University of Southern California Dissertations and Theses

Investigating bus route walkability: comparative case study in Orange County, California

(USC Thesis Other)

Investigating bus route walkability: comparative case study in Orange County, California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INVESTIGATING BUS ROUTE WALKABILITY: COMPARATIVE CASE STUDY
IN ORANGE COUNTY, CALIFORNIA

by

Stephanie Chen

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)

December 2012

Copyright 2012 Stephanie Chen
ii

Acknowledgements
I would like to express my gratitude to my committee chair, Dr. Robert Vos, for
his excellent guidance and support throughout the process of this research. I would also
like to thank my committee members, Dr. Katsuhiko Oda and Dr. Darren Ruddell,
without whose knowledge and advice this study would not have been successful. Special
thanks go to Dr. Su Jin Lee, who assisted me when I encountered problems during the
research.
I would also like to thank my family members and friends. They were always
supporting me and encouraging me to pursue this degree.

iii

Table of Contents
Acknowledgements ii

List of Tables v

List of Figures vi

Abstract viii

Chapter 1: Introduction 1
1.1 Motivation 3

Chapter 2: Background 11
2.1 Orange County Development 11
2.2 Previous Approaches to Bus Network Design Problems 13
2.3 Summary of Studies on Walkability 18

Chapter 3: Methodology 26
3.1 Three Types of Buffers 30
3.2 Variables, Data Sources and Calculation 32
3.3 Combined Score 37
3.4 Sensitivity Analysis 37

Chapter 4: Results 38
4.1 Scores for Half-Mile-Radii Buffers 38
4.2 Field Observations 42
4.3 Combined Score Comparison for the Tree Types of Buffers
(HMR, RA, and SARA) 45
4.4 Combined Score Comparison for Route 47 and Route 89 49
4.5 Sensitivity Analysis Results 51

iv

Chapter 5: Discussion and Conclusion 57
5.1 Limitations 59
5.2 Future Research 63

Bibliography 66

Appendix 71

v

List of Tables
Table 1: Measures for Determining Walkability Considered by the
Authors 20

Table 2: Route 47 and Route 89 Comparison Table 29

Table 3: Walkability Variables and Their Data Sources, Temporal
Sales, and Spatial Scales 33

Table 4: Tree Coverage “Value” and their Adjusted Values for
Calculating Tree Coverage Value for Buffers 36

Table 5: Difference of Means Test Results for the Tree Types of
Buffers (HMR, RA, and SARA) 47

Table 6: Descriptive Statistics of Route 47 and Route 89 for the Three
Types of Buffers (HMR, RA, and SARA) 50

Table 7: Difference of Means Test Results for Route 47 and Route 89
for the Three Types of Buffers (HMR, RA, and SARA) 51

Table 8: Comparison Table for Half-Mile-Radii Buffer Sensitivity
Analysis Results 56

vi

List of Figures
Figure 1: Boundary of Orange County located in Southern California 6

Figure 2: Grid vs. Cul-De-Sac Neighborhood Examples from
Orange County, California 27

Figure 3: Southbound Bus Stop Locations of Route 47 and Route 89 28

Figure 4: Route 89 Stop 6 and its Half-Mile-Radii Buffer 39

Figure 5: Half-Mile-Radii Buffers Results for Route 47 40

Figure 6: Half-Mile-Radii Buffers Results for Route 89 41

Figure 7: Photos of Route 47 Stops 43

Figure 8: Photos of Route 89 Stops 44

Figure 9: Histograms of Combined Bus Stop Scores for the Three
Types of Buffer (HMR, RA, and SARA) 47

Figure 10: Bus Stops with Adjusted and Non-Adjusted Buffers 48

Figure 11: Original and Alternative Locations for Route 47 Stop 10 53

Figure 12: Original and Alternative Locations for Route 89 Stop 38 54
vii

Figure 13: Original and Alternative Locations for Route 47 Stop 3 55

Figure 14: Manually Digitized Tree Coverage for Route 47 Stop 28’s
Half-Mile-Radii Buffer 62

Figure 15: Route 47 Stop 20 and its Three Types of Buffers
(HMR, RA, and SARA) 65

viii

Abstract
To improve bus route planning and understand walkability’s role in bus network design,
this study offers a method of evaluating the walkability of bus stops and provides a case
study for stops along two bus routes in Orange County, California. Having better
walkability for bus routes may both promote physical activity and encourage bus
ridership. Previous studies on bus route planning focus mostly on the passengers’ travels
on the bus and minimal attention is given to the bus riders’ experiences before reaching
the bus, after departing the bus, and during transfers between bus lines. This study shows
the relevance of considering the origin, destination, and walking paths for pedestrians
when approaching bus network design problems. The walkability of the southbound bus
stops along Route 47 and Route 89, operated by the Orange County Transportation
Authority (OCTA), were evaluated by calculating and combining the scores of four
variables within each bus stop buffer. The four variables evaluated were: population
density, street connectivity, steepness, and tree canopy. Results show that Route 47 has
higher overall walkability than Route 89, which is in accordance with the hypothesis that
a route that runs through grid neighborhoods (Route 47) would be more walkable than a
route that runs through cul-de-sac neighborhoods (Route 89). Sensitivity analyses
demonstrated that walkability scores may change when a stop is repositioned to a
hypothetical location further away from an arterial street and within a neighborhood.

ix

Although walkability will never be the sole factor in designing bus routes, future
modeling could weigh the importance of walkability as part of origin and destination
modeling and use the scoring of walkability to guide adoption of the “flexible-route” bus
lines. Future research should consider other methods of determining tree canopy scores
and explore other methods of identifying pedestrian “catchment” area of the bus stops.
1

Chapter 1: Introduction
Generally, bus network design problems involve optimization to improve system
performance requirements under given resource constraints (Fan and Machemehl 2006).
Working within various constraints and specific optimization goals, most studies
approach the bus network design problem only by considering the bus riders’ travels on
the bus. Such studies disregard the bus riders’ travels before and after the bus ride, which
are affected by factors such as length of waiting time, time spent walking, and bus route
walkability. This study investigates one such factor, bus route walkability, and discusses
its influence on the bus network design problems.
Walkability was first given attention in the post-modernist planning era starting in
the 1970s, due to its emphasis on human-scale, urban, and unique forms (Hirt 2005).
Post-modernist planners considered walkability as a crucial component of efficient,
accessible, equitable, sustainable, and livable communities (Lo 2009). Although many
studies tried to define “walkability”, they struggled to directly explain, define, and
measure the concept (Abley 2005). In addition, different technical disciplines, such as
engineering, planning, and health, define walkability differently to be in alignment with
their use and context (Abley 2005).

2

Even though there have been gaps and disagreements in defining walkability by the
different professional disciplines, this study describes walkability as a measure of how
easy it is to walk in a built environment (Abley 2005). A highly walkable environment
provides residents with accessibility to the transport network in addition to encouraging
community involvement and promoting healthy lifestyles.
As an initial exploration of the role of walkability in bus route planning, this
project is a case study that examines two bus routes in Orange County, California, with a
focus on residential neighborhoods as opposed to commercial and industrial areas. The
study develops a method to evaluate the walkability of bus routes using Geographic
Information System (GIS). As a case study on bus route walkability, two bus routes are
chosen for comparison: 1) a bus route that runs through grid neighborhoods, and 2) a bus
route that runs through cul-de-sac neighborhoods. The walkability of the bus routes is
determined using four variables: population density, street connectivity, steepness, and
tree canopy. In addition, several case studies on optimizing walkability by relocating bus
stops are performed.
Bus routes are currently designed to optimize cost and time of travel based on
ridership levels and traffic studies, among other factors (Chein, Dimitrijevic and Spasovic
2003). Redesigning the bus routes solely to optimize walkability to stops for the bus
riders would not likely become the ideal solution in terms of the overall efficiency of the
bus system; however, it is still a factor worthy of consideration (Ceder and Wilson 1986).
Since bus routes are designed specifically for the local and immediate population,
3

walkability of stops within or between bus routes varies depending upon a number of
considerations; some examples might include local mode of transportation preferences,
social and cultural norms, and specific geographic features, all of which make traveling
by buses easier in some areas than others. But what is less clear is whether such variation
is significant enough within existing routing parameters to make it effective and efficient
to consider walkability when locating stops.
The objective of this study is to investigate how GIS and concepts of walkability
contribute to bus network design using existing geospatial datasets in bus route planning.
In addition, by performing a case study evaluating the walkability of two bus routes in
Orange County, the project examines whether different levels of walkability can be seen
at bus stopS for routes in areas that have undergone different historical patterns of urban
development. Furthermore, this study also tries to understand whether walkability
improvements can be measured and reported when stops are moved short distances from
existing routes.
1.1 Motivation
California has long been known for its dense population growth and strained
transportation infrastructure. Since its statehood in 1850, California has grown to
approximately 37 million residents (United States Census Bureau 2012). The
development of the California interstate highway system since the 1910s and the
California State Routes since the 1930s has made Southern California, which was
previously mostly rural, accessible to anyone owning an automobile and thereby
4

invigorated the region’s sprawling growth (Wolch, Pastor and Dreier 2004). In addition,
from the 1950s until the 1990s, federal transportation funding explicitly favored the
continuous expansion of federal and state highway system (Lee and Rivasplata 2001).
The population growth and transportation development decisions in California resulted in
an urban sprawl that encourages dependency on the private vehicle for transportation
(Newman 1996).
As a case study focusing on bus route walkability in Orange County (Figure 1), it
is pertinent to investigate and understand the growth and development of Orange County
as a sprawling suburban community. There is no defined urban center in Orange County,
which vastly differs from most population centers that identify and surround a major city,
such as Greater Los Angeles or the San Francisco Bay Area (Halper and McKibben
2002). In the late 1800s and early 1900s, as an extension from the Greater Los Angeles
urban center, Orange County was mostly comprised of agricultural land, both livestock
and produce (Orange County Historical Society 2012). Aside from some traditionally
urban areas at the center of the older cities, including Anaheim, Santa Ana, and Fullerton,
Orange County remained mostly undeveloped in the early 1900s (Orange County
Historical Society 2012).
With the introduction of citrus fruits and its subsequent economic boom, which
continued until the 1950s and 60s, the region underwent substantial growth and older
cities sought to keep up with the population increase by expanding these traditional urban
areas (Orange County Historical Society 2012). These older city centers are based upon
5

the gridiron plan that originated from traditional urban planning, in which city streets are
laid out in an inter-connected grid running at 90-degree angles (Cozens and Hillier 2008).
In addition, with the federal transportation funding that directly resulted in the expansive
network of highway systems throughout Southern California starting in the 1910s, travel
by way of automobiles became easier than before (Lee and Rivasplata 2001). With the
increased use and popularity of cars, traffic engineers and urban planners sought to
respond to consumer trends and abandoned the grid neighborhoods in favor of cul-de-sac
neighborhoods when designing residential areas (Cozens and Hillier 2008). Cul-de-sacs
are essentially the end of the road, resulting in a dead-end street with only one inlet or
outlet; they are meant to reduce the amount of car traffic and crime rates (Cozens and
Hillier 2008). In regards to the suburban development of Orange County, northern cities
were established earlier with grid neighborhoods while southern cities in the county favor
cul-de-sac neighborhoods.

6

Figure 1: Boundary of Orange County located in Southern California

Sources: Esri Bing Maps Road (2012), Esri World Light Gray Base Layer (2012), Esri
USA Counties Layer (2012), Esri USA States Layer (2012)
7

In a discussion of the modern suburban community of Orange County, there lies a
high reliance on automobiles which ties into severe traffic congestions, high rates of
motor vehicle crashes, and urban air pollution of the overall region (Frumkin 2002). In
addition, personal transportation by means of the automobile indicates a lack of physical
activity as a means of transit, such as walking or bicycling (Frumkin 2002). This overall
lack of physical activity among people increases the chances of obesity, diabetes,
cardiovascular disease, and stroke, all of which lead to early mortality (Frumkin 2002).
Therefore, in order to improve quality of life and to accommodate California’s future
population growth, authorities have been searching for different approaches to
sustainable urban planning, design, and construction that would reduce air pollution,
minimize automobile-related injuries and deaths, encourage physical activity, and
promote mental health and a sense of community (The Strategic Growth Council 2012).
The importance of walkability of the built environment to physical health has
been a leading topic or research and advocacy (Handy et al. 2002). In an effort to
promote a healthy lifestyle, the Robert Wood Johnson Foundation created a national
program called Active Living Research that aims, “to support and share research on
environment and policy strategies that can promote daily physical activity for children
and families in the United States” (Active Living Research 2012).

8

Active Living Research emphasizes increasing physical activity to prevent obesity and
promote health in people (Active Living Research 2012). One of the ways for people to
get regular physical activity from walking and bicycling is by investing in sidewalks,
traffic-calming devices, greenways, trails, and public transit, which make it easier for
people to walk and bike to places they need to go (Active Living Research 2012).
In conjunction with promoting a healthy lifestyle, smart growth is a set of
planning practices and development principles that focuses on accessibility and aims for
more efficient land use and transport patterns (Litman 2012). Smart growth consists of
ten principles and two of these principles as articulated by the U.S. Environmental
Protection Agency (EPA) are to create walkable neighborhoods and provide a variety of
transportation choices (EPA 2012). Furthermore, Dan Burden, the Executive Director of
the Walkable and Livable Communities Institute, comments on walkability as follows:
Walkability is the cornerstone and key to an urban area’s efficient ground
transportation. Every trip begins and ends with walking. Walking remains
the cheapest form of transport for all people, and the construction of a
walkable community provides the most affordable transportation system
any community can plan, design construct and maintain. Walkable
communities put urban environments back on a scale for sustainability of
resources (both natural and economic) and lead to more social interaction,
physical fitness and diminished crime and other social problems.
Walkable communities are more livable communities and lead to whole,
happy, healthy lives for the people who live in them. (Mantri 2008)
In other words, by providing easily accessible transportation means, such as buses and
trolleys, and promoting walking, people’s reliance on automobiles could be decreased,
improving people’s health and the overall quality of life.
9

Seeing that walkability links the relationship between the average person, their
choice of transportation, and the surrounding built environment, it is important to employ
modern mapping techniques to visualize and conceptualize how bus route planning can
be improved. This particular case study utilizes the mapping technologies of GIS with
existing geospatial datasets. GIS became more popular in later 1980s and 1990s due to
the growth of GIS use on personal computers and the Internet as people recognize GIS’s
ability to visualize spatial information in accurate and flexible ways (Johnson 1993). It is
an integrated collection of computer software and data used to view and manage
geographic information, analyze spatial relationships, and model spatial processes (Esri
2012). GIS emerged as an excellent field to address views on urban planning; however,
being such a new technology, it is currently underutilized. This study therefore serves as
an example of using GIS to investigate the importance of walkability in bus route
planning.
In conclusion, most research literature on walkability has focused on investigating
the general walkability of a study area. There are few studies that examine walkability of
the walking paths of actual trips taken by pedestrians, which is the focus of this research.
In addition, by evaluating the walkability to bus stops along the routes in Orange County,
it can be determined whether the bus route locations, as planned by Orange County
Transportation Authority (OCTA), are in favor of facilitating walking for bus riders. The
result of the research can be a new source of valuable information for OCTA to re-
examine their bus routes designs.
10

Walkability may in fact be added as one new factor among many used in route models
that plan bus routes and placement of stops. Also, information on walkability can be used
to consider abandoning fixed route stops in certain areas in favor of “flexible-route transit
system.” This is a new bus transit concept introduced by Ouyang and Nourbakhsh (2012)
that is particularly suitable for low-demand or less walkable areas.

11

Chapter 2: Background
To recognize how walkability fits into the big picture of bus route planning, it is
important to understand the history of the development of Orange County in addition to
the background research on bus route planning and walkability that has been done prior
to this particular study. Section 2.1 analyzes and discusses how the development of
Orange County since the 1900s until now has impacted urban planning in the region.
Section 2.2 summarizes previous works in solving the bus network design problems.
Lastly, Section 2.3 presents the different approaches and variables researchers use to
evaluate walkability.
2.1 Orange County Development
Since its political division from Los Angeles County as a separate entity in 1889, Orange
County’s population settlement remained a relatively organic process (Orange County
Historical Society 2012). Northern regions of what is now Orange County tend to be
older city centers that feature buildings constructed in the early 1900s, such as the Old
Orange County Courthouse, built in 1901, and the Santa Ana Old City Hall constructed in
1935, both located in Downtown Santa Ana (Orange County Historical Society 2012).

12

With the opening of the Sana Ana Freeway in 1953, now the Interstate 5, an increased
ease of travel by automobile from Los Angeles to the region invited significant
residential development in Orange County (Orange County Historical Society 2012).
Looking at urban planning, older city centers in Northern Orange County expanded and
developed the pre-existing grid street pattern from hich these old city centers
accommodated the quickly growing population (Cozens and Hillier 2008).
On the other hand, the land that occupies the South Orange County region was
comprised of ranch lands that remained mostly agricultural until the introduction of
master planned communities in the 1960s (Orange County Historical Society 2012). One
such successful example is the development of the City of Irvine by the Irvine Company,
which was designed by architect and urban planner, William Pereira (Irvine Company
2012). Having a master plan, communities are carefully designed and thoughtfully
managed to minimize land use conflicts as well as optimize a variety of housing types,
job centers, shopping centers, recreation centers, and open space to promote quality of
life and economic growth (Irvine Company 2012). Specifically addressing residential
developments, master planned communities integrate cul-de-sacs into their street designs
because these types of streets appeal to consumers as ideal housing locations since they
reduce vehicle traffic as well as lower noise, localized air pollution and the probability of
accidents (Cozens and Hillier 2008).

13

In a general sense, modern Orange County urban planning maps exhibit grid
street patterns in older cities to the Northern side, while the newer cities in South Orange
County feature winding and twisting roads that incorporate cul-de-sacs into residential
neighborhoods. The historical and socio-economical influences of each region’s
development partially contribute to the income gap observed between Northern and
Southern Orange County. Furthermore, bus route planning is highly dependent on the
specific geographic features and street network designs found in the region of interest for
the proposed bus route, which is why it is pertinent to discuss the development of Orange
County and how the Northern and Southern side differ. As a case study of bus routes in
Orange County, this project examines two specific bus routes in which one is identified
with the grid street pattern in the North side while the other bus routes run through cul-
de-sac neighborhoods in the South side.
2.2 Previous Approaches to Bus Network Design Problems
Early work on bus network design problems can be generalized into two main groups: 1)
optimization model approaches predicated on idealization of the network, and 2) heuristic
approaches dealing with actual routes for more practical problems (Ceder and Wilson
1986). For the first group, bus network designs were formulated as analytical
optimization models that are applied to determine route design parameters on simplified
or regular shaped networks; parameters may include stop spacing, route spacing, route
length, bus size, and/or frequency of service (Fan and Machemehl 2006). Examples of
this type of bus route optimization model can be seen in the works of Newell (1979) and
14

Chang and Schonfeld (1991, 1993). These studies were based on the assumption of fixed
demand, limited design parameters, and the objective of minimizing the sum of passenger
and operator costs (Ceder and Wilson 1986). It has been shown that the analytical
methods are effective in solving optimization-related problems for small networks with
one or two decision variables but do not work well for bus network design problems with
realistic sizes that have many parameters to be determined (Fan and Machemehl 2006).
As a result, the optimization model approaches are more useful for screening or policy
analyses where approximate designs are adequate and are not recommended for tasks that
require route designs in real situations (Ceder and Wilson 1986).
For bus network design problems with larger network size and higher complexity,
the heuristic approach is preferred (Tom and Mohan 2003). This approach adopts the
rules by which the route network is built in a step-by-step procedure; therefore, it differs
from case to case (Tom and Mohan 2003). Furthermore, it primarily deals with
simultaneous design of the bus network and determination of its bus frequencies (Tom
and Mohan 2003). Examples of heuristic methods are found in Lampkin and Saalmans
(1967), Rea (1971), and Ceder and Wilson (1986). However, heuristic methods are tailor-
made for different applications and therefore lack adaptability to other contexts, unlike
the optimization model mentioned earlier (Tom and Mohan 2003).

15

Other studies that discuss solving bus network design problems include hybrid
models, experience-based models, simulation models, and genetic algorithms (Tom and
Mohan 2003). Hybrid models combine heuristic methods with accompanying methods
(e.g., analytical optimization methods and linear programming) (Tom and Mohan 2003).
Experience-based models are developed by capturing the experienced planners’
knowledge that has been acquired over a number of years in the form of rules that can be
used in the design (Tom and Mohan 2003). Simulation models are capable of
incorporating numerous variables that affect bus transit operation thereby demonstrating
different aspects of the bus systems (Tom and Mohan 2003).
All of the above methods are either only suitable for theoretical situations or are
case-specific (Tom and Mohan 2003). Therefore, genetic algorithm, which is a general
multipurpose optimization model for designing the bus transit network, emerged as an
alternative to many conventional approaches (Tom and Mohan 2003).
Studies have used the genetic algorithm to approach bus network design problems.
Tom and Mohan (2003) used genetic algorithms to select a solution route set and the
associated frequencies to achieve the desired objective, subject to the operational
constraint, which is the total system costs expressed as a function of bus operating cost
and the cost of passenger total travel time.

16

In another application, Fan and Machemehl (2006) used a genetic algorithm to
systematically examine the underlying characteristics of the optimal bus transit route
network design given variable transit demand. It is different from the work of Tom and
Mohan (2003) because it employs hybrid transit trip demand assignment models in the
genetic algorithm.
The study done by Fan and Machemehl (2006) is one of the few studies that take
the bus riders’ walking into consideration for optimizing the bus network design problem.
The objective of this project was to use a genetic algorithm to examine and understand
the underlying characteristics of the optimal bus network design problem through the
development of a multiobjective nonlinear mixed integer model (Fan and Machemehl
2006). To achieve the objective, the solution consists of three components: a route set
generation procedure that generates all feasible routes, a network analysis procedure used
to decide transit demand matrix, and a genetic algorithm procedure that combines these
two parts to guide the candidate solution generation process and select an optimal set of
routes (Fan and Machemehl 2006). The model’s objective function minimizes the sum of
user cost (walking cost, waiting time, transfer cost, and in-vehicle cost), operator cost
(cost to operate the required buses), and the unsatisfied demand cost for the bus network
(Fan and Machemehl 2006).

17

These three costs are determined based on planners’ experience and expert
judgment (Fan and Machemehl 2006). A such, it is important to note that these cost
estimates come from institutional knowledge that are held by a group of people at the
transit agency, which are difficult to be passed on and be considered for other studies.
As described previously, although Fan and Machemehl (2006) include the
pedestrians’ walking cost in its model, the walking cost is dependent on institutional
knowledge instead of being derived from empirical (spatial) variables (Fan and
Machemehl 2006). Therefore, to go beyond incorporating institutional knowledge, which
can be subjective and inaccurate, this study empirically investigates the walking aspects
of bus riding by analyzing spatial data through the use of GIS, taking advantage of its
ability to present the results systematically and objectively.
To summarize, walkability has not played a major or regular role in bus network
design problems in previous academic works and studies. From the literature on bus
network design problems, it can be observed that researchers have been mostly focused
on the passengers’ travels on the bus, and minimal attention is given to the bus riders’
travels before, after, and in between bus rides. However, it is apparent that people’s travel
behavior is greatly influenced by the built environment (Agrawal, Schlossberg and Irvin
2008).

18

There are various elements to the built environment that affect a person’s decision in
choosing to walk, bike, or drive to a desired destination. Thus, by investigating the
walkability of bus routes through the use of GIS, this study suggests a new variable that
can be optimized for the bus network design problems and also provides a new method to
evaluate redesigning bus routes.
2.3 Summary of Studies on Walkability
It is important to understand “walkability” and its many contributing factors when
developing a method for evaluating the walkability of the bus routes. There have been
numerous studies on the subject in the past, and eight different works are reviewed and
summarized in Table 1 to form the basis for identifying the measures of walkability used
in this study. These nine variables are population density, dwelling density, retail floor
ratio, street connectivity, safety, land use mix, access to facilities, steepness, and tree
canopy. Out of these nine variables, this particular study focuses on population density,
street connectivity, steepness, and tree canopy. Although not all of variables are
considered in this study, they are all pertinent to the understanding of walkability and
how it can be measured.

19

The first variable in Table 1 is population density; it is defined as the
measurement of population per unit area in which higher population density means there
are more people per unit area. This variable is a common measure in studies of the built
environment and transportation-based physical activity in which higher population
density correlates with higher walkability (Brownson et al. 2009). This variable was
considered in the studies of Smith et al. (2008) and Marshall, Brauer, and Frank (2009).
The main goal of the study of Smith et al. (2008) was to relate neighborhood walkability
to residents’ obesity. The authors emphasized that greater population density has been
associated with fewer weight problems and that it may encourage walking destination
development and discourage exclusive reliance on cars (Smith et al. 2008). Additionally,
the paper written by Marshall, Brauer, and Frank (2009) considers population density as
one of the four parameters that they used to estimate walkability in the study area,
implying that higher population density results in higher walkability.
The second variable in Table 1 is dwelling density; it is the number of residential
units per unit area (Frank et al. 2010). Dwelling density is similar to population density in
which higher densities indicate more people living in the area; therefore, greater dwelling
density also translates to higher walkability (Frank et al. 2010). This variable was
considered in the studies of Frank et al. (2010), Mantri (2008), and Marshall, Brauer, and
Frank (2009). In Frank et al. (2010)’s report of the Metro Vancouver Walkability Index
developed at the University of British Columbia, which measures neighborhood urban
form characteristics in Metro Vancouver, one of the five variables chosen was dwelling
density. Also, Mantri (2008) uses a GIS based approach to measure walkability of a
20

Central West End, a neighborhood in St. Louis, Missouri in which dwelling density was
also selected as one of the many measures used in formulating the GIS model (Mantri
2008). Marshall, Brauer, and Frank (2009) also consider dwelling density, in addition to
population density described previously, as one of the four parameters that they used in
their study on walkability.
Table 1: Measures for Determining Walkability Considered by the Authors

Population
Density
Dwelling
Density
Retail
Floor
Area
Ratio
Street
Connectivity
Safety
Land
Use
Mix
Access to
Facilities
Steep-
ness
Tree
Canopy
(Lo 2009) ✔ ✔ ✔ ✔
(Jaskiewicz
2000)
✔ ✔ ✔
(Smith, et al.
2008)
✔ ✔ ✔
(Reynolds, et
al. 2007)
✔ ✔ ✔
(Frank, et al.
2010)
✔ ✔ ✔ ✔
(Marshall,
Brauer and
Frank 2009)
✔ ✔ ✔
(Mantri 2008) ✔ ✔ ✔ ✔ ✔
(Krambeck
2006)
✔ ✔ ✔

21

The third variable in Table 1 is retail floor ratio, which is defined as the
proportion of the area designated for commercial use in the area of interest and a higher
ratio reflects better walkability (Frank et al. 2010). This variable indicates the proximity
between commercial destinations thereby showing the degree to which people can more
easily travel between places in the region (Frank et al. 2010). Retail floor ratio was also
considered in the studies of Frank et al. (2010) and Marshall, Brauer, and Frank (2009).
To determining how walkable the study area is, Frank et al. (2010) calculate the retail
floor ratio of Metro Vancouver as one of the variables in understanding the physical
environment characteristics. Marshall, Brauer, and Frank (2009) also included retail floor
ratio as one of the four parameters that they used for evaluating the built environment.
The fourth variable in Table 1 is street connectivity. This variable measures
walkability in a given area by showing the directness of pedestrian routes; higher street
connectivity correlates with high walkability (Brownson et al. 2009). It is often
determined by the number of intersections per area, the percentage of 4-way intersections,
or the number of intersections per length of street network (Brownson et al. 2009). This
variable was considered in the studies of Lo (2009), Jaskiewicz (2000), Krambeck (2006),
Smith et al. (2008), Frank et al. (2010), and Mantri (2008). Lo (2009) examined several
walkability indices from different sources, such as the Pedestrian Potential Index, the
Pedestrian Deficiency Index, and Kansas City pedestrian level of service matrices, in
which street connectivity was a key determining factor. Jaskiewicz (2000) outlined a
process where qualitative factors can be used to analyze pedestrian systems; the author
emphasized that a complex path network guarantees a high degree of connectivity
22

between activity centers and residential units that encourages pedestrians to walk as
compared to places with poor path network in which people are bound to the same route
all the time (Jaskiewicz 2000). Krambeck (2006) created the Global Walkability Index
that ranks cities across the world and street connectivity is one of the variables used in the
index (Krambeck 2006). Street connectivity was also deemed important in the study by
Smith et al. (2008). The authors determined that pedestrian-friendly street connectivity is
associated with fewer weight problems because people are more willing to walk. Frank et
al. (2010) suggested that greater degrees of street connectivity enables more direct travel
between places, which is why this measure is used in determining the walkability in
Metro Vancouver. Lastly, Mantri (2008) also incorporated street connectivity in the GIS
model built to evaluate walkability in a neighborhood in St. Louis, Missouri.
The fifth variable in Table 1 is the safety measure used to evaluate the level of
safety of the walking path, which can be determined with a variety of factors, such as:
crossing safety, traffic speed, traffic volume, road width, street lighting, sidewalk
conditions, freedom from crime, and the walking path’s separation from traffic
(Krambeck 2006). This measure was included in the studies of Reynolds et al. (2007), Lo
(2009), Jaskiewicz (2000), Mantri (2008), and Krambeck (2006). Reynolds et al.’s (2007)
study identified the environmental correlates of urban trail use by evaluating urban trails
in Chicago, Dallas, and Los Angeles using an instrument called Systematic Pedestrian
and Cyclist Environmental Scan (SPACES) in which safety is a crucial factor in
evaluating the trail. Safety is also considered among the various walkability indices
examined by Lo (2009). In the work of Jaskiewicz (2000), the author points out that the
23

pedestrian experience entails much more than simply a “commuting” function and that
safety is one of many important measures that distinguish a good pedestrian environment
from a poor one. Mantri (2008) also included the safety variable in the GIS model that
was developed to evaluate walkability. Lastly, safety is an essential component that
Krambeck (2006) used to develop the Global Walkability Index.
The sixth variable in Table 1 is the land use mix variable, which is often used to
estimate the ease of walkability between residences and neighboring businesses for a
given area (Brownson et al. 2009). This variable determines the evenness of square
footage distribution across the different types of land use; a higher land use mix indicates
a more even distribution of land between the land use type and higher walkability (Frank
et al. 2010). The land use mix measure was also considered in the studies of Smith et al.
(2008), Frank et al. (2010), and Mantri (2008). Smith et al (2010) suggested that a broad
mix of land use with walkable destinations increases walkability. Similarly, Frank et al.
(2010) stated that having a higher land use mix is associated with an increased likelihood
of getting sufficient physical activity. Mantri (2008) also incorporated the land use mix
measure into the GIS model as effective land use encourages residents to walk.

24

The seventh variable in Table 1 is the access to facilities measure. Shorter
distances indicate greater walkability; therefore, by measuring access to facilities for
residential areas using distances to schools, parks, transits, and shops, the ease of
walkability for the area can be determined (Brownson et al. 2009). This measure was also
included in the studies of Lo (2009), Reynolds et al. (2007), and Mantri (2008). The
access to facilities variable was one of the influencing factors of walkability that was
emphasized by the walkability indices examined by Lo (2009). Reynolds et al. (2007)
suggested that having greater access to facilities results in increasing trail use. Lastly,
Mantri (2008) also included the access to facilities variable into the GIS model to
evaluate walkability.
The eighth variable in Table 1 is steepness. This variable considers the slope that
pedestrians walk on; steep paths would be less desirable to walk on, as it requires more
control to support the body and muscles (Lay, Hass and Gregor 2006). Therefore, steeper
paths are associated with low walkability. The steepness variable is also included in the
studies of Lo (2009) and Reynolds et al. (2007).
The ninth variable in Table 1 is the tree canopy variable. Having more tree
canopy along a walking path results in shade, which protect pedestrians from the climate
(Jaskiewicz 2000). In addition, tree canopy creates shade over surfaces such as asphalt,
roofs, and concrete and prevents heating and storage of heat by these materials, which
reduces in the urban heat island phenomenon (NASA 1996). Due to these various reasons,
more tree canopy is associated with a cooler environment more pleasant for walking.
25

Tree canopy is measured in the studies of Jaskiewicz (2000) and Krambeck (2006).
Jaskiewicz (2000) recognizes how the presence of tree coverage enhances pedestrians’
experience; therefore, included tree canopy as an important factor in evaluating
walkability. Krambeck (2006) also included the number of trees as one of the indicators
in the Global Walkability Index.
Out of these nine variables, population density, street connectivity, steepness, and
tree canopy are chosen for this study. For the majority of the studies, walkability is
evaluated for the entire study area in which all nine of these variables are applied.
However, this project focuses on evaluating walkability for specific origination-
destination nodes in a transit network; therefore, many of the variables are less applicable
than others. For instance, retail floor ratio would make more sense when evaluated in
larger study area and would be difficult to be meaningful in this particular study, in which
only the walkability of short distances within a bus stop is examined. The same concept
applies to the land use mix and access to facilities measure. Dwelling density is not
incorporated into this study since it is very similar to population density and the results
for the two variables may be too similar to be included in the same study. Furthermore,
although safety is a crucial factor in walkability, it is a broad topic that is difficult to be
evaluated for the scope of this project. Street connectivity, steepness, and tree canopy are
variables that are directly related to the walking paths; therefore, they are chosen for
evaluating the bus route walkability in this study.

26

Chapter 3: Methodology
This study is a case study of two bus routes in Orange County, California that
aims to examine how GIS and ideas of walkability contribute to bus route planning using
existing geospatial datasets. By evaluating two bus routes in Orange County, the project
investigates whether different levels of walkability can be observed at bus stop for routes
in areas that have undergone different historical patterns of urban development. In
addition, sensitivity analyses are also performed to see whether moving stops short
distances from existing routes would improve walkability.
Two bus routes in Orange County are evaluated to develop and test a method for
indexing walkability. This study selects Bus 47 that runs through grid neighborhoods and
Bus 89 that runs through cul-de-sac neighborhoods to account for the two different types
of developments that took place in Orange County. Figure 2 shows the difference of a
grid and a cul-de-sac neighborhood and Figure 3 shows Route 47 and 89. The grid street
pattern is a traditional urban form that can be observed in earlier developments (Cozens
and Hillier 2008). It has highly connected linkages that incorporate commercial centers
along arterials streets, residential subdivisions by way of secondary streets, and major
transportation corridors and highways. This type of city planning encourages and
promotes pedestrian activity due its frequent intersections with the choice and directness
of route to desired destinations. In the latter part of the nineteenth century, city planners
and developers turned to the cul-de-sac pattern design that has winding streets, irregular
shapes, and dead-ends (Cozens and Hillier 2008). In a cul-de-sac neighborhood, the
27

streets are not inter-connected, which results in fewer route choices and longer distances
to travel, encouraging automobile dependency (Cozens and Hillier 2008). Since grid
neighborhoods have better street connections compared to cul-de-sac neighborhoods, this
study tests the effectiveness of the walkability scoring system by exploring the hypothesis
that stops along Route 47 would have an overall higher walkability than Route 89. The
bus stop locations in Orange County are maintained in an ArcGIS layer by OCTA
personnel who provided the data for the purpose of this study.
Figure 2: Grid vs. Cul-De-Sac Neighborhood Examples from Orange County,
California

Source: Google Maps (2012)

28

Figure 3: Southbound Bus Stop Locations of Route 47 and Route 89

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
29

Table 2 presents the historic and demographic data of the cities Route 47 and 89
serves. It can be seen that the cities which Route 47 passes through are relatively older
than those of Route 89, which is in alignment with the grid neighborhood type as
characterized by a more traditional city planning and development. In addition, the cities
that Route 47 runs through have a relatively lower average household income as
compared to those of the cities for Route 89, which could be a result of the historical and
socio-economical influences of the different developments in the Northern and Southern
Orange County.
Table 2: Route 47 and Route 89 Comparison Table
Route
Length
(miles)
Number of
Stops
Southbound
City (Year
Incorporated)
Mean Household Income
(U.S. Census Bureau 2012)
Route
47
~22 97 Fullerton (1904)
Anaheim (1870)
Garden Grove (1956)
Orange (1888)
Santa Ana (1886)
Costa Mesa (1953)
Newport Beach (1906)
$83,375
$70,436
$71,885
$90,125
$51,467
$80,480
$151,967
Average: $85,676
Route
89
~15 38 Lake Forest (1991)
Mission Viejo (1988)
Laguna Woods (1999)
Laguna Hills (1991)
Laguna Beach (1927)
$102,688
$109,510
$49,934
$123,968
$158,057
Average: $108,831

30

3.1 Three Types of Buffers
In this case study, a walkability score is given to each of the bus stops along the routes
that are then used to understand the overall walkability of the two routes. Buffers, which
identify pedestrians’ potential origins and destinations, are created for each of the bus
stops of the two routes; these buffers are where the variables are evaluated to determine
the walkability score for the designated bus stop. There are three types of buffers with
different logics of representing the pedestrians’ origins with which the walkability scores
are calculated and compared. The three types of buffers are: Half-Mile-Radii, Route-
Adjacent, and Stop-and-Route-Adjacent. With the street network data acquired from Esri,
the buffers are created using the Network Analyst tool in ArcGIS and setting the radius at
the desired distance, which traces the distance outward from each of the bus stops along
the streets and connects all the end points to form a polygon. The street network data
used as the basis for creating the buffers is developed by Esri and acquired through
ArcGIS Online.
According to the study by Agrawal, Scholossberg, and Irvin (2008), on average,
people are willing to walk approximately half a mile to transit; and following this logic,
the first buffer, the Half-Mile-Radii buffer, is created. Using ArcGIS’s Network Analysis
in which buffers of each of the bus stops are created by measuring half a mile along the
accessible paths from the bus stops and connecting the ends of each of the paths.

31

To create the second buffer, a Route-Adjacent (RA) buffer, the distance is
calculated from the subject bus stop to the closest transfer bus stop along a nearby route
since there are other bus routes also running in the same direction within close proximity
of Bus 47 and 89 several blocks away, as seen on the OCTA Bus System Map (OCTA
2012). If this distance is greater than half a mile, then the bus stop’s buffer would be
created with a half-mile radius as the maximum distance within this buffer. If this
distance is less than half a mile, then the bus stop’s buffer would be created with a radius
of half of this distance since passengers would choose to walk to a closer route in favor of
another bus route that was further away. The logic for not selecting more than half of this
distance as the radius is to prevent the likelihood of including possible passengers who
might be closer to the other bus stop.
When considering parameters for the third buffer, a Stop-and-Route-Adjacent
(SARA) buffer, it can be assumed that bus riders are more likely to take the bus from the
bus stop closest to their physical locations traveling in an appropriate route direction.
Therefore, it is logical to say that bus riders who are less than halfway between the
adjacent bus stops of the routes travelling in the same direction fall into the “catchment”
area of the bus route. To create a Stop-and-Route-Adjacent (SARA) buffer, the distance
is calculated between the bus stop of interest to the closest bus stop either on the same
route or of an adjacent nearby route carrying passengers in the same direction.

32

If this distance is greater than half a mile, then the bus stop’s buffer would be created
with a half-mile radius as the maximum distance within this buffer. If this distance is less
than half a mile, then the bus stop’s buffer would be created with a radius of half of this
distance. In this study, this distance is termed the “catchment” area for a given bus stop.
3.2 Variables, Data Sources and Calculations
Four variables from the nine main measures identified previously are chosen for this
project: population density, street connectivity, steepness, and tree canopy. Unlike most
of the other studies on walkability, this project evaluates walkability on an origination-
destination approach (i.e. pedestrian path instead of bus routes); therefore, retail floor
area ratio, land use mix, and access to facilities measures are not to be considered as they
are more applicable for more general large study areas. Safety measures, which are also
an important factor in determining walkability, are not measured as well due to the time
constraint and the scale of this project. Table 3 includes the data sources for each of the
four variables.

33

Table 3: Walkability Variables and Their Data Sources, Temporal Scales, and
Spatial Scales
Variable Data Source Temporal Scale Spatial Scale
Population
Density
U.S. Census 2010 Block Group
Street
Connectivity
U.S. Census 2010 Block
Steepness USGS National
Elevation Dataset
2012 30-meter resolution
Tree Canopy USGS National Land
Cover Dataset Tree
Canopy Layer
2001 30-meter resolution

Higher population density is associated with more walking (Forsyth et al. 2007).
Therefore, a buffer with a higher population density typically has a higher walkability
score. The population density is measured as the number of people per square mile for
each buffer. It is determined by creating an average for the overlapping block groups
within each buffer in ArcGIS. The population density of each buffer is the total
population divided by the total block group area and both the population data and the
block group area layer are acquired from the U.S. Census 2010. The results are classified
into 5 score levels using natural breaks (Jenks) classifications with 1 representing low
population density and 5 representing high population density.

34

Street connectivity indicates the directness of pedestrian routes and the more
direct the pedestrian routes the easier it is for people to walk (Brownson et al. 2009).
Street connectivity is determined by dividing the number of blocks the buffer overlaps
with by the buffer’s area in square miles. The blocks layer used in the calculation is
acquired from the U.S. Census 2010. Once the ratios of all the buffers along the bus route
are calculated, they are classified into 5 scores using natural breaks (Jenks) classifications
with 1 having the lowest ratios that correspond to low walkability and 5 having the
highest ratios that correspond to high walkability.
Even though the influence of route steepness has not been investigated much in
the walkability literature, aside from Reynolds’ study on auditing trails, it is considered
for this study since it is an important determining factor in evaluating the ease of walking
to a given bus stop. The slope of each of the paths is determined from the digital
elevation model at 10-meters resolution from USGS. Steep walking paths represent low
walkability and vice versa. The equation for the degree of slope ( ϴ) is:
Eq. 1: ϴ = arctan (
ris e
run
) (3.1)
In this formula, the range of the elevation in the buffer is the “rise” and the radius of the
buffer is the “run.” After the degree of slope is determined from the equation, they are
classified into 5 scores using natural breaks (Jenks) classifications with 1 having the
largest degrees of slope and 5 having the smallest degrees of slope.
35

Tree canopy provides pedestrians with protection from sunlight as well as blocks
heat absorbing materials along the paths from the sun to reduce urban heat island effect
(Jaskiewicz 2000; NASA 1996). The presence of tree coverage along a walking path is
especially important in Southern California, which has a high percentage of sun exposure
on any given day. The USGS National Land Cover Dataset (2001) Tree Canopy Layer at
30-meter resolution is used for determining tree canopy for the buffers. The tree canopy
coverage in the raster layer was determined by extrapolating calibrated density prediction
models derived with linear regression and regression tree techniques (Huang et al. 2001).
The layer consists of three attributes: “Rowid”, “Value”, and “Count”. “Rowid” is the
internal feature number, “Value” is the percent tree coverage, and “Count” is the total
number of cells in a grid for each unique value (USGS 2010). Using ArcGIS, the layer is
clipped against all the buffers in order to measure the coverage in each buffer. Then, for
each buffer, all the values for “Value” are converted to their corresponding adjusted
values according to Table 4. Afterwards, the adjusted values are multiplied by the “Count”
for each “Value” and summed together and divided by the total number of cells that are
within the buffer to become the new tree coverage value of the buffer. When the tree
canopy values of all the buffers are determined, they are classified into 5 scores using
natural breaks (Jenks) classifications with 1 having the lowest tree coverage and 5 having
the most tree coverage.

36

Table 4: Tree Coverage “Value” and their Adjusted Values for Calculating Tree
Coverage Value for Buffers
“Value” (Percentages) Adjusted Value
0-20 1
21-40 2
41-60 3
61-80 4
81-100 5

As mentioned in the definition and classification of the four variables above,
natural breaks (Jenks) classification is used to classify the walkability scores for each of
the four variables. This optimization partitions data into classes using an algorithm that
determines groupings of data values based on data distribution (Esri 2012). It is a
classification method that aims to reduce variance within groups and maximize variance
between groups (Esri 2012). By using natural break (Jenks) classification, the scores for
each variable would be classified with maximized variance between groups, allowing the
difference in the scores to be more obvious and distinguished among the stops.

37

3.3 Combined Score
All four variables are weighted equally and the calculated scores of each of the four
variables are summed together for a combined score of the individual buffers. These
combined scores represent the walkability of each of the bus stops and can be used to
understand the overall walkability of stops along Route 47 and Route 89. The theoretical
maximum combined score would be 20, where the highest score of 5 is obtained for each
variable; while the theoretical minimum combined score would be 4 by scoring the
lowest score of 1 for each variable.
3.4 Sensitivity Analysis
Sensitivity analyses on optimizing walkability by relocating bus stops are performed to
determine whether walkability improvements can be measured and reported when stops
are moved short distances from existing routes. In addition, by performing sensitivity
analyses, the project tries to determine whether current geospatial datasets offer a
potentially significant contribution for redesigning bus routes to increase walkability.
Sensitivity analyses are done on select bus stops in each route with the highest, lowest,
and average combined scores. For each of the stops, a hypothetical alternative bus stop is
created by picking an area out of arterial streets and plotting a point on a street roughly
half a mile away from the original stop. Then, the combined score for each of the
hypothetical alternative bus stop locations are compared with the ones for the original bus
stops.

38

Chapter 4: Results
This chapter examines the results of the analysis. It covers the following topics: (1)
results of Half-Mile-Radii buffers, (2) field observations, (3) combined score comparison
for the three types of buffers, (4) combined score comparison for Route 47 and 89, and (5)
sensitivity analysis results. First, some of the results of the walkability score calculations
for Half-Mile-Radii buffers are shown and compared with observations from field works.
Secondly, the results of the three types of buffers are compared to indicate whether
changing the buffers from Half-Mile-Radii buffers to Route-Adjacent and Stop-and-
Route-Adjacent buffers would make any difference. Then, the results of Route 47 and 89
are compared to see if the walkability scores for the two routes are different. Lastly, the
sensitivity analysis results are presented to show whether relocating the hypothetical
alternative bus stops makes any difference.
4.1 Scores for Half-Mile-Radii Buffers
The four variables were evaluated for each of the Half-Mile-Radii buffers. Figure 4
illustrates the Half-Mile-Radii buffer. For the Half-Mile-Radii buffers, Route 47 has an
overall higher combined bus stop score as compared to Route 89 (Figure 5 and 6). In
other words, the general walkability for the route that goes through grid neighborhoods is
better than that of the route which passes through cul-de-sac neighborhoods. The scores
for individual variables and the overall scores for each stop of the Half-Mile-Radii
buffers for both routes are listed in Table A.1 in the Appendix. The stops with the highest
score (score = 14) along Route 47 are stops 10-17, 25-27, and 39. The stop with the
39

lowest score (score = 6) is Route 89’s stop 38, which is also the last stop of the route. The
stops with the average score of 10 are Route 47’s stop 3, 4, 6, 20, 32, 34, 35, 57, 63, 64,
67, 68, 89, 90, and 93 and Route 89’s stop 2, 3, 8, 10-12, 19, 20, 29, and 31. For the
majority of the stops of Route 47, the scores for population density and steepness are
relatively high while the tree canopy scores are low throughout the route. Therefore, the
street connectivity score is the variable that is driving the high scores and low scores as it
varies for different stops along the route. By contrast, the scores for individual variables
for Route 89 all vary along the route; hence, it is difficult to determine one specific
variable that is driving the high scores and low scores.
Figure 4: Route 89 Stop 6 and its Half-Mile-Radii Buffer

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
40

Figure 5: Half-Mile-Radii Buffers Walkability Score Results for Route 47

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
41

Figure 6: Half-Mile-Radii Buffers Walkability Score Results for Route 89

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
42

4.2 Field Observations
The low scores for tree canopy for stops along Route 47 are confirmed from field
observations. In general, most of the stops along Route 47 do not have bus shelters as
shown in Figure 7 (a), (b), and (c) while (d) is one of the few stops that have shelters. In
addition, out of the stops that do not have shelters, there are some of them that only have
the bus stop sign and do not have any benches nearby. Even for the stops that have
benches, often the benches are not located near trees or buildings to provide tree canopy
for the passengers waiting for the bus (Figure 7 (a), (b), and (c)). On a sunny day, a
passenger may choose to sit on the curb near the stop that is protected by tree canopy
instead of the benches that are exposed completely to the sun as shown in Figure 7 (b).
Furthermore, the majority of the stops on Route 47 are on busy arterial streets, and the
greeneries are a lot of times more spread out and provide less shade for the pedestrians.
Comparing to Route 47, the majority of the stops along Route 89 have shelters as
shown in Figure 8 (a), (b), (c), and (d). Not only do most of the stops provide shelters and
benches, many of them also provide trashcans as shown in Figure 8 (a) and (d). The stops
along Route 89 are more likely to be located in places with a lot of tree; yet many times
the trees may be located too far away from the sidewalk to cast wide enough shadows to
provide tree canopy for the pedestrians as shown in Figure 8 (a), (b), and (d).

43

Figure 7: Photos of Route 47 Bus Stops

44

Figure 8: Photos of Route 89 Bus Stops

45

4.3 Combined Score Comparison for the Three Types of Buffers (HMR, RA, and
SARA)
Scores for the three different buffer types were calculated to allow for comparisons (see
Table A.4 in the Appendix for the complete scores). Figure 9 shows the histograms of the
score frequencies for the three types of buffers. From the histograms, it can be observed
that Route 47 has higher walkability scores than Route 89 for all three cases. Also, Half-
Mile-Radii buffers have overall higher combined scores compared to Route-Adjacent and
Stop-and-Route-Adjacent buffers.
The three difference of means test results of the three types of buffers against each
other indicate that the scores for Half-Mile-Radii buffers are significantly different from
those for Route-Adjacent buffers and Stop-and-Route-Adjacent (Table 5). However, the
scores for Route-Adjacent buffers and Stop-and-Route-Adjacent buffers are not
significantly different from each other (Table 5). In other words, the walkability scores
are significantly different for adjusted and non-adjusted buffers. However, it does not
make a significant difference whether the buffers are adjusted with the Route-Adjacent or
Stop-and-Route-Adjacent methods. Due to the proximity of the bus stops to each other,
all of the buffers for Route 47 stops and the majority of the Route 89 stops are adjusted to
a radius shorter than half a mile (Figure 10).

46

From the results, it can be observed that Half-Mile-Radii buffers consistently have
better scores than those of Route-Adjacent and Stop-and-Route-Adjacent buffers.
However, it is important to note this result should not be interpreted to mean that longer
walks to bus stops are more walkable than shorter walks, since length of walk to bus
stops was not measured as a variable. Instead, it appears that higher scores for the larger,
HMR buffers may be a sort of “statistical artifact” where the size of the buffer drives up
the scores on the four variables investigated in this study. The significance of having
different types of buffers is to account for how people would choose to ride the buses,
and in this particular study, it is assumed that people’s choices depend on the distances to
the bus stops relative to their locations to other routes and stops. Therefore, the different
sizes in buffers and their varying results indicate that it is a difficult challenge to create
buffers that are appropriate for potential bus stop locations in the bus route planning
models.

47

Figure 9: Histograms of Combined Bus Stop Scores for the Three Types of Buffers
(HMR, RA, and SARA)

Table 5: Difference of Means Test Results for the Three Types of Buffers (HMR,
RA, and SARA)
Buffer Types Mean Scores t Stat P one-tail
HMR vs. RA 11.02 vs. 9.82 5.19 2.09E-07
HMR vs.
SARA
11.02 vs. 9.70 5.51 4.29E-08
RA vs. SARA 9.82 vs. 9.70 0.46 0.32

48

Figure 10: Bus Stops with Adjusted and Non-Adjusted Buffers

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
49

4.4 Combined Score Comparison for Route 47 and Route 89
The combined scores of Route 47 and 89 for the three types of buffers are listed in Table
A.5 in the Appendix. Route 47 has higher “lowest score” and “highest score” than Route
89 for all three types of buffers except for the “highest score” for Stop-and-Route-
Adjacent buffer, in which the “highest score” for both routes are the same (Table 6). The
walkability score results support the hypothesis that stops along Route 47 are more
walkable than stops along Route 89. The difference of means tests of Route 47 and 89 for
each of the three types of buffers all indicate that the scores for Route 47 and 89 are 95%
certain to be at least 1 value but not 2 values different from one another (Table 7).
Overall, Route 47 has a higher walkability than Route 89 by approximately 1 score value
difference.

50

Table 6: Descriptive Statistics of Route 47 and Route 89 for the Three Types of
Buffers (HMR, RA, and SARA)
Route 47 Route 89
Half-Mile-Radii
Lowest Score 8 6
Highest Score 14 13
Median 12 9, 10
Mean 11.6 9.47
Standard
Deviation
1.70 1.48
Range 6 7
Route-Adjacent
Lowest Score 7 5
Highest Score 15 10
Median 11 8
Mean 10.5 7.95
Standard
Deviation
1.63 1.23
Range 8 5
Stop-and-Route-
Adjacent
Lowest Score 6 5
Highest Score 14 14
Median 11 8
Mean 10.4 8.03
Standard
Deviation
1.63 2.01
Range 8 9

51

Table 7: Difference of Means Test Results for Route 47 and Route 89 for the Three
Types of Buffers (HMR, RA, and SARA)
Buffer
Type
Mean Scores
47 vs. 89
Hypothesized
Mean Diff.
t Stat P one-
tail
HMR 11.6 vs. 9.47 0 7.30 1.22E-10
1 3.90 0.0001
2 0.52 0.301
RA 10.3 vs. 7.95 0 7.97 5.71E-13
1 4.64 4.59E-06
2 1.31 0.10
SARA 10.4 vs. 8.03 0 6.39 1.61E-08
1 3.66 0.0003
2 0.92 0.18

4.5 Sensitivity Analysis Results
Case studies on relocating bus stops to optimize walkability are done to examine whether
walkability improvements can be measured and reported when stops are moved short
distances from existing routes. Stop 10 of Route 47, Stop 38 of Route 89, and Stop 3 of
Route 47 were chosen for the sensitivity analysis because they represent the highest,
lowest, and average combined scores respectively (Figure 11, Figure 12, and Figure 13).
By adjusting the bus stop locations to a hypothetical alternative stop, the combined scores
increased for stops with the average and the lowest scores (i.e., Stop 38 of Route 89 and
Stop 3 of Route 7). However, for the stop with the highest score (i.e., Stop 10 of Route 47)
the hypothetical alternative stop’s combined score is lower than the original. From the
score breakdown of the highest combined scores of the original and the alternative stop, it
can be seen that lower street connectivity is the reason that this alternative stop has a
52

lower walkability score than the original stop (Table 8). It is reasonable that the street
connectivity score for the original stop is higher than that of the alternative stop since
arterial streets have more intersections and crossroads than the smaller roads in
neighborhoods. This might well be a common finding for high scoring stops along bus
routes in grid–style neighborhoods where current route design often places the stops on
arterial streets.

53

Figure 11: Original and Alternative Locations for Route 47 Stop 10

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
54

Figure 12: Original and Alternative Locations for Route 89 Stop 38

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
55

Figure 13: Original and Alternative Locations for Route 47 Stop 3

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
56

Table 8: Comparison Table for Half-Mile-Radii Buffer Sensitivity Analysis Results
Route_Stop
(Original/Alternative)
Population
Density
Street
Connectivity
Steepness Tree
Canopy
Combined
Score
47_10 (Original) 4 4 5 1 14
47_10 (Alternative) 5 2 5 1 13
89_38 (Original) 1 2 1 2 6
89_38 (Alternative) 2 2 1 4 9
47_3 (Original) 2 2 5 1 10
47_3 (Alternative) 5 2 5 1 13

57

Chapter 5: Discussion and Conclusion
The results for the three types of buffers employed in this study all support the hypothesis
that Route 47 has an overall higher walkability score than Route 89, which is drawn from
the fact that grid neighborhoods have highly connected streets while cul-de-sac
neighborhoods have less street linkages with lots of dead ends. Therefore, it is reasonable
to conclude that, based on the present study, grid neighborhoods have higher walkability
than cul-de-sac neighborhoods under normal circumstances. This research finding is
cause for concern due to the fact that modern neighborhoods (cul-de-sac) are being built
with less connectivity to public transportation systems compared to the traditional grid
neighborhood design. The sensitivity analysis results indicate that moving a bus stop with
an average or lower combined score to locations further from arterial streets and deeper
into neighborhoods would increase the walkability of the bus stop. However, as stated by
Ceder and Wilson (1984), there are real risks in redesigning the bus network that would
result in poorer bus system performances. In the hypothetical examples developed here,
each adjusted bus stop location would add approximately one mile to the original bus
route.

58

Thus, if OCTA were to make numerous adjustments to the bus routes in each case where
alternative bus stops’ scores are higher than those of the original stops, then a lot of
distances and time might be added to the already long bus rides. Therefore, although the
walkability of each of the adjusted bus stops would be increased, the overall
transportation efficiency, including time and fuel consumption, would be reduced. In
other words, even though walkability is an important factor for people to utilize public
transportation, rerouting to increase walkability may not be the optimal solution.
Although walkability may not be most significant factor in planning a bus route,
the scoring system and geospatial datasets introduced here may help transit engineers and
planners to weigh the significance of the walkability as part of origin and destination
modeling. For example, Ouyang and Nourbakhsh (2010) introduced an entirely new
transit concept called the “flexible-route transit system” that is suitable for low-demand
areas. To summarize the concept, the flexible-route bus would move within a coverage
area called a “bus tube” in which the bus moves back and forth picking up and dropping
off passengers at their precise origins and destinations while passing transfer points along
the way (Ouyang and Nourbakhsh 2010). The bus would only be dispatched when
potential passengers request service through a website (e.g., mobile GIS technology) or
phone calls (Ouyang and Nourbakhsh 2010). Passengers that want to travel beyond the
“bus tube” would be dropped off at transfer points and be picked up by another bus
(Ouyang and Nourbakhsh 2010). Since the bus would travel to the passengers’ precise
origins and destinations, passengers do not have to walk great distances to the nearest
stop (Ouyang and Nourbakhsh 2010).
59

By knowing which stops along a route have the lower walkability scores using
methods like the one demonstrated here, these places can be identified as the low-demand
areas. “Bus tubes” for entire routes or portions of routes can be planned accordingly. By
having the “flexible-route transit system” in which the buses would only be dispatched at
the requests of the passengers in the “bus tubes”, the demand for bus resources would be
reduced. Therefore, even if the result of this study indicates that redesigning bus routes to
optimize walkability would decrease the bus system’s efficiency, information on
walkability that is determined from this study can be used to consider abandoning fixed
route stops in certain cases in favor of variable route bus service.
The theories of both the fields of Active Living Research as well as Smart Growth
recognize the relationship between a walkable environment and public transportation. By
designing the public transits to be more accessible and convenient, people would be more
willing to travel with them regularly, thereby encourage walking and promote a healthy
lifestyle. This study shows the significance of considering the origin, destination, and
walking paths for pedestrians when approaching the bus network design problems. By
having more walkable routes, not only would bus ridership be encouraged, people’s
quality of life would be improved as well.

60

5.1 Limitations
As a specific case study done on understanding walkability’s role in the bus network
design problem, this project has several limitations. First of all, this study only
investigated two bus routes, which is a very small sample size considering the numerous
bus routes there are in Orange County, let alone the numerous bus systems worldwide.
Secondly, walkability is only evaluated with four variables due to the scale of this study;
yet there may be other crucial influencing factors that may contribute to walkability that
were not taken into account in this particular case study.
Furthermore, the USGS National Land Cover Dataset (2001) Tree Canopy Layer
(USGS Tree Canopy Layer) that was used to determine the tree canopy score may not be
the best representation of tree canopy for urbanized Orange County and resulted in
inaccurate tree canopy score results. The USGS Tree Canopy Layer was developed for
the USGS’s National Water-Quality Assessment Program to assess land-use change and
to allocate nutrient and pesticide loads to different land-use categories (USGS 2010). The
layer was created with a specific intention for use by the USGS instead of for the general
use of the public. In addition, the tree canopy coverage in the layer was determined by
extrapolating calibrated density prediction models that were derived using both linear
regression and regression tree techniques (Huang, et al. 2001).

61

Therefore, the data in the layer do not reflect the actual tree canopy locations but rather
are an approximation. However, for the purpose of this project, it is important for the tree
coverage data to be geographically accurate, which is why the USGS Tree Canopy Layer
is identified as a limitation in this particular case study.
Another method to determine the tree coverage percentage for each buffer is by
manually digitizing the trees. Figure 14 shows the manual digitization of the trees for
Route 47 Stop 28’s Half-Mile-Radii buffer. There is definitely the presence of trees
within the buffer of Stop 28; however, according to the USGS Tree Canopy Layer, there
are no trees within this region. The manual digitization method, although more accurate
than using the USGS Tree Canopy Layer, requires a lot more time and manpower to
develop. Furthermore, transit planning agencies will need to use ready-made GIS datasets
to incorporate walkability into planning models which renders this method as ideal for
understanding tree canopy’s effect in bus route planning.
Even though the USGS Tree Canopy Layer was unable to yield tree canopy
scores that reflect the actual tree coverage, it was used because other spatial data for the
analysis were not found. After much research, it has been concluded that the USGS Tree
Canopy Layer present in this study is the best option available for use in urbanized areas.

62

Figure 14: Manually Digitized Tree Coverage for Route 47 Stop 28’s Half-Mile-
Radii Buffer

Sources: Esri Bing Maps Road (2012)

63

5.2 Future Research
The Stop-and-Route-Adjacent buffers used in this study serve to identify pedestrian
“catchment” areas of the bus stops. However, there may be more dynamic and
sophisticated ways to calculate pedestrian “catchment” areas for transit stops based upon
the utilization rates, speeds, or connectivity of particular routes in the overall transit
network. Future research may consider building origin and destination demand models in
which the walking to stops aspect would be taken into consideration. Besides only
determining the “catchment” area by distance to the bus stops from pedestrians’ origins,
there are many other factors that contribute to people’s decision in choosing to ride the
bus from a certain stop than another. For instance, due to the variety of bus routes, people
might be willing to walk to further to an initial transit stop if the entire journey can be
made with fewer bus transfers. To consider the different possibilities, transit engineers
should work with GIS specialists to feed data on walkability into their origin and
destination demand models.
One topic that requires attention in future research is the problem that is
introduced by relating low population density with low walkability. Bus stops located
near concentrated nodes of shopping or industry may actually be quite convenient for
people to walk to retail stores or work, yet because of the low population densities in
these areas, they received lower walkability scores.

64

Therefore, the population density variable walkability results provide good
representations for walkability in residential areas, but not so much for commercial or
industrial areas. This limitation in the study needs refinement in future research that takes
into account the type of land use variation in areas where bus stops are located.
Another issue that should be addressed in future research is the modifiable areal
unit problem (MAUP) that may have affected the results of this study. In this study, three
types of buffers are used to represent three logics of choosing the bus stops. Due to the
way these buffers are determined, the Half-Mile-Radii buffers are generally much larger
than that of Route-Adjacent and Stop-and-Route-Adjacent buffers (Figure 15). The
walkability scores for the Half-Mile-Radii buffers are also higher than the other two
buffers. Therefore, sizes of the buffers may or may not be the reason behind the
differences in the scores. However, more data sampling and detailed field studies are
needed in order to draw a conclusion.

65

Figure 15: Route 47 Stop 20 and its Three Types of Buffers (HMR, RA, and SARA)

Sources: Esri Bing Maps Road (2012), OCTA Bus Stop Layer (2012)
66

Bibliography

Abley, Steve. "Walkability Scoping Paper." 2005.
Active Living Research. About Active Living Research. 2012.
http://www.activelivingresearch.org/about (accessed August 12, 2012).
—. Active Living Resources for Transportation. 2012.
http://www.activelivingresearch.org/taxonomy/term/218 (accessed August 12,
2012).
Agrawal, Asha Weinstein, Marc Schlossberg, and Katja Irvin. "How Far, by Which
Route and Why? A Spatial Analysis of Pedestrian Preference." Journal of Urban
Design, 2008: 81-98.
Brownson, Ross C., Christine M. Hoehner, Kristen Day, Ann Forsyth, and James F.
Sallis. "Measuring the FOod and Physical Activity Environments: State of the
Science." American Journal of Preventive Medicine, 2009: 99-123.
Ceder, Avishai, and Nigel H.M. Wilson. "Bus Network Design." Transpn. Res.-B, 1986:
331-344.
Chang, S. K., and P Schonfeld. "Multiple period optimization of bus transit systems."
Transportation research. Part B: methodological, 1991: 453-378.
Chang, S. K., and P Schonfeld. "Welfare maximization with financial constraints for bus
transit systems." Transportation Research Record 1395, 1993: 48-57.
Chein, Steven I-Jy, Branislav V. Dimitrijevic, and Lazar N. Spasovic. "Optimization of
Bus Route Planning in Urban Commuter Networks." Journal of Public
Transportation, 2003: 53-79.
Cozens, Paul, and David Hillier. "The Shape of Things to Come: New Urbanism, the
Grid and the Cul-De-Sac." International Planning Studies (International Planning
Studies), 2008: 37-41.
EPA. Smarth Growth. February 23, 2012. http://www.epa.gov/smartgrowth/about_sg.htm
(accessed March 27, 2012).
Esri. AcGIS Online Map and Geoservices. 2012.
http://www.esri.com/software/arcgis/arcgis-online-map-and-geoservices
(accessed April 28, 2012).
67

—. GIS Dictionary. 2012.
http://support.esri.com/en/knowledgebase/Gisdictionary/browse (accessed August
31, 2012).
Fan, Wei, and Randy B. Machemehl. "Optimal Transit Route Network Design Problem
with Variable Transit Demand: Genetic Algorithm Approach." Journal of
Transportation Engineering, 2006: 40-51.
Forsyth, Ann J., Michael Oakes, Kathryn H. Schmitz, and Mary Hearst. "Does
Residential Density Increase Walking and Other Physical Activity?" Urban
Studies, 2007: 44-679.
Frank, Lawrence D., Andrew Devlin, Shana Johnstone, and Josh van Loon.
Neighbourhood Design, Travel, and Health in Metro Vancouver: Using a
Walkability Index Executive Summary. Vancouver: UBC Active Transportation
Collaboratory, 2010.
Frumkin, Howard. Urban Sprawl and Public Health. Association of Schools of Public
Health, 2002.
Google. Google Maps. 2012. https://maps.google.com/ (accessed 4 20, 2012).
Gupte, Vaidehi Niteen. Designers' Perspective of Walkability and Accessbility of Dart's
Downtown Transitway Mall in Dallas, Texas. Ann Arbor: ProQuest LLC, 2009.
Halper, Evan, and Dave McKibben. "Orange County; Suburbs Dream of Downtown;
Planning: Officials are seeking ways to retrofit bedroom communities with what
Orange and other old cities have-- an urban core." Los Angeles Times, April 1,
2002: B.1.
Handy, Susan L., Marlon G. Boarnet, Reid Ewing, and Richard E. Killingsworth. "How
the Built Environment Affects Physical Activity." American Journal of
Preventive Medicine, 2002: 65-73.
Hirt, Sonia A. "Toward Post Modern Urbanism? Evolution of Planning in Cleveland,
Ohio." Journal of Planning Education and Research, 2005: 27-42.
Huang, Chengquan, Limin Yang, Bruce Wylie, and Collin Homer. A Strategy for
Estimating Tree canopy Density Using Landsat 7 ETM+ and High Resoluion
Images over Large Areas. Sioux Falls: USGS, 2001.
68

Irvine Company. Since 1864. 2012. http://www.irvinecompany.com/since-1864 (accessed
August 31, 2012).
Jaskiewicz, Frank. "Pedestrian Level of Services Based on Trip Quality." TRB Circular
E-C019: Urban Street Symposium, 2000.
Johnson, Maryfran. "GIS Popularity Growing." Computerworld, 1993: 41-42.
Krambeck, Holly Virginia. The Global Walkability Index. Massachusetts : Massachusetts
Institute of Technology, 2006.
Lampkin, W., and P. D. Saalmans. "The Design of Routes, Service Frequencies, and
Schedules for a Municipal Bus Undertaking: A Case Study." Operational
Research Quarterly, 1967: 375-397.
Lay, Andrea N., Chris J. Hass, and Robert J. Gregor. "The effect of sloped surface on
locomotion: A kinematic and kinetic analysis." Journal of Biomechanics, 2006:
1621-1628.
Lee, R.W., and C.R. Rivasplata. "Metropolitan Transportation Planning in the 1990s:
Comparisons and Contrasts in New Zealand, Chile and California." Pergamon,
2001: 47-61.
Litman, Todd. Evaluting Criticism of Smart Growth. Victoria: Victoria Transport Policy
Institute, 2012.
Lo, Ria Hutabarat. "Walkability: what is it?" Journal of Urbanism, 2009: 145-166.
Mantri, Anupama. A GIS Based Approach to Measure Walkability of a Neighborhood.
Master Thesis, Cincinnati: University of Cincinnati, 2008.
Marshall, Julian D., Michael Brauer, and Lawrence D. Frank. "Healthy Neighborhoods:
Walkability and Air Pollution." Environmental Health Perspectives, 2009: 1752-
1759.
NASA. What's hot in Huntsville and what's not: A NASA thermal remote sensing project.
February 14, 1996. http://wwwghcc.msfc.nasa.gov/land/heatisl/heatisl.htm
(accessed August 31, 2012).
Newell, G. F. "Some Issues Relating to the Optimal Design of Bus Routes."
Transportation Science, 1979: 20-35.
69

Newman, Peter. "Reducing Automobile Dependence." Environment and Urbanization,
1996: 67-92.
OCTA. OCTA Bus Stop. Orange, March 2012.
—. OCTA Map. 2012. http://www.octa.net/bus/feb11sysmap/index.html (accessed April
19, 2012).
Orange County Historical Society. A Brief History of Orange County California. 2012.
http://www.orangecountyhistory.org/history-brief.html (accessed August 31,
2012).
Ouyang, Yanfeng, and Seyed Mohammad Nourbakhsh. "A structured flexible transit
system for low demand areas." Elsevier, 2010: 204-216.
Pikora, Terri. Survey of the Physical Environment in Local Neighbourhoods SPACES
Instrument: Observers Manual. Nedlands: The University of Western Australia,
n.d.
Rea, John Charles. Designing urban transit system: An approach to the route-technology
selection problem. Dissertation, Uniervsity of Washington, 1971.
Reynolds, Kim D., et al. "Trail Characteristics as Correlates of Urban Trail Use." The
Science of Health Promotion, 2007: 335-345.
Smith, Ken R., Barbara B. Brown, Ikuho Yamada, Lori Kowaleski-Jones, Cathleen D.
Zick, and Jessie X. Fan. "Walkability and Body Mass Index Density, Design, and
New Diversity Measures." American Journal of Preventive Medicine, 2008: 237-
244.
The Strategic Growth Council. "California Strategic Growth Council Strategic Plan 2012-
2014." The Strategic Growth Council. 2012.
http://sgc.ca.gov/docs/workplan/strategicplan-01-24-12.pdf (accessed March 27,
2012).
Tom, V. M., and S. Mohan. "Transit Route Network Design Using Frequency coded
Genetic Algorithm." Journal of Transportation Engineering, 2003: 186-195.
U.S. Census Bureau. U.S. Census Bureau. 2012.
http://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml (accessed April 19,
2012).
70

United States Census Bureau. State & County QuickFacts. August 16, 2012.
http://quickfacts.census.gov/qfd/states/06000.html (accessed August 31, 2012).
USGS. Water Resources NSDI Node. 2010.
http://water.usgs.gov/GIS/metadata/usgswrd/XML/cnpy01_3.xml (accessed June
4, 2012).
Wolch, Jennifer R., Manuel Pastor, and Peter Dreier. Up Against the Sprawl: Public
Policy and the Making of Southern California. Minneapolis: University of
Minnesota Press, 2004.

71

Appendix
Table A.1 Combined Bus Stop Scores for Half Mile Radii Buffers
Stop
Population
Density
Street
Connectivity
Steepness
Tree
Canopy
Combined
Score
47_1 3 2 5 1 11
47_2 2 3 5 1 11
47_3 2 2 5 1 10
47_4 2 2 5 1 10
47_5 2 2 4 1 9
47_6 3 2 4 1 10
47_7 3 3 5 1 12
47_8 3 4 5 1 13
47_9 3 4 5 1 13
47_10 4 4 5 1 14
47_11 4 4 5 1 14
47_12 4 4 5 1 14
47_13 4 4 5 1 14
47_14 4 4 5 1 14
47_15 4 4 5 1 14
47_16 4 4 5 1 14
47_17 4 4 5 1 14
47_18 4 3 5 1 13
47_19 2 3 5 1 11
47_20 2 2 5 1 10
47_21 1 2 5 1 9
47_22 1 2 4 1 8
47_23 2 2 4 1 9
47_24 4 3 5 1 13
47_25 5 3 5 1 14
47_26 5 3 5 1 14
47_27 5 3 5 1 14
47_28 4 2 5 1 12
47_29 4 3 5 1 13
72

47_30 3 3 5 1 12
47_31 2 3 5 1 11
47_32 2 2 5 1 10
47_33 2 3 5 1 11
47_34 2 2 5 1 10
47_35 2 2 5 1 10
47_36 3 2 5 1 11
47_37 2 3 5 1 11
47_38 3 4 5 1 13
47_39 4 4 5 1 14
47_40 4 3 4 1 12
47_41 4 3 4 1 12
47_42 4 3 5 1 13
47_43 4 3 5 1 13
47_44 4 2 5 1 12
47_45 4 2 5 1 12
47_46 4 2 5 1 12
47_47 4 2 5 1 12
47_48 4 2 5 1 12
47_49 4 2 5 1 12
47_50 4 4 5 1 14
47_51 5 3 5 1 14
47_52 5 3 5 1 14
47_53 5 3 5 1 14
47_54 4 3 5 1 13
47_55 3 2 5 1 11
47_56 3 2 5 1 11
47_57 2 2 5 1 10
47_58 2 3 5 1 11
47_59 2 3 5 1 11
47_60 2 3 5 1 11
47_61 2 3 5 1 11
47_62 2 3 5 1 11
47_63 1 3 5 1 10
73

47_64 1 3 5 1 10
47_65 1 3 5 1 10
47_66 1 2 5 1 9
47_67 2 2 5 1 10
47_68 2 2 5 1 10
47_69 1 2 5 1 9
47_70 1 2 4 1 8
47_71 1 2 5 1 9
47_72 2 2 5 1 10
47_73 3 3 5 1 12
47_74 4 3 5 1 13
47_75 4 2 5 1 12
47_76 4 2 5 1 12
47_77 4 2 5 1 12
47_78 5 3 5 1 14
47_79 5 3 5 1 14
47_80 5 3 5 1 14
47_81 4 2 5 1 12
47_82 5 2 5 1 13
47_83 4 2 5 1 12
47_84 4 2 5 1 12
47_85 4 2 5 1 12
47_86 4 2 4 1 11
47_87 3 3 4 1 11
47_88 2 3 5 1 11
47_89 2 3 4 1 10
47_90 2 3 4 1 10
47_91 1 3 3 1 8
47_92 1 2 4 1 8
47_93 1 5 3 1 10
47_94 1 5 5 1 12
47_95 1 5 5 1 12
47_96 1 5 5 1 12
47_97 1 5 5 1 12
74

89_1 3 2 2 2 9
89_2 3 2 2 3 10
89_3 2 3 2 3 10
89_4 3 3 3 3 12
89_5 3 3 4 3 13
89_6 2 2 2 3 9
89_7 1 1 2 4 8
89_8 3 1 2 4 10
89_9 3 2 2 4 11
89_10 3 2 3 2 10
89_11 3 2 4 1 10
89_12 3 2 4 1 10
89_13 4 2 4 1 11
89_14 3 3 4 1 11
89_15 3 2 5 1 11
89_16 2 2 4 1 9
89_17 2 4 4 1 11
89_18 2 3 5 1 11
89_19 2 3 4 1 10
89_20 2 3 4 1 10
89_21 2 2 4 1 9
89_22 2 2 4 1 9
89_23 2 2 3 1 8
89_24 2 3 3 1 9
89_25 2 2 3 1 8
89_26 2 2 3 1 8
89_27 2 2 1 2 7
89_29 2 3 2 3 10
89_29 2 3 2 2 9
89_30 2 3 3 4 12
89_31 1 3 2 4 10
89_32 1 2 1 4 8
89_33 1 1 1 5 8
89_34 1 1 1 5 8
75

89_35 1 1 2 5 9
89_36 1 1 1 5 8
89_37 1 1 1 5 8
89_38 1 2 1 2 6

Table A.2 Combined Bus Stop Scores for Route-Adjacent Buffers
Stop
Population
Density
Street
Connectivity
Steepness
Tree
Canopy
Combined
Score
47_1 3 2 4 1 10
47_2 2 1 5 1 9
47_3 2 2 5 1 10
47_4 2 2 5 1 10
47_5 2 2 5 1 10
47_6 3 1 5 1 10
47_7 3 2 5 1 11
47_8 3 1 5 1 10
47_9 3 1 5 1 10
47_10 3 2 5 1 11
47_11 3 2 5 1 11
47_12 3 2 5 1 11
47_13 3 2 5 1 11
47_14 4 2 5 1 12
47_15 4 2 5 1 12
47_16 4 2 5 1 12
47_17 4 2 5 1 12
47_18 4 2 5 1 12
47_19 3 2 5 1 11
47_20 2 2 5 1 10
47_21 2 2 5 1 10
47_22 1 1 4 1 7
47_23 2 1 4 1 8
47_24 4 1 5 1 11
47_25 5 2 5 1 13
76

47_26 5 1 5 1 12
47_27 4 1 5 1 11
47_28 4 1 5 1 11
47_29 3 1 5 1 10
47_30 3 1 5 1 10
47_31 2 1 4 1 8
47_32 2 4 4 1 11
47_33 2 1 5 1 9
47_34 2 1 5 1 9
47_35 2 2 5 1 10
47_36 3 2 5 1 11
47_37 2 1 5 1 9
47_38 3 1 5 1 10
47_39 4 2 5 1 12
47_40 4 2 4 1 11
47_41 4 2 4 1 11
47_42 4 5 5 1 15
47_43 4 1 5 1 11
47_44 4 1 5 1 11
47_45 4 1 5 1 11
47_46 4 1 5 1 11
47_47 4 1 5 1 11
47_48 4 1 5 1 11
47_49 4 1 5 1 11
47_50 4 2 5 1 12
47_51 5 2 5 1 13
47_52 5 2 5 1 13
47_53 5 2 4 1 12
47_54 4 1 5 1 11
47_55 3 1 5 1 10
47_56 3 1 5 1 10
47_57 2 1 5 1 9
47_58 2 1 5 1 9
47_59 2 1 5 1 9
77

47_60 2 2 5 1 10
47_61 2 1 5 1 9
47_62 2 2 5 1 10
47_63 2 1 5 1 9
47_64 1 5 5 1 12
47_65 1 1 5 1 8
47_66 1 1 5 1 8
47_67 1 1 5 1 8
47_68 2 1 5 1 9
47_69 2 1 5 1 9
47_70 1 1 4 1 7
47_71 2 1 5 1 9
47_72 3 1 5 1 10
47_73 4 2 5 1 12
47_74 4 1 5 1 11
47_75 4 1 5 1 11
47_76 5 1 5 1 12
47_77 5 3 5 1 14
47_78 5 4 5 1 15
47_79 5 2 5 1 13
47_80 4 3 5 1 13
47_81 3 3 5 1 12
47_82 4 1 5 1 11
47_83 4 1 5 1 11
47_84 4 2 5 1 12
47_85 4 1 4 1 10
47_86 4 1 4 1 10
47_87 3 1 4 1 9
47_88 2 1 5 1 9
47_89 2 2 5 1 10
47_90 2 2 4 1 9
47_91 2 5 3 1 11
47_92 1 1 4 1 7
47_93 1 2 3 1 7
78

47_94 3 2 4 1 10
47_95 2 3 5 1 11
47_96 5 2 5 1 13
47_97 5 2 5 1 13
89_1 3 2 2 1 8
89_2 2 1 1 4 8
89_3 2 1 1 4 8
89_4 2 1 1 1 5
89_5 2 1 2 1 6
89_6 2 1 2 3 8
89_7 1 1 2 4 8
89_8 3 1 2 4 10
89_9 3 1 2 4 10
89_10 3 1 3 3 10
89_11 3 1 4 1 9
89_12 3 1 4 1 9
89_13 3 1 4 1 9
89_14 3 1 4 1 9
89_15 3 1 4 1 9
89_16 2 1 4 1 8
89_17 2 1 4 1 8
89_18 2 1 5 1 9
89_19 2 1 4 1 8
89_20 2 1 4 1 8
89_21 1 1 4 1 7
89_22 2 1 4 1 8
89_23 2 1 3 1 7
89_24 2 1 3 1 7
89_25 2 1 3 1 7
89_26 2 1 3 1 7
89_27 2 1 1 2 6
89_29 2 1 2 3 8
89_29 2 1 2 2 7
89_30 2 1 3 4 10
79

89_31 1 1 2 4 8
89_32 1 1 1 4 7
89_33 1 1 1 5 8
89_34 1 1 1 5 8
89_35 1 1 2 5 9
89_36 1 1 1 5 8
89_37 1 1 1 5 8
89_38 1 1 1 2 5

Table A.3 Combined Bus Stop Scores for Stop-and-Route-Adjacent Buffers
Stop
Population
Density
Street
Connectivity
Steepness
Tree
Canopy
Combined
Score
47_1 3 2 4 1 10
47_2 2 1 4 1 8
47_3 2 2 5 1 10
47_4 2 1 5 1 9
47_5 2 1 4 1 8
47_6 3 1 5 1 10
47_7 3 2 5 1 11
47_8 3 2 5 1 11
47_9 3 2 5 1 11
47_10 3 1 5 1 10
47_11 3 1 5 1 10
47_12 3 1 5 1 10
47_13 3 2 5 1 11
47_14 4 1 5 1 11
47_15 4 1 5 1 11
47_16 3 1 5 1 10
47_17 4 2 5 1 12
47_18 4 1 5 1 11
47_19 3 1 5 1 10
47_20 1 1 5 1 8
47_21 1 1 5 1 8
47_22 1 1 5 1 8
47_23 3 2 5 1 11
47_24 4 1 5 1 11
47_25 5 1 5 1 12
80

47_26 4 1 5 1 11
47_27 4 1 5 1 11
47_28 4 1 5 1 11
47_29 4 1 5 1 11
47_30 3 1 5 1 10
47_31 2 1 4 1 8
47_32 2 2 4 1 9
47_33 2 2 5 1 10
47_34 2 2 5 1 10
47_35 2 1 5 1 9
47_36 1 2 5 1 9
47_37 1 4 5 1 11
47_38 3 5 5 1 14
47_39 4 2 4 1 11
47_40 4 1 5 1 11
47_41 4 3 5 1 13
47_42 3 1 4 1 9
47_43 3 1 5 1 10
47_44 4 1 5 1 11
47_45 4 1 5 1 11
47_46 4 1 4 1 10
47_47 4 2 5 1 12
47_48 4 1 5 1 11
47_49 4 1 5 1 11
47_50 4 1 5 1 11
47_51 5 1 5 1 12
47_52 5 1 5 1 12
47_53 5 2 4 1 12
47_54 4 1 4 1 10
47_55 2 1 4 2 9
47_56 2 1 4 1 8
47_57 2 1 4 1 8
47_58 3 5 5 1 14
47_59 2 1 5 1 9
47_60 1 2 4 1 8
47_61 1 1 5 1 8
47_62 1 2 4 1 8
47_63 1 1 4 1 7
81

47_64 1 2 5 1 9
47_65 1 4 5 1 11
47_66 1 2 5 1 9
47_67 2 2 4 1 9
47_68 3 1 5 1 10
47_69 1 1 4 1 7
47_70 1 3 5 1 10
47_71 1 2 5 1 9
47_72 2 1 5 1 9
47_73 3 1 5 1 10
47_74 4 1 5 1 11
47_75 4 1 5 1 11
47_76 5 1 5 1 12
47_77 5 3 5 1 14
47_78 5 1 5 1 12
47_79 5 1 5 1 12
47_80 5 1 5 1 12
47_81 5 1 5 1 12
47_82 4 1 4 1 10
47_83 4 1 5 1 11
47_84 4 1 5 1 11
47_85 4 1 4 1 10
47_86 4 1 5 1 11
47_87 4 3 5 1 13
47_88 4 2 4 1 11
47_89 4 2 5 1 12
47_90 2 4 5 1 12
47_91 1 2 5 1 9
47_92 1 1 3 1 6
47_93 1 2 4 1 8
47_94 5 2 4 1 12
47_95 5 2 4 1 12
47_96 5 2 5 1 13
47_97 5 2 5 1 13
89_1 2 2 2 1 7
89_2 2 2 3 1 8
89_3 2 2 1 1 6
89_4 2 3 5 1 11
82

89_5 3 5 5 1 14
89_6 1 1 2 1 5
89_7 1 1 2 3 7
89_8 2 1 3 4 10
89_9 3 1 1 4 9
89_10 3 1 4 1 9
89_11 2 1 4 1 8
89_12 3 1 3 1 8
89_13 3 1 3 1 8
89_14 3 1 4 1 9
89_15 2 1 5 1 9
89_16 2 3 5 1 11
89_17 2 4 5 1 12
89_18 2 2 5 1 10
89_19 2 2 4 1 9
89_20 2 2 4 1 9
89_21 2 1 4 1 8
89_22 2 1 4 1 8
89_23 1 1 3 1 6
89_24 1 2 3 1 7
89_25 1 1 2 1 5
89_26 1 3 4 1 9
89_27 2 1 2 1 6
89_29 2 1 2 1 6
89_29 2 2 1 1 6
89_30 1 2 5 1 9
89_31 1 1 2 3 7
89_32 1 1 3 3 8
89_33 1 1 3 2 7
89_34 1 1 2 1 5
89_35 1 1 2 5 9
89_36 1 1 1 5 8
89_37 1 1 1 4 7
89_38 1 1 1 2 5

83

Table A.4 Combined Scores for the Three Types of Buffers
Stop Half-Mile-Radii Route-Adjacent Stop-and-Route-Adjacent
47_1 11 10 10
47_2 11 9 8
47_3 10 10 10
47_4 10 10 9
47_5 9 10 8
47_6 10 10 10
47_7 12 11 11
47_8 13 10 11
47_9 13 10 11
47_10 14 11 10
47_11 14 11 10
47_12 14 11 10
47_13 14 11 11
47_14 14 12 11
47_15 14 12 11
47_16 14 12 10
47_17 14 12 12
47_18 13 12 11
47_19 11 11 10
47_20 10 10 8
47_21 9 10 8
47_22 8 7 8
47_23 9 8 11
47_24 13 11 11
47_25 14 13 12
47_26 14 12 11
47_27 14 11 11
47_28 12 11 11
84

47_29 13 10 11
47_30 12 10 10
47_31 11 8 8
47_32 10 11 9
47_33 11 9 10
47_34 10 9 10
47_35 10 10 9
47_36 11 11 9
47_37 11 9 11
47_38 13 10 14
47_39 14 12 11
47_40 12 11 11
47_41 12 11 13
47_42 13 15 9
47_43 13 11 10
47_44 12 11 11
47_45 12 11 11
47_46 12 11 10
47_47 12 11 12
47_48 12 11 11
47_49 12 11 11
47_50 14 12 11
47_51 14 13 12
47_52 14 13 12
47_53 14 12 12
47_54 13 11 10
47_55 11 10 9
47_56 11 10 8
47_57 10 9 8
47_58 11 9 14
47_59 11 9 9
47_60 11 10 8
47_61 11 9 8
47_62 11 10 8
85

47_63 10 9 7
47_64 10 12 9
47_65 10 8 11
47_66 9 8 9
47_67 10 8 9
47_68 10 9 10
47_69 9 9 7
47_70 8 7 10
47_71 9 9 9
47_72 10 10 9
47_73 12 12 10
47_74 13 11 11
47_75 12 11 11
47_76 12 12 12
47_77 12 14 14
47_78 14 15 12
47_79 14 13 12
47_80 14 13 12
47_81 12 12 12
47_82 13 11 10
47_83 12 11 11
47_84 12 12 11
47_85 12 10 10
47_86 11 10 11
47_87 11 9 13
47_88 11 9 11
47_89 10 10 12
47_90 10 9 12
47_91 8 11 9
47_92 8 7 6
47_93 10 7 8
47_94 12 10 12
47_95 12 11 12
47_96 12 13 13
86

47_97 12 13 13
89_1 9 8 7
89_2 10 8 8
89_3 10 8 6
89_4 12 5 11
89_5 13 6 14
89_6 9 8 5
89_7 8 8 7
89_8 10 10 10
89_9 11 10 9
89_10 10 10 9
89_11 10 9 8
89_12 10 9 8
89_13 11 9 8
89_14 11 9 9
89_15 11 9 9
89_16 9 8 11
89_17 11 8 12
89_18 11 9 10
89_19 10 8 9
89_20 10 8 9
89_21 9 7 8
89_22 9 8 8
89_23 8 7 6
89_24 9 7 7
89_25 8 7 5
89_26 8 7 9
89_27 7 6 6
89_29 10 8 6
89_29 9 7 6
89_30 12 10 9
89_31 10 8 7
89_32 8 7 8
89_33 8 8 7
87

89_34 8 8 5
89_35 9 9 9
89_36 8 8 8
89_37 8 8 7
89_38 6 5 5

Table A.5 Combined Score Comparison Table for Route 47 and 89
Half-Mile-Radii Route-Adjacent Stop-and-Route-
Adjacent
Stop Route 47 Route 89 Route 47 Route 89 Route 47 Route 89
1 11 9 10 8 10 7
2 11 10 9 8 8 8
3 10 10 10 8 10 6
4 10 12 10 5 9 11
5 9 13 10 6 8 14
6 10 9 10 8 10 5
7 12 8 11 8 11 7
8 13 10 10 10 11 10
9 13 11 10 10 11 9
10 14 10 11 10 10 9
11 14 10 11 9 10 8
12 14 10 11 9 10 8
13 14 11 11 9 11 8
14 14 11 12 9 11 9
15 14 11 12 9 11 9
16 14 9 12 8 10 11
17 14 11 12 8 12 12
18 13 11 12 9 11 10
19 11 10 11 8 10 9
20 10 10 10 8 8 9
21 9 9 10 7 8 8
22 8 9 7 8 8 8
23 9 8 8 7 11 6
24 13 9 11 7 11 7
25 14 8 13 7 12 5
26 14 8 12 7 11 9
27 14 7 11 6 11 6
88

28 12 10 11 8 11 6
29 13 9 10 7 11 6
30 12 12 10 10 10 9
31 11 10 8 8 8 7
32 10 8 11 7 9 8
33 11 8 9 8 10 7
34 10 8 9 8 10 5
35 10 9 10 9 9 9
36 11 8 11 8 9 8
37 11 8 9 8 11 7
38 13 6 10 5 14 5
39 14 N/A 12 N/A 11 N/A
40 12 N/A 11 N/A 11 N/A
41 12 N/A 11 N/A 13 N/A
42 13 N/A 15 N/A 9 N/A
43 13 N/A 11 N/A 10 N/A
44 12 N/A 11 N/A 11 N/A
45 12 N/A 11 N/A 11 N/A
46 12 N/A 11 N/A 10 N/A
47 12 N/A 11 N/A 12 N/A
48 12 N/A 11 N/A 11 N/A
49 12 N/A 11 N/A 11 N/A
50 14 N/A 12 N/A 11 N/A
51 14 N/A 13 N/A 12 N/A
52 14 N/A 13 N/A 12 N/A
53 14 N/A 12 N/A 12 N/A
54 13 N/A 11 N/A 10 N/A
55 11 N/A 10 N/A 9 N/A
56 11 N/A 10 N/A 8 N/A
57 10 N/A 9 N/A 8 N/A
58 11 N/A 9 N/A 14 N/A
59 11 N/A 9 N/A 9 N/A
60 11 N/A 10 N/A 8 N/A
61 11 N/A 9 N/A 8 N/A
62 11 N/A 10 N/A 8 N/A
63 10 N/A 0 N/A 7 N/A
64 10 N/A 12 N/A 9 N/A
65 10 N/A 8 N/A 11 N/A
89

66 9 N/A 8 N/A 9 N/A
67 10 N/A 8 N/A 9 N/A
68 10 N/A 9 N/A 10 N/A
69 9 N/A 9 N/A 7 N/A
70 8 N/A 7 N/A 10 N/A
71 9 N/A 9 N/A 9 N/A
72 10 N/A 10 N/A 9 N/A
73 12 N/A 12 N/A 10 N/A
74 13 N/A 0 N/A 11 N/A
75 12 N/A 11 N/A 11 N/A
76 12 N/A 12 N/A 12 N/A
77 12 N/A 14 N/A 14 N/A
78 14 N/A 15 N/A 12 N/A
79 14 N/A 13 N/A 12 N/A
80 14 N/A 13 N/A 12 N/A
81 12 N/A 12 N/A 12 N/A
82 13 N/A 11 N/A 10 N/A
83 12 N/A 11 N/A 11 N/A
84 12 N/A 12 N/A 11 N/A
85 12 N/A 10 N/A 10 N/A
86 11 N/A 10 N/A 11 N/A
87 11 N/A 9 N/A 13 N/A
88 11 N/A 9 N/A 11 N/A
89 10 N/A 10 N/A 12 N/A
90 10 N/A 9 N/A 12 N/A
91 8 N/A 11 N/A 9 N/A
92 8 N/A 7 N/A 6 N/A
93 10 N/A 7 N/A 8 N/A
94 12 N/A 10 N/A 12 N/A
95 12 N/A 11 N/A 12 N/A
96 12 N/A 13 N/A 13 N/A
97 12 N/A 13 N/A 13 N/A

90

Table A.6 Tree Canopy Score Comparison Table for Route 47 and 89
Half-Mile-Radii Route-Adjacent Stop-and-Route-
Adjacent
Stop Route 47 Route 89 Route 47 Route 89 Route 47 Route 89
1 1 2 1 1 1 1
2 1 3 1 4 1 1
3 1 3 1 4 1 1
4 1 3 1 1 1 1
5 1 3 1 1 1 1
6 1 3 1 3 1 1
7 1 4 1 4 1 3
8 1 4 1 4 1 4
9 1 4 1 4 1 4
10 1 2 1 3 1 1
11 1 1 1 1 1 1
12 1 1 1 1 1 1
13 1 1 1 1 1 1
14 1 1 1 1 1 1
15 1 1 1 1 1 1
16 1 1 1 1 1 1
17 1 1 1 1 1 1
18 1 1 1 1 1 1
19 1 1 1 1 1 1
20 1 1 1 1 1 1
21 1 1 1 1 1 1
22 1 1 1 1 1 1
23 1 1 1 1 1 1
24 1 1 1 1 1 1
25 1 1 1 1 1 1
26 1 1 1 1 1 1
27 1 2 1 2 1 1
28 1 3 1 3 1 1
29 1 2 1 2 1 1
91

30 1 4 1 4 1 1
31 1 4 1 4 1 3
32 1 4 1 4 1 3
33 1 5 1 5 1 2
34 1 5 1 5 1 1
35 1 5 1 5 1 5
36 1 5 1 5 1 5
37 1 5 1 5 1 4
38 1 2 1 2 1 2
39 1 N/A 1 N/A 1 N/A
40 1 N/A 1 N/A 1 N/A
41 1 N/A 1 N/A 1 N/A
42 1 N/A 1 N/A 1 N/A
43 1 N/A 1 N/A 1 N/A
44 1 N/A 1 N/A 1 N/A
45 1 N/A 1 N/A 1 N/A
46 1 N/A 1 N/A 1 N/A
47 1 N/A 1 N/A 1 N/A
48 1 N/A 1 N/A 1 N/A
49 1 N/A 1 N/A 1 N/A
50 1 N/A 1 N/A 1 N/A
51 1 N/A 1 N/A 1 N/A
52 1 N/A 1 N/A 1 N/A
53 1 N/A 1 N/A 1 N/A
54 1 N/A 1 N/A 1 N/A
55 1 N/A 1 N/A 2 N/A
56 1 N/A 1 N/A 1 N/A
57 1 N/A 1 N/A 1 N/A
58 1 N/A 1 N/A 1 N/A
59 1 N/A 1 N/A 1 N/A
60 1 N/A 1 N/A 1 N/A
61 1 N/A 1 N/A 1 N/A
62 1 N/A 1 N/A 1 N/A
63 1 N/A 1 N/A 1 N/A
92

64 1 N/A 1 N/A 1 N/A
65 1 N/A 1 N/A 1 N/A
66 1 N/A 1 N/A 1 N/A
67 1 N/A 1 N/A 1 N/A
68 1 N/A 1 N/A 1 N/A
69 1 N/A 1 N/A 1 N/A
70 1 N/A 1 N/A 1 N/A
71 1 N/A 1 N/A 1 N/A
72 1 N/A 1 N/A 1 N/A
73 1 N/A 1 N/A 1 N/A
74 1 N/A 1 N/A 1 N/A
75 1 N/A 1 N/A 1 N/A
76 1 N/A 1 N/A 1 N/A
77 1 N/A 1 N/A 1 N/A
78 1 N/A 1 N/A 1 N/A
79 1 N/A 1 N/A 1 N/A
80 1 N/A 1 N/A 1 N/A
81 1 N/A 1 N/A 1 N/A
82 1 N/A 1 N/A 1 N/A
83 1 N/A 1 N/A 1 N/A
84 1 N/A 1 N/A 1 N/A
85 1 N/A 1 N/A 1 N/A
86 1 N/A 1 N/A 1 N/A
87 1 N/A 1 N/A 1 N/A
88 1 N/A 1 N/A 1 N/A
89 1 N/A 1 N/A 1 N/A
90 1 N/A 1 N/A 1 N/A
91 1 N/A 1 N/A 1 N/A
92 1 N/A 1 N/A 1 N/A
93 1 N/A 1 N/A 1 N/A
94 1 N/A 1 N/A 1 N/A
95 1 N/A 1 N/A 1 N/A
96 1 N/A 1 N/A 1 N/A
97 1 N/A 1 N/A 1 N/A

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Smart growth and walkability affect on vehicle use and ownership

PDF

Community gardens for social capital: a site suitability analysis in Akron, Ohio

PDF

Spatial analysis of urban built environments and vehicle transit behavior

PDF

Walkability study for school accessibility: case study of the San Juan, Puerto Rico elementary schools

PDF

Mapping future population impacts caused by sea level rise in Huntington Beach and Newport Beach: comparing the cadastral-based dasymetric system to past dasymetric mapping methods

PDF

Finding food deserts: a study of food access measures in the Phoenix-Mesa urban area

PDF

Surface representations of rainfall at small extents: a study of rainfall mapping based on volunteered geographic information in Kona, Hawaii

PDF

A site suitability analysis for an inland port to service the ports of Los Angeles and Long Beach

PDF

Visualizing historic space through the integration of geographic information science in secondary school curriculums: a comparison of static versus dynamic methods

PDF

An analysis of racial disparity in the distribution of alcohol licenses and retailers in Orange County, California

PDF

Site location suitability analysis for a smart grid network

PDF

Investigating the association of historical preservation and neighborhood status in Detroit, 1970-2015

PDF

Generating bicyclist counts using volunteered and professional geographic information through a mobile application

PDF

Drawing better lines: comparing commissions to legislatures on compactness and coterminosity

PDF

Historical temperature trends in Los Angeles County, California

PDF

Housing affordability near Metro’s light rail – Expo Line & Gold Line

PDF

Student engagement capabilities in mobile GIS: A framework for mobile GIS education

PDF

Wrong way driving in San Antonio, Texas: a transportation route study using Network Analyst

PDF

The role of amenities in measuring park accessibility: a case study of Downey, California

PDF

Improving the communication of water allocation decisions using interactive maps

Asset Metadata

Creator Chen, Stephanie (author)

Core Title Investigating bus route walkability: comparative case study in Orange County, California

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 09/25/2012

Defense Date 08/30/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bus route planning,GIS,OAI-PMH Harvest,Orange County

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Vos, Robert O. (committee chair), Oda, Katsuhiko (Kirk) (committee member), Ruddell, Darren M. (committee member)

Creator Email stea06@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-98075

Unique identifier UC11290154

Identifier usctheses-c3-98075 (legacy record id)

Legacy Identifier etd-ChenStepha-1214.pdf

Dmrecord 98075

Document Type Thesis

Rights Chen, Stephanie

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA