
in calculating transit crowding effects in between a specific origin point and a specific destination point.
In spite of these shortcomings, in those locations where transit vehicles reach capacity with significant
frequency, modeling the behavioral impacts of transit crowding could be an important aspect of testing
the potential outcomes of changes in the capacity of the transit system (Zorn, Sail, and Wu 2012).
In the study and modeling of nonmotorized transportation, different methods and considerations are
needed than in the modeling of automobile traffic and public transit use. Since nonmotorized modes
(especially walking) are consistently far slower than motorized modes, comparing motorized and
nonmotorized travel times would be a meaningless exercise. Also, since people of the same age and
height generally walk at similar speeds regardless of the respective environments in which they walk,
using travel distance, as opposed to travel time, in disutility functions makes more sense for travel by
foot, along with other barriers to nonmotorized transportation such as steep slopes, precipitation, the
dark of night, and environments that are generally perceived as less safe than others. Furthermore,
because the average nonmotorized trip is significantly shorter than the average motorized trip, the
analysis of nonmotorized travel requires a finer scale of analysis than what most transportation models
provide. Achieving an analysis with the necessary resolution to consider nonmotorized modes could
-------
45
potentially be done with the help of GIS software and data on the latitudes and longitudes of specific
origins and destinations, assuming the provision of sufficiently detailed demographic and employment
data. Finally, accounting must be made of such correlations as the fact that curvilinear streets and cul-
de-sacs (which discourage nonmotorized transportation) are more common in hilly terrain (which also
discourages nonmotorized transportation) (Cervero and Duncan 2003).
3.2.16	Data Availability for Modeling Travel Behavior
Many small and medium-sized towns, including those located near larger cities, do not have enough
money at their disposal to carry out the surveys necessary to gather data from which to calculate trip
generation rates for travel-behavior models. In such towns, identifying alternative data sources that
could be used with confidence can be helpful. Trip-generation rates applicable to various communities
were published by the National Cooperative Highway Research Program (NCHRP) in 1998, but only for
communities with at least 50,000 residents. For those communities that are smaller, the option exists of
adapting national-scale data to their purposes, such as from the Nationwide Personal Transportation
Survey (NPTS). Also, in some parts of the country, including South Carolina and Oregon, generalized
reusable transportation models have been created that are to be applied to many different small and
medium-sized communities. Finally, communities may choose to simulate travel data for their
jurisdiction on the basis of local socioeconomic data and patterns observed in travel surveys from other,
nearby, larger communities. If a small or medium-sized community that cannot afford a travel survey
and a large community that can afford such a survey are sufficiently close together and sufficiently
interconnected, data may be transferred from the larger community to the smaller one for the purpose
of creating trip generation rates. To determine if such a practice actually could be justifiable would
require additional research specific to these communities. If it does turn out to be a valid practice,
though, the implication would be that, at least sometimes, conducting a more extensive travel survey of
a smaller area (that can be applied to additional locations) may be better than collecting a less certain
dataset for a larger area (Horner, Stone, and Huntsinger 2008).
3.2.17	Specific Resources and Tools for Determining Travel Behavior Outcomes
•	The U.S. Environmental Protection Agency's Smart Growth Index (SGI) is a planning tool used for
modeling travel behavior and air quality, using Ewing and Cervero's 5 D's. SGI incorporates the
built-environment characteristic of density in the form of activity density, which is the sum of
the residents and jobs located in an area divided by the number of square miles in the area. The
built environment characteristic of land-use diversity is incorporated in the form of the ratio
between a given area's jobs-to-housing balance and the jobs-to-housing balance of a larger
region that the given area is within. The SGI also incorporates the transportation-system design
elements of street-network density, sidewalk coverage, and route directness, which is defined
as the ratio between the distance someone travels between two points and the straight-line
distance between those points (Ewing and Cervero 2010).
•	The Simultaneous Transportation Equilibrium Model (STEM) is a travel behavior model wherein
the trip generation rates are taken to be affected by the model's own outcomes. The Multiclass
Simultaneous Transportation Equilibrium Model (MSTEM) is based on the STEM model, but with
the added element of dividing travelers into classes according to their socioeconomic
-------
46
characteristics, their purposes for traveling, and the use of either single transportation modes or
combined transportation modes, with each class making transportation-related decisions on the
basis of different criteria. Because the cost attributed to using any particular link of the
transportation system is taken to be dependent upon the traffic flows on that and all other links,
different classes of travelers may be calculated to take different routes between the same origin
and destination. This model also combines the trip-generation step with estimating people's
decisions regarding departure time. The downside of the MSTEM model is that its use requires
a lot of time, money, and effort (Hasan and Dashti 2007).
•	CONTRAM is a route-choice model for given origin-destination pairs. It is used in many different
cities, in various countries (Kristoffersson and Engelson 2009).
•	The Unified Mechanism of Travel (UMOT) model is a transportation behavior model based
around the idea of unchanging household travel budgets that can be applied in situations across
both space and time. In this model, the constraints on travel activity are taken to be daily travel
time per traveler and money spent on transportation per household. The model assumes that
whenever someone experiences travel-time savings, they will use that extra time for more
travel. The UMOT model has the advantages of being simple and having low data requirements,
but also some limitations, including a car-ownership component that has been called
"simplistic" and the use of some questionable elasticities between the cost of using a given
mode of transportation and the rate at which that mode is actually used (Kitamura 2009).
•	Origin-Destination travel survey data exists for almost every metropolitan area in the United
States, often coming from up to three surveys per metropolitan region, taken at intervals of
approximately ten years. Typically, the Metropolitan Planning Organization (MPO) of a given
area (the local organization through which federal transportation funding is distributed to urban
areas with populations of 50,000 or more for the purpose of keeping transportation projects
and policies coordinated) will match up this survey data with information on their area's
transportation network and patterns of land use. Taken together, all of this information could
form a very useful database. However, the data is often not well-archived, not well-
documented, and not easily available to researchers (Kitamura 2009).
3.3 Air Quality Issues and Related Tools, Resources, and Indicators
Transportation sources significantly contribute to global, regional, and local air quality impacts. Air
emissions occur from fuel combustion, fuel and fluid evaporation through engine operations and leaks,
abrasion from brake use, tire wear, and re-entrainment of dust from road and other transportation
surfaces. Transportation sources emit criteria pollutants, air toxics, greenhouse gases, and multiple
chemical and physical forms of particulate matter.
Many methods have been employed to reduce air quality impacts from transportation sources. These
methods typically include reducing emissions through regulations and standards and/or reducing motor
vehicle activity. Investing in and deploying public transit, increasing costs for personal vehicle use (e.g.
congestion pricing), and improving facilities for walking and biking have all been employed in efforts to
reduce motor vehicle activity. Recent development practices have been initiated in many parts of the
-------
47
world to promote the reduction of vehicle activity, and subsequent impacts on air quality and the
environment, through compact growth and infill development. The goal of these "smart growth"
practices is to make transportation more accessible, convenient, and with a reduced impact on the
environment. However, these practices often also bring people in closer and longer contact with
transportation source emissions. Exact differences in exposure to near-road air pollution depend on
many variables, including the manner in which air pollutants are transported, which is an area of
significant research needs.
Recent research has demonstrated a link between exposures to air pollutants from traffic emissions
near large roadways and other transportation sources with adverse human health effects. The Health
Effects Institute (HEI) recently completed a review of a large number of health studies, concluding that
near-road exposures "are a public health concern." This is because of the toxicity of the small particles
and chemicals emitted and the emission of these particles and chemicals close to the ground, where
they are not well dispersed (http://pubs.healtheffects.org/view.php?id=334). Although the link
between adverse health effects and near-road exposures has been made, the science has not yet
progressed to an understanding of how some key elements affect these associations, such as the type
and size of roads of concern, the vehicle fleet mix, activities leading to highest exposures, and the
distance from the road at which near-road health impacts subside. Most studies on traffic and health
focus on roads with high levels of traffic (for example, counts of 100,000 annual average daily traffic
(AADT) or higher). A few studies have reported health effects associated with smaller traffic volumes,
with one study showing effects at volumes as low as 10,000 AADT in an area (HEI Panel on the Health
Effects of Traffic-Related Air Pollution 2010). While the health studies reviewed by HEI focused on
exposures to traffic emissions, other transportation sources such as rail yards, rail lines, airports, and
marine ports have comparable concerns due to similarities in the type and characteristics of air pollution
emissions.
For most transportation sources, air pollutant concentrations are generally highest closest to the source,
with concentrations lowering with distance from the facility. However, the magnitude and extent of
these increased air pollutant concentrations can vary based on a number of factors related to emissions,
including the source, meteorological and topographic conditions affecting pollutant transport and
dispersion, and the influence of roadway design and roadside features on pollutant transport and
dispersion. Traffic emissions may vary depending on the total number of vehicles using a road, the level
of congestion on the road, and the number of heavy-duty trucks present. For rail operations, the
number of trains, cargo weight, maintenance activities, and line/yard configuration will influence
emissions and exposures. Ports and airports will generate emissions from the ships/planes using the
facility, as well as support equipment permanently present. For marine ports, large numbers of heavy-
duty trucks may also be present on local roadways to move goods from the port; rail activity into and
out of a marine port may also be substantial. Air pollutant concentrations near transportation facilities
will also be affected by wind direction, wind speed, and atmospheric stability. Changes in local
topography from natural or roadway-design features will also affect air pollutant transport and
dispersion, which can lead to varying exposures for nearby populations. Thus, air quality may vary
-------
48
based on surrounding terrain and features, such as cut sections, noise walls, vegetation, or combinations
of these features (Baldauf et al. 2009).
International consensus has emerged that people living, working, and going to school near high-traffic-
volume roads face increased risks for a number of adverse health effects (HEI Panel on the Health
Effects of Traffic-Related Air Pollution 2010). These health effects have been attributed to acute and
chronic exposures to elevated levels of air pollution near these roads, including particulate matter (PM),
gaseous criteria pollutants, and air toxics (Karner, Eisinger, and Niemeier 2010). Field measurements
conducted throughout the world, including the U.S., have shown highly elevated air pollution levels near
high volume roadways. Pollutant concentrations are often highest within the first 100-150 meters, with
increased concentrations of some pollutants of as much as an order of magnitude. Pollutant
concentrations from traffic emissions can remain elevated 300-500 meters or more from the road
(Karner, Eisinger, and Niemeier 2010, HEI Panel on the Health Effects of Traffic-Related Air Pollution
2010).
With increased urbanization occurring world-wide, the number of people exposed to traffic emissions
near high-volume roadways will continue to increase. One factor contributing to this trend is that a
growing portion of public transportation and land use policies and practices supporting sustainable
development patterns promote compact growth in infill locations along major transportation corridors.
Transit-oriented development, offering a mix of housing and supportive land uses located near transit
and with accessibility to jobs and services, nonetheless offers a variety of benefits. This development
pattern is intended to capture the benefits of location efficiency, which is strongly correlated with
household transportation spending. As land use development patterns affect travel behavior, air quality
and global climate conditions are indirectly affected by urban form. While sustainable development
practices increase the population's access to services and transportation options and promote regional
reductions in vehicle miles traveled (VMT) and air pollution, they also often bring people closer to
sources of air pollutant emissions, including from traffic activity. Accordingly, there is a growing need to
reduce air-pollutant exposures for people residing and working near high-volume roadways.
The U.S. Environmental Protection Agency (EPA) is implementing a number of policies to address these
impacts of major roads on nearby air quality. Recent revisions to monitoring rules for the National
Ambient Air Quality Standards (NAAQS) require monitors for PM, carbon monoxide (CO), and nitrogen
dioxide (N02) near high-traffic roads in large metropolitan areas. The Environmental Protection
Agency's transportation conformity rule requires modeling of PM2.5 and/or PM10 "hot spot"
concentrations in the immediate vicinity of large federal highway or transit projects involving high levels
of heavy-duty diesel vehicle traffic. Projects are required to model concentrations at or below the
NAAQS, or show that concentrations with the project built are modeled lower than concentrations
without the project built.
The complexity and multitude of factors affecting air pollutant concentrations near transportation
sources makes it difficult to recommend a strict set of guidance for "safe" distances from these source
types, particularly given the potential for unintended consequences. Decision-makers considering one
-------
49
or more sites in close proximity to major transportation facilities should consider a range of approaches
to mitigate or avoid potential exposures. When evaluating potential sites that may be located near a
highway or other major transportation facility, several factors should be considered:
•	Are there other sites in the community at further distances from the source that are also being
considered? Urban areas may be limited in their ability to find appropriate sites away from
major roads, major goods movement facilities, and other transportation sources; thus, careful
consideration should be given to near-road and other transportation-source sites before
eliminating them if the only alternatives involve siting schools, for example, much further from
the communities being served. Unintended negative consequences of moving schools away
from these communities may include increased pollutant exposures during longer bus or
personal car commutes, increased traffic on local roads to access schools further from their
communities, lack of walking, biking, or other alternative commute options to school, and the
inability to meet many of the other smart growth objectives described in this document.
•	What options might be feasible for mitigating pollutant concentrations at the site from off-site
sources (Baldauf et al. 2009)?
o Studies suggest that roads in cut sections (i.e. road surface below surrounding terrain)
or that have combinations of noise barriers, vegetation, and/or buildings near the
roadside may reduce downwind air pollution concentrations,
o Building design techniques may be employed to reduce exposures at near-source
locations, such as encouraging activity as far from the source as possible (e.g.,
entrances, playgrounds, gathering places) and locating air intakes at locations not
affected by off- or on-site transportation-related air pollutant sources,
o Installing or preserving barriers such as trees, buildings, and noise barriers may reduce
air pollutant exposures,
o Filtration devices as part of HVAC design can be used to improve indoor air quality as
described in other sections of this guidance,
o Adding controls or redesigning transportation facilities to reduce pollutant emissions
and air concentrations. Examples of this practice include: replacing or retrofitting port
and rail engines/equipment with cleaner technologies, reducing idling at terminal
facilities, re-routing existing or projected traffic away from populated areas (e.g., truck-
only lanes), and adoption of high-density development and transit alternatives.
Transportation sources impact air quality through three major pathways: vehicle operating emissions of
gaseous and particulate contaminants, secondary formations during plume transport of gases and
particles, and mechanical processes that abrade particles from brakes, tires, and the road surface.
Carbon monoxide (CO), nitric oxides (NOx), and volatile organic compounds (VOCs), as well as PM
constituents such as polycyclic aromatic hydrocarbons (PAHs) and black carbon (BC) are among the
numerous compounds that have been identified at elevated concentrations near large roads
(Venkatram et al. 2007).
Emission reduction programs implemented by government agencies throughout the world have
significantly reduced emission rates of air pollutants from motor vehicles. Since 1970 in the U.S.,
-------
50
average per vehicle emissions have been reduced by over 90% for VOCs, and over 80% for PMio and
NOx. In spite of these reductions, motor vehicles still significantly contribute to pollution in urban areas,
often due to large increases in vehicle use offsetting per-vehicle emission reductions (Dallmann and
Harley 2010). Furthermore, emissions from some vehicle-associated sources (e.g. brake and tire wear)
are not regulated, and pollutants generated from these sources may also increase in the future with
increased vehicle use.
Populations near roads are exposed to this mixture of primary emissions and secondarily formed
pollutants. Approximately 30 to 45% of urban populations in the U.S. are likely exposed to elevated
pollution levels near roads (Zhou and Levy 2007, U.S. Environmental Protection Agency 2013c). In many
countries with densely populated urban areas this figure is likely to be even higher.
EPA's Office of Air Quality Planning and Standards recently promulgated rules requiring air quality
monitoring at near-road locations expected to experience maximum concentrations of criteria
pollutants emitted by motor vehicles. This requirement stemmed in part from concerns regarding the
adverse health effects experienced by populations spending significant amounts of time in a near-road
environment due to residence, school, or work. EPA's Office of Research and Development has initiated
the Sustainable and Healthy Communities (SHC) program to assist planners and community stakeholders
in analyzing and implementing sustainable development options. An area of particular interest is
understanding the environmental and health implications of different development futures being
considered by U.S. cities, such as compact growth with significant in-fill development. Such sustainable
community strategies can significantly reduce overall emissions by reducing vehicle trip lengths and
making it easier for people to walk, bike or take transit. However, development in dense urban
environments that is poorly planned and designed has the potential to increase some residents'
exposure to air pollution sources in their daily activities. As a result of this increased interest in near-
road exposures from a regulatory and development perspective, informing the general public affected
by near-road exposures is important in raising awareness of the issue, as well as identifying and
implementing best practices to reduce or eliminate these exposures. Although the transportation
planning process has a mechanism to consider near-roadway pollution in some cases ("conformity"
requirements), in many instances there are simply no mechanisms or tools that enable communities to
carefully assess and potentially mitigate near-road emissions exposure.
3.3.1 Specific Resources and Tools for Determining Air Quality Outcomes
• The Community Multiscale Air Quality (CMAQ) model version 5.0 employs a 3-dimensional
Eulerian modeling approach to address air quality issues such as tropospheric ozone, fine
particles, acid deposition, and visibility degradation (Byun and Schere 2006). CMAQ is process-
based and employs first-principal relationships to the greatest extent possible. The model is a
multiscale, multipollutant, "one atmosphere" system that includes a meteorological component
to describe atmospheric conditions, emission models for anthropogenic and natural emissions
that are released into the troposphere, and a chemical-transport model (CTM) to simulate
chemical transformations, atmospheric transport, and fate. Most anthropogenic and biogenic
emissions are parameterized as emission factors and activity rates, or are represented by hourly
-------
51
estimates of temporally and spatially allocated emissions from point, nonpoint, and mobile-
source inventories. CMAQ operates on a five-minute timestep, but results are usually produced
hourly. The model simulates regional-scale air quality issues via a set of rectangular grids
ranging in size from 1 km2 for small domains to 100 km2 for hemispheric-scale simulations. The
one-atmosphere approach supports a comprehensive, system-wide technique for analysis of
complex sustainability issues spanning air, land, and water media. CMAQ regional air quality
simulations at the community scale rely heavily on detailed inventories of transportation-related
emissions. A description of the mobile on- and off-road source emissions input to CMAQ is
provided at
http://epa.gov/ttn/chief/emch/2007v5/20Q7v5 2020base EmisMod TSD 13dec2012.pdf.
Transportation emissions include on-road vehicle, on-road refueling, nonroad (construction and
agriculture), rail, and marine emissions. Marine emissions are particularly important
contributors to overall particulate loads in cities that house large commercial seaports. Except
for California, all on-road vehicle and on-road refueling emissions numbers are generated using
a SMOKE-MOVES emissions modeling framework
(http://www.epa.gov/otaq/models/moves/index.htm) and hourly meteorology. This system
differentiates emissions by process (running of the engine, starting the engine, vapor venting,
etc.), vehicle type, road type, temperature, speed, hour of the day, etc. to produce a set of
emission factors. Emissions for a county are the result of multiplying these factors by vehicle
miles travelled or vehicle population (activity). California emissions are provided by the
California Air Resources Board (CARB) using their EMFAC model, designed specifically for
California fleets.
•	EPA maintains the MOtor Vehicle Emission Simulator (MOVES) model for estimating air pollution
emissions from all on-road motor vehicles including cars, trucks, motorcycles and buses. MOVES
allows motor vehicle emissions to be analyzed at various scales: national, county, and project,
using different levels of input data. The project scale allows the prediction of air emissions from
traffic activity on a specific road or intersection to be estimated. In addition to MOVES, EPA has
several other calculator-style tools that can be used to estimate emissions of specific types of
vehicles. EPA maintains guidance for using MOVES for regulatory purposes of state air quality
plans and transportation conformity determinations; however, MOVES is also EPA's best tool for
developing on-road greenhouse gas (GHG) emission inventories at the state and local level, with
guidance for using MOVES to estimate GHG inventories. For access to MOVES and these other
tools, as well as guidance for use and interpretation, please refer to
www.epa.gov/otaq/stateresources/tools.htm.
•	The EPA Office of Transportation and Air Quality (OTAQ) has guidance for estimating the air
quality benefits of various control measures and accounting for them in a state air quality plan
or an area's transportation conformity determination. Subjects covered by this guidance
include transportation pricing, land use, and commuter programs. Please refer to
www.epa.gov/otaq/stateresources/policv/pag transp.htm for a list of the full range of
measures on which guidance is available.
-------
52
•	EPA OTAQ has developed an approach for estimating emission reductions, of both criteria
pollutants and GHGs, from "travel efficiency strategies/' those emission reduction strategies
that affect travel activity, such as travel demand management (e.g., telecommuting, transit
subsidies), public transit fare changes and service improvements, road and parking pricing, and
land use/smart growth. This approach is described in a series of documents, found at
www.epa.gov/otaq/stateresources/ghgtravel.htm.
•	EPA OTAQ has guidance for completing quantitative hot-spot analyses for individual highway or
transit projects, required for certain projects in PM and CO nonattainment areas. The PM hot-
spot guidance covers estimating emissions using MOVES, estimating air quality concentrations
using EPA-approved dispersion models (e.g., AERMOD), determining the background air
pollutant concentration, and calculating the resulting design value for the project. Policy and
technical guidance for hot-spot analyses, as well as training and other resources, can be found
at http://www.epa.gov/otaq/stateresources/transconf/proiectlevel-hotspot.htm.
•	Fuel composition is regulated to achieve air-quality goals under the authority of the Clean Air
Act. Releases to the environment occur through the use of fuels for transportation, the
production of both fossil fuels and biomass-derived fuels, leaks from pipelines, spills during the
transport of fuels, and leaks from storage tanks. The EPA is developing a model to account for
the transport and fate of fuel components in soil gas and groundwater. Volatile components of
leaked fuels, including vehicle fuels, diffuse through the soil gas and in some cases cause indoor
air contamination (called vapor intrusion). The composition of the fuel, subsurface properties,
building properties, oxygen availability in the soil gas, and the location of the leaked fuel
determine the magnitude of the impact. Field experience has shown that most leak sites with
petroleum-vapor intrusion problems have strong sources in close proximity to building
foundations.
3.4 Energy Use and Climate Change Issues and Related Tools, Resources, and Indicators
Energy use in the transportation system, including both passenger and freight transportation, is of key
interest at the global, national, state, and local levels. At the national level, approximately 28% of total
energy consumption in the U.S. is for transportation, including light and heavy duty vehicles, airplanes,
buses, trains, barges, and ships (U.S. Energy Information Administration 2013a). Personal
transportation in light-duty vehicles (including cars, minivans, light-duty trucks, SUVs, motorcycles, etc.)
accounted for the majority (59%) of total transportation energy use (highway and non-highway) and
72% of highway transportation energy use in 2010 (from Table 2.5 in Davis, Diegel, and Boundy (2012)).
Energy use is also closely coupled to a number of other environmental outcomes, particularly air quality
and climate change. Economic factors also play an important role in assessing transportation energy
production and use, whether the consideration involves household and individual expenditures on
transportation fuels or broader national issues of energy security and reliability of the transportation
system in terms of fuel supply.
Transportation energy use can be broken down into a set of key indicators, which can then inform
policies and decision making regarding ways to reduce overall energy use and, by extension, greenhouse
-------
53
gas (GHG) emissions. These indicators include: end-use travel demand (often expressed as vehicle-
miles traveled or VMT), vehicle efficiency or fuel economy (gallons per mile), mode of travel (expressed
as a percentage modal split or VMT per mode, which can then be used to assess VMT and fuel economy
on the basis of the vehicle fleet of interest), and overall system efficiency of the surface transportation
system (such as congestion levels). Each of these will be discussed here, while highlighting selected
tools, resources and indicators related to travel demand, travel mode, and fuel economy. Taken
together, these elements can inform assessments of the key driving factors of transportation energy
use.
Climate change is closely interrelated with transportation energy use due to the combustion of
petroleum-based fuels. Given this close relationship, energy use, climate change, and the measures
taken to address these issues will be discussed together in this section. As a major consumer of fossil
fuels, the U.S. transportation sector is also a major source of GHG emissions (33% of 2011 C02 emissions
from fossil fuel combustion), the second largest contributor following the electricity sector (U.S.
Environmental Protection Agency 2013b). Most of these emissions are tailpipe emissions of C02,
although there are also non-C02 emissions such as methane (CH4) and nitrous oxide (N20), as well as
hydrofluorocarbons (HFCs) from vehicle air conditioning units. Methods and tools that reduce total
demand for transportation energy, whether in the form of liquid transportation fuels (fossil or
renewable) or electricity, will generally translate into reductions in carbon dioxide (C02) emissions.
However, different sources of transportation energy, namely gasoline, diesel and, increasingly,
electricity, are used to meet the total transportation end-use demand, and these fuels have different
carbon contents, affecting total C02 emissions. Switching to lower-carbon or biomass-based fuels can
represent an additional measure to reduce the carbon footprint of transportation fuels. However, when
assessing the use of renewable fuels derived from biomass or of electric vehicles powered from the grid,
additional factors need to be considered in order to capture the full greenhouse gas impacts of these
alternatives. In particular, in order to assess the greenhouse gas implications of alternative fuels,
quantifying the upstream impacts associated with the production of those fuels is important (U.S.
Environmental Protection Agency 2010e). Moreover, transportation fuel choices, whether gasoline,
diesel, compressed natural gas, propane, ethanol or biodiesel, will have important implications for the
vehicle fleets and fuel distribution infrastructure that are part of a community's landscape. While there
is a range of studies regarding specific issues (e.g., underground storage tanks for ethanol blends), there
do not appear to be tools or resources that look systematically at how transportation fuel infrastructure
fits within the broader sustainable communities context.
While research on life-cycle impacts (cradle to grave implications) has often focused on upstream
impacts of fuel and electricity production, there are also GHG and other environmental impacts related
to material flows in production. Although their total contribution to GHG impacts and energy use may
be small, the production of vehicles and vehicle components, as well as of larger transportation
infrastructure (e.g., concrete production), may need to be considered in terms of other environmental
endpoints, which would consequently be counted among the impacts of automobile dependence. To
illustrate, some studies have found that the GHG emissions for lithium-ion (Li-ion) battery materials and
their production may account for only 2-5% of total life-cycle GHGs (Samaras and Meisterling 2008).
-------
54
However, different battery systems (e.g., NiMH (Nickel Metal Hydride) or Li-ion) will have varying
impacts for other endpoints, such as metals depletion and ecotoxicity (Majeau-Bettez, Hawkins, and
Stromman 2011). Some of these related issues may become relevant to communities' materials
management and suggest the need for better end-of-life recycling for electric and plug-in hybrid vehicle
batteries. Another example of materials issues related to fuel production is the production of biodiesel
and the fate of the glycerol by-product produced (Steinmetz et al. 2013). This could be a significant
waste stream that communities would need to manage, but it may also have reuse potential in other
applications that could result in some additional air emissions. These examples represent the type of
potential problem-shifting that would need to be avoided as more sustainable transportation energy
solutions are pursued by communities, and argue for a greater need for systems approaches when
assessing solutions. The EPA's Sustainable and Healthy Communities Research Program is in the process
of producing another synthesis paper on sustainable waste and materials management, which may be
looked to for further insights on these issues.
3.4.1 Transportation Demand
One of the critical indicators of transportation energy consumption is vehicle-miles traveled (VMT) per
capita (or potentially average trip length, transportation mode shares, or some other descriptor of
people's transportation behavior) (Black, Paez, and Suthanaya 2002). At a very basic level, trends in
overall transportation demand for a particular region or community are driven by the combination of
population growth and person-miles traveled per capita. Population growth can be a major
fundamental driver increasing pressure on transportation energy use. For example, the Annual Energy
Outlook projects substantial increases in VMT in regions such as the Pacific Coast and the South Atlantic
Region, which includes most of the east coast states from Florida to the Carolinas and up to Maryland
(U.S. Energy Information Administration 2013b). On the other hand, VMT in the New England States
and Midwest is anticipated to grow at a more moderate pace. These projections result from trends in
population growth and changing demographics, including trends such as the migration of population to
regions such as the southeastern U.S. These are larger national and regional trends, which are often
beyond the scope of a community's decision making. However, the way in which these broader changes
in population growth and VMT play out at the community scale is closely linked to land use change and
urban form. Some communities have been able to document important energy and GHG reductions
achieved through urban design (Portland, OR is one example, see Rose and Burkholder (2009)). This
topic is addressed in greater depth in other sections of this document. However, some examples of
potential relationships between energy use and measures that are aimed at reducing transportation
demand in a given urban area are provided in the following paragraph and bullet points.
Past research suggests a number of correlations which provide potential performance measures for
reductions in energy use in the transportation sector in a given urban area based on linkages between
population density, mode share, and other factors affecting VMT, and thus demand for transportation
energy. These include the following elasticities taken from data on major cities around the world over a
period of multiple decades (Sinha 2003):
• A 1% increase in population density is associated with a 0.64% decline in private-transportation
energy use per person per year.
-------
55
•	On average, a 1% increase in transit boardings per person per year is associated with a decrease
in transportation energy use per person per year of 0.54%. However, this elasticity varies with
the magnitude of the input variable.
•	On average, a 1% increase in the number of parking spaces per worker in a region's Central
Business District is associated with a 1.27% decline in transit boardings per person per year.
However, this elasticity varies with the magnitude of the input variable.
The overall indication of the data is that increases in population density and employment density are
associated with fewer roads and parking spaces, decreases in the use of single-occupant motor vehicles,
and increases in public transit use and non-motorized transportation, all of which are trends that in turn
lead to less per-capita energy consumption in the transportation sector (Sinha 2003). However, these
associations do not distinguish between causation and correlation.
3.4.2	Travel Mode Choice and Public Transit
Closely related to travel demand reduction, there are a number of options available to communities to
influence their modal split, many of which are addressed in other sections of this paper. In general,
moving away from single-passenger vehicle travel to transit buses, rail, or non-motorized transportation
(walking, biking, etc.) can represent a gain in terms of reducing the energy intensity of travel (e.g.,
energy use per passenger-mile, Btu/mile). However, these options need to be assessed on a case-by-
case basis at the community level, given that there are wide variations in the energy and carbon
intensity of different passenger transportation options, which depend largely on the efficiency and
ridership of public transportation options. The energy intensity of light rail transit systems, for example,
can range from approximately 2,500 Btu/passenger-mile to over 30,000 Btu/passenger-mile (Davis,
Diegel, and Boundy 2012). In some cases, the energy intensity of a passenger-mile traveled on public
transit can actually be higher than for travel in a private vehicle. Therefore, the energy-use and GHG
reductions resulting from transit mode-share options must be carefully assessed to understand the true
magnitude of potential benefits, which requires understanding and modeling changes in both travel
demand and operational aspects of public transit. A number of methods are outlined in Transportation
Research Board of the National Academies (2013) that highlight tools for energy and GHG emissions
analysis, many of which incorporate transit options. Choices regarding public transit can also have
impacts on the deployment of alternative vehicle fleets and alternative fuels in a community, as will be
discussed more below. Cleaner-burning, alternative-fuel transit fleets can have important air quality
benefits. However, near-source issues, discussed in the previous section, must also be considered prior
to the implementation of public transportation options, as these options may in turn be constrained by
air and exposure issues affecting major transit corridors.
3.4.3	Fuel Economy and GHG Standards
While total travel demand and modal split are critical drivers of energy use, the most effective tool for
reducing the energy use and GHG emissions of the transportation sector is the setting of fuel economy
standards for light-duty vehicles, known as the Corporate Average Fuel Economy, or CAFE, standards.
Fuel economy standards in the U.S. were first set in 1975 as a mechanism to reduce imported oil in
response to the 1973 oil embargo. The standards have continued to evolve over the years, most notably
-------
56
in 2010 with the joint development of the Model Year 2012-2016 Light-Duty Vehicle Greenhouse Gas
Emissions Standards and Corporate Average Fuel Economy Standards by EPA and the National Highway
Traffic Safety Administration (NHTSA) (U.S. Environmental Protection Agency 2010d). This represented
the first time that GHG standards, under the authority of the Clean Air Act, were established along with
the fleetwide average fuel efficiency standards. In 2012, a joint rulemaking extended the GHG and fuel
economy standards to include model years 2017 to 2025 (U.S. Environmental Protection Agency 2012e,
U.S. Environmental Protection Agency and National Highway Traffic Safety Administration 2012).
These rules have important implications for the evolution of passenger vehicle fleets in communities
over the next 10-15 years. The standards are projected to result in an average industry fleetwide fuel
economy of 54.5 miles per gallon (mpg) (if achieved exclusively through fuel economy improvements)
for model year 2025 (U.S. Environmental Protection Agency 2012e). These fuel economy targets will
mean improvements in gasoline engines and transmissions, vehicle weight reduction, improved
aerodynamics, and other vehicle advancements. In addition, increased electrification of the fleet will
likely occur through the expanded production of hybrid vehicles, plug-in hybrid electric vehicles, and
electric vehicles. These fuel economy changes, and the potential increases in electrification, have
benefits for local air quality as well, given the large reductions in fuel use per vehicle-mile traveled.
Although these standards are set at the national level, any analyses or tools used to meet energy or GHG
reduction targets will need to account for changing vehicle fleets at the community level. Existing tools
used typically for air-quality purposes, such as the EPA's MOtor Vehicle Emission Simulator (MOVES)
model, as well as other emissions factor models such as California's EMFAC, can be applied to assess
GHG emissions and used in combination with travel demand models to develop a baseline of energy use
and GHG emissions (Transportation Research Board of the National Academies 2013). Perhaps more
important, however, is the need for communities to understand how they can influence changes in
vehicle and fuel technologies at the local level. A key knowledge gap and research need identified by
the Strategic Highway Research Program is "the effect of government interventions (e.g. pricing,
infrastructure deployment) on technology advancement...many local and state governments are
interested in how they can provide incentive to help accelerate the adoption of new vehicle
technologies" (Transportation Research Board of the National Academies 2013).
3.4.4 Operational Considerations Affecting Energy Use and Emissions
Changing vehicle fuel efficiency standards has a critical impact on reducing GHG emissions from the
transportation sector and from a research standpoint is perhaps the most studied aspect of
transportation energy use and GHG emissions. However, the real-world performance of vehicle
technologies and their resulting fuel use/GHG emissions will depend on their actual usage. A number of
factors can affect the usage and operation of vehicles, both conventional and advanced, and can be
considered broadly as system efficiency (Greene and Plotkin 2011). Some of these operational
considerations depend on the vehicle owner/driver, including maintenance issues that affect vehicle
efficiency (e.g., correct tire pressure) as well as driver behavior and resulting impacts on gas mileage.
For example, efficient driving behavior or "eco-driving" can encompass a number of approaches to safer
driving, such as smoothing out changes in vehicle speed, and avoidance of aggressive acceleration and
-------
57
starts/stops, as well as minimizing idling time. Understanding the potential for eco-driving in the U.S. to
reduce GHG emissions and energy use has been identified as a research need (Transportation Research
Board of the National Academies 2013). Support for an eco-driving "ethic" may be a strategy best
implemented in the context of sustainable communities, given that it would have other potential
community benefits, such as traffic congestion mitigation and safety. Some empirical evidence suggests
that this sort of improved driving can improve fuel economy between 5 and 20 percent (various studies
cited in Greene and Plotkin (2011)). System efficiency may also be achieved through more efficient trip-
making, possibly by utilizing GPS routing to find more efficient routes and to aid in trip-chaining, which
can reduce total VMT by combining trips and can reduce the number of cold starts.
There are also measures that can be taken by communities to improve traffic flow, both by promoting
speeds that improve fuel efficiency and by ensuring more constant speeds, instead of stop-and-go
traffic. These measures will also mitigate congestion, which can affect air quality, as well as quality of
life in terms of delays and time lost to commuting. Many congestion-mitigation and traffic-flow-
improvement measures are available to communities and can be implemented on roadways ranging
from freeways to smaller arterial streets. These can include ramp metering, incident management,
congestion pricing, speed limit enforcement, and improved traffic signalization and coordination. These
measures can have local, regional, and global benefits. For some heavily-congested facilities, such as Los
Angeles freeways, estimates suggest the potential of such measures to reduce C02 emissions by up to 30
percent when used in combination (Greene and Plotkin 2011). However, these measures must also be
used with caution, as the use of operational improvements may reduce travel times by increasing
speeds for single-occupant vehicles, potentially favoring travel in private vehicles over public transit.
3.4.5 Alternative and Renewable Fuels
In addition to measures affecting total demand for transportation energy and the roles of both vehicle
and system-level efficiency, looking at the fuel used by a number of different vehicle fleets in a
community is also important. At the federal level, the EPA is responsible for the Renewable Fuel
Standard (RFS) Program, ensuring that a minimum volume of renewable fuel, which can come from a
range of qualified biomass feedstocks and technology conversion pathways, is blended and sold in the
U.S. (U.S. Environmental Protection Agency 2010d). This program works to achieve multiple aims by
reducing lifecycle GHG emissions relative to petroleum fuels, reducing imported petroleum by displacing
gasoline and/or diesel use in the nation's light- and heavy- duty vehicle fleets, and expanding the
production of domestic renewable fuels and of feedstocks used to make them (U.S. Environmental
Protection Agency 2010d, e).
In additional to biomass-based renewable fuels, other alternative clean fuels can include natural gas,
propane, and hydrogen. However, the rates at which these alternative fuels are adopted in a given
community are influenced by a number of strategies that can be implemented by state and local
governments. These strategies generally involve either "provision of alternative fuels infrastructure" or
"direct purchase of alternative fuel vehicles for agency fleets" (Transportation Research Board of the
National Academies 2013). At the community level, much of what is needed is a detailed understanding
of different fuel and vehicle options, their associated energy use and GHG reduction benefits, and the
-------
58
factors that are involved in the implementation of new fuel and vehicle combinations. At a basic level,
one key implementation issue for a number of alternative-fuel fleets is the need for dedicated fueling
infrastructure and for people knowing where that fueling infrastructure is. The Department of Energy's
(DOE's) Alternative Fuels Data Center (AFDC) provides excellent information resources regarding
alternative fuels, vehicles, and related infrastructures. For example, the Alternative Fuel Station Locator
maps fueling stations, is accessible to the general public, and can be updated as new stations come
online regardless of the stations' ownership (www.afdc.energy.gov/locator/stations/) (U.S. Department
of Energy 2013a). Other resources include the Clean Cities coalitions of stakeholders, which can help
interested community stakeholders to understand alternative fuels and to receive technical assistance
supporting the deployment of alternative fuels and vehicle fleets (wwwl.eere.energy.gov/cleancities/)
(U.S. Department of Energy 2013b).
In addition to supporting infrastructure deployment, local governmental agencies are responsible for
large fleets of government vehicles and the management of public transit investments and operations
(Transportation Research Board of the National Academies 2011). As such, their decisions regarding
vehicle purchases can support the penetration of alternative fuels in a community. In fact, local and
state governments can often serve as leaders in the deployment of alternative fuels and vehicles by
using them for diverse fleets ranging from law enforcement, public transit, refuse collection, school
transportation, and shuttle services. Many already-successful efforts are documented as case studies by
the AFDC (www.afdc.energy.gov/case) (U.S. Department of Energy 2013a), and those case studies can
provide background information for communities interested in pursuing similar strategies.
3.4.6 Vehicle Electrification and Related Infrastructure
Infrastructure for alternative fuels such as biofuels or natural gas is one key link between the
transportation sector and energy system in a community. However, perhaps a more critical emerging
trend is the growth in vehicle electrification, which couples the transportation sector to the local electric
grid. The joint GHG and fuel economy standards, discussed above, will likely result in increased vehicle
electrification for light-duty vehicles. The questions of how and where the charging of electric or plug-in
hybrid vehicles will occur can be influenced in a number of ways by decisions and policies at the
community level. Many of these questions are being heavily researched, but because of electrically
powered vehicles' current small share of the transportation sector, there is relatively limited empirical
data regarding charging preferences and potential growth in the use of fully electric or plug-in hybrid
vehicles.
However, the build-out of electric charging stations is already occurring, with approximately 6,268
electric stations in the U.S., excluding private stations (www.afdc.energy.gov/locator/stations. last
accessed July 2013). This represents more than half of all alternative fuel stations in the U.S. (12,197
total), and therefore more stations than all other alternative fuels combined (including ethanol (E85),
biodiesel (B20 and higher), liquified natural gas (LNG), compressed natural gas (CNG), liquified
petroleum gas (propane), and hydrogen). Already, there are a number of government- and private-
sector-provided public charging spaces (836 of the electric-station owners documented by the AFDC in
-------
59
July 2013 were local governments) (U.S. Department of Energy 2013a). Given this growth in stations,
research on their impact on energy use and GHG emissions is needed.
There is the potential for synergies between the transportation-related choices made at the community
level and broader energy strategies. There are a number of research initiatives attempting to
understand the market for electric vehicles and charging patterns for this still-small but growing
segment of the fleet, as well as to understand potential impacts on electric-power dispatching.
However, beyond that is the question of how communities can promote synergistic solutions between
vehicle electrification and energy strategies, and how to quantify their benefits. For example, research
has been conducted to look at the question of whether plug-in vehicle buyers want green-power
electricity (Axsen and Kurani 2013). This could point to whether there are market preferences that
would support a combination of green power and plug-in hybrid or electric vehicles, as well as point to
policy and marketing strategies to advance these solutions (Axsen and Kurani 2013). However, there is a
lack of tools and resources available to communities to support these types of strategies. Another
question is that of at-home charging (occurring primarily at night), and how the built urban environment
can support, or perhaps even negatively affect, the prospects of electric-vehicle use and the ease of
charging at home. Therefore, vehicle electrification suggests a number of emerging questions related to
interactions between the built environment, private vehicle fleets, and the energy system, including the
question of what the relative impacts are of direct vehicle emissions and power-plant emissions that
result from generating electricity for plug-in vehicles.
3.4.7 Greenhouse Gas Footprint
Greenhouse gas footprint analysis and emissions calculators are approaches that involve estimation of
the GHG footprint of an entire system (or community), transportation service, or facility (Transportation
Research Board of the National Academies 2013). These can include registry- or inventory-based
calculators or life-cycle-analysis calculators (such as the Greenhouse Gases, Regulated Emissions, and
Energy Use in Transportation (GREET) model, the Lifecycle Emissions Model (LEM), GHGenius, or
Economic Input-Output Life Cycle Assessment (EIO-LCA)), many of which are spreadsheet-based, but
often are on the Internet or can utilize more complex software (see Chapter 4 of Transportation
Research Board of the National Academies (2013)). These calculations can then be used to set and track
C02 reduction targets such as: (a) annual percent reductions, (b) percent reductions in a future year
relative to a base year, or (c) absolute reductions in C02 or C02-equivalent, annually or for a target year
(Transportation Research Board of the National Academies 2013).
At the community level, another approach is visualization of the transportation carbon footprint in order
to compare transportation C02 emissions to the potential to biologically sequester that same amount of
C02. Under this methodology, total system fuel use can be used to estimate the amount of carbon
dioxide emitted over the course of one year. Data from sources such as state departments of natural
resources may then be used to determine how many acres of local forestland would be needed to
absorb and sequester that entire sum of C02, accounting for the likelihood and rate of sequestered C02
being rereleased through combustion or decay. Once this footprint has been established,
superimposing it on a map of the community allows a comparison of the carbon-footprint land area with
-------
60
the community's various other land-use needs. If the jurisdiction carrying out this analysis can meet all
of its other land use needs and still have enough natural areas to sequester all of the carbon dioxide
from its transportation system, that would be one potential indicator of sustainability (Chi and Stone
2005).
3.5 Water Issues and Related Tools, Resources, and Indicators
Transportation sources can impact surface and groundwater quality through deposited emissions from
fuel combustion, fluid leaks, and mechanical wear of brakes and tires. These emissions affect
watersheds through the atmospheric deposition of air pollutants and by being transported in
stormwater runoff along impervious surfaces (such as transportation infrastructure), which also
facilitate the movement of non-transportation pollutants into surface and groundwater.
Transportation-related projects also have the potential to significantly affect the sustainability of water
infrastructure systems.
3.5.1 Runoff from Impervious Surfaces
There has been a substantial body of research linking water quality and aquatic resources to
imperviousness, of which transportation infrastructure makes up a significant proportion. The
components of imperviousness are made up primarily of rooftops and the transport systems (roads,
driveways, parking lots) built to serve development. This transport component makes up the majority of
the impervious area created by development. For example, when a study measured imperviousness in
11 typical developments in the Olympia, Washington metropolitan area, the researchers separated the
percentage of a site's area devoted to roads, parking, and sidewalks and found that transportation-
related imperviousness comprised 63-70% of total impervious cover (City of Olympia 1995). Later,
Goetz et al. (2004) found in a GIS study in the Chesapeake Bay watershed that roads accounted for 36%
of all impervious surface area and the combination of roads, driveways, and parking lots accounted for
over 60% of all impervious surface area. The contribution of transportation infrastructure to
imperviousness is even more pronounced in rural and suburban areas, where roads must be longer to
reach homes that are farther apart, driveways must be longer to reach homes that are set back farther
from the road, and parking lots must be larger to accommodate more cars, since alternative modes of
transport are less common or nonexistent (Schueler 1994). However, rural areas still have the lowest
overall percentages of their land area covered in impervious surfaces.
Percent imperviousness in a watershed has been linked to overall watershed health in a number of
studies (Schueler, Fraley-McNeal, and Cappiella 2009). Currently, the most widely accepted "tipping
point" for watershed health degradation is 10% imperviousness. Once imperviousness reaches an
overall 10% share of land cover, there is noticeable degradation in the watershed. More recent surveys
of impervious-surface impacts on water quality generally find that adverse impacts are detectable when
percentage impervious surface is as low as 5% (Brabec, Schulte, and Richards 2002, Schueler, Fraley-
McNeal, and Cappiella 2009). More information on impervious land cover and its effects may be found
in the concurrent SHCRP Theme 4 synthesis paper on land use.
-------
61
Imperviousness caps are now being used in city and regional planning. For example, percent impervious
caps have been used as a planning tool in Montgomery County, Maryland for the Paint Branch
watershed (Montgomery County) and in Fairfax County, Virginia (Fairfax County). Land use controls and
stormwater best management practices that promote natural resource buffers and infiltration are now
commonplace in land-use site design practices.
3.5.1.1 Mitigation Strategies
There are two basic design strategies for mitigating stormwater runoff. The first strategy is to establish
a basin at the low point of a catchment area, holding rainwater until it either infiltrates into the ground,
evaporates, or is harnessed for some purpose (such as watering plants). This function may be served by
a rain garden, a constructed wetland, or a dry basin. The second design strategy for mitigating
stormwater runoff is to route the flow of water from upland areas to lowland areas across pervious
surfaces, which slow down the flow of surface water and allow more of it to infiltrate. To enact this
strategy, one may create vegetated buffers, including grass swales, downhill from a road or other
impervious surface, or one may replace an area of impervious pavement with an area of pervious
pavement (Guo et al. 2010).
There are many different types of pervious pavement (Gomez-Ullate et al. 2011, Scholz 2013). Both
concrete and asphalt may serve as materials for a pervious surface and water may either flow in
between paving blocks or through a series of pores in the pavement itself (Chai et al. 2012, Starke,
Gobel, and Coldewey 2010). Beneath the visible layer of a pervious pavement is a layer of aggregate,
the thickness of which must be greater in locations that receive more rainfall (Chai et al. 2012).
Potentially, pervious pavement could be used for the shoulders of a road, as opposed to replacing the
entire road, in which case the thickness of the layer of aggregate would also need to be proportional to
the width of the roadway that the pervious shoulders collect runoff from (Chai et al. 2012). Unlike other
design elements that mitigate runoff, pervious pavement does not require the use of any more land
than what the roadway currently occupies (Starke, Gobel, and Coldewey 2010), although some pervious
pavements may become clogged with sediment, reducing their effectiveness (Gomez-Ullate et al. 2011),
and the usefulness of pervious pavement is limited in places where the underlying soil is either already
saturated or a poor conductor of water (Chai et al. 2012). Like other runoff-mitigation tactics, pervious
pavement reduces flooding, improves water quality, aids in replenishing groundwater, and may be
equipped with a system for collecting (non-potable) stormwater for more immediate use by humans
(Gomez-Ullate et al. 2011, Starke, Gobel, and Coldewey 2010). Pervious pavement also results in more
evaporation than impervious pavement and causes evaporation to occur less rapidly following a rain
event, both contributing to the reduction of runoff and mitigating the urban heat island effect by
producing a density of moisture in the air that more closely resembles that which would be found in a
natural environment. However, pervious pavement is unable to produce as much beneficial evaporation
as natural soil and plants (Starke, Gobel, and Coldewey 2010). Currently, the use of pervious pavement
for transportation-related surfaces is mostly limited to parking lots, driveways, and some minor roads,
with no fully permeable pavement ever having been used on U.S. highways (Chai et al. 2012, Scholz
2013). Furthermore, there has so far been insufficient research, both experiment-based and simulation-
based, to determine the effect of pervious pavements on roadway design parameters (Chai et al. 2012).
-------
62
Another knowledge gap is whether or not it is advantageous to include a geotextile layer (a polymer-
based film) within the structure of a permeable pavement, with the theoretical benefit being the
filtering out of pollutants that take the form of suspended solids (Scholz 2013).
3.5.1.2 Assessing the Effectiveness of Mitigation Strategies
In a given catchment area or watershed, impervious surfaces may be divided between those that empty
onto a pervious surface and those that are contiguous with the low point and/or discharge point of the
catchment area/watershed, meaning that there is no pervious surface in between to mitigate runoff
effects. Likewise, pervious surfaces may be divided between those that receive runoff from an uphill
impervious surface and those that do not. Each of these two categories of impervious surface and each
of these two categories of pervious surface has a different effect on overall stormwater runoff, a fact
that is not reflected by simply calculating the percentage of an area that is impervious. Therefore,
calculating an area's "effective imperviousness" may be beneficial, wherein surface areas are weighted
according to whether or not they are part of a relationship where an impervious surface empties onto a
pervious surface and consideration is given to the ratio of uphill impervious area to downhill pervious
area in each of these relationships. Other important measures to consider are the capacity for runoff
storage in an area and the ratio of the local soil infiltration rate to rainfall intensity (Guo et al. 2010).
The EPA's Storm-Water Management Model (SWMM) is a useful tool for assessing stormwater runoff
impacts from impervious surfaces, able to model runoff volumes from impervious and pervious surfaces
in a network of catchments and subcatchments under many different circumstances. The SWMM is
primarily geared towards use in urban areas, but it also has nonurban applications (Guo et al. 2010).
Another program that simulates the effects of different kinds of impervious and pervious surfaces on
infiltration and runoff in a variety of contexts is the commercially-available HYDRUS model (Chai et al.
2012).
3.5.2	Groundwater Contamination from Leaked Fuel
Regardless of whether or not they are transported through stormwater runoff, transportation fuels
contain pollutants that leak from vehicles and storage facilities and enter the groundwater. As discussed
for vapor intrusion, a model is being developed by the EPA that accounts for the transport and fate of
fuel components in soil gas and groundwater. Groundwater contamination results from contact with
leaked vehicle fuels and other fuels. Constituents of fuels partition into groundwater and may form
plumes which impact either private or public drinking-water wells. The model being developed
addresses the mechanisms of weathering of gasoline and serves as an input for related groundwater
modeling work. The purpose of the model is to provide a basis for understanding the longevity and
behavior of difficult-to-remediate contaminated sites, which contribute to the current backlog of about
90,000 unresolved leaking-tank sites. The groundwater modeling is based on transport along
streamlines in flowing groundwater that connect sources to receptors.
3.5.3	Relationship between Transportation Infrastructure and Water Infrastructure
Automobile-based transportation and water-service infrastructures co-exist spatially in an urban
environment. Large diameter water mains are buried underground along main roads leading to
-------
63
population centers, where small diameter pipe branches serve subdivisions and households. Together
these two types of infrastructures in a metropolitan area form a spatial network that distributes urban
population and economic production, and hence defines physical centers of major production,
commercial, and living activities. One consequence of this relationship is that if transportation-system
expansion results in the rapid growth of an urban area, the water infrastructure necessary to serve that
area may become inefficient to operate. Another important connection between water and
transportation infrastructures is that the stormwater runoff collected by the gutters and drainage
ditches along roadways can significantly impact water quality.
3.5.3.1 Water Planning and Adaptation Methods: Implications for Transportation Choices
Water infrastructure in the U.S. has an extensive, far-reaching presence, providing service functions for
water supply - water treatment and distribution, wastewater collection and treatment, stormwater
drainage and urban flooding prevention. For example, drinking water infrastructure, a major asset of a
water utility, includes pipes, pumps, valves, storage tanks, reservoirs, meters, fittings, and other
hydraulic appurtenances that connect treatment plants to the taps of household, commercial and
industrial users. These drinking water systems are estimated to span almost 1 million miles in the
United States (Kirmeyer, Richards, and Smith 1994), with 3,200 miles (21,239 km) of new pipes installed
each year, including an estimated 154,000 finished water storage facilities (American Water Works
Association 2003). In addition to supplying potable water, water distribution systems must also be able
to provide water for non-potable uses, such as fire suppression, street watering, and irrigation of
landscaping. In wastewater management, 98% public wastewater treatment works are publically owned
and provide service to 190 million people or 73% of the population. Seventy-one percent serve less than
10,000 people and twenty-five percent of the population are not connected to centralized treatment,
but instead use some form of on-site treatment system such as onsite septic tanks. In total there are
approximately 600,000 miles of publically owned pipe in the U.S. (U.S. Environmental Protection Agency
2002b).
The planning and engineering practices for water infrastructure have been developed through many
decades, and are now embodied in the planning guidelines and engineering codes. In this long-held
practice, water supply and water management infrastructures are planned and designed for peak needs
in a projected future state of population and economic activities defined in urban master plans. Urban
master plans define how urban population and economic activity are distributed, setting up physical
boundaries for transportation choice and planning of water infrastructure services.
3.5.3.1.1 Current Practice
-------
64
The general community-development planning and engineering sequence begins with the development
of urban socioeconomic goals, followed by infrastructure planning and engineering, performance
monitoring, and assessment. A general planning-engineering-evaluation sequence is shown in Figure 4
Methods for transportation planning have been discussed in other sections.
For water infrastructure, master plans are commonly developed for a given set of land use and
economic projections. Major planning and engineering tools include EPANet and its commercial
derivatives (WaterCAD, H20map, etc.) for drinking water supply, EPA's SWMM and related stormwater
packages for stormwater management and urban drainage, and engineering software platforms (e.g,
SewerGems, H20Map/Sewer, HydraSewe, etc.)
General water infrastructure planning and engineering consists of three major steps, in sequence, as
land use and economic projection, analysis of spatial population distribution, and the projection of
water demand and wastewater generation. For the economy of scale, an initial monocentric urban form
favors a centralized water supply system, and mostly a single wastewater and stormwater management
network, if the hydrographic condition permits. A centralized water supply system delivers water from
treatment plants through a vast distribution network, in which energy is consumed and water quality
changes. In reverse, sewer systems collect wastewater from individual users to a central location for
Current Planning Framework
Growth factors
¦	Economic expansion
/ contraction
¦	Population change
" Life style change
Socio-physical
constraints
» Climate, carbon
allocation
¦	Water and land
availability
¦	Economics and
policy preference

Planning
Buildings -
Residential
8
5
c
E
o
?
Buildings -
Commercial
Buildings -
Industrial
Parks,
recreations
Green space &
eco. reservation
=>
Urban forms
1
Polycentnc
Low density
Green
development
Transformation
[ High density ]
Mixing
Energy source;
Source water
Material flow
Urban Sprawl
Evaluation assessment
¦	Carbon footpruit
(global & local)
¦	Water footprint
Efficiency
Economics
Satisfaction
Construction Phase
Operation Phase

<=>
Transportation Infrastructure
¦	Subway, mass transit
¦	Urban roads
J Mass transit
J Individual transit
•	Other roads (biking,
pedestrian, etc.)
¦	Tele-commuting
Water Infrastructure
•	Centralized
*	Water supply
J Storm water
S Wastewater
*	Waste and solid waste
Figure 4: General process of urban master planning and its relations to transportation and water infrastructure
engineering.
-------
65
treatment before discharge. Stormwater sewers drain an urban area and discharge overland runoff,
regulated by the National Pollutant Discharge Elimination System (NPDES) permit program. This
arrangement is the most cost-efficient for the water supply and management service of a mono-
centrically distributed population. Other factors such as topography, source water, available discharging
water bodies, and old combined-sewer systems can also affect specific engineering designs.
Monocentric urban formation is common in the U.S., wherein the population is distributed around a
single central business district (CBD), which contains high levels of economic activity. In the absence of
extensive transit systems, automobile-based mobility is a precondition to facilitate the form of urban-
suburban-exurban arrangement. Examples include numerous, mostly middle-to-large sized, urban
centers such as Las Vegas, Cincinnati, and Houston. As a city grows into a large metropolitan center, the
population becomes dispersed and the monocentric form evolves into a polycentric formation with
connected satellite cities. This is typical for very large metropolitan areas, such as New York City and Los
Angeles. Urban form transformation, and its implications for CBD formation, population distribution,
and transportation service, have been extensively investigated (e.g., Gordon, Richardson, and Wong
(1986), Small and Song (1994), Heikkila et al. (1989), Larson, Liu, and Yezer (2012), Garcia-Lopez (2012),
and Zhou et al. (2013)).
The transition of monocentric urban centers into polycentric forms is most relevant for U.S. cities and
metropolitan areas. As shown in Figure 4, the master planning process for urban development starts by
considering transportation choice and mobility; the improved mobility induces travel demand and
further facilitates urban expansion (Ewing 2008, Burchfield et al. 2006). This process commonly involves
planning, engineering, outcome monitoring, and simulation against a set of urban developmental goals.
Based on given developmental objectives, new developments continue to be planned, leading to further
urban sprawl, which can lead to uncontrolled expansion into exurbs and city perimeters (Figure 4).
Ewing (2008) examined the mechanisms responsible for this tendency, and concluded that public
transportation and improved highway systems are the primary enablers, accelerating urban expansion
along their routes or in a "leap-frog" pattern outward from urban centers. This type of "induced
demand" was discussed in an earlier section of this report.
Water infrastructure expands accordingly in the urban transformation to extend water services to new
developments (Figure 4). This process under the current planning framework is passive and in response
to transportation-facilitated urban expansion. Existing water infrastructure is expanded and adjusted in
its operation and management in order to meet the expansion-induced new water demand. Usually the
legacy of centralized configuration remains intact even after cities are transformed into a polycentric
form. Practical examples are numerous, such as the vast centralized water service infrastructure in Los
Angles and New York City. This tendency conforms to several notable attributes in urban planning:
¦ Water infrastructures, mostly buried in subsurface, are planned and designed to meet water
demand distribution given by the urban master plans. Once designed and built, the water
infrastructures and their functions have created a "lock-in" condition in heavily built urban
environment, making it difficult to change or modify the infrastructure framework in the future.
-------
66
¦	In centralized water systems, capital-intensive drinking water and wastewater treatment plants are
located away from urban centers. New developments can be served economically by expanding the
distribution and collection pipe network at relatively small capital costs, even at the expense of
operational and energy efficiency. However, there is a limit to this expansion before the existing
system configuration and operations become unsustainable without significant, fundamental
changes.
¦	Water infrastructure has its primary service functions in providing adequate capacity and reliability
to meet urban service needs, reduce capital and operational costs, and ensure compliance with
applicable Safe Drinking Water Act (SDWA) and Clean Water Act (CWA) regulations. System
efficiency and carbon footprints are often a secondary priority.
¦	Although subject to master development plans, urban water infrastructure is often engineered
independently from transportation infrastructures. The two types may become decoupled and
uncoordinated. This results in greater inflexibility of both infrastructures to change and adapt to
future functionalities and greater incompatibility between them.
Continuous urban expansion toward more dispersed, polycentric forms is a persistent trend leading to
uncontrolled urban sprawl. The trade-off in urban efficiency and sustainability of the two
infrastructures is under debate on subjects ranging from resource allocation and urban ecology to
engineering and operations (Small and Song 1994, Ewing 2008, Heikkila et al. 1989, U.S. Environmental
Protection Agency 2006, U.S. Environmental Protection Agency 2007a, Baynes 2009, Ostrom 2010).
Notable negative effects have been recognized, such as the challenges to reliable transportation and
mobility, as discussed in other chapters. For urban water services, centralized operation and
management allows for better control of water pollution and management over water regulations,
benefiting from economies of scale. Nevertheless, negative consequences of a centralized water
infrastructure can be found in multiple dimensions: excessive energy inefficiency and thus indirect
carbon emissions, barriers to resource recovery, and vulnerability to the impact of natural and man-
made incidences. Alternative approaches, especially in the form of decentralized water systems, have
been increasingly discussed. As monocentric urban form transforms into polycentric form, the
centralized water system faces technical and engineering challenges to evolve into a decentralized
framework. When this happens, the process requires coordinated urban planning and engineering
among transportation and water services.
3.5.3.1.2 Urban Form Transformation: Scenario Planning
The Intergovernmental Panel on Climate Change (IPCC) 3rd Working Group recently studied climate
change mitigation and adaptation in urban environments, and concluded that urban form
transformation has, by far, the single largest potential for urban efficiency improvement and carbon
emission reduction. Several common planning options are listed for urban transformation, including
infill, interior redevelopment, mixed land use, and employment centers. They have been applied across
U.S. cities (U.S. Environmental Protection Agency 2012d). These urban transformation measures,
capable of reversing or at least slowing urban sprawl, require changes in metropolitan transportation
and water services. Two approaches in urban form transformation have emerged — scenario planning
and urban smart growth.
-------
67
Specifically, each urban form has a set of characteristic physical layouts in water and transportation
infrastructure, with distinct operational properties. Thus each comes with economic costs, energy
consumption, carbon and water footprints, and the ability to provide desired services. In scenario
planning of existing metropolitan areas, the objective is often to determine the sustainability
parameters associated with each developmental option, and thus provide data for informed actions.
Water and transportation infrastructure, as the two major urban physical systems, can be planned and
engineered through coordinated scenario planning.
In principle, scenario-based urban planning is a systems-engineering approach to examine possible
urban development options. Such scenario-based analysis is conducted as a part of urban master
planning, involving planning, engineering, outcome assessment and re-planning for new development
(Figure 4). The outcome assists city planners and decision-makers to evaluate the capacity and
efficiency of existing transportation and water infrastructures, identify future improvement options, and
compare their benefits against a set of planning objectives. A technical plan consists of the following
major components:
¦	Population and land use planning and future projections
¦	Transportation analysis and planning, including a State Implementation Plan (SIP) on air
quality conformity
¦	Water infrastructure analysis and planning
3.5.3.1.2.1 Urban Population and Land Use Projection
Population and land use projections in planning scenarios are the most difficult and least quantifiable
among the three major components of scenario planning. Future population and land use is a function
of urban economic conditions and future economic policy initiatives that are less predictable or
quantifiable. As an approximation, cellular-automata Markov chain (CA-MC) simulation in GIS has been
used with model boundary conditions representing urban land policy restrictions. Tong, Sun, and Yang
(2012) and Sun et al. (2013) successfully projected future land use changes in the watersheds of the
Cincinnati and Las Vegas suburbs. Their modeling methodologies incorporated population and land use
variables as a GIS model filter in GIS CA-MC simulations.
Recently, ongoing EPA research attempted to simulate land use changes in the resolution of census
tracts in the Cincinnati metropolitan area. In this investigation, CA-MC methods supplemented by
restraints from water and land use policies were used to generate year 2030 land use patterns and
population distributions inside of Cincinnati for four development policy scenarios: 1) current
development or baseline; 2) infill; 3) high density development; and 4) mass-transit development. The
developed model was successfully calibrated against 1990 and 2000 urban land use and population
Census data. However, disruptive urban development policy and events can make the model
projections less accurate and useful. The disruption can violate spatial continuum assumptions
embedded in the semi-empirical CA-MC methodology. This potential problem and its challenge cannot
be under-estimated in urban land use and population projections.
-------
68
Separately, EPA has developed the Integrated Climate and Land-Use Scenario (ICLUS) tool and
projections of housing density and land use categories to 2100 under the IPCC Special Report on
Emission Scenarios (SRES). This large development scenario covering the entire U.S. is based on a pair of
models: a demographic model for population projection, and a spatial allocation model to distribute the
projected population into housing units at a 1-ha pixel resolution. Population allocation from a county
scale to census tract resolution is technically challenging, because of a set of model assumptions for
present, near-term and distant economic growth. For example, the spatially explicit regional growth
model (SERGoM) is used in population allocation to generate the projections at a spatiotemporal
resolution of 10 years and 1 ha. In generating high-resolution population maps, the allocation method
may produce model error and uncertainty excessive for infrastructure planning and engineering
purposes, although this potential has not been assessed. More details on the methodology can be
found in U.S. Environmental Protection Agency (2009b) and U.S. Environmental Protection Agency
(2010c).
3.5.3.1.2.2 Water Planning and Engineering
A wide range of tools and models are available for water infrastructure planning and engineering as the
result of decades of development. These include EPANET-based distribution modeling and design for
drinking water supply, EPA's SWMM and related commercial design packages (e.g., SewerCAD,
StormCAD) for storm water drainage, gravity and forced sewer systems. Essential to all tools and
models is GIS and topographic-information data-processing that describe urban spatial attributes of
infrastructures and their relationships. In urban planning, technical considerations are given to the
unique properties of each water infrastructure: gravity flow of wastewater and storm water systems,
pressured drinking water supply systems and pressure zone distribution, water treatment, monitoring,
and regulation compliance. All have significant implications to the physical components and layout,
service functions, energy consumption, and other sustainability attributes.
Noteworthy in water planning and design is that water system expansions are always built on the
existing infrastructure framework and physical footprints. The efforts are focused on component
optimization, system improvement and capacity expansion, while system-wide redesign and
reconstruction rarely happen. Urban form also evolves as the urban area grows, with some developing
into polycentric configurations. The transportation-induced redistribution in population and urban
activity can subsequently change the existing spatiotemporal configuration of water demand, for which
water service functions have to adjust. This continuum of urban growth is widespread across U.S.
metropolitan centers, with many experiencing urban sprawl (Figure 4). Under this condition, existing
centralized water infrastructure developed for a monocentric formation shows a mismatch with the new
water needs and existing water service functions and begins to experience difficulties with required
services.
3.5.3.1.3 Urban Form Transformation: Smart Growth and Water Systems Optimization
The second approach in urban transformation is based on smart growth, with the objectives of low-
carbon development, high urban density, and walkable and livable environments. The smart growth
concept initially developed for these general attributes is now propelled into specific urban and
-------
69
infrastructure planning activities. For example, U.S. Environmental Protection Agency (2012d) reported
national trends in smart growth adopted for urban development and urban renewal. Infill and green
field residential developments as means of smart urban growth have been gaining applications
throughout the U.S. So far, EPA has published a series of reports not only on residential smart
development, but also on transportation and water services of a sustainable community (e.g., U.S.
Environmental Protection Agency (2006), U.S. Environmental Protection Agency (2009a), and U.S.
Environmental Protection Agency (2011a)). Infill development and mixed transportation modes are
shown as being capable of reducing transportation-source emissions and improving system efficiencies.
For the urban water sector, infill development (within urban boundaries) and green infrastructure are
currently designed around water availability and the cost of providing reliable water services (U.S.
Environmental Protection Agency 2006). Less often they are considered as a part of smart urban growth
in coordination with transportation systems. Nevertheless, water system optimization is an essential
component to enable smart urban growth and reduce energy consumption for C02 emission avoidance.
EPA research has been developing planning support tools for regional water infrastructure planning
(Chang, Qi, and Yang 2012), water usage evaluation in urban planning scenarios (Wang, Burgess, and
Yang 2013), and energy optimization in water distribution (Yang, Chang, Neal, et al. 2013). Yang, Chang,
Li, et al. (2013) and Yang et al. (2011) further described evaluation criteria based on conjunctive use of
the carbon footprint and water footprint indices for urban infrastructure mitigation and adaptation.
These two footprint indices detail carbon emissions and water usage against the carbon allocation and
water availability of a given urban area.
Life-cycle analysis and Pareto optimization were the techniques used for planning scenario evaluation
and selection in Manatee County, Florida (Chang, Qi, and Yang 2012). A total of 20 developmental
scenarios in water supply infrastructure expansion were compared against their life-cycle carbon
emission and capital/operational cost. Individual structural components were analyzed for their life-
cycle impacts under different scenarios of infrastructure construction phases. The analysis was based on
a compromise non-linear programming model as the computation technique.
The EPA research programs (e.g., SHC, ACE, and SSWR) are developing other methods for individual
water infrastructure components. New developments, such as those in nutrient recovery, leak
detections, and green infrastructure, are focused on water infrastructure or its components alone.
Restraints from urban transportation, its mode and accessibility, are not yet considered. For integrated
planning, however, the smart growth methodology needs to simultaneously consider both types of
infrastructures in the evaluation of developmental options.
3.5.3.2 Known Effects of Transportation Decisions on Water Infrastructure Outcomes
In the current urban planning framework (Figure 4), water and transportation infrastructure
improvements are frequently made in the period between two adjacent master planning events. The
uncoordinated infrastructure work potentially creates a "lock-in" condition in urban water
infrastructure, which can hamper future water service optimization and sustainability. These inter-
infrastructure interactions were briefly described in the preceding subsections. Alternatively, smart
-------
70
urban growth shifts the development paradigm to adaptive planning, readjusts developmental goals,
and enables water service function changes through a combination of gray and green water
infrastructures. This systems-approach-process is shown in Figure 5.
Here the impacts of transportation decisions on water services and infrastructure sustainability are
outlined. Discussion is made in reference to the existing development framework and the smart-growth
transformation shown in Figures 4 and 5, respectively.
Adaptive Planning Framework
Planning
Growth factors
¦	Economic expansion
/ contraction
¦	Population change
¦	Life style change
Socio-physical
constraints
¦	Climate, carbon
allocation
¦	Water and land
availability
¦	Economics and
policy preference
Buildings -
Residential
Buildings -
Commercial

Buildings -
Industrial
Parks,
recreations
Green space &
eco. reservation
Urban Adaptation
Energy source;
Source water
Material flow
4>
Monocentric
Transformation
districts
Polycentric
J
j" High density j
Mixing
Green
development
Minimize New
Developments

Economic and policy adjustment
Evaluation assessment
¦	Carbon footprint
(global & local)
¦	Water footprint
Efficiency
Economics
Satisfaction
Construction Phase
Operation Phase


Transportation Infrastructure
-	Subway, mass transit
¦ Urban roads
S Mass transit
•s Individual transit
-	Other roads (biking,
pedestrian, etc.)
• Tele-commuting
Water Infrastructure
Centralized
S Water supply
s Storm water
S Waste water
^ Waste and solid waste
Figure 5: General process of adaptive urban planning and engineering. Compared to traditional master planning (Figure
4), adaptive planning promotes economic and policy adjustment on urban development goals and urban adaptation for
high-density, polycentric form through transformation and proper transportation and water infrastructure adjustment.
3.5.3.2.1 City Development Examples
3.5.3.2.1.1 Rapid Urban Sprawl
Many metropolitan areas with rapid growth experience undesired impacts that were facilitated by their
transportation choices. The experience of Las Vegas illustrates this. Las Vegas has rapidly increased its
physical urban footprint over the past decades. The urban population increased from 273,288 in 1970
to 741,368 in 1990, and approached 1,900,000 by 2010. Correspondingly the urban area increased by
2.9 times from 58 square miles to 170 square miles from 1970 to 1990. The majority of the urban
expansion occurred along major highways, US-95, 1-15, and 1-215 extending from the valley toward Lake
Meade. Accompanying this rapid growth, the metropolitan water distribution system has expanded,
consisting of >4,500 miles of water pipes, 65 pump stations, and 68 water storage facilities. The
-------
71
network is currently expanding into the Summerlin area to the west, responding to new road and
housing developments. Through this history of development, the water distribution system remains
centralized, with two water treatment plants and a single distribution network, at significant energy cost
for pumping. Lake Mead to the southeast is the primary water source for the city.
Rapid expansion of the centralized water infrastructures met the urban development needs, but also
brought technical challenges to the required water service objectives. During the economic depression
starting in the late 2000s, vacancy rates were high and water demand was chronically low compared to
the preceding housing boom periods. This change led to over-capacity in the water distribution system,
and excessively long water ages or water residence time in many parts of the distribution network. A
direct consequence was high levels of disinfection by-products (DBPs) (e.g., chloroform) in the tap
water, risking violation of the Safe Drinking Water Act (SDWA) DBP Stage-ll regulations.
Many metropolitan areas with rapid growth have experienced problems of a similar nature in water
supply and water sanitation. A lack of adaptability of centralized water systems to further urban
development or changes is a common cause. Such an observation can be made in old industrial cities in
the Midwest, such as Detroit, Cleveland, Youngstown, etc., where long periods of economic depression
in urban centers are followed by renewed redevelopment efforts in recent years. As an extreme
international example, unprecedented urban sprawl in Beijing, China is accelerated by the rapid
expansion of car and mass-transit transportation systems along the continuously expanding ring roads
around the city center, leading to ecological and water environmental problems (Wang et al. 2007).
Because of flat topography in the Beijing metropolitan area, urban drainage and stormwater
management becomes a daunting challenge as the monocentric city expands. Inner-city flooding
occurred in the last three years with over 86 people killed in flooded roads during a flash flood in
October 2012. Although less severe in magnitude, urban drainage problems and investment in runoff
control have been a major issue for many U.S. metropolitan areas.
3.5.3.2.1.2 Leap-Froa Development and Small/Community Water Systems
Highway transportation and mass transit have resulted in "leap-frog" new urban development centers
— isolated developments, mostly in exurban areas, that exist before further development fills the
undeveloped areas between. Sun et al. (2013) show through land use modeling the likelihood and
evolution of such development in the Las Vegas metropolitan area. In general, leap-frog development is
characteristic of small-community settlement, propelled by highway development and mass transit, in
areas away from the urban continuum (Figure 6). Such development can be for the purpose of
affordable land, natural scenery, or valuable but low-density estate properties.
Leap-frog development, fueled by transportation infrastructure, has direct implications for the form of
water services. Such development often has no existing water and wastewater services from water
utilities, tending to be served by small water systems or on-site treatment, like septic tanks. Examples
can be found outside Washington, DC. Small-scale, operationally simple, decentralized community or
household systems are preferable choices in these areas. As described in U.S. Environmental Protection
Agency (2002a) and U.S. Environmental Protection Agency (2007c), these small systems employ less
-------
72
robust water treatment technologies, receive little operational and maintenance attention, and
frequently are vulnerable to performance upsets. The environmental consequences have been well
documented, particularly nutrient pollution from small wastewater systems. For mitigation, the states
of Maryland and Virginia have enacted laws and regulations for these small community water and
wastewater systems.
[|High Density Sprawl
I I Medium Density Sprawl
| | Low Density Sprawl
Radial Sprawl	Ribl,on	L"pf[00
Sprawl	Sprawl
Figure 6: Urban sprawl and leap-frog development. Reprinted from Sudhira et al. (2005).
3.5.3.2.1.3 High-density Urban Centers
Urban areas of high population density are developed with the assistance of mass transport. They often
are the focus of water master planning and operational adjustment, due to high water consumption and
vulnerability to service disruptions. Unique requirements include steady and reliable water supply for
domestic consumption by a large population, daily and seasonal variation in commercial services, and
high-rise fire fighting, in addition to wastewater and storm water management in regulation compliance.
Transportation-supported high-density urban forms can significantly affect water services. In the U.S.,
many high-density urban centers are also old development districts with aged water infrastructures and
compromised pipe integrities. These high-density urban centers are characteristic of old business and
urban centers, such as those in New York City. Constant transportation improvement in these cities
reinforces their high population density in a geographically restricted area. Many historically deprived
high-density urban centers are now redeveloped with the assistance of improved transportation
services. These urban centers are characterized by high population concentrations, a high fraction of
impervious surface, nearly 100% land coverage, and frequently old and structurally-compromised water
infrastructures. Old structures for combined sanitary and stormwater sewers are common in U.S.
Northeastern, Eastern, and Midwestern urban centers. Such combined sewers are vulnerable to
overflow in heavy rain events, sending raw sewage into waterways, a problem which is exacerbated by
runoff from impervious surfaces that is funneled through roadside curb-and-gutter systems.
These conditions make it difficult to improve or redevelop underground water infrastructures. For this
reason, mitigation and correction of combined sewer overflow (CSO) problems have proven to be
difficult in high-density urban areas.
Another problem for old high-density urban districts occurs in the operation of drinking-water piping,
wherein persistently high rates of water loss from drinking-water pipes occurs, particularly between
-------
73
points with a large elevation difference. For these areas, pressure zone management is a challenge to
meet the varying water demand in a day, pressure requirement in building fire code, and reduce the risk
of damaging aged and structurally compromised underground water pipes.
Urban water and transportation infrastructure improvement can be coordinated to avoid or minimize
these problems in high-density urban centers. Urban redevelopment presents an opportunity for
integration of the two systems (U.S. Environmental Protection Agency 2009a). One example is the use
of green infrastructure (vegetated natural systems instead of "grey infrastructure," or engineered
concrete systems) as a part of smart urban growth measures, described below.
3.5.3.2.2 Smart Growth
Smart growth in urban development is often defined differently in literature than in practice. In terms
of urban development, smart growth refers to high population density, low carbon intensity, walkability,
and green infrastructure, thus creating a livable environment. This premise requires the
accommodation of mass transit and other transportation infrastructures. The transportation policies
contained in U.S. Environmental Protection Agency (2011a) all aim to increase urban population density,
create a walkable and pedestrian-friendly community, and maximize the use of existing infrastructure
through infill and redevelopment. U.S. Environmental Protection Agency (2011a) outlined major types
of transportation modes with the following sustainability metrics:
¦	Transit accessibility, for reduced VMT and increased public transportation
¦	Bicycle and pedestrian mode share
¦	VMT per capita, to reduce automobile reliance and commute distances
¦	Carbon intensity (of transportation) per capita
¦	Percentage land-use mixture in terms of residential and commercial activities
How these changes toward smart growth affect urban water infrastructures and, in return, how the
water services facilitate or impede smart growth have received little attention. Many questions persist,
ranging from paradigm changes in the water supply and water management, the use and application of
green infrastructure, and transitioning the existing grey infrastructure to the new developmental
paradigm. Green stormwater infrastructure has been a focus area, wherein EPA has conducted a series
of research and site studies, as well as provided significant information on its website. In this context,
green infrastructure improves surface water quality (U.S. Environmental Protection Agency 2009a),
retards transportation pollutant transport (Baldauf et al. 2008), and enhances the livability of an urban
environment.
3.5.4 Transportation Fuel Choice and Impacts on Water Resources
Water usage in biofuel production for automobile fleets also links water resources to urban
transportation sustainability. These two categories, on the surface, do not reveal much of the inter-
dependency. The relationship, however, may be revealed through close examination of upstream water
usage in alternative transportation fuel production. The transportation choice of renewable fuels is
driven by U.S. national energy policy. More extensive biofuel use in the future will consume a significant
portion of available-but-limited water resources for energy-related biomass production in competition
-------
74
with other beneficial uses. Additionally, associated new transportation technologies will require
modification to current fuel supply chains and distribution, travel patterns, VMT distribution, and,
importantly, the emission characteristics at the road and in near-road environments. This future state
of the transportation system has been discussed and is summarized in preceding sections. Information
and data, however, are limited at the present, hampering an informed decision on future urban
developmental sustainability.
The use of renewable and alternative fuels in the transportation sector has significant implications in
multiple dimensions of water sustainability. EPA Office of Transportation and Air Quality (OTAQ) has
been conducting an Integrated Planning Model (IPM) assessment to analyze fuel change impacts both in
upstream and downstream environments, with a focus on the power sector. The upstream impacts on
water are largely in the form of water usage and wastewater generation during biomass and fuel
production. Downstream, the impact is anticipated to be in the form of automobile emission changes
for future fleets, urban forms (i.e., buildings, road network, population and activity distribution), and
climate conditions.
Water consumption in biomass production for renewable fuels has been extensively investigated and
discussed (e.g., National Research Council of the National Academies (2008), U.S. Environmental
Protection Agency (2011a), Dominguez-Faus et al. (2009), Berndes (2002), and de Fraiture, Giordano,
and Liao (2008)). In a triennial report to Congress, the U.S. Environmental Protection Agency (2011b)
described significant water quality and quantity impacts in the production of bioethanol and biomass-
based biodiesel. This nationwide large-scale assessment is further supplemented by ongoing research
work in EPA's Air, Climate, and Energy (ACE) research program, including detailed region-specific
analysis using MARKAL (a model that can optimize energy mixes to accomplish any given function)
(Dodder et al. 2011), and by other work on water usage and wastewater generation in the biofuel
production process (Lingaraju, Lee, and Yang 2013). This existing pool of knowledge points to a larger
water footprint of biofuels than of conventional petroleum counterparts. This impact on water
resources, particularly in the form of water-availability competition against both urban water supplies
and ecological water needs, is a prominent issue in water-poor regions.
The downstream impacts of biofuel usage on emissions have been studied at national scales (U.S.
Environmental Protection Agency 2010a), while the upstream impacts have been reported as well. One
significant recent study to address the emission impacts of biofuels is the EPAct/V2/E-89 Tier 2 Gasoline
Fuel Effects Study (http://www.epa.gov/otaq/models/moves/epact.htm) (U.S. Environmental Protection
Agency 2013e). Numerous published studies (e.g., U.S. Environmental Protection Agency (2002b)) have
shown that biofuel combustion in current-fleet engines leads to a meaningful reduction in some
pollutants (e.g., particulate matter (PM), carbon monoxide (CO), and hydrocarbons), but an increase in
others (e.g., aldehydes and nitrous oxides (NOx)). Chemical and physical properties of PM emissions also
change. These changes are important to near-road air pollutant monitoring and mitigation options
(Baldauf et al. 2008). Their potential effects on land use decisions, and hence water infrastructure
planning, have not been thoroughly investigated.
-------
75
3.6 Human Weil-Being Issues and Related Tools, Resources, and Indicators
Human well-being consists not just of physical health outcomes, but also of psychological and social
health outcomes, economic outcomes (discussed in the next section), and equity outcomes. More
information on the effects of transportation projects and policies on human health and well-being may
be obtained from the Centers for Disease Control and Prevention (CDC) at
http://www.cdc.gov/transportation/default.htm.
3.6.1 Physical Health Outcomes
The transportation sector is increasingly recognized as a critical determinant of human health outcomes.
In addition to public health's traditional focus on vehicle emissions and injuries as sources of morbidity
and mortality, new areas of investigation are enhancing our understanding of the diverse pathways
through which the design and use of transportation systems affect human health. These impacts can be
direct or indirect and exert their influence over a wide range of temporal and spatial scales. They not
only affect human well-being, but represent a costly externality of the transportation sector that adds
additional burden to the rising cost of healthcare in the United States.
3.6.1.1 Direct Impacts
Direct health impacts from the transportation sector result primarily from exposure to environmental
contamination or unintentional injuries. Effects may be felt by users of the transportation system or
those in proximity to it or its supporting industries.
3.6.1.1.1 Air Pollution
A large body of literature has characterized the adverse health outcomes associated with particulate
matter, volatile organic compounds (VOCs), SOx, NOx, carbon monoxide, and ozone that result from
tailpipe emissions produced from the combustion of fossil fuels. These pollutants, addressed elsewhere
in this synthesis paper, contribute to significant cardiovascular and respiratory morbidity and mortality.
While exposure carries well-described risk, factors that determine level of exposure are less clearly
defined. Two important considerations for this exposure route are that: children are disproportionately
affected, because they breathe in more air for their body size than do adults; and economically-
disadvantaged populations are often disproportionately affected by living closer to roadways and other
transportation sources (e.g., ports, rail yards) than wealthier populations.
Growing interest in shifting a greater percentage of mode share to active forms of transportation
(primarily walking and bicycling) has raised questions about the impacts of such a shift on exposures to
air pollutants. While the overall effect of decreased automobile use might be reduced emissions of
criteria pollutants and particulates, and thus lower levels of overall population exposure, some have
suggested that individuals who choose to walk or bike might receive disproportionately high exposures
when compared to motor vehicle or transit passengers. Existing research is inconclusive (Knibbs, Cole-
Hunter, and Morawska 2011, Zuurbier et al. 2010, McNabola, Broderick, and Gill 2009). However, some
studies have shown that while walkers and cyclists are actually exposed to lower or similar
concentrations of pollutants than motor vehicle passengers, the increased gas exchange from physical
exertion and travel time associated with active transportation significantly modifies total inhaled or
-------
76
deposited doses for these mode choices (Int Panis et al. 2010). Given strong inverse associations
between distance from tailpipe emissions and level of exposure, there is evidence that providing greater
separation between bicycle/pedestrian facilities and roadways could modify this effect. Further
research is needed to more effectively quantify relative exposure by mode of transportation while
properly accounting for associated levels of gas exchange and duration of exposure, in addition to
accounting for the substantial health benefits of walking and cycling (see below).
Transportation-associated air pollution extends beyond tailpipe emissions. Pollution, and its associated
health impacts, can occur throughout the production of any vehicle fuel source. For example, while
increasing dependence on electricity as an energy source for personal and commercial vehicle use (all-
electric or plug-in hybrid) may reduce harmful tailpipe emissions, the overall impact on human health
depends largely on the fuel stock used to generate electricity. Use of renewable energy sources could
reduce air pollution while continued reliance on coal could merely transfer the source of harmful
emissions and change the affected population. Comparing the relative benefits and harms to human
health of any transportation energy source requires consideration of emissions from the full lifecycle of
energy production, as previously discussed in the context of greenhouse gas emissions, and the
populations affected.
Air pollution occurs within both the operational and non-operational components of the transportation
sector. Operational components refer to pollution stemming directly from vehicle use while non-
operational components refer to the manufacture, distribution, and disposal of vehicles and their
supporting infrastructure (including roads). Comparison of the relative health impacts of
transportation-associated air pollution requires evaluation of emissions from both components. Energy
consumption and production of criteria air pollutants from non-operational components of
transportation can exceed that of operational components by orders of magnitude. The health impacts
of air pollution from non-operational components of the transportation sector could therefore be
significant, although this effect will be mediated by both the fuel stock used for energy production and a
population's level of exposure.
3.6.1.1.2 Noise
A growing body of evidence has found significant associations between environmental noise and
adverse impacts on human health. A World Health Organization study in western Europe found that
traffic-related noise was responsible for over 1 million disability-adjusted life years (DALYs), a measure
of healthy life years lost, primarily due to sleep disturbance and annoyance (Fritschi et al. 2011).
Despite these conclusions, evidence linking sleep disturbance and annoyance caused by road-traffic
noise to specific health outcomes is limited (Hume, Brink, and Basner 2012). However, positive
associations between road-traffic noise and ischemic heart disease and hypertension have been
demonstrated with greater certainty, even when controlling for air pollution as a potential confounder
(Davies and van Kamp 2012).
3.6.1.1.3 Injuries
-------
77
Transportation-related injuries are a leading cause of morbidity and mortality in the United States, with
32,885 individuals killed and 2.24 million injured in traffic crashes in 2010. Fatal and nonfatal injuries
among motor vehicle occupants have declined - per 100,000 population and per vehicle mile traveled
(VMT) - over the past two decades, largely due to advances in roadway and vehicle engineering as well
as interventions targeting seat belt use and drunk driving. Fatal and nonfatal injuries have also declined
among pedestrians and cyclists during this time period. Nonetheless, traffic crashes remain the leading
cause of death among individuals aged 5 to 24, one of the top ten leading causes of death among all age
groups, and the fourth leading cause of nonfatal injury treated in emergency departments (Centers for
Disease Control and Prevention). In addition to their devastating emotional toll, traffic crashes cost $99
billion annually in medical expenses (Naumann et al. 2010).
Critical to evaluation of the relative health benefits of motor vehicle transportation versus active forms
of transportation is an understanding of their associated risks for injury. Unfortunately, comparing
these risks is challenging. While many more motor vehicle occupants are killed and injured than
pedestrians and cyclists each year, the extent to which this reflects higher levels of exposure versus
increased risk is unclear. The absence of an agreed upon, consistent, and appropriately comparable
metric for standardizing level of exposure limits current efforts to assess relative safety. Vehicle miles
traveled has commonly been used to standardize fatality and injury rates for automobiles. Although
various measures have been proposed - including number of walkers and cyclists, number of trips, time,
and distance traveled - none is commonly accepted for pedestrians and cyclists.
Experience from Europe demonstrates that pedestrian and bicyclist safety can be enhanced through a
combination of engineering, educational, and policy interventions (Fischer et al. 2010). In Germany,
Denmark, and the Netherlands, the risk of cycling - as measured by number of fatalities per distance
cycled - is three to five times lower than in the United States (Pucher and Buehler 2008). Perhaps more
impressively, risks associated with cycling have been so successfully reduced in these European
countries that, despite rising levels of cycling over the past four decades, the total number of cyclist
fatalities has decreased. The frequently observed inverse association between absolute numbers of
pedestrians and cyclists and risk for injury is commonly known as safety-in-numbers (Jacobsen 2003).
Some have suggested that the safety-in-numbers phenomenon reflects a change in driver behavior in
the presence of increased numbers of pedestrians and cyclists, and thus provides support for policies
that promote walking and cycling as a method by which to enhance safety. However, such interventions
may not be justified based upon the primarily cross-sectional data that underlies the safety-in-numbers
principle (Bhatia and Wier 2011). Increases in levels of walking and cycling may have occurred in
response to interventions that enhanced pedestrian and cyclist safety, or perceptions thereof.
Alternatively, if increased numbers of pedestrians and cyclists reflect a shift in mode share from driving
to active transportation, decreases in injuries may be due to reductions in traffic volume, a known
predictor of crash frequency.
Enhancing pedestrian and cyclist safety is not only possible, but is likely essential to achieving significant
shifts in mode share. Although beyond the scope of this review, numerous studies have demonstrated
-------
78
that the design of pedestrian and bicycle infrastructure can both reduce injuries and influence
transportation decisions through individuals' perception of safety (de Nazelle et al. 2011).
3.6.1.2 Indirect Impacts
The transportation sector affects human health indirectly through diverse pathways. Transportation
system design and access to transportation services influence health behaviors, as well as other critical
social determinants of health, such as access to essential services, personal finance, and social
interaction. Indirect health impacts also occur through environmental effects operating at a global
scale, such as climate change.
3.6.1.2.1 Health Behaviors
Regular physical activity can help prevent or treat many chronic diseases, including hypertension,
coronary artery disease, stroke, type 2 diabetes, breast and colon cancer, and osteoporosis and may be
protective against depression (Warburton et al. 2010, Teychenne, Ball, and Salmon 2008). Furthermore,
physical activity protects against obesity, which is a risk factor for many chronic health conditions. Yet,
despite these known health benefits, nearly 40% of US citizens fail to meet recommended levels of
exercise through a combination of leisure-time, occupational, and transportation-associated physical
activity (Tucker, Welk, and Beyler 2011). The individual and societal price is high. Physical inactivity is
responsible for nearly 11% of all deaths in the United States (Murray et al. 2012), and has been
estimated to cost the US $250 billion annually in healthcare expenses, workers' compensation, and
productivity losses (Chenoweth and Leutzinger 2010).
Recently, many have suggested that decisions within the transportation sector have contributed to and
could offer a solution for high rates of physical inactivity in the United States. Greater reliance on the
automobile for transportation at the expense of public transit and active forms of transportation has
reduced population levels of transportation-associated physical activity. Concerns about safety, tied
closely to individuals' perceptions of protection afforded by pedestrian and bicycle infrastructure, are an
important deterrent to increased levels of walking and cycling for transportation (de Nazelle et al. 2011).
A growing body of literature suggests that efforts to shift mode share from motor vehicles to transit and
active transportation could have impacts on physical activity levels and associated health outcomes.
Although high quality intervention studies are limited, observational studies of active transportation
have identified associations with reductions in cardiovascular disease outcomes (Hamer and Chida 2008)
and obesity (Wanner et al. 2012). Walking to and from public transportation can help individuals obtain
recommended amounts of physical activity (Besser and Dannenberg 2005). In particular, populations in
urban areas with access to rail transit are more likely to meet guidelines for physical activity that other
populations (Freeland et al. 2013). Modeling exercises suggest that reducing automobile usage by 1%
would result in a 0.4% reduction in the chance of obesity, a 0.3% reduction in the chance of high blood
pressure, a 1.3% reduction in the chance of high cholesterol, and a 1% reduction in the chance of heart
attack (Samimi and Mohammadian 2010).
Development density, street connectivity, traffic calming features, and pedestrian safety measures are
all aspects of the built environment that are associated with higher rates of active transportation.
-------
79
However, research has yet to establish the degree to which these factors affect the overall health of
community members (Samimi and Mohammadian 2010). In fact, the amount that people walk and ride
bicycles is more greatly affected by the factors of inhospitable topography, darkness, inclement
weather, and the demographics of the travelers. The built-environment feature that most strongly
predicts more walking is the presence of retail establishments near people's homes. The built-
environment features that most strongly predict more cycling are density, mixed land uses, and
favorable neighborhood design characteristics, especially when these things are found at the home end
of people's trips (Cervero and Duncan 2003).
People are also more likely to travel by active modes in neighborhoods with good aesthetics, safety
features, and amenities. Conducive aesthetic features include green space, attractive architecture, and
the absence of noise, air pollution, and physical disorder. A safe walking environment is one where both
crime and traffic accidents occur at low levels. Amenities conducive to walking include public benches
and sidewalk retail and restaurants. Also, pedestrians generally prefer an environment where numerous
other pedestrians are around, whereas drivers generally prefer an environment where there are as few
other drivers as possible. An advantage of using aesthetics and safety upgrades to promote physical
activity is that it takes less time to implement than increasing urban densities or changing land-use
patterns. Another advantage is that improvements to aesthetics and safety are easy to convey to
community members. Research is still wanting on the relative impacts that various aesthetic and safety
features have on physical activity. More research is also needed on how people's demographic and
socioeconomic characteristics affect their walking behavior when all other factors are controlled for.
Finally, research could shed more light on the degree to which pedestrians regard public spaces as either
a destination or an obstruction (Neckerman et al. 2009).
3.6.1.2.2 Social Determinants of Health
A large body of research has characterized the importance of social and economic factors on health and
well being, broadly referred to as social determinants of health (Braveman, Egerter, and Williams 2011).
Transportation - itself a determinant of health (Wilkinson and Marmot 2003) - influences other social
determinants of health such as access to healthcare, personal finance, employment opportunity, and
social interaction. Decisions about transportation, and their ultimate health effects, reflect tradeoffs
between the need to access housing, services, and opportunities and the money and time people spend
on transportation. The magnitude of this impact is often inequitably distributed, creating a greater
burden on those with more limited transportation options such as minority, economically
disadvantaged, disabled, and rural populations.
Among those with a usual source of outpatient health care, lack of transportation can be a major factor
causing individuals to seek care in an emergency department (ED) rather than with a regular provider
(Rust et al. 2008). Seeking care for nonurgent conditions in the emergency department can contribute
to worse patient outcomes, ED crowding, and higher healthcare costs (Hoot and Aronsky 2008).
Expenditures on transportation represent the second largest household expense behind housing
(Surface Transportation Policy Project 2003). Transportation costs consume a disproportionately large
-------
80
percentage of economically disadvantaged populations' income, an important social determinant of
health (U.S. Department of Labor 2012). Those who rely on personal automobiles for commuting are
particularly affected (U.S. Department of Transportation 2003a). In efforts to obtain affordable housing,
many households accept greater transportation costs associated with commuting, as opposed to
choosing to save money on transportation through location efficiency. In addition to higher commuting
expense, greater spatial separation from places of employment results in increased commute times that
can compromise individuals' ability to find and hold jobs, as well as their time available for other
priorities, such as accessing child care. Limited transit options or frequency place a greater burden on
those dependent upon public transportation for commuting (Roberto 2008).
Increased commute time can also have an impact on social interactions, another well-described social
determinant of health (Holt-Lunstad, Smith, and Layton 2010). Longer commutes have been associated
with fewer trips taken for social purposes and decreased time spent with family (Besser, Marcus, and
Frumkin 2008, Christian 2012a), as well as decreased time spent in health-related activities (Christian
2012b). Access to transportation also affects levels of social interaction and social isolation, both of
which have been associated with health outcomes (Fujiwara and Kawachi 2008, Holt-Lunstad, Smith,
and Layton 2010, Dickens et al. 2011). Older populations are at particularly high risk of social isolation
and its attendant health risks. Disabled individuals disproportionately report being unable to leave their
homes due to transportation difficulties (U.S. Department of Transportation 2003b).
3.6.1.2.3 Air Pollution and Climate Change
As previously discussed, exposure to air pollution emitted by vehicles has been shown to increase risks
of adverse health effects for a number of outcomes, including respiratory, cardiovascular, cancer, and
premature mortality. These exposures can occur from being close to transportation sources and the
emissions from vehicles using transportation facilities, or from more regional air-quality deterioration
from the cumulative effect of vehicle emissions within an urban or larger area. These regional impacts
can increase concentrations of air pollutants that are directly harmful to human health. While exposure
to air pollution increases public health risks, two studies suggest that the risks from air pollution and
injury are outweighed by the health benefits of increased physical activity from walking or biking (de
Hartog et al. 2010, Rojas-Rueda et al. 2011), although these studies focused on regional air pollution
impacts and did not account for elevated exposures near roadways and other transportation sources.
Global climate change is increasingly recognized as a significant public health threat, and, as discussed
previously, the transportation sector is a key contributor to greenhouse gas emissions. Climate change
is anticipated to affect human health through diverse mechanisms including heat waves, extreme
weather events, changes in air quality, impacts on ecological systems that affect food production,
vectorborne disease transmission, water quality and quantity, and short- and long-term population
displacement due to changing environmental conditions (McMichael, Montgomery, and Costello 2012,
Confalonieri et al. 2007). In the year 2000, the World Health Organization estimated that climate
change was responsible for 150,000 deaths globally each year (Patz et al. 2005). In the United States
between 2000 and 2009, six climate-related events of the type projected to increase with rising global
-------
81
temperatures were cumulatively responsible for an estimated $14 billion in health damages (Knowlton
etal. 2011).
3.6.1.3 Tools
With growing recognition of and interest in the diverse health impacts of the transportation sector, a
number of tools and processes have been developed to evaluate the potential health effects of
transportation planning decisions.
Health Economic Assessment Tool (HEAT) for Walking or Cycling
The World Health Organization's Health Economic Assessment Tool (HEAT) for walking or cycling
(Kahlmeier et al. 2011, World Health Organization) can be used to calculate the magnitude of mortality
reduction from a given level of active transportation in a population and determine the associated
economic value. The tool was designed for transportation planners, among others, in order to inform
decision-making around transportation infrastructure projects. HEAT can be used to determine the
value of specific levels of walking or cycling at a point in time or before and after an intervention.
HEAT combines existing epidemiological data quantifying the mortality risk reduction associated with
active transportation with local data. Users enter the number of walkers or cyclers and time spent per
year in either activity. Default parameters such as baseline mortality rate, projected mode share, time
over which to average benefits, and locally accepted estimates of value of a statistical life can be
modified. Cost information can be incorporated in order to derive cost-benefit ratios.
Currently HEAT only considers savings from mortality risk reductions associated with active
transportation since epidemiological data characterizing associated reductions in morbidity is less
robust. HEAT does not incorporate potential costs due to changes in numbers of injuries or morbidity
and mortality from exposure to air pollution that could be associated with increased walking and cycling.
Recognizing this limitation, the tools' creators state that information about air pollution risks to walkers
and cyclists is currently limited and the observed 'safety in numbers' phenomenon suggests injury risk
could decrease with increased walking and cycling. They cite the two studies mentioned earlier that
found that risks from air pollution and injury are far outweighed by the health benefits of increased
physical activity (de Hartog et al. 2010, Rojas-Rueda et al. 2011). Given an absence of strong
epidemiological data in children or by gender, HEAT can only be applied to adult populations.
Integrated Transport and Health Impact Modelling Tool (ITHIM)
The Integrated Transport and Health Impact Modelling Tool (ITHIM), created by the Centre for Diet and
Activity Research (CEDAR), has been developed to evaluate the health impacts of different
transportation scenarios and policies via their effects on physical activity, injuries, and air pollution
exposure (Centre for Diet and Activity Research). ITHIM uses a comparative risk assessment
methodology to model changes in population risk associated with changes in transportation mode share
and/or vehicle emissions. Unlike HEAT, which is based on changes in all-cause mortality associated with
active transportation, ITHIM models transportation impacts on both morbidity and mortality from
-------
82
numerous chronic conditions, providing an output in disability adjusted life-years (DALYs). The output
also includes changes in greenhouse gas emissions with under modeled scenarios.
ITHIM has been used to model the health impacts of transportation scenarios in England and Wales
(Woodcock, Givoni, and Morgan 2013) and in the San Francisco Bay Area (Maizlish et al. 2013).
Although not currently available as a tool for public use, the Excel-spreadsheet-based model can be
obtained from CEDAR.
Health Impact Assessment
Health impact assessment (HIA) is rapidly gaining acceptance as a tool with which to ensure the health
implications of decisions like those in transportation planning are taken into consideration during the
initial design stages. HIA is grounded in the observation that many health outcomes are influenced less
by individuals' interactions with the healthcare sector than by policies and interventions implemented in
other sectors. The HIA process is used to evaluate the potential health impacts of a proposed plan or
policy and the distribution of those impacts within a population. The process culminates in
recommendations by which to minimize and mitigate adverse effects and maximize health benefits.
These recommendations are provided to decision-makers in order to inform the design process prior to
finalization and implementation (National Research Council of the National Academies 2011a).
HIA is rooted in a process of active stakeholder engagement and, as such, provides an opportunity for
communities to participate in the planning process. Although the process consists of numerous
standardized steps, the HIA is a flexible tool that can be applied to a wide variety of projects and policies
using diverse methodologies and analytic tools. The ability to quantify the direction and magnitude of
potential health outcomes is often a reflection of the availability of necessary population data, existing
epidemiological data relevant to health impacts of interest, and the feasible scope of analysis that can
be conducted within project constraints.
Community Cumulative Assessment Tool (CCAT)
The Community Cumulative Assessment Tool (CCAT) is a computerized, step-by-step assessment
method that leads users through a guided, yet flexible, process to sort through information and develop
a "to-do" list of actions to address social, environmental, and economic impacts in their community,
primarily focused around human health and ecological/environmental concerns. CCAT was developed
by the EPA as part of its Sustainable and Healthy Communities Research Program. CCAT is a component
of a larger tool, the Community-Focused Exposure and Risk Screening Tool (C-FERST), discussed earlier in
this paper, which provides a wide range of information, datasets, GIS maps, and guidance to help inform
users of potential health and environmental issues in their community.
CCAT provides guidance on how to relate stressors to impacts, and provides an evidence-based scoring
method to evaluate and compare a wide range of issues simultaneously and help to prioritize solutions
based on the weight of a given impact. CCAT is not an exposure or risk model, but rather a decision-
making support tool. CCAT includes a conceptual model builder in its computer framework, where users
can assemble cause-effect flowcharts of stressors and impacts related to a particular issue, such as the
-------
83
range of impacts potentially associated with landfills. Transportation-related issues (focused around
health and ecological/environmental concerns), for example, could include near-road air, water, or soil
quality; mobile-source emissions impacts on sensitive land-use areas related to daycare facilities,
schools, or healthcare centers; or an exploration of the benefits and potential impacts of a given
transportation-infrastructure expansion project.
When viewed from a local-scale, community-based perspective, sustainable transportation relates to a
wide variety of stakeholders and considerations. CCAT allows users to examine and compare a diverse
range of issues at the same time, and therefore makes it easier to assemble a to-do list of actions to
address these issues. The CCAT framework is general enough to be transferable across communities,
yet specific enough to be tailored to each community and application. CCAT draws from cumulative-risk
assessment methodologies established by the EPA and the broader scientific community (U.S.
Environmental Protection Agency 2003, U.S. Environmental Protection Agency 2007b, National Research
Council of the National Academies 2009). The ability to include and compare multiple social,
environmental, and economic issues across a range of stressors and impacts provides a suitable context
in which to weigh sustainability options when considering solutions to various problems (National
Research Council of the National Academies 2011b).
Eco-Health Relationship Browser
The Eco-Health Relationship Browser is an interactive, web-based tool that displays published linkages
between ecosystem services (benefits supplied by nature) and many aspects of public health and well-
being. The Browser provides information about several of the nation's major ecosystems, the natural
benefits they provide, and how those benefits or their absence may affect human health and well-being.
Ecosystems such as wetlands and forests provide a wide variety of goods and services, many of which
we use every day. However, some of these goods and services, such as air and water filtration, are not
only free of direct financial cost, but also "out of sight," and so can be difficult to appreciate in terms of
their relevance to people's daily lives. The Eco-Health Relationship Browser reflects a detailed review of
the recent scientific literature, with more than 300 citations in its bibliography. The most compelling
research findings are summarized for dozens of public health issues. Many of the featured studies do
not address causality, but document statistical associations. Plausible, proposed, and known causal
mechanisms are also presented. The Eco-Health Relationship Browser is accessible through EnviroAtlas,
discussed in an earlier section, and at
http://www.epa.gov/research/healthscience/browser/introduction.html.
HYGIEIA
HYGIEIA is a modeling framework created by the EPA, the CDC, and various state and local government
agencies in Texas. The model's purpose is to examine links between ecosystem services and human
health and well-being and then to project the possible health outcomes of various courses of action that
a community might take. HYGIEIA is able to produce time-lapse maps of health outcomes in different
scenarios, representing predictions across both space and time. HYGIEIA is also able to predict health
outcomes for members of different demographic categories within an area, distinguishing people by
such factors as age, race, and annual income. Most of the input data required, such as census data,
-------
84
land-cover data, and climate-change projections, are already available in standardized formats in the
United States. Any part of the country for which these data are available may benefit from HYGIEIA.
Although it is able to model a wide variety of health-related inputs and outputs, HYGIEIA has so far only
been used, in the city of Austin, TX, to predict the human mortality and morbidity effects of heat stress,
such as from urban heat islands.
3.6.2 Social Interaction Outcomes
Transportation infrastructure and the ways in which people use it are able to produce substantial
effects, both positive and negative, on how much and in what ways the members of a community
interact with one another. Even though theoretical connections have been established between a
community's transportation system and social interaction/isolation, the magnitudes of these impacts
are challenging to quantify (Kennedy 2002).
If roadways or rail lines are designed with the primary goal of moving people and freight efficiently and
at minimum cost, this system is likely to have a significant negative effect on the perceived quality of the
public space in a community, leading to less use of that space, and hence less social interaction. This
phenomenon is due to traffic noise, pollution, the unaesthetic appearance of the infrastructure, and the
tendency of major transportation routes to divide communities into segments by forming a physical
and/or psychological barrier to movement by pedestrians and cyclists perpendicular to a major roadway
or rail line. Alternatively, transportation infrastructure (including bridges, public transit stops, and
bicycle paths) may be designed in accordance with architectural principles that encourage social
interaction. A particularly memorable, aesthetically pleasing, and enjoyable-to-use piece of
transportation infrastructure may serve as a common reference point within the geography of a
community, with the result that people will tend to use it as a gathering place. Regardless of whether a
community's transportation infrastructure promotes or discourages social interaction, the effects are
liable to be very long-lasting, as it is difficult for major infrastructure elements to be removed in their
entirety and replaced with something dramatically different, as opposed to merely being modified or
expanded (Meyboom 2009).
Transportation mode choice may have a significant effect on social interaction within a community, as
well. Particularly, personal automobiles, more than other modes, represent a choice of private space
over public space. As a result, people may become more isolated from one another and have less sense
of community, and socioeconomic segregation may be more easily facilitated (Mercier 2009).
Furthermore, the noise produced by automobile traffic has been found to discourage sidewalk activity in
residential neighborhoods, meaning that automobiles may increase social isolation both for people that
use them and for people who do not use them (Kennedy 2002). However, if someone travels for the
express purpose of taking part in social activities, the speed, flexibility, security, and comfort of personal
automobile travel serve to increase that person's access to opportunities for social interaction,
especially in areas that have been developed at too low of a density for other modes to be practical
(Kennedy 2002). Regardless, a study in Miami, FL found that residents of neighborhoods with more
automobile commuters per land area tend to experience more depressive symptoms and also found
-------
85
that residents of neighborhoods with higher housing densities tend to experience fewer depressive
symptoms (Miles, Coutts, and Mohamadi 2012).
Senior citizens eventually reach a point in time when they must either stop driving or greatly reduce the
amount that they drive; hence, losing the access to social interaction that driving provided, senior
citizens may especially benefit from the availability of alternative transportation modes, especially if
they are unable to meet their travel needs by riding with friends and family members (Waldorf 2003). If
the non-driving transportation alternatives in a community happen to include good walking
environments, the resultant physical activity may also benefit the psychological well-being of older
residents (Kim and Ulfarsson 2008).
3.6.3 Equity Outcomes
Issues related to equity are discussed in various sections of this synthesis paper, such as in discussions of
pollution exposure, health outcomes, and transportation funding mechanisms. Rather than discuss
equitability in the distribution of the effects of externalities, this subsection is dedicated to equitability
in the distribution of transportation as a product that the government provides to its constituents. For
the purpose of assessing equity among the various users of a community's transportation system, the
most appropriate measure by which to compare those users is destination accessibility. Destination
accessibility is a function of how fast people are able to travel, how cheaply they can travel, and the
number and variety of desirable destinations that are located relatively close to their home or place of
business. Destination accessibility is distinct from potential mobility, which merely consists of being able
to easily travel a long distance at a great speed. Measures of potential mobility do not consider whether
or not the journeys a person is able to make lead to places that are worth the trip. Discussing
destination accessibility only in terms of a single mode of transportation would be misleading, as most
people have more than one modal option (even if those options are not created equally) and may
choose different modes under different circumstances or mix modes within a trip. Furthermore, while a
transportation agency may directly influence the potential-mobility component of people's destination
accessibility, the distribution of origins and potential destinations within a community (the other
component of people's destination accessibility) is mostly outside the ability of transportation agencies
to directly influence, making it necessary to also consider as inputs to any model of destination-
accessibility equity the local land-use regime, economic-development activities, and the provision of
various other services. However, transportation-agency actions still indirectly affect the distribution of
origins and destinations, since changes in accessibility produce reactions in the real estate market, which
then alter land-use patterns, producing more changes in accessibility. The resulting feedback loop
makes long-term transportation equity outcomes difficult to predict. In addition, individual community
members make different tradeoffs between accessibility, the cost of housing, and other considerations
when they make decisions on where to live, where to work, and how to travel, meaning that a situation
that may seem equitable or inequitable through the lens of destination accessibility might actually be
the opposite when a wider view is taken. This represents yet another complication to account for in the
evaluation of transportation equity (Martens 2012).
-------
86
While one might be inclined to define an equitable transportation system as one where every user's
destination accessibility is roughly equal, the fact that urban areas tend to organize themselves into
cores and peripheries (with the core usually having access to more destinations than the periphery)
suggests that transportation planners must accept the existence of at least some variation in the
destination accessibility of different individuals and groups of individuals. In that case, seeking to
simultaneously maximize the average level of accessibility among community members and keep the
difference between worst accessibility and best accessibility under a certain limit may be advisable.
Alternate ways of achieving an equitable distribution of accessibility include distribution according to
need and distribution according to merit (wherein destination accessibility is considered a "reward" for
some "burden," such as living someplace undesirable). Neither of these ways of looking at equity is
well-suited to transportation questions, though. Distribution by need would encounter the complication
of distinguishing between "need" and "want," as well as differences in need among people who would
equally benefit from a transportation project and changes in people's relative levels of need that occur
faster than the transportation sector can respond. Meanwhile, merit-based distribution would require
making an inherently subjective judgment of what is a "burden" and what constitutes an appropriate
"reward" (Martens 2012).
One of the most important indicators contributing to measures of destination-accessibility equity within
a transportation system is how much it costs to use the system. Because traveling by automobile is
usually costlier than traveling by public transit, low-density cities with automobile-dependent
transportation systems are not very equitable; low-income travelers will have to either spend a large
percentage of their earnings on automobile transportation or travel on an underfunded and inefficient
public transit system (Mercier 2009), bearing in mind that even the public transit systems in cities that
are not particularly automobile-dependent do not necessarily serve all of the destinations that a given
transit-dependent individual may require (Berechman and Paaswell 1997). Furthermore, the societal
costs of traveling by private automobile are not as internalized as those of other modes of
transportation, which is another form of inequity (Sinha 2003) (i.e., drivers do not "pay for" the cost to
the broader tax base of motor-vehicle infrastructure on their own, or for negative externalities like
pollution).
Even when two neighborhoods provide the same amount of access to modes of transportation other
than private automobiles, there may still be transportation inequalities between them. If two
neighborhoods are equally walkable in terms of their density, land-use mix, street connectivity, and
access to transit, the richer neighborhood is likely to have better aesthetics, better pedestrian safety
features, and other amenities that make pedestrian travel a more attractive prospect. Though the
walkability of a neighborhood (measured in terms of density, land-use mix, street connectivity, and
access to transit) is strongly correlated with the amount of physical activity its residents get, reduced
aesthetics, safety, and amenities dampen this effect. This may partially account for higher rates of
health problems that are associated with insufficient physical activity in high-poverty neighborhoods
(Neckerman et al. 2009). In addition, residents of different neighborhoods with comparable access to
transportation do not necessarily have comparable access to destinations.
-------
87
Another dimension of transportation equity is the international dimension. For example, even though
North America is only home to 5% of the world's population, it accounts for 40% of worldwide
transportation energy use (Mercier 2009). Also, the negative externalities of transportation policy
decisions are often not felt until a later generation than the one in which they are implemented (Loo
and Chow 2006).
More research is still needed on how accounting for racial segregation would change models of future
population and employment distribution and models of future land use patterns. More research is also
needed on how a given geographic area may derive an economic advantage over another geographic
area from the particular mix of jobs that it has, as opposed to its total level of employment (Carruthers
and Vias 2005). If verifiable answers are provided to these research questions, which have clear equity
implications, those answers could provide additional guidance in the setting of transportation policies
that have equitable outcomes.
3.7 Economic Issues and Related Tools, Resources, and Indicators
Before considering the specific case of how transportation-related government actions affect the
economy, familiarity with several concepts that may be applied to analyzing the economic impacts of
any government action is advisable. The following are some such concepts (U.S. Environmental
Protection Agency 2010b):
•	Benefit-Cost Analysis (BCA)\ A method of calculating the net benefits to society produced by a
given action, regardless of which specific individuals within society experience the benefits of
the action and which ones experience the costs. The results of a benefit-cost analysis are
considered to support a particular government action if they conclude that the benefits accrued
from the action are greater than the costs imposed, regardless of who experiences the benefits
and who experiences the costs. In isolation, this kind of analysis is blind to issues of equity.
•	Willingness to Pay (WTP): The greatest amount of money that either a person or a collection of
people would be willing to part with in order to either bring about a "favorable" government
action or prevent an "unfavorable" government action.
•	Willingness to Accept (WTA): If a particular action would be detrimental to a given individual or
group, this is the minimum amount of compensation that they would agree to receive in
exchange.
•	Baseline: Describes the future scenario that would occur if no changes are made to the status
quo (i.e., "business as usual"). Preferably, any scenario would be projected out to a time when
the adjustment periods related to any changes in the status quo have passed and a new state of
economic equilibrium has been reached.
•	Kaldor-Hicks Criterion: This criterion describes a situation that is favorable to initiating a given
change in the status quo. The expected beneficiaries of a proposed change may incentivize the
expected cost-bearers to compromise and allow the change to take place; in this case, the
minimum amount of compensation must be less than the maximum the beneficiaries of the
change are willing to provide. Likewise, the expected cost-bearers may incentivize the expected
-------
88
beneficiaries to maintain the status quo; in this case, the minimum value of the incentive must
be greater than the maximum the people providing it are willing to sacrifice.
•	Economic Impact Analysis (EIA): Disaggregates the analysis of the benefits and costs of a
government action to the scale of individual sectors of the economy or individual entities and
institutions. This may include looking at effects on specific industries, specific governmental
units, nonprofit organizations, individual companies, and the consumers and suppliers of given
products. This kind of analysis delineates which parties benefit from a government action and
which ones lose from it.
•	Equity Assessment: An analysis of the benefits and costs of a government action that are
experienced by specific subpopulations, especially those that are regarded as disadvantaged.
•	Standing: If the potential effects of a particular government action on a given party are
significant enough to be considered in the economic analyses of that government action, the
party in question is regarded as having "standing." Ideally, an economic analysis would grant
standing to all individuals who have any potential of being affected. However, for practical
purposes, standing may not be granted to those who are assumed to only be very remotely
affected, such as residents of foreign countries.
•	Externalities: Benefits and costs of an action that are experienced by those who are not directly
connected to it (e.g., air pollution is a negative externality, while a pleasant view of green
infrastructure would be a positive externality).
•	Opportunity Cost: The value that would be derived from the best of any of the alternatives to a
given course of action (not always reducible to monetary terms). This is the value that someone
accepts missing the opportunity to receive whenever they choose between mutually exclusive
alternatives. Choosing a given course of action is only advisable if the value derived from that
course of action is greater than the opportunity cost (i.e., the opportunity provides more value
than any alternative would).
•	Discounting: An analytical operation that equates the value of some amount of future economic
consumption to the value of some lesser amount of present-day consumption. Due to
considerations of uncertainty about the future, inflation, and opportunity costs, people
generally prefer to receive a given quantity of benefits in the present than to wait to receive
that same quantity of benefits sometime in the future, meaning that benefits that occur in the
distant future are valued less, or discounted, relative to benefits that occur in the near future.
Since the buying power of a dollar generally goes down over time, spending and investing both
provide more value than simply saving. If someone owns an object that they do not use and
which physically degrades over time (such as an unoccupied house), they receive more value
from selling it now than from waiting to do so. An investment that takes one year to turn a
profit is more attractive than an investment that takes twenty years to turn a profit, since the
profits from the former investment could be reinvested sooner, resulting in more profits
(opportunity cost). The longer it takes for an investment to pay out, the more opportunities
there are for unforeseen events to reduce the odds of making a profit (natural disaster disrupts
operations, economy goes into recession, etc.). However, if anticipated future benefits from a
given action sufficiently exceed a theoretical set of near-term benefits, people will regard the
-------
89
two as equally attractive. Since different alternative courses of action yield their benefits and
costs over different amounts of time after they are initiated, translating all of them to the
amount of present-day benefit or cost they are regarded as being equivalent to enables direct
comparisons. They are all converted to their Net Present Value (NPV). When it comes to natural
resources that either are nonrenewable or renew very slowly, this practice could become
problematic.
•	Cost-Effectiveness Analysis (CEA): If a given government action is intended to produce a specific
quantifiable benefit, the result of a cost-effectiveness analysis is the ratio of dollars spent on the
government action in question per unit of change in whatever measure is used for the
quantifiable benefit under consideration (i.e., bang per buck).
When considering a new, altered, or expanded transportation system, several basic questions of an
economic nature must be considered, including (Sastry 1973):
•	How many people will use the system?
•	How will changes to the transportation system alter forecasts of future population distributions
and land use patterns?
•	Will business activity near the transportation infrastructure be discouraged, encouraged, or
otherwise shifted?
•	If business activity is shifted, what neighborhoods will respectively suffer and benefit? What will
be the effect on the local real estate market?
•	What environmental and aesthetic impacts will the transportation system produce that might
affect the economy?
Going beyond just the local area around new transportation connections, what impact will those
new connections have on the state of the regional economy?
•	What impact will a transportation project or policy have on the size of the tax base for a given
jurisdiction and how does that impact compare with any governmental expenditures resulting
from the project or policy?
When economic analyses are carried out to address the above questions for a particular transportation
project under consideration, varying the extensiveness of the analysis in proportion to the size and cost
of the project being proposed is well-advised. Upon carrying out such analyses, which generally require
very large datasets, the step of determining economic costs and benefits must be preceded by the step
of defining the population and employment bases for which costs and benefits will be considered. Then,
costs and benefits must be considered that are both of a quantifiable and of an abstract nature. The
most quantifiable items in an analysis of a transportation infrastructure project are the costs of capital,
maintenance, and administration, which must be paid by the agency responsible for the project. More
abstract costs may consist of environmental and societal impacts, while major components of a project's
benefits include increases in travelers' access to various types of destinations (which is independent of
how far away those destinations are), their overall mobility across the landscape (which is independent
of how many worthwhile destinations occupy that landscape), safety improvements, and the reduction
of environmental impacts. However, the most critical source of economic costs and benefits from a
-------
90
transportation project is its effect on area-wide traffic patterns, preferably expressed in the form of link-
by-link forecasted traffic volumes for each alternate scenario contemplated. Not only must costs and
benefits be identified for the entire area impacted by a transportation project, but also for the entire
timeframe of the project's useful life (Shadewald, Hallmark, and Souleyrette 2001).
The potential benefits and costs of transportation projects and policies that are listed above may be
referred to as direct economic impacts, which may produce any number of indirect impacts. Both direct
and indirect effects must be considered in order for an economic analysis to be complete. One of the
most noteworthy indirect effects of transportation projects is when development is attracted to an area
by the reduced transportation expenditures that follow from increased accessibility. However, such
increases in development activity often result in air pollution, noise pollution, and other such deterrents
to development in close proximity to heavily-used transportation infrastructure (Hof, Heyma, and van
der Hoorn 2012, Jha and Kim 2006). Land use impacts from transportation projects and policies, as well
as other indirect effects, typically come about over a period of years and are difficult to estimate. The
long-term, indirect effects of new rail transit projects are especially difficult to predict, since very few
case studies exist of such projects that have been in existence long enough for all of their economic
impacts to have manifested (Polzin 1999, Szeto, Jaber, and O'Mahony 2010). One reason that the
overall economic impacts of transportation projects are so uncertain is that performing benefit-cost
analyses runs the risk of some effects being double-counted, as indirect benefits and costs are not
necessarily additive (Hof, Heyma, and van der Hoorn 2012). Meanwhile, some effects are simply
neglected in the analysis or not quantified.
According to a study of economic models from the Netherlands (Hof, Heyma, and van der Hoorn 2012),
direct economic effects from transportation projects may be significantly greater than the indirect
effects of the same projects. However, even if this conclusion holds true in other settings and with
other models, identifying direct and indirect impacts would still help to keep effects from being double-
counted. Models would be useful that feature a "quick scan" function that estimates whether or not a
proposal's indirect effects would be great enough to warrant taking the extra time necessary to produce
a detailed estimate of those effects (Hof, Heyma, and van der Hoorn 2012). Other beneficial model
attributes include:
•	The consideration of travel into and out of the study area (Hof, Heyma, and van der Hoorn
2012).
•	The representation of both personal travel and freight travel (Hof, Heyma, and van der Hoorn
2012).
•	The anticipation of as many of the reasons for which people travel as possible (Hof, Heyma, and
van der Hoorn 2012).
•	Accounting for people valuing their time in different ways (Szeto, Jaber, and O'Mahony 2010).
•	Showing how changes to the transportation network may result in changes in the distribution of
land uses, which results in changes in the distribution of traffic volumes, which may necessitate
further changes to the transportation system (Szeto, Jaber, and O'Mahony 2010).
-------
91
•	Modeling the elasticities of housing demand and housing supply for the determination of land
use effects (Szeto, Jaber, and O'Mahony 2010).
•	Allowances for uncertainties in demand and supply (Szeto, Jaber, and O'Mahony 2010).
•	The calculation of tradeoffs between the interests of various parties, including unequal changes
in landowners' profits (Szeto, Jaber, and O'Mahony 2010).
•	Modeling activity on the various links within a transportation network instead of modeling all of
the paths that may be traveled between origins and destinations within that network, as the use
of path-based modeling on large networks is burdensomely complex (Szeto, Jaber, and
O'Mahony 2010).
•	Making it so that transportation models that address different geographic scales have consistent
units, inputs, and assumptions, so that the outputs of large-scale models may be used as inputs
for the small-scale models (Hof, Heyma, and van der Hoorn 2012).
•	Modeling that is based on empirical observations and describes the path that all relevant
measures take from the starting year to the ending year of the model simulation, in contrast to
many current models of the economic effects of transportation projects and policies (Hof,
Heyma, and van der Hoorn 2012).
3.7.1 Accessibility
Changes in accessibility are considered to be one of the most fundamental economic impacts stemming
from transportation projects and policies; the primary purpose of most transportation projects and
policies is to decrease the amount of time and money required by people and businesses to meet their
travel requirements. A few basic methods exist by which accessibility may be enhanced. First, a
transportation project may increase the number of people who have the option of traveling to a given
location, such as by extending a new transit line to a previously unserved commercial district and
allowing transit-dependent individuals to become customers and workers there. Second, improvements
to the transportation system may increase the percentage of the local population that is able to reach a
given location within a given travel time, such as by reducing traffic congestion. Third, a transportation
project may provide travelers with value in the form of an additional option for how to travel. Even if
the new option is not necessarily superior to existing options, simply having an expanded range of
choices may be regarded as beneficial (Polzin 1999). Finally, a transportation project or policy may open
up a travel option that is not faster, more flexible, or more comfortable than preexisting options, but
which is cheaper than those other options and frees up the opportunity cost of paying for them
(Kennedy 2002). At the same time, though, travel is not purely a derived demand. While arriving at
given destinations in a timely and cost-effective manner is still the primary reason that people travel,
they also derive at least some value from the act of traveling itself, meaning that people may be willing
to trade some amount of accessibility in exchange for a more pleasant travel experience (Loo and Chow
2006, Polzin 1999). Furthermore, value may be derived from working (or doing some other worthwhile
activity) while traveling, such as during the commute to and from an individual's place of employment,
effectively making the commute part of the workday (Gripsrud and Hjorthol 2012).
3.7.2 Announcement Effect
-------
92
The "announcement effect" is when land prices around a transportation infrastructure project increase
before anything has actually been built, in anticipation of a future increase in accessibility. So far, there
has been little research on how to calculate the timing of the announcement effect (Tsutsumi and Seya
2008).
3.7.3	Non-Accessibility Benefits from Transportation Projects
The primary intended economic benefits from transportation projects usually consist of providing faster,
cheaper, and/or more convenient travel for a greater number of people and businesses, with fewer
externalities. However, such projects may also produce other economic benefits, whose influence may
be difficult to distinguish from that of the transportation-related benefits. Government intentions to
invest in the transportation infrastructure of a given area may convince businesspeople to invest their
own money in the area as well, with the first wave of new investors inspiring additional investors. This
effect may be enhanced by efforts of the government to promote the economic advantages of the
transportation project. However, if the transportation infrastructure in a particular area is built up at
too much faster of a pace than nearby land development, the apparent mismatch may discourage
private-sector investment (Polzin 1999).
Some transportation infrastructure projects are initiated in conjunction with other policy actions that
serve the purpose of making investment in the area around the transportation project more attractive.
For example, the government may choose to expedite the approval process for nearby development
proposals, reduce relevant fees, change the local land-use zoning map, increase maximum building
densities, alter parking requirements, provide tax incentives, or locate other types of government
facilities in the same area. The government may also choose to use the power of eminent domain to
buy properties near the transportation project for resale to developers who would not have been able
to gather the land needed for their planned buildings on their own. Furthermore, the presence of a new
transportation project may serve as a rallying point to inspire government actors to institute unrelated
improvements in the area. This may include beautification projects, historic preservation efforts,
increased public lighting, increased police presence, or increased trash service. In some cases, the
complementary policies stemming from a transportation project will affect a jurisdiction far larger than
the immediate area around the project (Polzin 1999).
Other non-transportation economic effects may arise unintentionally from the implementation of a
transportation infrastructure project. For example, when the government buys land for the right-of-way
of a transportation project, the entire property must be purchased, even if only a portion is needed.
Selling the part of the property that they do not need can produce an economic benefit. Similarly,
public building projects may require that some plots of land be purchased strictly for the temporary
purpose of holding construction equipment and materials. This land, too, is likely to be resold when the
project is finished (Polzin 1999).
3.7.4	Agglomeration Effects
When a transportation project increases the profitability of doing business in a given area, either by
reducing the cost of transportation or other means, more money will likely be invested in that particular
-------
93
area, with new businesses moving in and preexisting businesses increasing their presence. In addition,
as a particular kind of business increases its presence in an area, businesses that are complementary
may be motivated to locate in the same area, as proximity reduces transportation costs. As businesses
become agglomerated in a particular area, whether around a transportation facility or some other initial
focal point, greater competition arises between businesses, resulting in lower prices and smaller profit
margins. However, if improvements to the transportation system increase the number of businesses
within range of the same customer base, those businesses have greater opportunities to differentiate
their products from their competitors' products, potentially allowing higher prices and higher profits.
Therefore, the same initial event may produce drivers of low prices and drivers of high prices at the
exact same time (Hof, Heyma, and van der Hoorn 2012). The concentration of businesses in a given area
also increases land-use diversity and destination accessibility within that area, which, as discussed
earlier in this paper, are factors that encourage the use of modes of transportation other than the
private automobile. However, if businesses become concentrated in one part of a region at the expense
of others, regional land-use diversity and destination accessibility may be reduced.
3.7.5 Regional Adjustment Model
Both jobs attract people to an area and people attract jobs to an area, although much is still unknown
about the exact nature of this relationship (Sukkoo 2007). Consequently, assessing the economic
sustainability of transportation decisions would be greatly aided by the use of a tool that can untangle
the connections between population changes and employment changes. A regional adjustment model
consists of two separate equations describing the migration of people and jobs in and out of a given
area, with changes in population and changes in employment both being functions of the starting values
of population and employment and any number of additional variables, such as measures of ecosystem
services, infrastructure provision, and beneficial government policies. This type of model uses a positive
feedback loop, where jobs and people both follow each other, as opposed to employers locating jobs
wherever they like and then counting on workers to move to the area. One of the implications of this
dynamic is that non-job-related amenities (which elements of the transportation system may either
create or destroy) in a given area may attract more residents and subsequently more jobs (Carruthers
and Vias 2005, Kennedy 2002). Under this theoretical construct, the economy is taken to be in a state of
partial equilibrium, constantly readjusting towards an ideal spatial distribution of people and jobs, which
is never actually reached, on account of shocks to the economic system (Carruthers and Vias 2005).
Furthermore, because some people may be willing to accept worse job prospects in exchange for good
location-specific amenities (such as natural environments and vibrant communities) and other people
may be willing to accept living in an amenity-poor area in exchange for a high income, a state of
equilibrium in the distribution of people and jobs would not necessarily appear homogenous (Carruthers
and Vias 2005). The mutual causality between population and employment levels has not been firmly
established at all geographic and time scales, with the greatest evidence found at the very-large-region
scale (i.e., consisting of several states or provinces within a country) (Carruthers and Vias 2005). Most
likely, the interactions between population and employment levels vary from region to region, meaning
that different regions need to have different models. Ideally, each region's model would consider
interrelationships with other regions (Carruthers and Vias 2005).
-------
94
3.7.6	Models of Transportation-Infrastructure Land Use and Other Land Uses
The Solow and Vickrey model describes the relationship between traffic volumes, real estate values, and
the amount of land dedicated to transportation infrastructure in an urban area in simplified terms.
Using economic principles, this model concludes that land rents do not typically fully reflect the
transportation advantage enjoyed by properties at the center of an urban area. This subsequently
affects the distribution of transportation demand throughout the urban area, with the result of an
unnecessarily large number of lane-miles of roads being built, most especially at the center of the urban
area. A variation on the Solow and Vickrey model, the Legey, Ripper, and Varaiya model, goes on to
predict that in a city where a central authority does not allocate land in a manner contrary to market
forces, not only will too many resources be dedicated to transportation at the urban center, but there
will also be too many resources dedicated to housing along the outer edges of the urban area, a finding
which one may interpret as describing market-based drivers of urban sprawl. In another variation on
the Solow and Vickrey model, attention is given to the case of a city laid out in a circular pattern. In this
scenario, the transportation system would be optimized by a balance between radial roads leading out
from the center of the urban area and several progressively larger ring roads that encircle the center
and intersect with the radial roads. Meanwhile, another study adapted the Solow and Vickrey model to
the scenario of a square city with streets that all intersect at right angles. In this case, the model
showed that the amount of land dedicated to transportation infrastructure would be excessive in all
parts of the city, with the worst of the overbuilding of roads occurring along the edges of the urban
area, instead of the center; at the same time, the center of the urban area would still experience the
highest rents. Finally, adaptation of the Solow and Vickrey model shows that if an urban area is
polycentric, whichever of its multiple centers is the most central will experience the greatest amount of
traffic. Consequently, that particular center will most likely be the one that is the most heavily
developed and the one where the greatest amount of land is required for transportation infrastructure
(Medda, Nijkamp, and Rietveld 2003).
3.7.7	Market Imperfections
An imperfect market is one in which the price of a product, such as transportation, is either unequal to
the cost of supplying it or unequal to the value derived from it by the buyer. Most markets are
imperfect and may come to be that way either through government actions (subsidies, taxes, etc.) or
through private-sector actors exerting market power that comes from economies of scale. When
market imperfections exist, they prevent the theoretical economic effects of a transportation project
from being realized in their entirety. For example, zoning laws and other land-use regulations constrict
the reaction of the real estate market to changes in the transportation system. Also, if a company either
has a monopoly or is part of an oligopoly, they are less likely to pass on to customers the savings from a
reduction in transportation-related expenses. However, transportation improvements may also
mitigate existing market imperfections. If a transportation project triggers an agglomeration effect,
those companies that are consequently placed in more direct competition with one another are forced
to make the prices of their products more closely reflect the cost of providing them. Furthermore, a
transportation project may create market imperfections, for either good or bad, by way of benefits
being derived from something in a non-transaction manner. An example is when a transportation
system benefits someone who is not part of the tax base that supports the system, or whose tax
-------
95
payments going to the transportation system are not proportional to their use of that system (Hof,
Heyma, and van der Hoorn 2012).
3.7.8	Import Substitution
If one is primarily concerned with the economic well-being of a specific geographic region, such as a
metropolitan area, considering the distinction between transportation expenditures that are paid to
parties within the same local economy and those that are paid to outside parties is important (for
example, depending on location-specific circumstances, one may drive a plug-in electric vehicle whose
energy source is a potentially-local power plant instead of driving a gasoline-powered vehicle whose fuel
may come from a refinery in a different state, or one may buy a car at a local dealership instead of going
out of town to shop for a vehicle). If the latter expenditures are too great, it may contribute to a trade
imbalance, hurting the local economy. Furthermore, even if a given area currently has a trade balance
with the outside world, a lack of diversity in its exports or an inability to meet its own needs by way of
internal sources may still produce a state of inflexibility that will endanger the local economy in the
future. If the area in question contains a car factory, for example, a portion of the money that residents
spend on automobiles will eventually contribute to the wages of local workers. When it comes to
import substitution, public transit has an advantage over travel by private automobile. Wages paid to
employees living in the transit system's service area may represent a significant portion of the money
that the local government spends on public transit, and the local government may choose to favor
locally-based companies when deciding where to buy transit vehicles and other products necessary to
the operation of the transit system (Kennedy 2002).
3.7.9	Cost-Effective Use of Government Budgets
In order to spend transportation-related funds in a cost-effective manner, tradeoffs must be considered
between different kinds of costs. An example of such a tradeoff is when a government entity spends
money to build a bridge across a body of water and is consequently able to end a previous practice of
subsidizing ferry service across that body of water (Hof, Heyma, and van der Hoorn 2012). Another
important area for tradeoffs is the Level of Service (LOS) that exists on freeways and arterial streets.
Urban freeways are expensive to build, requiring large amounts of land and building materials, and are
often designed with the primary goals of achieving a high free-flow speed and accommodating heavy
truck traffic. Achieving these design goals requires the creation of shoulders on the road and wide
traffic lanes. The result is fewer total traffic lanes and a smaller traffic volume that can be
accommodated, potentially resulting in economic costs from traffic congestion. If a particular urban
area generates very high traffic volumes during the peak travel periods of the day, and a relatively small
percentage of overall vehicle traffic consists of trucks, having a greater number of narrower traffic lanes
may be more cost-effective than a smaller number of wider lanes, choosing the economic benefits of
optimizing peak-travel-period traffic volumes over the economic benefits of optimizing off-peak traffic
speeds. Going farther, in some highly congested areas, the travel-time savings of having freeways
instead of non-limited-access, unsignalized arterial streets may not produce enough economic benefit to
offset the added expense of building freeways. However, freeways also tend to have lower accident
rates than ordinary arterial streets, which may save the government enough additional money to make
freeways cost-effective (Ng and Small 2012).
-------
96
Maintaining a transportation system will be far more cost-effective for the government if it takes
advantage of economies of scale. If a large number of destinations and transportation-system users are
densely concentrated in a small geographic area, their travel needs can be met at a low cost per capita
or per ton of freight, such as when companies agglomerate around a major seaport (Hof, Heyma, and
van der Hoorn 2012). The cost-effectiveness benefits of high development densities are especially
noticeable in the case of public transit systems. There are certain thresholds of development density
required for economically feasible operation of fixed-route transit systems (Kennedy 2002). However,
more research is still warranted to arrive at better estimates of these thresholds. The question of what
minimum development density may justify introducing public transit service may have different answers
in different contexts, such as monocentric and polycentric urban areas, or different-size of jurisdictions
(Kennedy 2002).
Other economy-of-scale benefits can be derived from the size of the government agency in charge of a
given transportation network. If an organization has a frequent need for a specific asset (such as
concrete or vehicle-fleet maintenance services), the organization will seek to reduce the uncertainty and
transaction costs associated with purchasing the asset. To do this, the organization will tend to favor
transitioning from buying the asset on the open market to buying it from a contractor on a standing
basis and then transitioning to providing the asset in-house. However, the benefits of reducing
uncertainty and transaction costs in this manner are offset by the overhead costs associated with
supplying an asset in-house. If an organization is large enough, economies of scale can reduce the
overhead cost per unit of a repeatedly needed asset. As a result, providing or producing the asset
within the organization becomes more economical, or, if in-house provision is not practical, the
organization uses the promise of a large-volume order to get a contractor to agree to a low per-unit
price to supply the same asset (Mercier 2009).
3.7.10 Equity in Transportation Funding Mechanisms
Two of the most common ways of perceiving the equitability of how a transportation system is funded
include: (1) how proportionate someone's contribution is to the amount of value they get from the
system and (2) how regressive the funding scheme is in terms of the percentage of a person's income
that is required. Aside from a progressive income tax, most of the ways in which a government entity
may collect revenue are regressive, likely requiring low-income people to pay a larger percentage of
their earnings than what high-income people must pay. Therefore, if potential transportation funding
mechanisms are to be analyzed in terms of their regressiveness, asking whether a given mechanism is
more or less regressive than its alternatives may be more appropriate than simply asking whether the
cost is regressive at all. Meanwhile, the question of whether or not the amount people pay for the use
of a transportation system and the amount of use they receive from it are proportional becomes more
relevant in cases where use of the transportation system produces significant negative externalities,
such as noise, air pollution, or accidents. If the amount someone pays for a service is little affected by
how much they use it, they are likely to use it more than they otherwise would, resulting in an increase
in any associated externalities. Common ways of funding surface transportation infrastructure include
fuel taxes, vehicle registration taxes, vehicle sales taxes, general sales taxes, and tolls and fares. All of
-------
97
these options are regressive to some extent, but fuel taxes, tolls, and fares are mostly proportional to
people's usage of the transportation system. Fuel taxes, though, have a difficult time keeping pace with
the demand for transportation-infrastructure revenue, thanks to increases in motor vehicles' fuel
efficiency and the phenomenon of fuel taxes being set as a fixed amount of money per gallon instead of
a percent of the price of the fuel purchase. General sales taxes are both regressive and blind to the
relative amounts that people use the transportation system, placing the greatest cost burden on
infrequent users of the system. However, because sales taxes spread the cost of transportation
infrastructure across a very large number of individuals, are paid in numerous small increments over the
course of a year, and are easy for the government to collect, these are often seen as a more attractive
option than road tolls, even by people who would ultimately pay more in transportation-related sales
taxes than they otherwise would in tolls and fees. Nonetheless, if a roadway system is funded through
tolls and low-income people retain the option of traveling by cheaper means than a toll road or toll lane
(such as riding on public transit), most tolls will be paid by middle-income and upper-income individuals,
making road pricing potentially less income-regressive than other funding schemes; however, this could
still leave low-income individuals traveling by slower means than other people (Schweitzer and Taylor
2008).
3.7.11 Economic Analysis Software for Transportation Planning
The Federal Highway Administration (FHWA) has put out several different software programs that
analyze the economic impacts of transportation project alternatives, including the Spreadsheet Model
for Induced Travel Estimation (SMILE) and the Sketch Planning Analysis Spreadsheet Model (SPASM).
The ITS Deployment Analysis System (IDAS) and the Surface Transportation Efficiency Analysis Model
(STEAM) focus on the estimation of system-wide impacts and include default values for the economic
costs of parameters such as vehicle emissions, fuel consumption, and traffic accidents (Shadewald,
Hallmark, and Souleyrette 2001).
Iowa State University's Center for Transportation Research and Education (CTRE) created an interface
between the GIS platform ArcView and the transportation demand model Tranplan. In combination
with a program such as FHWA's STEAM, an interface like this can be used in the early planning stages of
a transportation project to compare alternative and base design scenarios through visual
representations. One may map the economic benefits that an alternative scenario will produce for
different districts within a study area, allowing analysis of economic equity issues (Shadewald, Hallmark,
and Souleyrette 2001).
4 EXAMPLE APPLICATION OF SUSTAINABILITY PRINCIPLES AND ASSESSMENT TOOLS: COMPLETE
STREETS
This section assesses the sustainability of transportation policies and practices related to "Complete
Streets" design principles and compares them with more typical urban designs, oriented around
personal vehicle use, that have predominated development in the United States for many years. This
comparison highlights many of the sustainability principles presented in this paper, as well as the
assessment tools described. This comparison also demonstrates the consideration of transportation and
-------
98
other land-use decisions through a systems approach to account for the multiple benefits, and potential
costs, of transportation-related decisions, including the indication of tradeoffs, co-benefits, and
mitigating factors (with the aid of causal-loop diagrams). The complete streets principles examined here
are increasingly used in communities to address growth management, economic development, and
multi-modal transportation issues.
When a transportation corridor is designed in accordance with the principle of "Complete Streets," it is
designed to accommodate multiple modes of transportation, typically including private automobiles,
public transit vehicles, cycling, and walking. This type of design may also include other benefits and
amenities, such as green stormwater infrastructure, park-like elements, sidewalk dining, or on-street
parking. Design elements that may serve this purpose include frequent intersections, traffic calming
measures, other non-automobile-oriented safety features, features that make travel by a given mode
more comfortable even if they do not make it safer, a compact development pattern, mixed land uses,
and having building set back as little as possible from the property line. However, the most central
element of Complete Streets is how much of a transportation corridor's cross-section is available for use
by each specific mode and to what degree those modes are kept out of conflict with one another. Due
to the development of adjoining private properties, most transportation corridors cannot be widened
indefinitely; therefore, the amount of transportation-corridor width dedicated to each mode must be
traded off with one another. In Figure 7, below, one may see a system-dynamics diagram illustrating
these trade-offs in the case of non-arterial streets in a theoretical urban neighborhood.
gutter/drainage ditch % of
corridor (neghborhood
streets)
Width (neighborhood
streets )>
• (neighborhood streets)
Width (neighborhood
streets)>
# of car lane(s)
(neighborhood streets).
Width (neighborhood
streets)>
tdtal car lane wi&th
(neighborhood streets)
-gutters/drahage ditches
\ (neighborhood streets)
Width (neighborhood
streets )>
# of parking lanes
(neighborhood streets)
(neighborhood streets)
ZZ- (neighborhood streets)
park rig % of corridor +
(neighborhood streets)
total parking width
(neighborhood streets)'
Width (neighborhood
streets)>
width ofparkhg lanes
-I- (neighborhood streets).
Width (neighborhood
streets)>
# of turning \
lane(s)/median(s) \
. (neighborhood streets)
width of bike lane(s)
' (neighborhood streets).
Width (neighborhood
streets)>
Width (neighborhood
streets)>
>~*totaltumhg lane/median
width (neighborhood
streets)
(neighborhood streets)
turning lane/median % of
corridor (neighborhood
streets)
width of turning
lane(s)/median(s)
(neighborhood streets)
frike lane % of corridor
(neighborhood streets)
# of transit ROW (s)
(neighborhood streets).
"# ofbke lane(s)
(neighborhood streets)
width of transit ROW(s)
(neighborhood streets)
Width (neighborhood
streets)>
Width (neighborhood
streets)>
Width (neighborhood
streets)>
Width (neighborhood
streets )>
(neighborhood streets)
Figure 7: System-dynamics diagram of the trade-offs among different uses of a neighborhood transportation corridor. Red
arrows are negative relationships and blue arrows are positive relationships.
-------
99
One of the most prominent benefits attributed to Complete Streets is safety from vehicle-pedestrian
and vehicle-cyclist collisions. Roads that do not include facilities for pedestrians (sidewalks) or cyclists
(bike lanes and shoulders) will likely still be used by at least some pedestrians and cyclists, especially
those who do not have the option of motorized travel. Pedestrians and cyclists who travel along
roadways that are designed only for motor vehicles face a greater risk of being involved in traffic
accidents than they would if the transportation corridor's cross-section included segregated facilities for
each mode. Furthermore, the replacement of ordinary vehicle traffic lanes with medians, turning lanes,
and/or on-street parking makes it safer for pedestrians and cyclists to cross intersections, as it reduces
the amount of time they must spend crossing motor-vehicle lanes. These concepts are illustrated in
Figure 8. The effects of transportation policies and practices on accident rates, as well as other public
health outcomes, can be assessed using the Integrated Transport and Health Impact Modelling Tool
(ITHIM).



+ Odds of given mile of vehicle driving
—~~ leading to collision w/ another veh. in
+ peak period (neighborhood streets)
<# of turning
lane(s)/median(s)
(neighborhood streets)>

+ Odds of given mile of vehicle driving
^ leading to collision w/ another veh. in
+ off-peak period (neighborhood streets)


<$ per intersection spent on
veh. traffic control
(neighborhood streets)>

/ <# of sidewalk(s)
(neighborhood streets)>
Odds of given mile of
walking leading to auto
collision (neighborhood
streets)


<# of bike lane(s)
(neighborhood streets)>






$ per intersection spent
on ped./bike traffic
control (neighborhood
	streets)	
Odds of given mile of
cycling leading to auto
collision (neighborhood
streets)
Figure 8: Inputs to traffic safety for pedestrians, cyclists, and drivers.
Since reductions in the safety, comfort, and convenience of a given mode inhibit the destination
accessibility of individuals who already have limited modal options (for reasons such as household
budget or physical disability/infirmity), a transportation system that accommodates as many modes as
possible provides an equity benefit.
If Complete Streets practices succeed in making travel by modes other than the private automobile
safer, easier, and more pleasant, the result may be increased usage of these alternative modes and
decreased usage of private automobiles, especially if capacity constraints reduce the average speed of
automobile travel (Figures 2, 9, 10, and 11). Verification of these effects may be aided by various
-------
100
transportation behavior models, some common (but not universal) elements of which include travel
time budgets (Figure 3) and the conventional four-step transportation modeling process of first
determining the number of trips that people will make, then those trips' destinations, then mode
choices, then route choices. Some examples of such models include the EPA's Smart Growth Index (SGI),
the Multiclass Simultaneous Transportation Equilibrium Model (MSTEM), CONTRAM, and the Unified
Mechanism of Travel (UMOT).

<# ofsidewak(s)
(neighborhood streets)>

Perceived comfort
of walking facilities
(neighborhood
streets)





Perceived comfort
of cycling facilities
(neighborhood
streets)

<# of bike lane(s)
(neighborhood streets)>

Figure 9: Inputs to the perceived comfort of walking and cycling facilities.

cycling popularity
waking popularity
Concentration of
local/regional air pollution in
neighborhood>
Perceived comfort
of cycling facilities
(adjacent
neighborhoods)
Perceived coml oil
of walking facilities
(adjacent
neighborhoods)
<% ot neighborhood
land in conservations
Adj acent-neighborhood
job density
Adj acent- neighb omood
intersection density

cy clmg-only-tnp-mile s/
^ resident/day
walk-onty-tnp-mile s/
resident/day


Neighborhood-street
intersection density

frequency ot connections
btwn. arterial &
neighborhood streets


<% of neighborhood within
traffic noise zones (off-peak
periods)>
<% of neighb oihood within
traffic noise zones (peak
periods)>

Perceived comfort of cycling facilities
(arterial streets in neighborhood) times
of neighborhood cycling that is on arterial
streets
<% otneighbomood
affected by near-road air
pollution>


Adjoining-neighborhood
crime rate
Neighborhood
crime rate
Perceived comfort of cycling facilities
(neighb oihood streets) times % of
neighborhood cycling that is on
neighborhood streets
Perceived comfort ot walking facilities
(neighborhood streets) times % of
neighborhood walking that is on
neighborhood streets


law enforcement $
per resident or
worker
Ave. travel distance to school of
neighborhood students times % of
population that is children

<% ot population
that is children>
Perceived comfort of walking facilities
(arterial streets in neighborhood) times %
of neighborhood walking that is on arterial
streets
Figure 10: Inputs to the distance per day that people travel by walking and cycling.
-------
101
Odds of given driving trip to a transit
stop leading to collision w/ another veh.
times % of travel to transit stops that is
transit-itinera ly-miles/
resident/day
Average transit
itineraiy distance
from neighboiiiood
Odds of being hurt
while on transit veh.

Convenience of transit
info, services


transit popularity+
transit trips/
resident/day

Ave. speed of
transit itineraries
by residents
Comfort of transit
veh.s/facilities

Frequency of scheduled transit-veh.
stops during peak periods (regardless
of location w/in neighborhood)
Perceived non-time hardship of
walking to transit times % of travel
to transit stops that is walking
Duration of transit
y service day
Frequency of scheduled transit-veh.
stops during off-peak periods
(regardless of location w/in
neighborhood)
Ave. time to
reach transit
stop
Perceived non-time hardship of
cycling to transit times % of travel
to transit stops that is cycling
Figure 11: Inputs to public transit usage by the residents of a neighborhood, including characteristics of the overall transit
system and inputs related to traveling to and from transit stops within the neighborhood.
As discussed throughout this paper, numerous benefits would follow from such a shift in modal shares,
including decreased transportation-energy use (Figure 12), decreased tailpipe emissions (including both
greenhouse gases (Figure 13) and chemicals that more directly affect human health), increased levels of
physical activity, reduced traffic-related noise pollution, and decreased levels of social isolation
(especially if the resulting pedestrian facilities are aesthetically pleasing and conducive to personal
interaction). However, even if overall emissions of air pollution from motor vehicles are reduced,
Complete Streets practices may simultaneously increase exposure to those emissions by encouraging
the construction of buildings closer to roadways and placing pedestrians and cyclists (who breathe in
more air per period of time than do motor-vehicle occupants) in close proximity to vehicles' tailpipe
emissions for the duration of trips that take more time to complete than they would if taken by motor
vehicle (Figure 14). Changes in energy use and greenhouse gas emissions resulting from a given shift in
transportation behavior may be assessed using the EPA's MOtor Vehicle Emission Simulator (MOVES)
model or California's EMFAC, models that consider the composition of the vehicle fleet whose use is
being affected. Also useful in this task are assessment tools that use greenhouse-gas footprint
methodologies and/or consider life-cycle emissions impacts, such as the Greenhouse Gases, Regulated
Emissions, and Energy Use in Transportation (GREET) model, the Lifecycle Emissions Model (LEM),
GHGenius, and Economic Input-Output Life Cycle Assessment (EIO-LCA). Meanwhile, emissions and
transport of a variety of other transportation-related airborne pollutants may be assessed using the
EPA's Community Multiscale Air Quality (CMAQ) model. Evaluations of the health impacts of modals
-------
102
shifts to walking and cycling may be aided by the Health Economic Assessment Tool (HEAT) and by
ITHIM. Useful tools for assessing a broader range of health impacts from transportation policies and
practices may include Health Impact Assessments and the EPA-developed tools of HYGIEIA and the
Community Cumulative Assessment Tool (CCAT).


Private-vehicle fuel
consumption of trips
generated by the
neighborhood per day
total
public-transit-related
energy use attributable
to trips generated by the
neighboihood per day
Energy use of
private-vehicle trips
generated by the
neighboihood per
	day	
Energy per gallon of
private-veliicle fiiel
Total transportation
energy use
attributable to
neighboihood
fe \ o n-1 i-,i ns r> i it;i t i o n
i- energy use in
neighboihood per
	day	
energy use in

Total
neighboihood
eneigy use per
day
Neighboihood
eneigy use per
resident or worker
	per day	
Figure 12: Inputs to neighborhood energy use.
-------
103


GHG emissions
pe r gallon of
non-transit veh.
fuel
\ -/
Private-vehicle fuel
cons umption/re s ide nt/
day

Total private-vehicle
fuel consumption by
residents
+
Fuel consumption per day of
car trips with non-home end in
neighborhood
Total GHG emissions from car
trips with non-home end in
neighborhood per day
non-transit transportation
GHG emissions/resident/day
+ 4
total non-transit
GHG emissions
from resident travel
per day
Total non-transit GHG
emissions attributable to
trips generated by the
neighborhood pe r day

T rans portation-re late d
GHG
e mis s ions /re s ide nt/day

total public-transit-related GHG
emissions attributable to
neighborhood residents per day
+
total trans portation

GHG emissions

attributable to

neighborhood residents

per day
N on-trans portation
GHG emissions in
neighborhood per day
Total transportation-related
GHG emissions from trips
generated by the
ne ighboriiood pe r day
z:
total publie-transit-related
GHG emissions
attributable to trips
generated by the
neighborhood per day
Total neighborhood
GHG emissions per day


+
+
Total residents and
woi^ers in
neighborhood
Neighborhood GHG
emissions per
re s ide nt or worke r
per day
Figure 13: Inputs to neighborhood greenhouse gas emissions.
-------
104
Non-transit motor-veliicle
Total area affected
by near-road air
pollution in the
neighboihood
% of neighboihood
affected by
near-road air
pollution

Figure 14: Inputs to near-road air pollution exposure rates.
A community may choose to enact Complete Streets policies and practices for economic reasons (Figure
15). If the number of available modal options for reaching a given commercial district is increased, a
greater number of people will have an easy time getting there, theoretically increasing their patronage
of the businesses within that district. Furthermore, since Complete Streets principles encourage mixed
land uses and minimal setbacks of buildings from their property lines, the benefits of agglomeration
economies may also be realized. However, if dedicating more of a transportation right-of-way to non-
automobile modes too greatly reduces the convenience of automobile travel, travel to a given district by
that mode could conceivably be reduced by a greater amount than use of other modes is increased. In
that event, economic activity in that particular district could suffer. The assessment of economic
outcomes may be aided by the programs SMILE, SPASM, IDAS, and STEAM, all put out by the FHWA, and
by the tools of benefit-cost analyses, economic impact analyses, equity assessments, and cost-
effectiveness analyses.
person-miles within
neighborhood per day

in neighborhood>	\



^ Transit passengers passing
through neighborhood per
day
<% of neighborhood
land in conservation^

# of people affected by
near-road air pollution in
the neighborhood per day

<# ofjobs at schools
in neighborhood>


<# ofjobs on non-school
institutional land in
neighborhood>

-------
105

<# of jobs w/in 30-minute
offpeak-period drive of
neighborhood>
<# of jobs w/in 30-minute
peak-period drive of
neighborhood>

Cost of doing
business in the
metropolitan area
Competition from
other businesses
Cost ot
comme rcial-ve hicle
transportation for
businesses in the
neighborhood
Cost to businesses
in neighborhood of
one hi*, of labor
Net attractiveness
of the
neighborhood as
place to do
business

tloor space
per com. acre
Busmees
taxes
collected
Business revenue
attributable to
location in
neighborhood
Cost of doing
business in the
neighborhood
commercial
Business
vacancy
rl rate
tax rate

ave. price of
square ft of com.
floor space

Attractiveness of other locations m
the same metropolitan area to
businesses' potential customers
ave. price ot land
in com. use
(non-parking)
commercial
property
taxes/square It
of floor space
<% of neighborhood
affected by near-road air
pollution?5
<% of neighborhood
land in conservation'-
commercial
property
tax rate

<% of neighborhood within
traffic noise zones (off-peak
periods)>
<% of neighborhood within
traffic noise zones (peak
periods)>

Concentration of
local/regional air pollution in
neighborhood>
Figure 15: Inputs to neighborhood business activity.
Depending on what design elements are incorporated, Complete Streets has the potential to produce
benefits that are independent of how people use the transportation system. Trees, grass, and other
forms of vegetation may be incorporated into medians and pedestrian facilities. Such features may
mitigate the detrimental effects of stormwater runoff from impervious surfaces (such as paved
transportation facilities) and absorb a portion of the greenhouse gases emitted during the combustion
of transportation fuels. Furthermore, if it would not be practical to use pervious paving materials in the
construction of motor-vehicle lanes on a particular roadway, those materials might instead be used for
other surfaces within the transportation corridor, such as shoulders, sidewalks, and bike lanes (Figure
16). Finally, because Complete Streets encourages a compact development pattern, it may lead to
greater efficiency in the provision of various public services, such as the distribution of drinking water,
the collection of wastewater, and the provision of fire and police services. The effects of particular
Complete Streets designs on stormwater runoff effects could be evaluated using the EPA's Storm-Water
Management Model and the commercially-available HYDRUS. The assessment of water-infrastructure
outcomes may be aided by commercial design packages such as SewerCAD and StormCAD, as well as the
development of a water footprint index, comparing water usage with water availability. The degree to
which vegetation in a transportation right-of-way mitigates C02 emissions may be assessed using
greenhouse gas footprint analysis.
-------
106
Other tools discussed in this paper that may aid in the assessment of sustainability outcomes in
scenarios that do or do not feature Complete Streets practices include, but are not limited to, LEED-ND,
INVEST, and the EPA-developed DOSII, DASEES, C-FERST, Tribal-FERST, TRACI, EnviroAtlas, and Green
Communities Program.
; bike lane impervious
surface (neighborhood
streets)
transit ROW impervious
surface (neighborhood
streets)
on-street parking impervious
surface (arterial streets in
neighborhood)
sidewalk impervious surface
(arterial streets in
neighborhood)
turning lane/median
impervious surface
(neighborhood streets)
car lane impervious surface
(arterial streets in /
neighborhood) /
turning lane/median impervious
surface (arterial streets in
neighborhood)
sidewalk impervious
surface (neighborhood
streets) ,
gutter/drainage ditch
impervious surface
(neighborhood streets)
bike lane impervious surface
(arterial streets in
neighborhood)
on-street parking impervious
surface (neighborhood
	v a r—ju
N e ighborhood-s tre e t-corridor
impervious surface
transit ROW impervious
surface (arterial streets in
neighborhood) +
: car lane impervious
surface (neighborhood
streets)
gutter/drainage ditch
impervious surface (arterial
streets in neighborhood) +

Off-street
parking
Total impervious
surface within the
neighborhood
Arte rial-s tre e t-corridor
impervious surface
Transportation-corridor
impervious surface
within the neighborhood
Trans portation-re la te d
impervious surface
within the
neighborhood
Non-trans porta tion-re la te d
impervious surface
% of total
neighborhood area
that is covered in
impervious surface
% of neighborhood
land area not in
conservation use that
is covered in
impervious surface
Figure 16: Inputs to neighborhood imperviousness.
5 IMPORTANT INFORMATION GAPS
The following is a compilation of important information gaps described by cited researchers and
highlighted throughout this report. This section has been organized to highlight issues raised in the
previous discussions rather than consolidate and integrate the research needs in order to provide the
summary in order of topics presented and avoid confusion or oversimplification that may occur when
consolidating these issues.
From "Tools, Resources, and Indicators for Assessing Sustainability" (Opening) Section
•	Insufficient understanding of the measurement of sustainability in the regular functions of a
transportation agency (Ramani et al. 2011).
•	Uncertainty in the inputs of transportation and land use models (Duthie et al. 2010).
•	Insufficient understanding of how the built environment affects pedestrian and bicycle travel, as
opposed to motorized transportation (Cervero and Duncan 2003).
-------
107
From "Drivers of Transportation Behavior" Section
•	Research to date has given little attention to the travel-behavior effects of the land uses that
surround a person's place of work, as opposed to those that surround their place of residence
(Frank et al. 2008).
•	The phenomenon of induced demand is not well understood, would require very complex
theories to model, and is not incorporated into standard travel demand forecasts (Kitamura
2009).
•	The greatest source of doubt that exists regarding the magnitude of the effect of the built
environment on travel behavior is residential self-selection, the idea that a neighborhood built
around a particular kind of transportation behavior will simply attract residents who already had
a predisposition towards that kind of transportation behavior rather than motivating other
people to change their transportation behavior (Ewing and Cervero 2010).
•	Little research has so far been conducted on the manner in which parents decide how their
children will travel to school (Faulkner et al. 2010).
From "Transportation-Behavior Modeling Techniques" Subsection
•	Most conventional travel behavior models consist of four sequential steps, based on the
assumption that people first decide to make a trip, then decide what the destination of that trip
will be, then decide what mode they will make the trip by, then decide what route the trip will
follow. However, a better reflection of reality might be modeling these decisions as being
simultaneous. Unfortunately, just like most other modeling techniques that produce a closer
representation of reality, modeling people as making multiple decisions at the same time would
be very complex, difficult, and time-consuming, resulting in a necessary tradeoff between
accuracy and expediency (Hasan and Dashti 2007, Newman and Bernardin Jr. 2010).
•	A better understanding is needed of systems approaches to integrating behavior choices with
other factors in the decision-making process.
•	Few studies have so far examined the possibility of modeling mode choices before destination
choices (Newman and Bernardin Jr. 2010).
•	The theoretical link between transportation system capacity and trip generation has not yet
been proven (Kitamura 2009).
•	Most travel behavior models do not give much consideration to the time of day at which a
person departs on a trip, and most of the models that do give this consideration only narrow
down departure times to one of a series of broad time periods, usually a couple hours in
duration. Greater precision than this would be required in order to study variations in traffic
congestion within a given peak travel period, which is currently considered a high-priority
research area (Kristoffersson and Engelson 2009).
From "Energy Use" Section
•	While there is a range of studies regarding specific issues (e.g., underground storage tanks for
ethanol blends), there do not appear to be tools or resources that look systematically at how
transportation fuel infrastructure fits within the broader sustainable communities context.
-------
108
•	While research on life-cycle impacts has often focused on upstream impacts of fuel and
electricity production, there are also GHG and other environmental impacts related to material
flows. Although their total contribution to GHG impacts and energy use may be small, the
production of vehicles and vehicle components, as well as of larger transportation infrastructure
(e.g., concrete production), may need to be considered in terms of other environmental
endpoints.
•	A key knowledge gap and research need identified by the Strategic Highway Research Program
is "the effect of government interventions (e.g. pricing, infrastructure deployment) on
technology advancement...many local and state governments are interested in how they can
provide incentive to help accelerate the adoption of new vehicle technologies."
•	Understanding the potential for eco-driving in the U.S. to reduce GHG emissions and energy use
has been identified as a research need.
•	There are a number of research initiatives attempting to understand the market for electric
vehicles and charging patterns for this still-small but growing segment of the fleet, as well as to
understand potential impacts on electric-power dispatching. However, beyond that is the
question of how communities can promote synergistic solutions between vehicle electrification
and energy strategies, and how to quantify their benefits.
•	Research has been conducted to look at the question of whether plug-in vehicle buyers want
green-power electricity. This could point to whether there are market preferences that would
support a combination of green power and plug-in hybrid or electric vehicles, as well as point to
policy and marketing strategies to advance these solutions. However, there is a lack of tools and
resources available to communities to support these types of strategies.
From "Water Infrastructure" Subsection
•	Methods for projecting population growth and activity modeling.
•	The EPA research programs (e.g., SHC, ACE, and SSWR) are developing analysis methods for
individual water infrastructure components. New developments, such as those in nutrient
recovery, leak detection, and green infrastructure, are focused on water infrastructure or its
components alone. Restraints from urban transportation, accounting for mode and
accessibility, are not yet considered. For integrated planning, however, the smart growth
methodology needs to simultaneously consider both types of infrastructures in option
evaluation.
•	Smart growth in urban development is often defined differently in literature than in practice. In
light of urban development, smart growth refers to high population density, low carbon
intensity, and green infrastructure, thus creating a livable environment. This premise requires
the accommodation of mass transit and other transportation infrastructures. The
transportation policies contained in U.S. Environmental Protection Agency (2011a) all aim to
increase urban population density, create a walkable and pedestrian-friendly community, and
maximize the use of existing infrastructure through infill and redevelopment. U.S.
Environmental Protection Agency (2011a) outlined major types of transportation modes with
the following sustainability metrics:
-------
109
¦	Transit accessibility, for reduced VMT and increased public transportation
¦	Bicycle and pedestrian mode share
¦	VMT per capita, to reduce automobile reliance and commute distances
¦	Carbon intensity (of transportation) per capita
¦	Percentage land-use mixture in terms of residential and commercial activities
•	How changes toward smart growth affect urban water infrastructure, and in return how water
services facilitate or impede smart growth, has received little attention.
•	Numerous published studies (e.g., U.S. Environmental Protection Agency (2002b)) have shown
that biofuel combustion in current fleet engines leads to a meaningful reduction in some
pollutants (e.g., particulate matter (PM), carbon monoxide (CO), and hydrocarbons), but an
increase in others (e.g., aldehydes and nitrous oxides (NOx)). Chemical and physical properties
of PM emissions also change when biofuels are used. These changes are important to near-road
air pollutant monitoring and mitigation options (Baldauf et al. 2008). Their potential effects on
land use decisions, and hence water infrastructure planning, have not been thoroughly
investigated.
From "Physical Health" Subsection
•	Research is still wanting on the relative impacts that various aesthetic and safety features have
on physical activity.
•	More research is also needed on how people's demographic and socioeconomic characteristics
affect their walking behavior when all other factors are controlled for.
•	Research could shed more light on the degree to which pedestrians regard public spaces as
either a destination or an obstruction.
As the SHCRP program develops, these and other research needs raised should be considered for further
evaluation both as individual activities and integrated programs to enhance communities in their drive
to achieve sustainable transportation developments.
-------
110
BIBLIOGRAPHY
American Water Works Association. 2003. Waterstats 2002 Water Utility Distribution Survey CD-ROM.
Denver, CO: American Water Works Association.
Axsen, J., and K. Kurani. 2013. Do plug-in vehicle buyers want green electricity? A survey of U.S. new car
buyers. In Transportation Research Board92nd Annual Meeting. Washington, DC.
Baldauf, R., E. Thoma, A. Khlystov, V. Isakov, G. Bowker, T. Long, and R. Snow. 2008. "Impacts of noise
barriers on near-road air quality." Atmospheric Environment no. 42 (32):7502-7507. doi:
10.1016/j.atmosenv.2008.05.051.
Baldauf, R. W., T. A. Cahill, C. R. Bailey, A. Khlystov, K. M. Zhang, J. R. Cook, . . . G. Bowker. 2009. Can
roadway design be used to mitigate air quality impacts from traffic? Environmental Manager
(EM), August.
Bare, Jane. 2011. "TRACI 2.0: The tool for the reduction and assessment of chemical and other
environmental impacts 2.0." Clean Technologies and Environmental Policy no. 13 (5):687-696.
doi: 10.1007/sl0098-010-0338-9.
Bare, Jane C. 2012. Tool for the Reduction and Assessment of Chemical and Other Environmental
Impacts (TRACI), Software Name and Version Number: TRACI version 2.1 - User's Manual. U.S.
Environmental Protection Agency, Office of Research and Development, National Risk
Management Research Laboratory, Sustainable Technology Division, Systems Analysis Branch.
Bare, Jane C., Patrick Hofstetter, David W. Pennington, and Helias A. Udo de Haes. 2000. "Midpoints
versus endpoints: The sacrifices and benefits." The International Journal of Life Cycle Assessment
no. 5 (6):319-326. doi: 10.1007/bf02978665.
Bare, Jane C., Gregory A. Norris, David W. Pennington, and Thomas McKone. 2002. "TRACI: The Tool for
the Reduction and Assessment of Chemical and Other Environmental Impacts." Journal of
Industrial Ecology no. 6 (3-4):49-78. doi: 10.1162/108819802766269539.
Baynes, Timothy M. 2009. "Complexity in urban development and management: Historical overview and
opportunities." Journal of Industrial Ecology no. 13 (2):214-227. doi: 10.1111/j.l530-
9290.2009.00123.x.
Berechman, J., and R. E. Paaswell. 1997. "The implications of travel profiles for transportation
investment: The Bronx Center Project." Transportation no. 24 (l):51-77. doi:
10.1023/a: 1017912602000.
Berndes, Goran. 2002. "Bioenergy and water - the implications of large-scale bioenergy production for
water use and supply." Global Environmental Change-Human and Policy Dimensions no. 12
(4):253-271. doi: 10.1016/s0959-3780(02)00040-7.
Besser, Lilah M., and Andrew L. Dannenberg. 2005. "Walking to Public Transit: Steps to Help Meet
Physical Activity Recommendations." American Journal of Preventive Medicine no. 29 (4):273-
280. doi: 10.1016/j.amepre.2005.06.010.
-------
Ill
Besser, Lilah M., Michele Marcus, and Howard Frumkin. 2008. "Commute time and social capital in the
U.S." American Journal of Preventive Medicine no. 34 (3):207-211. doi:
10.1016/j.amepre. 2007.12.004.
Bhatia, Rajiv, and Megan Wier. 2011. ""Safety in Numbers" re-examined: Can we make valid or practical
inferences from available evidence?" Accident; analysis and prevention no. 43 (l):235-240. doi:
10.1016/j.aap.2010.08.015.
Black, J. A., A. Paez, and P. A. Suthanaya. 2002. "Sustainable urban transportation: Performance
indicators and some analytical approaches." Journal of Urban Planning & Development no. 128
(4):184-209.
Brabec, Elizabeth, Stacey Schulte, and Paul L. Richards. 2002. "Impervious surfaces and water quality: A
review of current literature and its implications for watershed planning." Journal of Planning
Literature no. 16 (4):499-514.
Braveman, Paula, Susan Egerter, and David R. Williams. 2011. "The Social Determinants of Health:
Coming of Age." Annual Review of Public Health no. 32 (l):381-98. doi: 10.1146/annurev-
publhealth-031210-101218.
Burchfield, Marcy, Henry G. Overman, Diego Puga, and Matthew A. Turner. 2006. "Causes of sprawl: A
portrait from space." The Quarterly Journal of Economics no. 121 (2):587-633. doi:
10.1162/qjec.2006.121.2.587.
Byun, Daewon, and Kenneth L. Schere. 2006. "Review of the governing equations, computational
algorithms, and other components of the models-3 Community Multiscale Air Quality (CMAQ)
modeling system." Applied Mechanics Reviews no. 59 (l-6):51-77. doi: 10.1115/1.2128636.
Carruthers, John I., and Alexander C. Vias. 2005. "Urban, suburban, and exurban sprawl in the Rocky
Mountain West: Evidence from regional adjustment models." Journal of Regional Science no. 45
(l):21-48. doi: 10.1111/j.0022-4146.2005.00363.x.
Centers for Disease Control and Prevention. 2013. Injury Prevention & Control: Data & Statistics
(WISQARS) [cited March 22 2013], Available from
http://www.cdc.gov/iniury/wisqars/index.html.
Centers for Disease Control and Prevention. 2013. CDC Transportation Recommendations, August 2 2012
[cited April 11 2013], Available from http://www.cdc.gov/transportation/default.htm.
Centre for Diet and Activity Research. 2013. Integrated Transport and Health Impact Modelling Tool
(ITHIM), 2013/04/02/16:35:47 [cited April 2 2013], Available from
http://www.cedar.iph.cam.ac.uk/research/modelling/ithim/.
Cervero, Robert, and Michael Duncan. 2003. "Walking, bicycling, and urban landscapes: Evidence from
the San Francisco bay area." American Journal of Public Health no. 93 (9):1478-1483.
Chai, Lin, Masoud Kayhanian, Brandon Givens, John T. Harvey, and David Jones. 2012. "Hydraulic
performance of fully permeable highway shoulder for storm water runoff management." Journal
-------
112
of Environmental Engineering-Asce no. 138 (7):711-722. doi: 10.1061/(asce)ee.l943-
7870.0000523.
Chang, Ni-Bin, Cheng Qi, and Y. Jeffrey Yang. 2012. "Optimal expansion of a drinking water infrastructure
system with respect to carbon footprint, cost-effectiveness and water demand." Journal of
Environmental Management no. 110:194-206. doi: 10.1016/j.jenvman.2012.06.004.
Chang, Shu Chun. 2007. "Measuring and assessing economic activities of central business districts for
planning purposes." Journal of Urban Planning & Development no. 133 (2):95-98. doi:
10.1061/(asce)0733-9488(2007)133:2(95).
Chenoweth, David, and Joe Leutzinger. 2010. "The Economic Cost of Physical Inactivity and Excess
Weight in American Adults." Journal of Physical Activity and Health no. 3 (2).
Chi, Guangqing, and Brian Stone. 2005. "Sustainable transport planning: Estimating the ecological
footprint of vehicle travel in future years." Journal of Urban Planning & Development no. 131
(3):170-180. doi: 10.1061/(asce)0733-9488(2005)131:3(170).
Christian, Thomas J. 2012a. "Automobile commuting duration and the quantity of time spent with
spouse, children, and friends." Preventive Medicine no. 55 (3):215-218. doi:
10.1016/j.ypmed. 2012.06.015.
Christian, Thomas J. 2012b. "Trade-offs between commuting time and health-related activities." Journal
of Urban Health: Bulletin of the New York Academy of Medicine no. 89 (5):746-757. doi:
10.1007/sll524-012-9678-6.
City of Olympia, Public Works Department, Water Resources Program. 1995. Impervious Surface
Reduction Study: Final Report. Olympia, WA.
Confalonieri, Ulisses, Bettina Menne, Rais Akhtar, Kristie L. Ebi, Maria Hauengue, R. Sari Kovats, . . .
Alistair Woodward. 2007. "2007: Human Health." In Climate Change 2007: Impacts, Adaptation
and Vulnerability. Contribution of Working Group II to the Fourth Assessment of the
Intergovernmental Panel on Climate Change., edited by M. L. Parry, O. F. Canziani, J. P. Palutikof,
P. J. van der Linden and C. E. Hanson, 391-431. Cambridge, UK: Cambridge University Press.
Dallmann, Timothy R., and Robert A. Harley. 2010. "Evaluation of mobile source emission trends in the
United States." Journal of Geophysical Research-Atmospheres no. 115. doi:
10.1029/2010jd013862.
Davies, Hugh, and Irene van Kamp. 2012. "Noise and cardiovascular disease: A review of the literature
2008-2011." Noise & Health no. 14 (61):287-291. doi: 10.4103/1463-1741.104895.
Davis, Stacy C., Susan W. Diegel, and Robert G. Boundy. 2012. Transportation Energy Data Book: Edition
31. U.S. Department of Energy, Oak Ridge National Laboratory.
de Fraiture, Charlotte, Mark Giordano, and Yongsong Liao. 2008. "Biofuels and implications for
agricultural water use: Blue impacts of green energy." Water Policy no. 10 (Sl):67-81. doi:
10.2166/wp.2008.054.
-------
113
de Hartog, Jeroen Johan, Hanna Boogaard, Hans Nijland, and Gerard Hoek. 2010. "Do the Health
Benefits of Cycling Outweigh the Risks?" Environmental Health Perspectives no. 118 (8):1109-
1116. doi: 10.1289/ehp.0901747.
de Nazelle, Audrey, Mark J. Nieuwenhuijsen, Josep M. Anto, Michael Brauer, David Briggs, Charlotte
Braun-Fahrlander, . . . Erik Lebret. 2011. "Improving health through policies that promote active
travel: A review of evidence to support integrated health impact assessment." Environment
International no. 37 (4):766-777. doi: 10.1016/j.envint.2011.02.003.
Dickens, Andy P., Suzanne H. Richards, Colin J. Greaves, and John L. Campbell. 2011. "Interventions
targeting social isolation in older people: A systematic review." BMC Public Health no. 11. doi:
10.1186/1471-2458-11-647.
Dodder, Rebecca, Tyler Felgenhauer, Carey King, and William Yelverton. 2011. Water and greenhouse
gas tradeoffs associated with a transition to a low carbon transportation system. In 2011
International Mechanical Engineering Congress and Exposition. Denver, CO: American Society of
Mechanical Engineers.
Dominguez-Faus, R., Susan E. Powers, Joel G. Burken, and Pedro J. Alvarez. 2009. "The water footprint of
biofuels: A drink or drive issue?" Environmental Science & Technology no. 43 (9):3005-3010. doi:
10.1021/es802162x.
Duthie, Jennifer, Avinash Voruganti, Kara Kockelman, and S. Travis Waller. 2010. "Highway improvement
project rankings due to uncertain model inputs: Application of traditional transportation and
land use models." Journal of Urban Planning & Development no. 136 (4):294-302. doi:
10.1061/(asce)u p. 1943-5444.0000026.
El-Diraby, T. E., B. Abdulhai, and K. C. Pramod. 2005. "The application of knowledge management to
support the sustainable analysis of urban transportation infrastructure." Canadian Journal of
Civil Engineering no. 32 (1):58-71. doi: 10.1139/104-115.
Ewing, Reid, and Robert Cervero. 2010. "Travel and the built environment." Journal of the American
Planning Association no. 76 (3):265-294. doi: 10.1080/01944361003766766.
Ewing, Reid H. 2008. "Characteristics, causes, and effects of sprawl: A literature review." In Urban
Ecology, edited by John M. Marzluff, Eric Shulenberger, Wilfried Endlicher, Marina Alberti,
Gordon Bradley, Clare Ryan, Ute Simon and Craig ZumBrunnen, 519-535. Springer US.
Fairfax County. 2013. Chesapeake Bay Preservation Ordinance [cited May 29 2013], Available from
http://www.fairfaxcounty.gov/dpwes/environmental/cbav/.
Faulkner, Guy E. J., Vanessa Richichi, Ronald N. Buliung, Caroline Fusco, and Fiona Moola. 2010. "What's
"quickest and easiest?": Parental decision making about school trip mode." International Journal
of Behavioral Nutrition and Physical Activity no. 7. doi: 10.1186/1479-5868-7-62.
Fischer, Edward L., Gabe K. Rousseau, Shawn M. Turner, Ernest J. Blais, Cindy L. Engelhart, David R.
Henderson, .. . Charles V. Zegeer. 2010. Pedestrian and Bicyclist Safety and Mobility in Europe.
Washington, DC: US Department of Transportation.
-------
114
Frank, Lawrence, Mark Bradley, Sarah Kavage, James Chapman, and T. Keith Lawton. 2008. "Urban form,
travel time, and cost relationships with tour complexity and mode choice." Transportation no.
35 (l):37-54. doi: 10.1007/sllll6-007-9136-6.
Freeland, Amy L., Shailendra N. Banerjee, Andrew L. Dannenberg, and Arthur M. Wendel. 2013.
"Walking associated with public transit: Moving toward increased physical activity in the United
States." American Journal of Public Health no. 103 (3):536-542. doi: 10.2105/ajph. 2012.300912.
Fritschi, Lin, A. Lex Brown, Rokho Kim, Dietrich Schwela, and Stelios Kephalopoulos. 2011. Burden of
disease from environmental noise: Quantification of healthy life years lost in Europe.
Copenhagen, Denmark: World Health Organization.
Fujiwara, Takeo, and Ichiro Kawachi. 2008. "Social capital and health - A study of adult twins in the US."
American Journal of Preventive Medicine no. 35 (2):139-144. doi:
10.1016/j.amepre. 2008.04.015.
Garcia-Lopez, Miquel-Angel. 2012. "Urban spatial structure, suburbanization and transportation in
Barcelona." Journal of Urban Economics no. 72 (2-3):176-190. doi:
http://dx.doi.Org/10.1016/i.iue.2012.05.003.
Gim, Tae-Hyoung Tommy. 2012. "A meta-analysis of the relationship between density and travel
behavior." Transportation no. 39 (3):491-519. doi: 10.1007/sllll6-011-9373-6.
Goel, Rahul, and Mark W. Burris. 2012. "Hot lane policies and their implications." Transportation no. 39
(6):1019-1033. doi: 10.1007/sllll6-011-9382-5.
Goetz, Scott J., Claire A. Jantz, Stephen D. Prince, Andrew J. Smith, Dmitry Varlyguin, and Robb K. Wright.
2004. "Integrated analysis of ecosystem interactions with land use change: The Chesapeake Bay
watershed." In Ecosystems and Land Use Change, edited by R. S. DeFries, G. P. Asner and R. A.
Houghton, 263-275. Washington: American Geophysical Union.
Gomez-Ullate, Elena, Elena Castillo-Lopez, Daniel Castro-Fresno, and Joseba Rodriguez Bayon. 2011.
"Analysis and contrast of different pervious pavements for management of storm-water in a
parking area in northern Spain." Water Resources Management no. 25 (6):1525-1535. doi:
10.1007/sll269-010-9758-x.
Gordon, P., H. W. Richardson, and H. L. Wong. 1986. "The distribution of population and employment in
a polycentric city: The case of Los Angeles." Environment & Planning A no. 18 (2):161-173. doi:
10.1068/a180161.
Greene, David L., and Steven E. Plotkin. 2011. Reducing Greenhouse Gas Emissions from U.S.
Transportation. Arlington, VA: Pew Center on Global Climate Change.
Gripsrud, Mattias, and Randi Hjorthol. 2012. "Working on the train: From 'dead time' to productive and
vital time." Transportation no. 39 (5):941-956. doi: 10.1007/sllll6-012-9396-7.
Guo, James C. Y., Gerald E. Blackler, T. Andrew Earles, and Ken MacKenzie. 2010. "Incentive index
developed to evaluate storm-water low-impact designs." Journal of Environmental Engineering-
ASCE no. 136 (12):1341-1346. doi: 10.1061/(asce)ee.l943-7870.0000270.
-------
115
Hamer, Mark, and Yoichi Chida. 2008. "Active commuting and cardiovascular risk: A meta-analytic
review." Preventive Medicine no. 46 (1):9-13. doi: 10.1016/j.ypmed.2007.03.006.
Hasan, Mohamad K., and Hussain M. Dashti. 2007. "A multiclass simultaneous transportation
equilibrium model." Networks & Spatial Economics no. 7 (3):197-211. doi: 10.1007/sll067-006-
9014-3.
Haustein, Sonja. 2012. "Mobility behavior of the elderly: An attitude-based segmentation approach for a
heterogeneous target group." Transportation no. 39 (6):1079-1103. doi: 10.1007/sllll6-011-
9380-7.
HEI Panel on the Health Effects of Traffic-Related Air Pollution. 2010. Traffic-Related Air Pollution: A
Critical Review of the Literature on Emissions, Exposure, and Health Effects. Boston: Health
Effects Institute.
Heikkila, E., P. Gordon, J. I. Kim, R. B. Peiser, H. W. Richardson, and D. Dale-Johnson. 1989. "What
happened to the CBD-distance gradient?: Land values in a policentric city." Environment and
Planning A no. 21 (2):221-232. doi: 10.1068/a210221.
Hof, Bert, Arjan Heyma, and Toon van der Hoorn. 2012. "Comparing the performance of models for
wider economic benefits of transport infrastructure: results of a Dutch case study."
Transportation no. 39 (6):1241-1258. doi: 10.1007/sllll6-011-9383-4.
Holt-Lunstad, Julianne, Timothy B. Smith, and J. Bradley Layton. 2010. "Social relationships and mortality
risk: A meta-analytic review." PLOS Medicine no. 7 (7). doi: 10.1371/journal.pmed.1000316.
Hoot, Nathan R., and Dominik Aronsky. 2008. "Systematic review of emergency department crowding:
Causes, effects, and solutions." Annals of Emergency Medicine no. 52 (2):126-136. doi:
10.1016/j.annemergmed. 2008.03.014.
Horner, John W., John R. Stone, and Leta F. Huntsinger. 2008. "Data reuse methods for transportation
planning in small- and medium-sized towns." Journal of Urban Planning & Development no. 134
(4):149-152. doi: 10.1061/(asce)0733-9488(2008)134:4(149).
Hume, Kenneth I., Mark Brink, and Mathias Basner. 2012. "Effects of environmental noise on sleep."
Noise & Health no. 14 (61):297-302. doi: 10.4103/1463-1741.104897.
Int Panis, Luc, Bas de Geus, Gregory Vandenbulcke, Hanny Willems, Bart Degraeuwe, Nico Bleux, . . .
Romain Meeusen. 2010. "Exposure to particulate matter in traffic: A comparison of cyclists and
car passengers." Atmospheric Environment no. 44 (19):2263-2270. doi:
10.1016/j.atmosenv.2010.04.028.
Jacobsen, P. L. 2003. "Safety in numbers: More walkers and bicyclists, safer walking and bicycling." Injury
Prevention no. 9 (3):205-209. doi: 10.1136/ip.9.3.205.
Jacobson, Justin, and Ann Forsyth. 2008. "Seven American TODs: Good practices for urban design in
Transit-Oriented Development projects." Journal of Transport and Land Use no. 1 (2):51-88.
-------
116
Jha, Manoj K., and Eungcheol Kim. 2006. "Highway route optimization based on accessibility, proximity,
and land-use changes." Journal of Transportation Engineering no. 132 (5):435-439. doi:
10.1061/(asce)0733-947x(2006) 132:5(435).
Kahlmeier, Sonja, Nick Cavil I, Hywell Dinsdale, Harry Rutter, Thomas Gotschi, Charlie Foster, .. .
Francesca Racioppi. 2011. Health economic assessment tools (HEAT) for walking and for cycling.
Methodology and user guide. Copenhagen: World Health Organization.
Karner, Alex A., Douglas S. Eisinger, and Deb A. Niemeier. 2010. "Near-roadway air quality: Synthesizing
the findings from real-world data." Environmental Science & Technology no. 44 (14):5334-5344.
doi: 10.1021/esl00008x.
Kennedy, C. A. 2002. "A comparison of the sustainability of public and private transportation systems:
Study of the Greater Toronto Area." Transportation no. 29 (4):459-493. doi:
10.1023/a: 1016302913909.
Kim, Hong Sok, and Eungcheol Kim. 2004. "Effects of public transit on automobile ownership and use in
households of the USA." Review of Urban & Regional Development Studies no. 16 (3):245-262.
doi: 10. llll/j.l467-940X. 2005.00090.x.
Kim, Sungyop, and Gudmundur F. Ulfarsson. 2008. "Curbing automobile use for sustainable
transportation: Analysis of mode choice on short home-based trips." Transportation no. 35
(6):723-737. doi: 10.1007/sllll6-008-9177-5.
Kirmeyer, G., W. Richards, and C. D. Smith. 1994. An Assessment of Water Distribution Systems and
Associated Research Needs. Denver, CO: American Water Works Association Research
Foundation.
Kitamura, Ryuichi. 2009. "The effects of added transportation capacity on travel: A review of theoretical
and empirical results." Transportation no. 36 (6):745-762. doi: 10.1007/sllll6-009-9245-5.
Knibbs, Luke D., Tom Cole-Hunter, and Lidia Morawska. 2011. "A review of commuter exposure to
ultrafine particles and its health effects." Atmospheric Environment no. 45 (16):2611-2622. doi:
10.1016/j.atmosenv.2011.02.065.
Knowlton, Kim, Miriam Rotkin-Ellman, Linda Geballe, Wendy Max, and Gina M. Solomon. 2011. "Six
Climate Change-Related Events In The United States Accounted For About $14 Billion In Lost
Lives And Health Costs." Health Affairs no. 30 (11):2167-2176. doi: 10.1377/hlthaff.2011.0229.
Koo, Dae-Hyun, Samuel T. Ariaratnam, and Jr Edward Kavazanjian. 2009. "Development of a
sustainability assessment model for underground infrastructure projects." Canadian Journal of
Civil Engineering no. 36 (5):765-776. doi: 10.1139/109-024.
Kristoffersson, Ida, and Leonid Engelson. 2009. "A dynamic transportation model for the Stockholm
area: Implementation issues regarding departure time choice and OD-pair reduction." Networks
& Spatial Economics no. 9 (4):551-573. doi: 10.1007/sll067-009-9104-0.
-------
117
Landers, Dixon H., and Amanda M. Nahlik. 2013. Final Ecosystem Goods and Services Classification
System (FEGS-CS). Washington, DC: U.S. Environmental Protection Agency, Office of Research
and Development.
Larson, William, Feng Liu, and Anthony Yezer. 2012. "Energy footprint of the city: Effects of urban land
use and transportation policies." Journal of Urban Economics no. 72 (2-3):147-159. doi:
http://dx.doi.Org/10.1016/i.iue.2012.05.001.
Lemp, Jason D., Bin Zhou, Kara M. Kockelman, and Barbara M. Parmenter. 2008. "Visioning versus
modeling: Analyzing the land-use-transportation futures of urban regions." Journal of Urban
Planning & Development no. 134 (3):97-109. doi: 10.1061/(asce)0733-9488(2008)134:3(97).
Levinson, Herbert S. 2004. "Highways, people, and places: Past, present, and future." Journal of
Transportation Engineering no. 130 (4):406-411. doi: 10.1061/(asce)0733-947x(2004)130:4(406).
Lingaraju, Bala P., Joo-Youp Lee, and Y. Jeffrey Yang. 2013. "Process and utility water requirements for
cellulosic ethanol production processes via fermentation pathway." Environmental Progress &
Sustainable Energy no. 32 (2):396-405. doi: 10.1002/ep.11604.
Loo, Becky P. Y., and S. Y. Chow. 2006. "Sustainable urban transportation: Concepts, policies, and
methodologies." Journal of Urban Planning & Development no. 132 (2):76-79. doi:
10.1061/(asce)0733-9488(2006)132:2(76).
Maizlish, Neil, James Woodcock, Sean Co, Bart Ostro, Amir Fanai, and David Fairley. 2013. "Health
Cobenefits and Transportation-Related Reductions in Greenhouse Gas Emissions in the San
Francisco Bay Area." American Journal of Public Health no. 103 (4):703-709. doi:
10.2105/AJPH.2012.300939.
Majeau-Bettez, Guillaume, Troy R. Hawkins, and Anders Hammer Stromman. 2011. "Life cycle
environmental assessment of lithium-ion and nickel metal hydride batteries for plug-in hybrid
and battery electric vehicles." Environmental Science & Technology no. 45 (10):4548-4554. doi:
10.1021/esl03607c.
Martens, Karel. 2012. "Justice in transport as justice in accessibility: Applying Walzer's 'Spheres of
Justice' to the transport sector." Transportation no. 39 (6):1035-1053. doi: 10.1007/sllll6-012-
9388-7.
McDonald, Noreen C. 2008. "Children's mode choice for the school trip: The role of distance and school
location in walking to school." Transportation no. 35 (l):23-35. doi: 10.1007/sllll6-007-9135-7.
McMichael, T., H. Montgomery, and A. Costello. 2012. "Health risks, present and future, from global
climate change." BMJ no. 344 (I):el359-el359. doi: 10.1136/bmj.el359.
McNabola, Aonghus, Brian M. Broderick, and Laurence W. Gill. 2009. "A principal components analysis
of the factors effecting personal exposure to air pollution in urban commuters in Dublin,
Ireland." Journal of Environmental Science and Health, Part A no. 44 (12):1219-1226. doi:
10.1080/10934520903139928.
-------
118
Medda, Franceses, Peter Nijkamp, and Piet Rietveld. 2003. "Urban land use for transport systems and
city shapes." Geographical Analysis no. 35 (l):46-57.
Mercier, Jean. 2009. "Equity, social justice, and sustainable urban transportation in the twenty-first
century." Administrative Theory & Praxis (M.E. Sharpe) no. 31 (2):145-163. doi:
10.2753/atpl084-1806310201.
Metropolitan Area Planning Council. 2013. Creating a "Park Once" District, February 3 2010 [cited
August 8 2013], Available from http://www.mapc.org/resources/parking-toolkit/parking-issues-
questions/create-park-once-district.
Meyboom, Annalisa. 2009. "Infrastructure as practice." Journal of Architectural Education no. 62 (4):72-
81. doi: 10.Ill 1/j. 1531-314X.2009.01006.x.
Miles, Rebecca, Christopher Coutts, and Asal Mohamadi. 2012. "Neighborhood urban form, social
environment, and depression." Journal of Urban Health-Bulletin of the New York Academy of
Medicine no. 89 (1):1-18. doi: 10.1007/sll524-011-9621-2.
Montgomery County, Department of Environmental Protection. 2013. The Paint Branch Watershed
[cited May 29 2013], Available from
http://www6.montgomervcountvmd.gov/dectmpl.aspPurh/content/dep/water/sub paint.asp.
Murray, Christopher J. L., Theo Vos, Rafael Lozano, Mohsen Naghavi, Abraham D. Flaxman, Catherine
Michaud,. . . Alan D. Lopez. 2012. "Disability-adjusted life years (DALYs) for 291 diseases and
injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study
2010." The Lancet no. 380 (9859):2197-2223. doi: 10.1016/S0140-6736(12)61689-4.
National Research Council of the National Academies, Division on Earth and Life Studies, Board on
Environmental Studies and Toxicology, Committee on Health Impact Assessment. 2011a.
Improving Health in the United States: The Role of Health Impact Assessment. Washington, DC:
The National Academies Press.
National Research Council of the National Academies, Division on Earth and Life Studies, Board on
Environmental Studies and Toxicology, Committee on Improving Risk Analysis Approaches Used
by the U.S. Environmental Protection Agency. 2009. Science and Decisions: Advancing Risk
Assessment. Washington, DC: The National Academies Press.
National Research Council of the National Academies, Division on Earth and Life Studies, Water Science
and Technology Board, Committee on Water Implications of Biofuels Production in the United
States. 2008. Water Implications of Biofuels Production in the United States. Washington, DC:
The National Academies Press.
National Research Council of the National Academies, Policy and Global Affairs Division, Science and
Technology for Sustainability Program, Committee on Incorporating Sustainability in the U.S.
Environmental Protection Agency. 2011b. Sustainability and the U.S. EPA. Washington, DC: The
National Academies Press.
Naumann, Rebecca B., Ann M. Dellinger, Eduard Zaloshnja, Bruce A. Lawrence, and Ted R. Miller. 2010.
"Incidence and Total Lifetime Costs of Motor Vehicle-Related Fatal and Nonfatal Injury by Road
-------
119
User Type, United States, 2005." Traffic Injury Prevention no. 11 (4):353-360. doi:
10.1080/15389588.2010.486429.
Neckerman, Kathryn M., Gina S. Lovasi, Stephen Davies, Marnie Purciel, James Quinn, Eric Feder,. . .
Andrew Rundle. 2009. "Disparities in urban neighborhood conditions: Evidence from GIS
measures and field observation in New York City." Journal of Public Health Policy no. 30:S264-
S285. doi: 10.1057/jphp.2008.47.
Newman, Jeffrey P., and Vincent L. Bernardin Jr. 2010. "Hierarchical ordering of nests in a joint mode
and destination choice model." Transportation no. 37 (4):677-688. doi: 10.1007/sllll6-010-
9277-x.
Ng, Chen Feng, and Kenneth A. Small. 2012. "Tradeoffs among free-flow speed, capacity, cost, and
environmental footprint in highway design." Transportation no. 39 (6):1259-1280. doi:
10.1007/sllll6-012-9395-8.
Ostrom, Elinor. 2010. "Beyond markets and states: Polycentric governance of complex economic
systems." American Economic Review no. 100 (3):641-672. doi: 10.1257/aer.l00.3.641.
Patz, Jonathan A., Diarmid Campbell-Lendrum, Tracey Holloway, and Jonathan A. Foley. 2005. "Impact of
regional climate change on human health." Nature no. 438 (7066):310-317. doi:
10.1038/nature04188.
Polzin, Steven E. 1999. "Transportation/land-use relationship: Public transit's impact on land use."
Journal of Urban Planning & Development no. 125 (4):135-151.
Popuri, Yasasvi, Kimon Proussaloglou, Cemal Ayvalik, Frank Koppelman, and Aimee Lee. 2011.
"Importance of traveler attitudes in the choice of public transportation to work: Findings from
the Regional Transportation Authority Attitudinal Survey." Transportation no. 38 (4):643-661.
doi: 10.1007/sllll6-011-9336-y.
Pucher, John, and Ralph Buehler. 2008. "Cycling for Everyone: Lessons from Europe." Transportation
Research Record: Journal of the Transportation Research Board no. 2074 (-l):58-65. doi:
10.3141/2074-08.
Ramani, Tara L., Josias Zietsman, William E. Knowles, and Luca Quadrifoglio. 2011. "Sustainability
enhancement tool for state departments of transportation using performance measurement."
Journal of Transportation Engineering no. 137 (6):404-415. doi: 10.1061/(asce)te.l943-
5436.0000255.
Roberto, Elizabeth. 2008. Commuting to Opportunity: The Working Poor and Commuting in the United
States. In Transportation Reform Series. Washington, DC: Brookings Institution, Metropolitan
Policy Program.
Rojas-Rueda, David, Audrey de Nazelle, Marko Tainio, and Mark J. Nieuwenhuijsen. 2011. "The health
risks and benefits of cycling in urban environments compared with car use: health impact
assessment study." British Medical Journal no. 343. doi: 10.1136/bmj.d4521.
-------
120
Rose, Eliot, and Rex Burkholder. 2009. "CO(2) reduction through better urban design: Portland's story."
In Reducing Climate Impacts in the Transportation Sector, edited by Daniel Sperling and James S.
Cannon, 139-157. Dordrecht, Netherlands: Springer.
Rust, George, Jiali Ye, Peter Baltrus, Elvan Daniels, Bamidele Adesunloye, and George Edward Fryer.
2008. "Practical barriers to timely primary care access - Impact on adult use of emergency
department services." Archives of Internal Medicine no. 168 (15):1705-1710. doi:
10.1001/archinte.l68.15.1705.
Samaras, Constantine, and Kyle Meisterling. 2008. "Life cycle assessment of greenhouse gas emissions
from plug-in hybrid vehicles: Implications for policy." Environmental Science & Technology no.
42 (9):3170-3176. doi: 10.1021/es702178s.
Samimi, Amir, and Abolfazl Kouros Mohammadian. 2010. "Health impacts of urban development and
transportation systems." Journal of Urban Planning & Development no. 136 (3):208-213. doi:
10.1061/(asce)u p. 1943-5444.0000020.
Santos, A., N. McGuckin, H. Y. Nakamoto, D. Gray, and S. Liss. 2011. Summary of Travel Trends: 2009
National Household Travel Survey. U.S. Department of Transportation, Federal Highway
Administration.
Sastry, M. V. Rama. 1973. "Systems approach to cost-benefit analysis of urban transportation."
Transportation Journal (American Society of Transportation & Logistics Inc) no. 12 (3):39-45.
Scheiner, Joachim, and Christian Holz-Rau. 2013. "Changes in travel mode use after residential
relocation: A contribution to mobility biographies." Transportation no. 40 (2):431-458. doi:
10.1007/sllll6-012-9417-6.
Scholz, Miklas. 2013. "Water quality improvement performance of geotextiles within permeable
pavement systems: A critical review." Water no. 5 (2):462-479. doi: 10.3390/w5020462.
Schueler, T. R. 1994. "The importance of imperviousness." Watershed Protection Techniques no. 1
(3):100-111.
Schueler, Thomas R., Lisa Fraley-McNeal, and Karen Cappiella. 2009. "Is impervious cover still
important? Review of recent research." Journal of Hydrologic Engineering no. 14 (4):309-315.
doi: 10.1061/(asce) 1084-0699(2009) 14:4(309).
Schumaker, Nathan. 2013. Welcome to the HexSim Project [cited June 12 2013], Available from
http://www.hexsim.net/.
Schweitzer, Lisa, and Brian D. Taylor. 2008. "Just pricing: The distributional effects of congestion pricing
and sales taxes." Transportation no. 35 (6):797-812. doi: 10.1007/sllll6-008-9165-9.
Shadewald, Jerry K., Shauna Hallmark, and Reginald R. Souleyrette. 2001. "Visualizing system-wide
economic impacts of transportation projects." Journal of Urban Planning & Development no. 127
(4):158-168.
-------
121
Sinha, Kumares C. 2003. "Sustainability and urban public transportation." Journal of Transportation
Engineering no. 129 (4):331-341.
Small, Kenneth A., and Shunfeng Song. 1994. "Population and employment densities: Structure and
change." Journal of Urban Economics no. 36:292-313.
Starke, P., P. Gobel, and W. G. Coldewey. 2010. "Urban evaporation rates for water-permeable
pavements." Water Science and Technology no. 62 (5):1161-1169. doi: 10.2166/wst.2010.390.
Steinmetz, Scott A., Jason S. Herrington, Christopher K. Winterrowd, William L. Roberts, Jost O. L.
Wendt, and William P. Linak. 2013. "Crude glycerol combustion: Particulate, acrolein, and other
volatile organic emissions." Proceedings of the Combustion Institute no. 34:2749-2757. doi:
10.1016/j.proci.2012.07.050.
Stockton, T., B. Dyson, W. Houghteling, K. Black, M. Buchholtz ten Brink, T. Canfield, ... A. Rehr. 2011.
Decision Support Framework Implementation of DASEES: Decision Analysis for a Sustainable
Environment, Economy, and Society. Cincinnati, OH: U.S. Environmental Protection Agency.
Sudhira, H. S., T. V. Ramachandra, Andreas Wytzisk, and C. Jeganathan. 2005. Framework for Integration
of Agent-based and Cellular Automata Models for Dynamic Geospatial Simulations. Bangalore:
Indian Institute of Science, Centre for Ecological Sciences.
Sukkoo, Kim. 2007. "Changes in the nature of urban spatial structure in the United States, 1890-2000."
Journal of Regional Science no. 47 (2):273-287. doi: 10.1111/j.l467-9787.2007.00509.x.
Sun, Yu, Susanna T. Y. Tong, Mao Fang, and Y. Jeffrey Yang. 2013. "Exploring the effects of population
growth on future land use change in the Las Vegas Wash watershed: An integrated approach of
geospatial modeling and analytics." Environment, Development and Sustainability. doi:
10.1007/sl0668-013-9447-z.
Surface Transportation Policy Project. 2003. Transportation Costs and the American Dream: Why a Lack
of Transportation Choices Strains the Family Budget and Hinders Home Ownership. Washington,
DC: Surface Transportation Policy Project.
Sweet, Matthias N., and Mengke Chen. 2011. "Does regional travel time unreliability influence mode
choice?" Transportation no. 38 (4):625-642. doi: 10.1007/sllll6-011-9335-z.
Szeto, W. Y., Xiaoqing Jaber, and Margaret O'Mahony. 2010. "Time-dependent discrete network design
frameworks considering land use." Computer-Aided Civil & Infrastructure Engineering no. 25
(6):411-426. doi: 10.1111/j.l467-8667.2010.00654.x.
Teychenne, Megan, Kylie Ball, and Jo Salmon. 2008. "Physical activity and likelihood of depression in
adults: A review." Preventive Medicine no. 46 (5):397-411. doi: 10.1016/j.ypmed.2008.01.009.
Tomalty, Ray. 2009. "Urban tipping point." Alternatives Journal no. 35 (5):12-16.
Tong, S. T. Y., Y. Sun, and Y. J. Yang. 2012. "Generating a future land use change scenario with a modified
population-coupled Markov Cellular Automata model." Journal of Environmental Informatics no.
19 (2):108-119. doi: 10.3808/jei.201200213.
-------
122
Transportation Research Board of the National Academies, Committee for a Study of Potential Energy
Savings and Greenhouse Gas Reductions from Transportation. 2011. Policy Options for Reducing
Energy Use and Greenhouse Gas Emissions from U.S. Transportation. Washington, DC: National
Academy of Sciences.
Transportation Research Board of the National Academies, Second Strategic Highway Research Program.
2013. Incorporating Greenhouse Gas Emissions into the Collaborative Decision-Making Process.
Washington, DC: National Academy of Sciences.
Tsutsumi, Morito, and Hajime Seya. 2008. "Measuring the impact of large-scale transportation projects
on land price using spatial statistical models." Papers in Regional Science no. 87 (3):385-401. doi:
10.1111/j.l435-5957.2008.00192.x.
Tucker, Jared M., Gregory J. Welk, and Nicholas K. Beyler. 2011. "Physical Activity in U.S. Adults:
Compliance with the Physical Activity Guidelines for Americans." American Journal of Preventive
Medicine no. 40 (4):454-461. doi: 16/j.amepre.2010.12.016.
Tuomisto, Jouni T., and Marko Tainio. 2005. "An economic way of reducing health, environmental, and
other pressures of urban traffic: A decision analysis on trip aggregation." BMC Public Health no.
5:123. doi: 10.1186/1471-2458-5-123.
U.S. Department of Energy, Energy Efficiency and Renewable Energy. 2013. Alternative Fuels Data
Center 2013a [cited July 2013], Available from http://www.afdc.energy.gov/.
U.S. Department of Energy, Energy Efficiency and Renewable Energy. Clean Cities 2013b. Available from
http://wwwl.eere.energy.gov/cleancities/.
U.S. Department of Labor, Bureau of Labor Statistics. 2012. Consumer Expenditures in 2010: Lingering
Effects of the Great Recession.
U.S. Department of Transportation, Bureau of Transportation Statistics. 2003a. Commuting Expenses:
Disparity for the Working Poor.
U.S. Department of Transportation, Bureau of Transportation Statistics. 2003b. Transportation
Difficulties Keep Over Half a Million Disabled at Home.
U.S. Department of Transportation, Federal Highway Administration. 2013. Context Sensitive
Solutions.org [cited August 7 2013], Available from http://contextsensitivesolutions.org.
U.S. Department of Transportation, Federal Highway Administration. 2012. INVEST: Sustainable
Highways Self-Evaluation Tool [cited October 2 2012], Available from
https://www.sustainablehighways.org/.
U.S. Energy Information Administration, Office of Energy Statistics. 2013a. Monthly Energy Review June
2013. Washington, DC: U.S. Department of Energy.
U.S. Energy Information Administration, Office of Integrated and International Energy Analysis. 2013b.
Annual Energy Outlook 2013 with Projections to 2040. Washington, DC: U.S. Department of
Energy.
-------
123
U.S. Environmental Protection Agency. 2013. Green Communities [cited June 12 2013], Available from
http://www.epa.gov/greenkit/index.htm.
U.S. Environmental Protection Agency. 2007a. Measuring the Air Quality and Transportation Impacts of
Infill Development.
U.S. Environmental Protection Agency. 2009a. Water Quality Scorecard: Incorporating Green
Infrastructure Practices at the Municipal, Neighborhood, and Site Scales.
U.S. Environmental Protection Agency. 2011a. Guide to Sustainable Transportation Performance
Measures.
U.S. Environmental Protection Agency. 2013. Ecosystem Goods and Services Production Function Library,
May 31 2012a [cited June 14 2013], Available from
http://www.epa.gov/research/ecoscience/eco-goods-services-library.htm.
U.S. Environmental Protection Agency. 2013. National Ecosystems Goods and Services Classification
System (NEGSCS), May 31 2012b [cited June 14 2013], Available from
http://www.epa.gov/research/ecoscience/eco-negscs.htm.
U.S. Environmental Protection Agency. 2013. Community-Focused Exposure and Risk Screening Tool (C-
FERST), April 29 2013a [cited July 12 2013], Available from http://www.epa.gov/heasd/c-ferst/.
U.S. Environmental Protection Agency. 2013b. Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2011. Washington, DC.
U.S. Environmental Protection Agency. 2013c. "National Ambient Air Quality Standards for Particulate
Matter." Federal Register no. 78 (10):3085-3287.
U.S. Environmental Protection Agency. 2013. Tribal-Focused Environmental Risk and Sustainability Tool
(T-FERST), April 29 2013d [cited July 12 2013], Available from
http://www.epa.gov/heasd/research/tferst.html.
U.S. Environmental Protection Agency, and National Highway Traffic Safety Administration. 2012. "2017
and later model year light-duty vehicle greenhouse gas emissions and corporate average fuel
economy standards." Federal Register no. 77 (199):62623-63200.
U.S. Environmental Protection Agency, Office of Air and Radiation, Office of Mobile Sources,
Transportation Air Quality Center. 1998. Transportation Control Measures: Work Schedule
Changes.
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Air Quality
Assessment Division. 2010a. Air Quality Modeling Technical Support Document: Changes to the
Renewable Fuel Standard Program.
U.S. Environmental Protection Agency, Office of Policy Economics and Innovation, Development,
Community, and Environment Division. 2006. Growing Toward More Efficient Water Use:
Linking Development, Infrastructure, and Drinking Water Policies.
-------
124
U.S. Environmental Protection Agency, Office of Policy, National Center for Environmental Economics.
2010b. Guidelines for Preparing Economic Analyses.
U.S. Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment. 2007b. Concepts, Methods and Data Sources for Cumulative Health
Risk Assessment of Multiple Chemicals, Exposures and Effects: A Resource Document.
Cincinnati, OH.
U.S. Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment. 2011b. Biofuels and the Environment: First Triennial Report to
Congress. Washington, DC.
U.S. Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment, Global Change Research Program. 2009b. Land-Use Scenarios:
National-Scale Housing-Density Scenarios Consistent with Climate Change Storylines.
Washington, DC.
U.S. Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment, Global Change Research Program. 2010c. Integrated Climate and
Land-Use Scenarios (ICLUS) vl.3 User's Manual: ArcGIS Tools and Datasets for Modeling US
Housing Density Growth. Washington, DC.
U.S. Environmental Protection Agency, Office of Research and Development, National Risk Management
Research Laboratory. 2007c. Small Drinking Water Systems: State of the Industry and Treatment
Technologies to Meet the Safe Drinking Water Act Requirements.
U.S. Environmental Protection Agency, Office of Research and Development, National Risk Management
Research Laboratory, Sustainable Technology Division. 2012c. A Framework for Sustainability
Indicators at EPA.
U.S. Environmental Protection Agency, Office of Research and Development, Office of Water. 2002a.
Onsite Wastewater Treatment Systems Manual.
U.S. Environmental Protection Agency, Office of Sustainable Communities, Smart Growth Program.
2012d. Residential Construction Trends in America's Metropolitan Regions: 2012 Edition.
U.S. Environmental Protection Agency, Office of Transportation and Air Quality. 2010d. EPA Finalizes
Regulations for the National Renewable Fuel Standard Program for 2010 and Beyond.
U.S. Environmental Protection Agency, Office of Transportation and Air Quality. 2010e. EPA Lifecycle
Analysis of Greenhouse Gas Emissions from Renewable Fuels.
U.S. Environmental Protection Agency, Office of Transportation and Air Quality. 2012e. EPA and NHTSA
Set Standards to Reduce Greenhouse Gases and Improve Fuel Economy for Model Years 2017-
2025 Cars and Light Trucks.
U.S. Environmental Protection Agency, Office of Transportation and Air Quality. 2013. EPAct/V2/E-89
Tier 2 Gasoline Fuel Effects Study, May 30 2013e [cited September 25 2013], Available from
http://www.epa.gov/otaq/models/moves/epact.htm.
-------
125
U.S. Environmental Protection Agency, Office of Transportation and Air Quality, Assessment and
Standards Division. 2002b. A Comprehensive Analysis of Biodiesel Impacts on Exhaust Emissions.
U.S. Environmental Protection Agency, Risk Assessment Forum. 2003. Framework for Cumulative Risk
Assessment. Washington, DC.
U.S. Green Building Council. 2012. LEED [cited November 26 2012], Available from
https://new.usgbc.org/leed.
U.S. Green Building Council, Congress for the New Urbanism, and Natural Resources Defense Council.
2012. LEED 2009 for Neighborhood Development, October 2012 [cited October 5 2012],
Available from https://new.usgbc.org/sites/default/files/LEED%202009%20Rating_ND_10-
2012_9c.pdf.
Venkatram, Akula, Vlad Isakov, Eben Thoma, and Richard Baldauf. 2007. "Analysis of air quality data
near roadways using a dispersion model." Atmospheric Environment no. 41 (40):9481-9497. doi:
10.1016/j.atmosenv.2007.08.045.
Waldorf, Brigitte. 2003. "Automobile reliance among the elderly: Race and spatial context effects."
Growth & Change no. 34 (2):175-201. doi: 10.1111/1468-2257.00213.
Wallbaum, Holger, Sabrina Krank, and Rolf Teloh. 2011. "Prioritizing sustainability criteria in urban
planning processes: Methodology application." Journal of Urban Planning & Development no.
137 (l):20-28. doi: 10.1061/(asce)up.l943-5444.0000038.
Wang, Guangmei, Gaoming Jiang, Yunlong Zhou, Quanru Liu, Yanshou Ji, Shixiong Wang, . . . Hui Liu.
2007. "Biodiversity conservation in a fast-growing metropolitan area in China: A case study of
plant diversity in Beijing." Biodiversity and Conservation no. 16 (14):4025-4038. doi:
10.1007/sl0531-007-9205-3.
Wang, Xinhao, Amy Burgess, and Jeff Yang. 2013. "A scenario-based water conservation planning
support system (SB-WCPSS)." Stochastic Environmental Research and Risk Assessment no. 27
(3):629-641. doi: 10.1007/s00477-012-0628-3.
Wanner, Miriam, Thomas Gotschi, Eva Martin-Diener, Sonja Kahlmeier, and Brian W. Martin. 2012.
"Active Transport, Physical Activity, and Body Weight in Adults: A Systematic Review." American
Journal of Preventive Medicine no. 42 (5):493-502. doi: 10.1016/j.amepre.2012.01.030.
Warburton, Darren E. R., Sarah Charlesworth, Adam Ivey, Lindsay Nettlefold, and Shannon S. D. Bredin.
2010. "A systematic review of the evidence for Canada's Physical Activity Guidelines for Adults."
International Journal of Behavioral Nutrition and Physical Activity no. 7. doi: 10.1186/1479-
5868-7-39.
Wilkinson, Richard, and Michael Marmot, eds. 2003. Social determinants of health: the solid facts. 2nd
ed. Copenhagen: World Health Organization.
Wong, Bonny Yee-Man, Guy Faulkner, and Ron Buliung. 2011. "GIS measured environmental correlates
of active school transport: A systematic review of 14 studies." International Journal of
Behavioral Nutrition and Physical Activity no. 8. doi: 10.1186/1479-5868-8-39.
-------
126
Woodcock, James, Moshe Givoni, and Andrei Scott Morgan. 2013. "Health impact modelling of active
travel visions for England and Wales using an Integrated Transport and Health Impact Modelling
tool (ITHIM)." PLoS ONE no. 8 (1). doi: 10.1371/journal.pone.0051462.
World Health Organization. 2013. Health economic assessment tool, 2013/04/01/21:48:23 [cited April 1
2013], Available from http://www.heatwalkingcvcling.org/.
Yang, Y. J., J. Neal, S. Liang, Z. Yao, T. C. Keener, H. Wei, and X. Wang. 2011. On the black carbon related
carbon footprints in urban transportation. Paper read at 104th Air and Waste Management
Association Annual Conference and Exhibition, June 21-24, at Orlando, FL.
Yang, Y. Jeffrey, Ni-Bin Chang, Zhiwei Li, Steven G. Buchberger, Susanna Tong, Xinhao Wang, .. . James
A. Goodrich. 2013. A systems approach to manage drinking water quality through integrated
model projections, adaptive monitoring and process optimization. In World Environmental &
Water Resources Congress 2013. Cincinnati, OH: American Society of Civil Engineers.
Yang, Y. Jeffrey, Ni-Bin Chang, Jill Neal, Heng Wei, Marissa S. Liang, and Timothy C. Keener. 2013. Water
and carbon footprints for sustainability analysis of urban infrastructure. In World Environmental
& Water Resources Congress 2013. Cincinnati, OH: American Society of Civil Engineers.
Zhou, Jian, Jianyi Lin, Shenghui Cui, Quanyi Qiu, and Qianjun Zhao. 2013. "Exploring the relationship
between urban transportation energy consumption and transition of settlement morphology: A
case study on Xiamen Island, China." Habitat International no. 37:70-79. doi:
http://dx.doi.Org/10.1016/i.habitatint.2011.12.008.
Zhou, Ying, and Jonathan I. Levy. 2007. "Factors influencing the spatial extent of mobile source air
pollution impacts: A meta-analysis." BMC Public Health no. 7. doi: 10.1186/1471-2458-7-89.
Zorn, Lisa, Elizabeth Sail, and Daniel Wu. 2012. "Incorporating crowding into the San Francisco activity-
based travel model." Transportation no. 39 (4):755-771. doi: 10.1007/sllll6-012-9405-x.
Zuurbier, Moniek, Gerard Hoek, Marieke Oldenwening, Virissa Lenters, Kees Meliefste, Peter van den
Hazel, and Bert Brunekreef. 2010. "Commuters' Exposure to Particulate Matter Air Pollution Is
Affected by Mode of Transport, Fuel Type, and Route." Environmental Health Perspectives no.
118 (6):783-789. doi: 10.1289/ehp.0901622.
-------
SEPA
United States
Environmental Protection
Agency
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Printed on 100% recycled/recyclable paper
with a minimum 50% post-consumer
fiber using vegetable-based ink.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
-------