University of Southern California Dissertations and Theses

Access to active play parks for youth segments in Alexandria, Virginia

(USC Thesis Other)

Access to active play parks for youth segments in Alexandria, Virginia

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Access to Active Play Parks for Youth Segments in Alexandria, Virginia

by

Alexander Fox

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)

August 2018

Copyright © 2018 by Alexander Fox

To my wife Erin, for putting up with me while I did this study

iv
Table of Contents
List of Figures ............................................................................................................................... vii
List of Tables ............................................................................................................................... viii
List of Equations ............................................................................................................................ ix
Acknowledgements ......................................................................................................................... x
Abstract .......................................................................................................................................... xi
Chapter 1 Introduction .................................................................................................................... 1
1.1. Definitions...........................................................................................................................1
1.2. Motivation ...........................................................................................................................2
1.3. Study Area ..........................................................................................................................3
1.4. Uses for this Study ..............................................................................................................5
1.5. Organizational Framework .................................................................................................7
Chapter 2 Related Work.................................................................................................................. 9
2.1. Accessibility ........................................................................................................................9
2.2. Park Accessibility .............................................................................................................10
2.3. Connections between Parks and Well-being .....................................................................12
2.4. Limitations of Walking Distances Selected in Previous Studies ......................................14
Chapter 3 Methods ........................................................................................................................ 17
3.1. Data Selection and Preparation .........................................................................................19
3.1.1. Census Block Shapefile ...........................................................................................20
3.1.2. Population Table ......................................................................................................20
3.1.3. Parcel Shapefile .......................................................................................................21
3.1.4. Building Shapefile ...................................................................................................21
3.1.5. Housing Unit Table ..................................................................................................22
3.1.6. Park and Road Shapefiles ........................................................................................22
v
3.2. Calculations and Processes ...............................................................................................23
3.2.1. Selecting Age-Appropriate Travel Distances ..........................................................24
3.2.2. Assigning Housing Unit Data to Parcels .................................................................24
3.2.3. Assigning Census Population Data to Parcels .........................................................25
3.2.4. Creation of Active Play Features .............................................................................28
3.2.5. Creation of Park Access Points ................................................................................29
3.2.6. Creating Park Service Areas ....................................................................................32
3.2.7. Calculating Park Load ..............................................................................................33
3.2.8. Calculating Accessible Active Park Acreage per Parcel .........................................34
3.2.9. Mapping Techniques ................................................................................................35
3.3. Conclusion ........................................................................................................................35
Chapter 4 Results .......................................................................................................................... 38
4.1. Analyzing Service Areas...................................................................................................38
4.1.1. Youth Ages 0 to 4 and the 0.25-Mile Service Area .................................................39
4.1.2. Youth Ages 5 to 9 and 0.5-Mile Service Area .........................................................41
4.1.3. Youth Ages 10 to 17 and One-Mile Service Area ...................................................43
4.2. Areas with No Active Park Accessibility .........................................................................45
4.3. Analyzing Park Congestion ..............................................................................................48
4.4. Conclusion ........................................................................................................................50
Chapter 5 Discussion and Conclusion .......................................................................................... 52
5.1. Lessons Learned................................................................................................................52
5.1.1. Lessons Learned from Dasymetric Mapping ...........................................................52
5.1.2. Lessons Learned in Identifying Park Features and Access Points ...........................54
5.2. Limitations ........................................................................................................................55
5.3. Suggestions for Future Work ............................................................................................56
vi
5.4. Future Applications ...........................................................................................................58
5.5. Overall Conclusions ..........................................................................................................59
References ..................................................................................................................................... 61
Appendix A ................................................................................................................................... 67

vii
List of Figures
Figure 1 Alexandria, Virginia Location .......................................................................................... 3
Figure 2 Alexandria Park Size Histogram ...................................................................................... 4
Figure 3 Alexandria, Virginia Parks ............................................................................................... 5
Figure 4 Youth Population Density in 2015 by Census Block Group in Alexandria, VA ........... 18
Figure 5 Model Showing Workflow for Calculating Population per Parcel ................................ 27
Figure 6 Total Youth Population Density in Alexandria Virginia based on Residential Parcels . 28
Figure 7 Map of Access Points at Beach Park .............................................................................. 30
Figure 8 Access Point Types at the African American Heritage Park .......................................... 31
Figure 9 Active Parks within 0.5-mile of Alexandria ................................................................... 36
Figure 10 Accessible Active Park Acres per Youth Aged Zero to Four within 0.25-Mile .......... 40
Figure 11 Histogram of Accessible Active Park Acres for Youth Ages Zero to Four ................. 41
Figure 12 Accessible Active Park Acres per Youth Aged Five to Nine within 0.5-Mile ............. 42
Figure 13 Histogram of Accessible Active Park Acres for Youth Ages Five to Nine ................. 43
Figure 14 Accessible Active Park Acres per Youth Aged Ten to Seventeen within One-Mile ... 44
Figure 15 Histogram of Accessible Active Park Acres for Youth Ages Ten to Seventeen.......... 45
Figure 16 Residential Parcels with No Age-Appropriate Active Park Access ............................. 46
Figure 17 Residential Parcels with No Age-Appropriate Park Access for Youth Ages Zero to
Four ............................................................................................................................................... 47
Figure 18 Park Congestion for Youth Ages Zero to Four ............................................................ 49
Figure 19 Park Congestion in Downtown for Youth Ages Zero to Four...................................... 50
Figure 20 Closeup of Buildings in Multiple Parcels ..................................................................... 53
Figure 21 No Active Park Access for Youth Aged Five to Nine .................................................. 67
Figure 22 No Active Park Access for Youth Aged Ten to Seventeen .......................................... 68
Figure 23 Park Congestion for Youth Ages Five to Nine ............................................................. 69
Figure 24 Park Congestion for Youth Ages Ten to Seventeen ..................................................... 70

viii
List of Tables
Table 1 Source Data and Variable Fulfilled ................................................................................. 19
Table 2 Age-Appropriate Active Play Features by Age Group .................................................... 23

ix
List of Equations
Equation 1 Calculating Expected Population per Parcel .............................................................. 26
Equation 2 Calculating Expected Youth Population Rate per Housing Unit ............................... 26
Equation 3 Calculating the Expected Youth Population per Parcel ............................................. 27
Equation 4 Calculating Youth per Park Acre ............................................................................... 33
Equation 5 Calculating Accessible Active Park Acres per Expected Youth ................................ 34

x
Acknowledgements
I am grateful to my mentor, Professor Vos, for the direction I needed and my other faculty who
assisted me when I needed it. I would like to thank my wife, Erin, who helped with the editing of
the study.

xi
Abstract
Park accessibility is important for city planners because the accessibility of parks can impact
people throughout the community. Youth park accessibility is especially important, as parks
positively impact physical, emotional, and social development. This study uses dasymetric
mapping of census block group population data to estimate segments of youth population at each
residential parcel, and then associates those segments with age-appropriate active play features at
each park. Network analysis connects parcels to parks and their amenities, providing a more
precise accessibility rating at the city-level than studies based solely on geodesic buffers from
park centroids.
This study shows that while Alexandria, Virginia has many parks throughout the city, the
distribution of age-appropriate active play features is not uniform. Most children in Alexandria
have access to at least one active-play park. Only 132 parcels have zero access to age-
appropriate, active-play parks, a rate of less than one-hundredth of a percent. There are areas for
improvement, but the City of Alexandria has done an excellent job ensuring children have access
to active play parks. For other cities, this sort of accessibility analysis could help planners to
target areas to increase funding for fitness amenities and programs within parks, establish new
parks, or add pedestrian paths to improve walkability to existing park resources.

1

Chapter 1 Introduction
This study analyzes the active play park features in Alexandria, Virginia and their accessibility to
youth ages zero to seventeen. The study uses residential parcels as a more accurate estimation of
population distribution than the customarily used census tract of census block group. The higher
level of detail enables a more accurate assessment of park accessibility for youths walking along
the road network.
1.1. Definitions
Accessibility is generally defined as how well people can travel to a type of location.
Paez et al. (2012) define accessibility as, “the potential for reaching spatially distributed
opportunities.” Paez et al. divides accessibility into two components, travel cost and quality of
opportunities. There are two ways to analyze accessibility, from the origin and to the destination.
Examples of accessibility from the origin include the number of supermarkets within one mile of
the population. Accessibility to the destination includes the population within five miles serviced
by a hospital.
Well-being is used in many ways, depending on the field of study. The Centers for
Disease Control and Prevention defines well-being in its simplest terms as, “judging life
positively and feeling good” (Centers for Disease Control and Prevention 2016). They further
recognize the primary aspects of well-being researched as physical, economic, social, emotional,
psychological, life satisfaction, development and activity, and engaging activities and work.
Much of the previous study centering around parks and children assess the physical, social, and
developmental aspects of well-being.
Pauleit et al. (2003) define greenspace to include woodlands, farmlands, parks, playing
fields, open spaces, playgrounds, and gardens. For the purpose of this study, greenspace is

2

defined as publicly maintained parks, playgrounds, and playing fields with no admission criteria
beyond hours of operation.
1.2. Motivation
Parks provide many positive impacts to society in general and specifically children. Parks
provide opportunities for play, exercise, and social development. Play is vital in children’s
physical, cognitive, social and emotional development (Little and Wyver 2008). Social
interaction between children increases with use of sports and outdoor environments. The
opportunities to meet and interact with other children promotes social development, face-to-face
communication skills, and making friends. (Seeland, Dübendorfer and Hansmann 2009). Formal
park activities combined with the presence of other active children results in an increased level
of physical activity (Floyd, et al. 2011).
Public parks and green space provide locations for physical activity through play or
exercise. This study focuses on youth access to parks. Youth accessibility is different from adult
accessibility. Adults in this metropolitan area may have access to automobile transportation.
When traveling alone, children are limited to walking or biking distances. As such, youth
accessibility is less than that of adults. Also, adults may have the option of moving to another
area if they wish to increase their accessibility, while children are limited in that their parents
decide where the family should live.
Identifying areas that have low accessibility to parks for youth can help city planners
either develop new parks or encourage other well-being programs for those areas between parks.
Parks are an excellent location for the development of physical and social skills as well as
improve emotional well-being. If a park is not available to the youth of an area, other programs

3

or efforts should be developed to target that area to make up for the lost opportunity represented
by the lack of an active play park.
1.3. Study Area
The city of Alexandria, Virginia is directly south of the United States’ capital and is
home to an estimated 150,000 residents. The city has over 900 acres of protected open space and
over 560 acres of city-owned parks out of 9,920 acres, a rate of 9%. Figure 1 below shows the
location of Alexandria in relation to the state of Virginia and its capital, Richmond. Alexandria is
bounded to the east by the Potomac River, to the south by Interstate 495, to the west by the city
of Annandale, and the north by the city of Arlington.
Figure 1 Alexandria, Virginia Location

(Maps of World n.d.)
Alexandria has 103 parks listed on their website. The parks range in size from 0.05-acres
to 65-acres. Typical features include basketball, tennis, and volleyball courts, seating areas, dog
areas, picnic tables, playgrounds, athletic fields, skateboard parks, and swimming pools. The city
boasts of 49 multi-use athletic fields, 36 playgrounds, 36 tennis courts, and four pools. (City of

4

Alexandria, Virginia 2017). Only one park is not included in the GIS Department’s data. Figure
2 below shows that 81% of parks are less than 9.75 acres. In fact, 49% of parks are less than
2.25-acres. These many small parks give the opportunity to have equitable distribution
throughout the city. There are still a significant number of parks with over 40-acres. These large
parks are more difficult to distribute throughout the city
Figure 2 Alexandria Park Size Histogram

Figure 3 below shows that although there is a multitude of parks and other green spaces
in the city, the parks are not uniformly distributed. This can lead to unequal access for the
residents. Analyzing the access to green space helps planners determine the need for new parks
or expansion of public transportation to make the current parks more accessible to residents.
Adding further unequal access is the difference in active play features. Also, Figure 3 might
appear to have an adequate distribution of parks, but it does not take active play features into
account.

5

Figure 3 Alexandria, Virginia Parks

Parks with different sizes are not evenly distributed across the city. The more extensive
parks are located mainly in the western part of the city with a couple of exceptions along the
waterfront to the east. The western half of the city also has a different structure to its road
network than in the east. In the east, a road grid is standard. The road network in the western
portion of the city has a higher percentage of cul-de-sacs and has fewer road intersections. This
can affect accessibility by limiting the available road branches within a specified distance.
1.4. Uses for this Study
Current academic green space studies focus on many areas throughout the world, but no
published study focuses on Alexandria, Virginia. Building an accessibility model can help

6

provide planners with the data and information necessary to support new public works projects to
expand, improve, or provide easier public transportation to the park system.
It is unrealistic to expect the city of Alexandria to purchase large amounts of existing
property to change the use of the parcels to a new park, as it is prohibitively expensive. Instead,
by focusing on opportunities for physical activity within parks, this study might serve as
evidence to provide a targeted audience for fitness marketing campaigns. Knowing which areas
have more park access allows the city to focus on efforts in the areas with reduced access to
parks. These efforts could be the NFL’s Play60, Alexandria’s My Gym, or the Presidential
Youth Fitness Program. Such efforts encourage kids to become active and keep moving.
Focusing these programs on areas with less access to parks could provide more equitable access
to fitness and contribute to youth well-being.
Another alternative to purchasing new, large parks is micro-parks. The city has several
“tot lots” that are less than 0.5-acres and simply have a playground. Plots of land like alleys,
small empty spaces between larger parcels, and parking lots can provide access to active play
parks for new areas, especially for physical play for the youngest age groups, with less
investment requirements than developing bigger, more traditional parks (Nordh et al. 2009).
Finally, even without developing new park spaces, adding age-appropriate, active-play
features are something that the city can control at a reasonable cost. By installing age-appropriate
features in areas with no access to those features, the city can provide the opportunity for parents
to take their children to a park where the children can be encouraged to play actively. Playing
fields, skateboard parks, and playgrounds are appropriate for children and youth at different
stages of development and should be analyzed as appropriate for specific age and play/physical
activity segments. Breaking the population and parks into different age bands and their

7

associated active play parks allows for the use of different approaches to more accurately analyze
the distribution of age-appropriate active play features. This study also identifies those parks
with active play features that have a significantly higher number of expected users within their
service area. This potential overcrowding of active play features may reduce the positive effect
of accessible active play parks. If the parks are always crowded, it can discourage youth and their
parents from using the park.
Several studies have shown that distances between one-third and one-quarter of a mile
are “walkable.” Dunton et al. and Wolch et al. use 500-meters (0.31-mile) as a travel distance to
define what parks are within a child’s proximity (Dunton et al. 2014; Wolch et al. 2011). These
studies view travel distances as sensitive to a child’s age, looking at three travel distances, 0.25-,
0.5-, and 1-mile to determine the variable’s sensitivity. The study breaks down the population by
appropriate age groups for those travel distances with parents of older children being expected to
be willing allow or assist their children in traveling long distances to a park. This study looks at
how much accessibility is improved by increasing travel distance as youths age.
Obesity is a growing problem for youth in the United States. While there are many
factors, one factor that municipalities can affect is access to parks for active play. Dr. Heidi
Blanck shows that parks and playgrounds provide many benefits for youths. The social aspect is
helpful to development as well as the physical. Multiple research studies show that access to
parks is directly related to activity and inversely related to rates of obesity and overweight
population. (Blanck et al. 2012; Cohen et al. 2007; Jennings et al. 2016).
1.5. Organizational Framework
This study is organized into five chapters. Chapter 2 reviews available studies and peer-
reviewed articles to discuss standard techniques for analyzing accessibility. This study’s

8

methodology was based on the benefits and disadvantages of these methods. Chapter 3 describes
the methodology of assigning appropriate population and housing data to parcels, creating active
play park features, building park service areas, and calculating park accessibility and park
congestion. Chapter 4 discusses the results of the study and presents the outcomes and identify
strengths and weaknesses of this methodology. Chapter 5 examines the implications of this
method and identifies possible directions for future research.

9
Chapter 2 Related Work
Measurement of physical accessibility of populations to facilities, and particularly park
accessibility, has been studied extensively throughout the world. Using GIS allows for a more
detailed analysis by quickly developing service areas and providing more detailed population
distributions. Conventional approaches include applying a standard service area distance, either
through Euclidean or network approaches, and treating all parks as equally desirable.
This chapter discusses accessibility in general, park accessibility, the connection between
parks and well-being, and limitations of previous works. This study’s differentiation is discussed
in each of the sections. Unlike many other studies, it uses dasymetric mapping at the residential
parcel level and separates age groups by different distances and park amenities.
2.1. Accessibility
Many accessibility studies focus on large, rural areas. This was done to analyze access to
limited resources. A frequent topic is the access to healthcare facilities which affect the well-
being of distributed populations. Multiple studies focus on access to health services in extremely
rural areas. These areas include New Zealand, the Philippines, and Bhutan. (Bagheri, Holt and
Benwell 2009; Delgado and Canters 2011; Jamtsho, Corner and Dewan 2015). Most of these
studies rely on vehicles for transportation. Not many studies use walking as the primary mode of
transportation. While such general observations are intuitive, evidence based on carefully
analyzed spatial data provides a stronger argument and details areas with the most extreme
accessibility challenges.
Euclidean or geodesic distance is often used for accessibility. However, Pedigo and Odoi
(2010) found that network analysis is more accurate and better suited for determining
accessibility. This is especially true for rural areas where road networks are the only way of

10
moving. In a dense urban environment, network analysis can also be a useful means of
estimating travel distances, but may not always be the most accurate method, especially where
walking is being assessed. The GIS road network may not include shortcuts that residents use
when walking. These shortcuts include alleys, parking lots, and cutting through other’s property.
In a dense urban environment, network analysis is also stronger when roads are assigned
different speed limits or have data that considers traffic congestion.
2.2. Park Accessibility
Socio-economic factors directly affect park accessibility. Wang et al. (2015) show that
low-income groups have lower access to parks than more affluent residents in both Brisbane,
Australia and Zhongshan, China. According to research literature, this is broadly true in the
United States as well (Rigolon and Flohr 2014; Jennings and Gaither, 2015). Part of this is due to
property valuations and taxes. Wolch et al. (2005) shows that high-income areas have larger park
systems compared to low-income areas. There are some studies with detailed and to some degree
contradictory findings, including Boone et al. (2009) who studied Baltimore, Maryland. This
study showed that African Americans had better walking access to parks than white residents,
but fewer acres compared to whites due to the spatial pattern and timing of suburbanization.
They also studied needs-based assessment, which focused on children, the elderly, the carless,
and low-income neighborhoods. They found that areas with more high-need people in the
population had a lower mean distance of 239-meters (0.15-miles) to the nearest park than areas
with fewer high-needs residents, where residents had a mean distance of 864-meters (0.54-
miles). These areas have high park congestion, but good accessibility (Boone et al. 2009).
Comber et al. (2008) built a GIS database to analyze greenspace accessibility based on
ethnicity and religion in Leicester, England. One of the key exclusions highlighted by Comber et

11
al. is that of school playing fields and golf courses. Since these are not open to the public, it
makes sense to separate out these green spaces. Another idea to consider is the insertion of
access points in the network analysis to determine travel times. Rather than just walking into the
park at any point along the perimeter, this would normalize where the entrances are located
(Comber et al. 2008). Alessandro Rigolon and Travis Flohr research the effect of economic
divisions and access to parks in youth. They focus their attention on the Denver area, where there
are different income and racial backgrounds. Their study concludes that the low-income
neighborhoods have the lowest access to parks. When amenities are included as variables, the
difference in access between low- and high-income areas is even more pronounced (Rigolon and
Flohr 2014). Gary Higgs developed a GIS framework to analyze access to public sporting
locations in Wales (Higgs et al. 2015).
Tijs Neutens writes two papers that focus on how traffic congestion affects accessibility.
The time penalty for travel by vehicle is different at different times of day, thus affecting
accessibility (Neutens et al. 2014). Hours of operation might affect accessibility (Neutens et al.
2010). While these studies focus on government offices in Ghent, Belgium, they could also
affect the parks of Alexandria, Virginia. Since most of the parks in Alexandria are open sunrise
to sunset, this could influence accessibility for residents.
Many studies focus on regions, large cities, or even nationally. Using Euclidean distances
between census tracts and park centroids, Zhang et al. (2011) analyzed the entire United States
using a container approach. They also used a weighted average of the nearest seven parks to
account for resident choice. Residents may not always choose the closest park; their choice often
revolves around amenities. Zhang et al. (2011) selected the most frequently chosen technique of
using census tracts and park centroids. This approach is efficient for work at a nationwide scale,

12
but for a city such as Alexandria, this method would not have sufficient detail to provide useful
information to city officials.
2.3. Connections between Parks and Well-being
Access to parks increases the likelihood of physical activity. Kaczynski et al. (2009)
showed that increased park size corresponded positively with odds of adults conducting 150-
minutes of moderate-to-strenuous physical activity (MSPA) for one week in a mid-sized
Canadian city. An increase in available park size led to an increase of 2% in the odds of
conducting 150-minutes of MSPA. However, additional parks within one-kilometer (0.6-mile)
increase the odds of MSPA by 17%. They analyzed proximity to the nearest park, number of
parks within 1-kilometer, and total park area to determine if there was a significant predictor of
MSPA. The study states that only the number of parks within one-kilometer was a significant
predictor of MSPA among residents. They identified other variables that were predictors of high
MSPA including children in the household and residents above 55-years of age. Park
accessibility was not a significant predictor for the age group of 35-54, as a higher percentage of
the age group worked and traveled outside of their immediate neighborhood. Their study
collected data through surveys and did not attempt to determine causality, only correlation. Other
socio-economic factors beyond age were not analyzed (Kaczynski et al. 2009)
Andrew Oftedal and Ingrid Schneider state that while many studies have been conducted,
there is not always a strong positive relationship between outdoor recreation and physical health.
In their study in Minnesota, they found that the number of recreation opportunities was more
consistently related to health than per capita opportunities (Oftedal and Schneider 2013). This
means that the opportunity to exercise at a park is more important than the congestion of the
parks.

13
Larson et al. (2016) show that the percentage of parkland within in a community is a
strong predictor of overall and physical well-being. They studied the physical, social,
community, financial, and purpose components of well-being to develop an overall Gallup-
Healthways well-being index. While they used the percent of the population within 0.5-miles of
a park, they did not conduct a GIS analysis to determine this population ratio. The study drew
from the Trust for Public Land’s Park Score Index (Larson et al. 2016). The Trust for Public
Land’s website provides the rationale for their 0.5-mile delineation, but does not discuss their
methodology for calculating the population percentage living within 0.5-miles of a park (The
Trust for Public Land n.d.)
Oftedal and Scheider highlight the study by Cohen et al. (2007) that shows that
approximately two-thirds of park users were sedentary, or of too low an intensity to provide
significant health benefits. Cohen et al. (2007) limited their description of park users’ effort to
sedentary, walking, and vigorous exercise. The authors’ observations and resident surveys
showed that people within 1-mile of a park were four times more likely to visit the park at least
weekly than those that lived further than one-mile from a park (Cohen et al. 2007).
This study also segregates parks based on their features. Each age group has an
associated group of parks that have age-appropriate active play features. This has not occurred in
previous literature. By breaking the parks apart based on age-appropriateness, the study does not
assign the accessibility of a teenager to a playground the same weight as the accessibility of a
teenager to a sports field.
Floyd et al. (2011) shows through direct observation higher levels of physical activity are
linked to courts and formal activities. Active play features like basketball and tennis courts are
associated with higher levels of activity than picnic tables. They also found that the type of

14
active play feature matters. Baseball and softball fields resulted in lower energy expenditures
compared to basketball and tennis courts. The authors did find some differences in age groups,
with preschool-age children preferring more spontaneous play than older children.
Little and Wyver (2008) identify that in a world of decreasing outdoor play driven by
concern for children’s safety, parks provide a good area for children to engage in risk-taking in a
relatively controlled environment which allows children to gain confidence, refine locomotive
skills, and understand themselves and others. They also suggest that parks can promote life-long
physical activity in pursuit of an active, healthy lifestyle. This contributes to obesity prevention.
Low movement skills can lead to lower self-esteem and fewer friends. Seeland et al. (2009)
concludes that parks promote social inclusion through communication and recreation in parks.
Seeing and interacting with other children promotes friendships and is a way to help promote
multicultural environments.
2.4. Limitations of Walking Distances Selected in Previous Studies
Identifying appropriate distances for children to walk to parks is little studied, but many
studies look at travel to and from school (Bejleri et al. 2011; Lopez and Wong 2017; Schlossberg
et al. 2006). These studies can give a general perspective on children’s travel distances.
Several studies specify the lack of supporting evidence on selecting distances. Common
distances selected for children walking are 0.5- and one-mile. Schlossberg et al. (2006) studied
various transportation methods for taking children to and from school. They utilize 0.5-mile
intervals from one- to 3.5-mile. Their rationale for beginning at one-mile is that the 1.5-mile
break is used by many school districts to delineate where bus service begins. (Schlossberg et al.
2006). Alexandria City Public Schools defines students who live within one-mile of their

15
elementary schools and those living within 1.5-mile from their secondary schools as walkers
(Alexandria City Public Schools n.d.).
School is a mandatory activity, but children are less likely to travel that far for day-to-day
entertainment or play. Schoeppe et al. (2016) surveyed adults in Queensland, Australia on their
perception of safe distances for eight to twelve-year-old children’s unsupervised travel and play.
The survey results showed that a majority, or 74%, of adults wanted to restrict eight to twelve-
year-old children independent play to less than five hundred meters, or approximately 0.3-mile.
An additional 14% were comfortable with the children playing within 0.3- to 0.6-mile and the
remaining 12% would allow distances of over 0.6-mile. (Schoeppe et al. 2016). This is
significantly less than the distances studied for children walking to school. A central difference is
that schools do not provide school buses for those children living within one- to 1.5-mile of their
school.
Several factors affect how adults perceive the safety of their children when traveling to
school. M. C. Lopez and Y. D. Wong determined that while distance was the primary factor,
perceptions of neighborhood accessibility, connectivity, traffic safety, and personal safety can
have a significant effect on children walking to school. A neighborhood that is perceived as
accessible, with quality connectivity (pathways), and in low traffic and low crime area would
normally have more children whose parents allow them to walk to school. (Lopez and Wong
2017). Schlossberg et al. (2006) found that traffic danger, distance, and personal safety were not
the top reasons for driving children to school. The top three reasons for driving were that school
was on the way to work, the child’s backpack was too heavy, and bad weather. (Schlossberg et
al. 2006). The literature suggests that the decision of whether and how far to walk is complex,
and so it is not easy to give determine walking distances by relying on previous studies.

16
This study builds on previous work by segregating park accessibility by both age group
and distance. It is unreasonable to judge toddlers and teenagers by the same travel distance. For
this study, age groups have been assigned park transportation distances based on their age in the
age bands defined by the census. For newborns to children four years old, 0.25-mile is a
reasonable distance for parents to take their children to a park. For ages five to nine, 0.5-mile is
used. For youth from ten to seventeen, one-mile is used due to the increased likelihood of access
to bicycles and automobiles. Ten- to seventeen-year-olds also are more apt to engage in team
sports and would be willing to travel further to participate.

17
Chapter 3 Methods
This study analyzes the green space in Alexandria, Virginia and its accessibility to various socio-
economic segments, focused primarily on children up to age 17. The study provides a detailed
coverage-based method by computing the distance to the nearest park from all residential parcels
in the city. This provides a more accurate analysis of park accessibility based on residential
parcels than is typical in the literature, which is often based on census tracts as noted in Chapter
2 above.
Figure 4 below underscores this point by showing the population density of children aged
five to seventeen in Alexandria, Virginia at the census block level. The census blocks cover
many neighborhoods and do not provide an effective representation of where the children live. It
would be difficult to determine the accessibility of the park system when painting the city with
such a broad brush. Many census block groups have at least one park adjacent to or within its
boundaries. Other census block groups, often with small populations, are larger than the majority
of the parks in the city. Without estimating where people actually live within the census block
groups, it is impossible to accurately assess park access. By breaking down the population into
expected population per residential parcel, the study observes the different accessibilities as
population density increases or decreases near parks.

18
Figure 4 Youth Population Density in 2015 by Census Block Group in Alexandria, VA

Census block groups include many types of parcels including industrial and commercial
that do not have residents. These parcels can also provide a false sense of distance to parks.
Figure 4 above does not have sufficient detail to determine if the area around a park is home to
any children. Several parks in Alexandria have commercial or industrial lots adjacent to the park.
To overcome this issue, this study employs dasymetric mapping to estimate locations of youth by
distributing the city’s youth population across residential parcels. The study used ArcMap 10.5.1
as the GIS software.
The data and methods used to dasymetrically map the youth population and to calculate
access to parks based on walking networks are discussed in this chapter. The chapter is organized

19
into the following main sections: Data Selection and Preparation, Calculations and Processes and
Conclusion.
3.1. Data Selection and Preparation
Data for this study comes from the US Census Bureau and the City of Alexandria GIS
Department. This data is free for download through the internet. Population data and census
block group data was from 2015; parcel, building, and road data were from 2017; and housing
data was from 2015. This section describes the data; the following section discusses how the data
was processed to produce results. Table 1 below shows the feature classes used for this study,
their source, and the variable fulfilled by the feature class.
Table 1 Source Data and Variable Fulfilled
Feature Class Source Variable Fulfilled
2015 Census Block
Group Polygons
US Census Bureau (US Census
Bureau 2017)
Census Block Groups
2015 Population Data –
American Community
Survey 5-year estimates
by census block group
US Census Bureau (US Census
Bureau 2017)
Total Population, Population
by Age, Population by Sex
2017 Parcel Data City of Alexandria GIS (City of
Alexandria GIS Department 2017)
Residential Parcels
Building Polygons City of Alexandria GIS (City of
Alexandria GIS Department 2017)
Housing Units on Residential
Parcels
2015 Housing Data –
American Community
Survey 5-year estimates
US Census Bureau (US Census
Bureau 2017)
Number of Housing Units per
Census Block Group
Park Polygons City of Alexandria GIS (City of
Alexandria GIS Department 2015)
Park Location and Size
Road Line Segments City of Alexandria GIS (City of
Alexandria GIS Department 2017)
Road Network
Railroad Line Segments City of Alexandria GIS (City of
Alexandria GIS Department 2017)
Railroad Network

20
3.1.1. Census Block Shapefile
Data from the Census Bureau are commonly available at the census block group level.
This is the most precise geography at which the government breaks down population by age.
This served as the table to which all other data was related. Population data broken down by age
was linked to the census block shapefile. The study performed a simple selection by attribute of
the census block group shapefile based on the county number 510, which indicates the city of
Alexandria and clips the statewide block group shapefile to the study area.
3.1.2. Population Table
This study focused on park accessibility for youths. Based on the population breakdown
by age provided by the US Census Bureau, the age groups are divided for this study into Male 0-
4, Male 5-9, Male 10-14, Male 15-17, Female 0-4, Female 5-9, Female 10-14, Female 15-17, and
total population ages 0-17 years. Age groups 18 and over do not fall within the scope of this
study and were ignored.
The Census Bureau provided the age breakdown in its census block groups. The study
initially utilized census tracts but recalculated with census block groups as the data was more
precise for estimating the distribution of people to residential parcels. The data downloaded from
the Census Bureau was not formatted for importing into ArcMap. The data was reformatted with
field titles in the first row and block groups in the first column. Field titles were simplified and
combined in Microsoft Excel as per the following example: “Male 5 to 9 years” and “Female 5
to 9 years” became “T0_4.” The Excel file also had two rows of titles that were reduced to one
row for use in ArcMap.

21
3.1.3. Parcel Shapefile
The parcel shapefiles from the City of Alexandria GIS Department include information
about the location and owner of the parcel. It is the primary shapefile used for this study. All
other shapefiles and tables eventually relate to the parcel shapefile. The parcel data is from 2017.
It was the only available timeframe for the data. This mismatch in years between the Census data
and parcel data is not a significant source of error in the study, as very few parcels changed uses
in the two years since the census estimate. Unlike many cadastral datasets for other cities, the
parcel shapefile for Alexandria does not include an attribute for how many housing units are in
each parcel. Therefore, an alternate method of determining the number of housing units present
was developed. Spatially joining the parcel shapefile with a building shapefile from the GIS
Department brings the housing unit field to the parcel shapefile.
3.1.4. Building Shapefile
As mentioned above, this study makes use of spatial data provided by the City of
Alexandria GIS Department that is a polygon shapefile showing all buildings over one hundred
square feet in size and providing the number of housing units in each building. It also indicates
the building use. People live in residential and multi-use buildings, which were the relevant types
of building selected to estimate housing units per parcel for this study. The building data is from
2017. Ideally, the building shapefile would match the population data, but the GIS Department
only has the most recent data on their website. According to the United States Census Bureau’s
American Community Survey 5-year estimates in 2015 and 2016, the total number of housing
units in Alexandria Virginia only increased by 770-units or one percent of the city’s total.
Extrapolating this rate to 2017, a two percent difference in the number of housing units likely
does not introduce a significant level of error into the study.

22
3.1.5. Housing Unit Table
The Census Bureau website provided the estimated number of housing units within each
census block group using the American Community Survey (ACS) 5-year estimates. This
detailed housing report included the total number of housing units, owner versus rental
quantities, size of structures by housing unit, age, rooms, type of heating fuel, and value of the
housing unit. The only relevant field for this study was the total number of housing units. Data
from 2015 was selected in order match the year of the census block group population data and
shapefile.
Preparation of the housing data consisted of removing the second header row and
replacing periods with underscores for use in ArcMap. The relevant field code for the number of
housing units was “HD01_VC01.” This data was used to calculate expected youth population in
residential parcels.
Both the housing unit table and building shapefiles are necessary for this study because
either one alone does not provide the required information. The housing unit table does not show
how the housing units are distributed. The building shapefile allows the study to identify how
many housing units are on each parcel when spatially joined with the parcels. The housing unit
table only provides the total housing units in the census block.
3.1.6. Park and Road Shapefiles
The park and road shapefiles were also downloaded from the Alexandria GIS
Department. The park shapefile was manually updated with the park size from the Alexandria
Parks Department website. Each park was identified if it had active play features for further
comparison later. The park size was not adjusted for the size of the active play feature. This
would have required identifying the actual size of just the active play features and was beyond

23
the scope of this study. This does introduce some error into the calculations and will be further
discussed in Section 5.2. Table 2 below shows which active play features this study considers
age-appropriate. Active-play features for newborns to four-year-olds are playgrounds. Ages five
through nine add playing fields and swimming pools. Ages ten through seventeen have the most
active play feature types including basketball courts, playing fields, skateboard parks, swimming
pools, tennis courts, and volleyball courts.
Table 2 Age-Appropriate Active Play Features by Age Group
Age Group
Basketball
Court
Playground
Playing
Field
Skateboard
Park
Swimming
Pool
Tennis
Court
Volleyball
Court
0-4 X
5-9 X X X
10-17 X X X X X X

A series of Select by Attribute tools created shapefiles that included only active play
parks for each age group, from zero- to four-, five- to nine-, and ten- to eighteen-years-old.
The roads shapefile from the city GIS department did not include a field for sidewalks.
Pedestrians prefer to walk on sidewalks. A Selection by Attribute was run to select all roads with
a speed limit of thirty-five miles per hour or less. This new shapefile serves as a proxy for
pedestrian routes. Those roads with a speed limit of less than twenty-five miles per hour might
not have a sidewalk but are still considered part of the pedestrian transportation network.
3.2. Calculations and Processes
There are four main steps in the workflow for this study. Further preparation of the data
included assigning housing unit data to parcels, assigning census population data to parcels,

24
creation of active play features, and creation of park access points. These steps ensure that the
data is correctly formatted for analysis. Conducting network analysis creates service areas and
begins the analytical section of the study. Calculating park load and calculating available park
acreage per parcel complete the analytics to compare parks and parcels to determine accessibility
of age-appropriate active play park feature for youths in Alexandria, Virginia.
3.2.1. Selecting Age-Appropriate Travel Distances
Many studies use common, round numbers for travel distances. Some studies
acknowledge that these values were selected for ease of use without any reference to supporting
evidence. This study shares this weakness, as no definitive reference was found for the distance
bands that are standard in the walkability literature. The author’s life experience and logical
process resulted in using round numbers for travel distance. This may not be the most accurate
representation of how far a child can be expected to travel to access a park. It does seem
reasonable to assume that older children would walk or perhaps bike up to the 0.5-mile and one-
mile distances. The actual distances may vary from year to year, from region to region, and
within the age groups used for this study. For example, a seventeen-year-old typically has an
increased travel area due to the freedom given by parents as a result of their maturity and access
to automobiles than a ten-year-old. However, not every seventeen-year-old child has these
mobility enhancers. The distances used in this study should be suitable representations of how
far an average youth in the age group is likely to travel.
3.2.2. Assigning Housing Unit Data to Parcels
Since the parcel data provided by the City of Alexandria did not include the number of
housing units, steps were taken to calculate this data. The building shapefile from the City of
Alexandria did include the number of housing units in the building. The buildings were spatially

25
joined to the residential parcels to pass along the housing unit attribute. The parcels collected the
sum of the attributes of all buildings that intersected the parcel. This meant that if a parcel had
two five-unit buildings on it, the parcel would have the attribute of ten units. A small potential
error occurred if a building straddled two parcels. In this case, the units were counted twice. This
occurred infrequently and would not affect the results significantly. The result of this step is that
every residential parcel had an attribute with the number of housing units in the parcel.
3.2.3. Assigning Census Population Data to Parcels
There were five inputs required to calculate expected population per parcel; housing units
per census block group, census block groups, population data by census block group, housing
units per building, and parcels. It was unreasonable to expect exact knowledge of how many
people reside in each parcel in the city using only census data. While the ACS 5-year estimates
provide total housing units per census block, data from the city of Alexandria is more accurate.
The 5-year estimate is exactly that, an estimate. The GIS Department likely cooperates with the
permitting and building inspection processes to update their files more frequently than the
Census Bureau. The city will also have the distribution of housing units within the census block.
The 5-year estimate only provides a total number of housing units whereas this study requires the
number of housing units per parcel. To best simulate this, the total population in each census
block group was divided by the number of housing units in that block group and then multiplied
by the number of housing units on the parcel to predict the number of residents living on the
parcel as shown in Equation 1 below:

26
𝑃𝑜𝑝 𝑝 =
𝑃𝑜𝑝 𝐵𝐺
𝐻𝑈
𝐵𝐺
∗ 𝐻𝑈
𝑝
Equation 1 Calculating Expected Population per Parcel
𝑃𝑜𝑝 𝑝 is shown to be equal to 𝑃𝑜𝑝 𝐵𝐺
, the total population in the block group, multiplied
by 𝐻𝑈
𝑃 , the total number of housing units in the parcel, divided by the number of housing units
in the block group, 𝐻𝑈
𝐵𝐺
. The next step is calculating the expected youth population per parcel
using Equation 2:

𝑃𝑜𝑝 𝑌𝑟
=
(𝑃𝑜𝑝
𝑀 0−4
+𝑃𝑜𝑝 𝑀 5−9
+𝑃𝑜𝑝
𝑀 10−14
+𝑃𝑜𝑝
𝑀 15−17
+𝑃𝑜𝑝
𝐹 0−4
+𝑃𝑜𝑝
𝐹 5−9
+𝑃𝑜𝑝 𝐹 10−14
+𝑃𝑜𝑝
𝐹 15−17
)
𝐻𝑈
𝐵𝐺

Equation 2 Calculating Expected Youth Population Rate per Housing Unit
The rate of youth population in the parcel, 𝑃𝑜𝑝 𝑌𝑟
, is equal to the sum of the male and
female populations of youths ages 0-4, 5-9, 10-14, and 15-17 in the census block divided by the
number of housing units in the block group, 𝐻𝑈
𝐵𝐺
. The census block group shapefile was joined
to the population data by the block group identifier. The census block group shapefile was then
spatially joined to the parcel shapefile to transfer the rates to the parcels. If a parcel was in
multiple block groups, such as a large apartment complex, the average of the rates was stored in
the attribute.
A similar process used a portion of Equation 2 to calculate the expected youth in each age
group on the parcel. For example, adding the male and female total population for children ages
zero to four and dividing that sum by the number of housing units in the block group and
multiplied by the number of housing units in the parcel resulted in the expected youth population
ages zero to four in that parcel.

27
𝑃𝑜𝑝 𝑌𝑃
= 𝑃𝑜𝑝 𝑌𝑟
∗ 𝐻𝑈
𝑝
Equation 3 Calculating the Expected Youth Population per Parcel
To determine the final expected youth population in the parcel, Equation 3 was used. The
expected rate of youth population, 𝑃𝑜𝑝 𝑌 𝑟 , was multiplied by the number of housing units on the
parcel as calculated in Section 3.2.2 above, 𝐻𝑈
𝑝 . This process was followed for each age group
as well as all youths.
The process for calculating expected youth population per parcel in ArcMap is shown in
Figure 5 below. Housing data and population data were joined to the census block group
shapefile. Buildings were spatially joined to parcels to assign each building a parcel. Only
residential and multi-use buildings were selected for inclusion in the next steps. The population
data was then joined to the parcel data. The expected population was calculated as shown in
Equation 1 and Equation 2 above.
Figure 5 Model Showing Workflow for Calculating Population per Parcel

Figure 6 shows the population density of all youths under seventeen years old. Several
parcels stand out. These are large apartment complexes, primarily in the west part of the city
with four complexes in the northeast section.

28
Figure 6 Total Youth Population Density in Alexandria Virginia based on Residential Parcels

3.2.4. Creation of Active Play Features
To analyze how accessible parks are based on their age-appropriate, active-play features,
they must be classified according to active play. Floyd et al. (2011) show that children from
zero- to twelve- years-old are significantly more likely to use playgrounds and older adolescents
use amenities dedicated to sports. Table 2 above shows age-appropriate active play features for
this study. The presence or absence of these active play features in given parks was determined
through the City of Alexandria’s park information website. A field was created in the parks
shapefile to indicate the presence of an active play feature. Separate shapefiles were created for
each age group with active features for that age group using the Select by Attribute tool.

29
3.2.5. Creation of Park Access Points
To calculate accessibility to parks, access points are required for network analysis. It was
not realistic to use the centroid of the park for calculations, as some parks are vast, and the
centroid is not an accurate representation of where users enter the park.
Conversely, the corner points of the park were not always good substitutes for access
points, as there are sometimes surrounding parcels that block pedestrian traffic. An example of
this is a park that abuts a residential property. Pedestrians cannot traverse private property to
access the park. As such, overhead imagery was used to determine common-sense access points
as shown in the example in Figure 7 below. These are the points that provide pedestrians the
closest possible access from the road network in approaching the park from every direction.

30
Figure 7 Map of Access Points at Beach Park

Since the centroid of a park would not suffice for the detailed level of analysis required,
access points for each park were created. The first step was to use the ArcMap tool Feature
Vertices to Point, creating points at each vertex of the 103 parks in Alexandria. As a next step,
for each of the parks in Alexandria, the study used basemap world imagery to look for distinct
access points to parks. In cases where points at vertices did not make for appropriate park access
(as noted above), these extra vertices were deleted manually. New points were created for
obvious access points not located on vertices. Obvious access points included sidewalk entry and
access points near a road intersecting with the park.

31
As a general rule, the corner and road intersection access points were used to identify the
easiest way to get into the park from an adjoining or intersecting road. For example, Figure 8
below shows eight access points for the African American Heritage Park. The top right corner
access point allows for pedestrians arriving along Jamieson Avenue from the east. The top
middle access point is easily defined by the sidewalk entering the park. The top left access point
is both a corner point and an easily identified sidewalk entry. Along the west side of the park,
several access points are either sidewalk entries or access points near a T-intersection to allow
access from the west along Ballenger Avenue or Emerson Avenue, the two east-west streets in
the western portion of the map.
Figure 8 Access Point Types at the African American Heritage Park

32
After adding access points, the park shapefiles were then segmented based on active play
features. This created three sets of park polygons with access points: all parks, parks with active
play features, and parks with no active play features. Each park had its access points broken out
through the United States Geological Survey tool Split by Attribute (United States Geological
Survey n.d.). The USGS built a customization of ArcMap 10 to help divide a shapefile into
multiple shapefiles based on unique values of an attribute. Using this tool, the all-parks access
points shapefile was split into 103 shapefiles, each associated with the name of the park. This
split is used for creating service areas for each park.
3.2.6. Creating Park Service Areas
Using the ArcGIS Network Analyst toolbox allows for the creation of walksheds that
show the distance traveled from a starting location along the network. Although the travel
distance will vary significantly from family to family based on a wide variety of factors, this
study simplifies the travel distances to three bands based on age of the youth. The following
steps were repeated for all parks, active parks, and active parks for ages zero to four, five to nine,
and ten to seventeen. Distances used were one-quarter, one-half, and one-mile for all parks and
all active parks. Age-appropriate active parks used distances considered proper for the age group.
Newborns to four-years-old utilized 0.25-mile, children aged five- to nine-years-old used 0.5-
mile, and youth ten- to seventeen-years-old used one-mile.
Importing the street dataset from the City of Alexandria into the tool builds the
transportation network. The study then selected a new service area and loaded the appropriate
access point shapefiles as the facilities. The appropriate distance was selected with a matching
break to ensure only a single ring. Key option selections included U-turns allowed, as pedestrians
do not have to follow vehicle turning restrictions. To avoid multiple overlapping service areas

33
for the same park based on multiple access points to that park, merge by break value was
selected. This shapefile confirms if a residential parcel is within the park’s service area.
Overlapping service areas were used later in the process when calculating available park acreage.
The resulting service areas show what area is covered within a certain walking distance
of the park following the road network. This is used to estimate which parcels are in within given
walking distances of specific parks and then to calculate the number of expected children that
can be serviced by that park, including expected youth users per acre.
3.2.7. Calculating Park Load
Park load is a useful calculation showing the total number of people within the park’s
service area divided by the park’s area. It indicates the degree to which a park may potentially be
congested with users. This method normalizes the number of park users for comparison across
parks. Park load is defined in Equation 4 below:
Equation 4 Calculating Youth per Park Acre
𝐿 𝑃 =
𝑃𝑜𝑝 𝑌𝑝
𝐴 𝑝
Here park load (𝐿 𝑃 ) is shown to be equal to youth (𝑃𝑜𝑝 𝑌𝑝 )
divided by the area of the park in
acres (𝐴 𝑃 ). Two parks with the same number of people in their service area but with differing
park sizes will have different park loads, indicating park congestion.
To determine whether a residential parcel falls within a park service area, parcel centroids
were created. These centroids take on all the attributes of the parcel (except its geometry), which
simplifies the determination if the parcel is within the service area. Either the centroid is in or out
of the service area whereas a parcel polygon could be partially inside the service area. Each of
the park service areas were spatially joined with the parcel centroids while summing the

34
attributes. This ensured that the park service area now had the sum of all expected youth in the
residential parcels within the service area. All park service areas with the same distance (0.25-,
0.5-, and one-mile) were then merged to create one shapefile for the 103 parks. The service area
was manually edited with the park’s FeatureID to prepare for the next step.
To show each park’s load, the park shapefile was joined to the merged service areas
based on the park’s FeatureID. New fields were created for the density of the various age groups
at differing distances, for example, “Hd0_17” for the density of all youth within 0.5-mile of the
park. These fields were calculated using Equation 4 above. This allows for the geographical
representation of the park’s density within its boundaries.
3.2.8. Calculating Accessible Active Park Acreage per Parcel
The final step was calculating active park acreage per youth. This calculation shows which
residential parcels have the highest and lowest accessibility to active parks. The calculation uses
Equation 5 below.
𝐴 𝐴𝑃
=
𝐴 𝑃 𝑃𝑜𝑝 𝑌𝑝

Equation 5 Calculating Accessible Active Park Acres per Expected Youth
The accessible active park area in acres, 𝐴 𝐴𝑃
, is shown to be equal to the active park area
in acres, 𝐴 𝑃 , divided by the expected youth population in the parcel, 𝑃𝑜𝑝 𝑌𝑝
. Dividing by the
expected youth population normalizes the accessible active park acres to enable comparison of
different parcels. Since different parcels can have a significantly different number of expected
youths, ranging from less than one to over one hundred, normalization is necessary. This process
highlights those parcels and neighborhoods with good access to active parks and those with
inferior access to active parks.

35
To determine the active park area for each parcel, a spatial join was conducted using the sum
criteria. By joining the merged service area shapefile (as described in Section 3.2.7 Calculating
Park Load above) with the residential parcels, it brings the total park acreage accessible within
the specified distance from each parcel. Then all parcels with a non-zero expected youth
population and a non-zero active park area were selected by attribute. A field calculation was
conducted using Equation 5 to determine the number of park acres available per expected youth
living at that parcel. The expected youth population used for each distance corresponded with the
earlier stated reasonable walking distance; youth aged zero to four were paired with active parks
within 0.25-mile, youth aged five to nine with 0.5-mile, and youths aged ten to seventeen paired
with one-mile.
3.2.9. Mapping Techniques
Displaying accessible park acres per expected youth on a map was a two-step process. A Jenks
natural break classification minimizes the variance within a class while maximizing the variance
from other classes. Five classifications were selected for this study. Because the study wanted to
highlight those parcels with zero active park access, a sixth classification was necessary. To keep
the variance between classes high and not have an extra-large classification at the lower end of
the spectrum, a five classification Jenks natural break was conducted on each park acreage per
parcel attribute. The breakpoints were noted and then transferred to a manual break with six
classifications. This resulted in a classification for no access and five classifications for the
remaining data.
3.3. Conclusion
One concern when conducting an accessibility analysis is handling borders. In some
cases, an amenity is located outside the official boundaries of the study area but can affect

36
accessibility. In line with this study, a park might lie just over the city boundary. The city
boundary would not prevent residents from crossing to use the park. There are two options for
handling this problem; treating the boundary as a “hard” boundary in which people are assumed
to not travel across it and including the road network and amenities just beyond the study area in
the study. Figure 9 shows an analysis of overhead imagery surrounding the boundaries of
Alexandria, Virginia. It shows only two small parks with active features within 0.5- mile. These
parks do not affect the analysis due to their distance from Alexandria residential parcels. The
school in the north central part of the map is over 0.5-mile along roads from the closest parcel.
Figure 9 Active Parks within 0.5-mile of Alexandria

37
Further boundaries such as highways and railroads with few crossing points also
restricted the southern and northern borders of the city. To the east, the Potomac River bounds
the city. Parks in adjacent cities do not affect the analysis in this study because the small handful
within 0.5-mile of the city boundaries is blocked by highways or rivers.

38
Chapter 4 Results
The goal of this study is to analyze the distribution of age-appropriate active play features in
Alexandria, Virginia parks. This study used techniques that have not been publicized widely and
represent a step forward in correctly analyzing actual accessibility.
This chapter reviews the study’s process and describes lessons learned, presents the
results of the study, and analyzes how well distributed the active parks are in the City of
Alexandria. It looks at each age group and their related service areas to show what parts of the
city have better access to active parks. Finally, the chapter identifies the residential parcels with
no age-appropriate access to active play features at parks. This analysis can then be included in
further studies to serve the youth of Alexandria better.
4.1. Analyzing Service Areas
The three age groups show different levels of accessibility, but there are some common
themes. There are common areas with poor access to the west, north central, and south central.
These areas are bounded by highways and or railroads that impede pedestrian accessibility.
These provide effective boundaries to travel, thus squeezing the service area into long, skinny
areas unlikely to intersect parks. While some of the higher values are skewed due to easy access
to large parks, several areas stand out with no access to active parks. In the west, an intersection
of highways reduces the available road network. A visual inspection of overhead imagery shows
that there are no parks nearby outside the city limits, making this an accurate representation of
available parks. A similar situation occurs along the northern border of the city, as highways
block access and smaller parks reduce available acres. Parcels around the larger parks show
higher access as expected. However, this does not directly address the size of the active play
features. A park may have over sixty acres, but if fifty-nine acres are woods with only one-acre

39
available for open area and a playground, this gives a false indication of higher level of access.
This is addressed in Chapter 5.3 Suggestions for Future Work.
Areas with a dense road network appear to be more accessible, as the route options
multiply with each intersection. Homes around schools have excellent access to parks. Many
schools have multiple types of active play features, primarily playgrounds and playing fields.
The maps showing accessible active-play acres per expected youth were classified using
Jenks classifications with an added classification for parcels with no accessible active play parks.
The maps for the older age groups have significantly different breaks than the youngest group
due to the higher maximum acres.
4.1.1. Youth Ages 0 to 4 and the 0.25-Mile Service Area
As expected, youth ages zero to four had the least access to active parks. This is primarily
a consideration of the shortened distance in the service area. Figure 10 below shows the
accessible park acres per youth in each residential parcel. The common areas of poor access in
the west, north central, and south central are joined by neighborhoods in the center of the city.
Some of these areas are residential cul-de-sacs. This design, while sought after for quieter traffic
patterns, significantly decrease walking networks along roads. Since there is only one way in or
out of the neighborhood, homes at the end of the cul-de-sac have inadequate access to parks. The
primary driver behind lack of accessibility for youth ages zero to four appears to be the shortened
travel distance. Many of the poor accessibility areas would have better access if the distance
were increased. There are also several poor accessibility areas that are within 0.25-mile of a park.
These parks in the north-central and southwestern sections of the map are prime candidates for
installation of a playground. This would only affect a small number of parcels, as most of the
zero-access parcels are over 0.25-miles from a park.

40
Figure 10 Accessible Active Park Acres per Youth Aged Zero to Four within 0.25-Mile

The distribution of accessibility is skewed towards fewer available acres per youth as
shown in Figure 11 below. The large number of small parks throughout the city means that a
parcel is more likely to be within 0.25-mile of the small park. The larger parks do not have as
many parcels within their combined service areas because there are fewer large parks than small
parks.

41
Figure 11 Histogram of Accessible Active Park Acres for Youth Ages Zero to Four

4.1.2. Youth Ages 5 to 9 and 0.5-Mile Service Area
Overall, youth aged five to nine have better accessibility than the younger age group. The
middle age group has higher maximum available acres and fewer poor access areas. Figure 12
below shows the accessibility of the middle age group. Surprisingly, there are more parcels in the
downtown area to the southeast portion of the map between the railroad tracks and the eastern
side of the city with poor accessibility when compared to the zero to four age group. This is due
to the specialization of the parks in that area. Several “tot lots” in the downtown area have small
areas that are just playgrounds. As the age group progresses, the space required for an active play
feature increases, thus moving most active play areas out of the downtown area. The exceptions
to this are the parks along the river to Alexandria’s east. These large parks service much of
downtown.

4,161
7,900
3,212
2,775
1,361
1,165
361
455
196
544
145 180
51
274
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
Frequency

42
Figure 12 Accessible Active Park Acres per Youth Aged Five to Nine within 0.5-Mile

Figure 13 below shows the distribution of accessible active park acres for children ages
five to nine based on residential parcels. The high number of parcels with less than 200-acres
available shows that accessibility is not equitable throughout the city. While the skewing is not
as bad as that of newborns to four-year-old shown in Figure 11 above, it is still distorted towards
less access. 200-acres of access per youth is remarkable. This histogram does not show that there
is a significant amount of poor access in the city, it just highlights the inequity of access.

43
Figure 13 Histogram of Accessible Active Park Acres for Youth Ages Five to Nine

4.1.3. Youth Ages 10 to 17 and One-Mile Service Area
The oldest age group has the best access to active play features. This is primarily due to
the increased travel distance. Figure 14 below shows the accessible active park acreage for youth
ten to seventeen. The age group also has the most type of active play features. The lowest Jenks
break for youth aged ten to seventeen includes up to 370 acres per expected youth. This is still a
significant level of accessibility.
When compared to the youngest age group, this value would be above average. Even at
the one-mile distance, some of the same poor access areas as other age groups can be seen.
Again, the lack of a road network is the primary cause. As with the other age groups, railroads
and highways provide effective linear barriers to pedestrian travel.
1,922
1,492
2,700
3,435
3,939
2,469
1,382
1,317
2,011
754
588
292
80
399
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
Frequency

44
Figure 14 Accessible Active Park Acres per Youth Aged Ten to Seventeen within One-Mile

The distribution of park accessibility for youth ages 10-17 in Figure 15 below is much
more even than the younger age groups. There is no skewing to the low end of the range. This is
due to the increased service area provided by the longer travel distance. It does not highlight a
significant need for changes, as even the lowest non-zero grouping still has sufficient park
access.

45
Figure 15 Histogram of Accessible Active Park Acres for Youth Ages Ten to Seventeen

4.2. Areas with No Active Park Accessibility
The study also identified residential parcels with no active park accessibility. These areas
stand out more clearly when they are the only parcels mapped as in Figure 16 below. There are
132 residential parcels in Alexandria with no access to active play parks for youth of any age.
There are no age-appropriate active play parks for youth ages zero to four within 0.25-mile,
within 0.5-mile for youth from five to nine, and within one-mile for youth ten to seventeen.
These 132 parcels result from selecting the parcels which have no active park acres within their
age-appropriate service areas. Considering that there are over 22,000 residential parcels, this is a
rate of less than one-hundredth of a percent. These parcels only have an expected youth
population of 110 children. This shows how well the park locations were planned to enable
access. Highways bound most of the parcels with low access, reducing the available service area.
1,336
1,721
2,717
2,749
2,585
1,732
2,980
2,165 2,146
1,054
642
174
150
629
0
500
1,000
1,500
2,000
2,500
3,000
3,500
Frequency

46
Figure 16 Residential Parcels with No Age-Appropriate Active Park Access

The youngest age group has the most children with no active park access. Figure 17
below shows the location of the 4,161 parcels with no access to active play features. These
parcels represent 3,086 youth. As discussed earlier, the parcels with no access are clustered
based on road networks designed for low traffic volume at the expense of pedestrian
accessibility. Cul-de-sacs and highways restrict the service area significantly with limited routes
for travel.

47
Figure 17 Residential Parcels with No Age-Appropriate Park Access for Youth Ages Zero to
Four

Children aged five to nine have a total of 1,922 parcels and 584 children with no access
to active play features. The oldest age group has the best accessibility with only 1,336 parcels
and three expected children without access to active play parks. Figures Figure 21 and Figure 22
in Appendix A show the residential parcels with no age-appropriate access to active play park
features for the three age groups. Each of the age groups follow a similar pattern. Residential
parcels with no access are concentrated in areas with poor road networks. Highways and cul-de-
sacs prevent few options for pedestrian travel.

48
4.3. Analyzing Park Congestion
Park congestion measures how many youths are within the service area per park acre. It
allows for the comparison of differently-sized parks by normalizing to the park acre. Two factors
affect park congestion: youth serviced by the park and park size. The more highly congested
parks have many serviced youth and a small park size. Large, spread out parks will be inherently
less congested as their size reduces the congestion calculation.
As expected, the smaller parks in Alexandria have the most congestion for youth from
zero to four-years-old. Figure 18 below shows that the large parks have low congestion while
some of the smaller parks in the southeast have higher congestion. This is because some parks
are as small as a quarter acre, which lends to higher congestion. The highly congested parks in
the northeast section of the map are directly next to large apartment complexes with a high
number of children. Although there are some densely populated parcels in the west directly next
to parks with active play features, the methodology of this study used the parcel centroid as the
determining factor if the parcel was within the parks service area. While those living on the side
of the apartment complex nearer to the park would be within walking distance, the parcel was
ruled outside the service area.

49
Figure 18 Park Congestion for Youth Ages Zero to Four

Figure 19 below shows a closer view of the smaller parks within the downtown area.
While there are not many children living within their service areas, these parks are small and
thus have a higher congestion calculation.

50
Figure 19 Park Congestion in Downtown for Youth Ages Zero to Four

Figures Figure 23 and Figure 24 in Appendix A show the park congestion maps for youth
ages five to nine and ten to seventeen. They follow similar patterns as the youngest age group.
The total numbers of youth serviced increase due to the expanded travel distance.
4.4. Conclusion
Alexandria has distributed its active play features through its parks well. This leads to
more opportunities for children to get outside and play. Good access to parks has been shown to
be one method of fighting the youth obesity that is a rising problem in the United States.
Alexandria can further improve park accessibility with the addition of playgrounds in zero access

51
areas, giving 1,500 children easy access to active play features. Alexandria can also work to
encourage active play by children through outreach programs.
These 1,500 children are not completely without access to active play features. While
they do not have park access from their homes, their schools likely have some type of active play
features. Encouraging the children to play while at or after school can help encourage positive
well-being, social interaction, and development.

52
Chapter 5 Discussion and Conclusion
The purpose of this study was to analyze the active play park accessibility for youth in the City
of Alexandria, Virginia. The study dasymetrically mapped the youth population across the
residential parcels with three age groups: zero- to four-years-old, five- to nine-years-old, and ten-
to seventeen-years-old. Parks were classified into different active park classifications based on
age-appropriate active play features. The study created service areas for each park at three
distances corresponding to the three age groups of the study: 0.25-mile, 0.5 mile, and one-mile.
Combining the service areas with the residential data allowed for calculating the park acreage
accessible for each age group. This final park acreage accessibility calculation shows that
Alexandria has distributed its parks and active play features well, where only 132 parcels have
no age-appropriate access to active play parks. This is less than one-hundredth of a percent of all
residential parcels in the city.
5.1. Lessons Learned
This study expanded into several areas that had not been combined in previous works. As
such, there were two main categories of lessons learned: dasymetric mapping and park features
and access point selection.
5.1.1. Lessons Learned from Dasymetric Mapping
There were several lessons learned throughout the study. Using dasymetric mapping to
better represent where the city youth were expected to live provided many bumps along the way.
The data available did not support the concept as initially conceived. Techniques to work around
the problem had to be developed. The first problem was that the data from the GIS Department
did not have the desired fields, specifically the housing units in each residential parcel. This

53
made the process more complex. The workaround developed by using the building shapefile to
identify the number of housing units worked well. One problem that was highlighted was that
some building occurred in multiple parcels. This introduced some small error as the housing
units were then counted twice. Figure 20 below highlights a building that is mapped in two
parcels. The highlighted building is a single-family home that crosses parcel lines. The process
of spatially joining residential buildings to parcels brought a housing unit to both parcels. The
parcel on the left ended up with two housing units. The time required to verify the location of
each building in the over 22,000 residential parcels in the city was beyond the scope of this
study.
Figure 20 Closeup of Buildings in Multiple Parcels

54
This study began by using census tract population data from 2010. The data was initially
selected because it was the last complete census conducted. Although this data was not the best
data for the study, it served a useful purpose in developing the methodology by which the
dasymetric mapping would be executed. The readily available dasymetric mapping literature did
not translate directly to this project, as the data structures were different. The early census tract
data allowed for experimentation on the order of tools applied and available options for each
tool. The study used ACS population estimates from 2015 in the final version. While the margin
of error might have increased slightly due to the uncertainty of the Census Bureau’s American
Community Survey 5-year estimates, it is more current than the ten-year census. The five-year
estimate is the most precise estimate published by the Census Bureau. Using the data from 2015
also enabled the close matching of housing and parcel data. The housing data was also published
in 2015.
Changing data sources from census tracts to census block groups also increased the
precision of the study. While this caused work to be repeated, it resulted in a more accurate
representation of where the youth were expected to live. Since the purpose of the study is to
analyze how accessible active play parks are to the youth population, census block groups
provided the best source data. Manipulating the larger quantities of data for the block groups data
went smoothly as the process had been developed and tested with the census tract data.
5.1.2. Lessons Learned in Identifying Park Features and Access Points
The availability of overhead imagery significantly helped with identifying park access
points. Chapter 5.3 Suggestions for Future Work addresses ways in which this process can be
improved. However, comparing the background imagery in ArcMap with Google Earth imagery
was sufficient to identify likely access points. Still, it was not possible to identify every access

55
point without site visits. Overhead imagery and available road shapefiles do not always show the
existence of sidewalks, paths through yards, or holes in fences, which can be used to access, the
park from non-standard points. Extensive field work would be required to accurately place all
access points to each of Alexandria’s parks and was beyond the scope of this study.
5.2. Limitations
This study used several assumptions and accepted limitations. The study only used parks
as identified by the Alexandria Parks Department website. These public areas have equal access
for all. Private parks and features such as playgrounds on apartment complex grounds were not
included.
A second limitation was the inclusion of the full size of the park in calculations for
access. This assumes that the entire park is dedicated to the active play feature. A more accurate
representation would be to provide several area fields within the park data file. Additional fields
should include acres of active play features by age group. By calculating the available acres with
the full size of the park instead of the actual size of the play features, active park access is
overstated.
A major limitation is the difficulty of generalizing entire age groups behavior. While
necessary for this study, stratifying the distance a youth and/or their parent would travel to a park
is inherently inaccurate. The distance relies on a multitude of factors. It is likely that parents are
willing to put their very young children who cannot walk far on their own in a stroller and walk a
further distance as exercise for the adult or social interaction with other parents. Each family
makes its own decision on how far they are willing to or allow their children to travel to a park to
play. A single parent family may have different emphasis on parks than a family with a stay at
home parent or a family with two working parents and a caretaker for the child. The age of the

56
caretaker, whether a parent, older sibling, other relative, or hired help will also affect the travel
distance. A grandparent with their own health issues will not be willing to take a child as far as a
young, stay at home, fitness-enthusiast parent.
The ages of other children, either in the household or a playgroup, also will affect what
parks and distances are available. If a family has two children with a five-year age gap, they will
likely concentrate of visiting parks with play features for both children. They may also focus on
one child to ensure that they have access to their preferred play activity. Personal preference of
both the youth and parents will also affect where they choose to visit.
A final limitation of the study is a common limitation with all accessibility studies. This
study looks at a single point in time, specifically 2015, based on the population data. It is
impossible to obtain the data on where each youth lives in the city in near real-time. This study is
based on the expected population at a single point in time. It relies on the assumption that youth
are evenly distributed across housing units it the city.
5.3. Suggestions for Future Work
There are several ways to improve this process. Due to time constraints, several
assumptions were made during this study. One problem with how this study was conducted is
that the study assumed that the entirety of the park was available as an active play feature when
calculating available active play acres per expected youth. A 60-acre park may include sports
fields, basketball courts, and a playground while remaining primarily woods by acreage. It is not
accurate to calculate that the entire 60-acres are available for toddlers as a playground. In this
example, the playground may only cover 0.25-acres, the basketball courts 0.5-acres, and four-
acres of playing fields. The fields for total size, active play size for zero to four, active play size
for five to nine, and active play size for ten to seventeen would be: 0.25-, 4.25-, and 4.5-acres

57
respectively. This would ensure that the calculated available acres per expected child would be
more accurate than this study. The outliers of residential parcels with thousands of acres of
accessible active parks would be removed, and a more realistic output would result. This would
require measurement of the space taken up by active play features in each park. Two techniques
are easily identified; mapping through overhead imagery and site-visits and spatial data
collection. The best balance between accuracy and time required would dictate the ratio of
technique used. An additional benefit of site visits is verification of the park data stated by the
Parks Department. Playgrounds and playing fields can be identified more accurately based on
age appropriateness.
A second topic for future work is to include the results from this study as inputs in a site
selection study for new parks or new active play park features. As retail stores analyze
population data to ensure that a new store services a previously underserved community, city
planners should use areas of poor accessibility to help plan new parks and features.
A third interesting future topic is to stratify the expected youth population by other
socioeconomic factors. These studies should consider if race, religion, or family income affect
accessibility. Since some of the areas with poor access have large numbers of housing units, this
suggests that lower-income families, who are more likely to live in large apartment complexes,
might have lower access to active play parks. This study does not have sufficient evidence to
support this hypothesis, but it would be easy to follow a process similar to this study. The
primary difference is when conducting the dasymetic mapping, future authors should segment
based on race and family income in addition to age. This can be accomplished by selecting
census data that includes race and family income.

58
This study used roads with a speed limit of less than or equal to thirty-five miles per hour
as a proxy for the pedestrian network in the city. The data from the city GIS Department is good
for vehicular transportation but does not include sidewalks or other common pedestrian routes
such as pedestrian bridges or tunnels. A method to gain a more accurate assessment of
accessibility to active play features is to build a pedestrian transportation network for the city. A
potential source for this information is OpenStreetMap. This open source website has many
updated shapefiles and other data. OpenStreetMap has sidewalk information on many cities,
including Washington, District of Columbia. Through several spot checks throughout the city, it
does not have sidewalk data for streets in Alexandria (OpenStreetMap contributors n.d.).
Repeating this study’s process in other cities with sidewalk or pedestrian network data should
result in a more accurate representation of pedestrian accessibility.
A final option for future study is the inclusion of private play features. Some apartment
complexes include playgrounds, pools, and basketball or tennis courts as amenities. This would
affect accessibility to active play features. Some apartment buildings from this study that have no
access could have access to playgrounds in reality. Adding private active play features as an
alternative to parks could change the results of the study. A method of accounting for the fact
that these features are not public would have to be developed.
5.4. Future Applications
The primary application for this study is the analysis of active park accessibility in
Alexandria, Virginia for youth ages zero to seventeen. The results of the study show the areas
with no or poor accessibility. The parks and recreation department can consider this in planning
and upgrading active play features throughout the city. This study identified 211 residential
parcels with no age-appropriate active play features available and over 4,000 parcels with no

59
active play access for newborns to four-years-old. This study completed the first step to fix the
problem, identifying the areas of most need. Since several clusters readily stood out, the parks
department should look at adding active play features for toddlers to the parks within 0.25-miles
of these clusters. This simple fix can immediately add access for over 1,300 children under four.
Since playgrounds are the only active play feature acceptable to this age group in this study,
adding playgrounds will have an immediate impact. Some playgrounds require less than twenty
feet on a side.
Another application of this specific study is for the city to analyze the service areas. For
long, skinny service areas, the opportunity exists to expand the reach by changing the pathways
open to children. This could include sidewalks running from the back of a cul-de-sac to another
neighborhood. This study focused only on using roads as effective travel networks.
The framework for this study can be replicated across the country. Some changes might
be necessary based on the structure of the available data. Some cities may already include the
number of housing units in their parcel files. In such a case, several steps in this study would be
unnecessary.
5.5. Overall Conclusions
While there are still some refinements necessary to the process developed by this study,
they are focused on local differences in data. In Alexandria, the number of housing units was not
included as attributes in the parcel shapefile. This may not be the case throughout the country
where this study process can be applied. This study recommends that local governments include
this attribute in their parcel shapefiles. It will make analysis more accurate and faster to perform.
This process can be replicated in any community that has or can derive the required data:
residential parcels with housing unit data, population data at the census block group or census

60
tract level, and park data including active play features. The results of this process show where
communities can focus their park planning. It highlights areas that may already have a park, but
no active play features, or areas where a new park with active play features can affect the
outdoor play of many children.

61
References
Alexandria City Public Schools. n.d. Transportation / Overview. Accessed November 5, 2017.
https://www.acps.k12.va.us/transportation/.
Bagheri, Nasser, Alec Holt, and George L Benwell. 2009. "Using Geographically Weighted
Regression to Validate Approaches for Modelling Accessibility to Primary Health Care."
Applied spatial analysis and policy 2 (3): 177-194. doi:10.1007/s12061-009-9021-0.
Bejleri, Ilir, Ruth L Steiner, Allison Fischman, and Jeffrey M Schmucker. 2011. "Using GIS to
analyze the role of barriers and facilitators to walking in children's travel to school."
Urban Design International 16 (1): 51-62. doi:10.1057/udi.2010.18.
Blanck, Heidi M, Diana Allen, Zarnaaz Bashir, Nina Gordon, Alyson Goodman, Dee Merriam,
and Candace Rutt. 2012. "Let's Go to the Park Today: The Role of Parks in Obesity
Prevention and Improving the Public's Health." Childhood Obesity 8 (5): 423-428.
doi:10.1089/chi.2012.0085.blan.
Boone, Christopher G, Geoffrey L Buckley, J Morgan Grove, and Chona Sister. 2009. "Parks
and People: An Environmental Justice Inquiry in Baltimore, Maryland." Annals of the
Association of American Geographers 99 (4): 767-787.
doi:10.1080/00045600903102949.
Centers for Disease Control and Prevention. 2016. Health-Related Quality of Life (QRQOL).
May 22. Accessed March 18, 2018. https://www.cdc.gov/hrqol/wellbeing.htm.
City of Alexandria GIS Department. 2017. Buildings. City of Alexandria GIS Department.
Alexandria, VA, June 6. Accessed July 9, 2017. http://cityofalexandria-
alexgis.opendata.arcgis.com/datasets/buildings.
City of Alexandria GIS Department. 2017. Parcels. City of Alexandria GIS Department.
Alexandria, VA, April 5. Accessed April 7, 2017. http://cityofalexandria-
alexgis.opendata.arcgis.com/datasets/parcels.
City of Alexandria GIS Department. 2015. Park. City of Alexandria GIS Department.
Alexandria, VA, November 25. Accessed April 7, 2017. http://cityofalexandria-
alexgis.opendata.arcgis.com/datasets/park.
City of Alexandria GIS Department. 2017. Roads. City of Alexandria GIS Department.
Alexandria, VA, January 31. Accessed April 7, 2017. http://cityofalexandria-
alexgis.opendata.arcgis.com/datasets/roads.
City of Alexandria, Virginia. 2017. Park Information. January 13. Accessed July 30, 2017.
https://www.alexandriava.gov/recreation/info/default.aspx?id=12342.
Cohen, Deborah A. MD, Thomas L. PhD McKenzie, Amber Sehgal, Stephanie Williamson,
Daniela PhD Golinelli, and Nicole MD Lurie. 2007. "Contribution of Public Parks to

62
Physical Activity." American Journal of Public Health 97 (3): 509-514.
doi:10.2105/AJPH.2005.072447.
Comber, Alexis, Chris Brunsdon, and Edmund Green. 2008. "Using a GIS-based network
analysis to determine urban greenspace accessibility for different ethnic and religious
groups." Landscape and Urban Planning 86 (1): 103-114.
doi:10.1016/j.landurbplan.2008.01.002.
de Vries, Sjerp, Sonja van Dillen, Peter Groenewegen, and Peter Spreeuwenberg. 2013.
"Streetscape greenery and health: Stress, social cohesion and physical activity as
mediators." Social Science & Medicine 94: 26-33. doi:10.1016/j.socscimed.2013.06.030.
Delgado, Marc Elgin M, and Frank Canters. 2011. "Measuring the accessibility of different
household income groups to basic community services in upland Misamis Oriental,
Northern Mindanao, Philippines." Singapore journal of tropical geography 32 (2): 168-
184. doi:10.1111/j.1467-9493.2011.00427.x.
Dunton, Genevieve Fridlund, Estela Almansa, Michael Jerrett, Jennifer Wolch, and Mary Ann
Pentz. 2014. "Neighborhood Park Use by Children: Use of Accelerometry and Global
Positioning Systems." American Journal of Preventative Medicine 46 (2): 136-142.
doi:10.1016/j.amepre.2013.10.009.
ESRI. 2017. About the ArcGIS Network Analyst extension tutorial. June 13. Accessed July 23,
2017. http://desktop.arcgis.com/en/arcmap/latest/extensions/network-analyst/about-the-
network-analyst-tutorial-exercises.htm.
Floyd, Myron F. PhD, Jason N. PhD Bocarro, William R. PhD Smith, Perver K. PhD Baran,
Robin C. MCP Moore, Nilda G. PhD Cosco, Michael B. PhD Edwards, Luis J. PhD
Suau, and Kunsheng MPA Fang. 2011. "Park-Based Physical Activity Among Children
and Adolescents." American Journal of Preventative Medicine 41 (3): 258-265.
doi:10.1016/j.amepre.2011.04.013.
Ford, Alistair C, Stuart L Barr, Richard J Dawson, and Philip James. 2015. "Transport
accessibility analysis using GIS: Assessing sustainable transport in London." ISPRS
International Journal of Geo-Information 4 (1): 124-149.
doi:http://dx.doi.org.libproxy1.usc.edu/10.3390/ijgi4010124.
Haugen, Katarina, Einar Holm, Magnus Stromgren, Bertil Vilhelmson, and Kerstin Westin.
2012. "Proximity, accessibility and choice: A matter of taste or condition?" Papers in
regional science 91 (1): 65-84. doi:10.1111/j.1435-5957.2011.00374.x.
Higgs, Gary, Mitch Langford, and Paul Norman. 2015. "Accessibility to sport facilities in Wales:
A GIS-based analysis of socio-economic variations in provision." Geoforum 62: 105-120.
doi:10.1016/j.geoforum.2015.04.010.
Jamtsho, Sonam, Robert Corner, and Ashraf Dewan. 2015. "Spatio-Temporal Analysis of Spatial
Accessibility to Primary Health Care in Bhutan." ISPRS International Journal of Geo-
Information 4 (3): 1584-1604. doi:10.3390/ijgi4031584.

63
Jennings, Viniece, and Cassandra J Gaither. 2015. "Approaching Environmental Health
Disparities and Green Spaces: An Ecosystem Services Perspective." International
Journal of Environmental Research and Public Health 12 (2): 1952-1968.
doi:10.3390/ijerph120201952.
Jennings, Viniece, Lincoln Larson, and Jessica Yun. 2016. "Advancing Sustainability through
Urban Green Space: Cultural Ecosystem Services, Equity, and Social Determinants of
Health." International Journal of Environmental Research and Public Health 13 (2): 196.
doi:10.3390/ijerph13020196.
Jimenez, Edgar H. 2016. "The Role of Amenities in Measuring Park Accessibility: A Case Study
of Downey, California." Masters Thesis. Los Angeles: Institute of Spatial Sciences,
University of Southern California.
Kaczynski, Andrew T, Luke R Potwarka, Bryan J A Smale, and Mark E Havitz. 2009.
"Association of Parkland Proximity with Neighborhood and Park-based Physical
Activity: Variations by Gender and Age." Leisure Sciences 31 (2): 174-191.
doi:10.1080/01490400802686045.
Larson, Lincoln R, Viniece Jennings, and Scott A Cloutier. 2016. "Public Parks and Wellbeing in
Urban Areas of the United States." PLoS ONE 11 (4). doi:
10.1371/journal.pone.0153211.
Lee, Hedwig Ph.D., Dohoon Ph.D. Lee, Guang Ph.D. Guo, and Kathleen Ph.D. Harris. 2011.
"Trends in Body Mass Index in Adolescence and Young Adulthood in the United States:
1959-2002." Journal of Adolescent Health 49: 601-608.
doi:10.1016/j.jadohealth.2011.04.019.
Lindsey, Greg, Maltie Maraj, and SonCheong Kuan. 2001. "Access, Equity, and Urban
Greenways: An Exploratory Investigation." The Professional Geographer 53 (3): 332-
346. doi:10.1111/0033-0124.00288.
Little, Helen, and Shirley Wyver. 2008. "Outdoor play: does avoiding the risk reduce the
benefits." Austrailian Journal of Early Childhood 33 (2): 33-40.
Liu, Suxia, and Xuan Zhu. 2004. "An Integrated GIS Approach to Accessibility Analysis."
Transactions in GIS 8: 45-62. doi:10.1111/j.1467-9671.2004.00167.x.
Lopez, M C Rojas, and Y D Wong. 2017. "Children's active trips to school: a review and
analysis." International Journal of Urban Sustainable Development 9 (1): 79-95.
doi:10.1080/19463138.2016.1264405.
Maps of World. n.d. "Where is Alexandria, Virginia." Maps of World. Accessed July 16, 2017.
https://www.mapsofworld.com/location-maps/usa/alexandria-va.html.
Neutens, Tijs, Sarah Daniels, Joeri Minnen, Ignace Glorieux, Philippe de Maeyer, and Nico van
de Weghe. 2014. "Spatial and temporal fluctuations in individual accessibility: a

64
comparative analysis among subgroups of the population." Geografisk Tidsskrift-Danish
Journal of Geography 114 (2): 119-131. doi:10.1080/21662282.2013.863547.
Neutens, Tijs, Tim Schwanen, Frank Witlox, and Philippe de Maeyer. 2010. "Evaluating the
Temporal Organization of Public Service Provision Using Space-Time Accessibility
Analysis." Urban Geography 1039-1064. doi:10.2747/0272-3638.31.8.1039.
Nordh, H, T Hartiq, C M Hagerhall, and G Fry. 2009. "Components of small urban parks that
predict the possibility for restoration." Urban Forestry & Urban Greening 8 (4): 225-
235. doi:10.1016/j.ufug.2009.06.003.
Oftedal, Andrew, and Ingrid Schneider. 2013. "Outdoor Recreation Availability, Physical
Activity, and Health Outcomes: County-Level Analysis in Minnesota." Journal of Park
and Recreation Administration 31 (1): 34-56. doi:10.1016/j.ypmed.2013.11.010.
OpenStreetMap contributors. n.d. Open Street Map. Accessed March 17, 2018.
http://www.openstreetmap.org.
Páez, Antonio, Darren Scott, and Catherine Morency. 2012. "Measuring accessibility: positive
and normative implementations of various accessibility indicators." Journal of Transport
Geography 25: 141-153. doi:10.1016/j.jtrangeo.2012.03.016.
Páez, Antonio, Steven Farber, Ruben Mercado, Matthew Roorda, and Catherine Morency. 2013.
"Jobs and the Single Parent: An Analysis of Accessibility to Employment in Toronto."
Urban Geography 815-842. doi:10.1080/02723638.2013.778600.
Parker, Nathan H, Daniel P. PhD O'Connor, Dennis T. PhD Kao, and Rebecca E. PhD Lee. 2016.
"Do Neighborhood Physical Activity Resources and Land Use Influence Physical
Activity among African American Public Housing Residents?" Journal of Health Care
for the Poor and Underserved 27: 1330-1344. doi:10.1353/hpu.2016.0135.
Parsons, Jonathan E. 2015. "Mapping Uniformity of Park Access using Cadastral Data within
Network Analyst in Wake County, NC." Masters Thesis. Los Angeles: Institute of Spatial
Sciences.
Pauleit, Stephan, Paul Slinn, and John Hondley. 2003. "Promoting the Natural Greenstructure of
Towns and Cities: English Nature's "Accessible Natural Greenspace Standards" Model."
Built Environment 29 (2): 157-170.
http://www.jstor.org.libproxy2.usc.edu/stable/23288814.
Pedigo, Ashley S, and Agricola Odoi. 2010. "Investigation of Disparities in Geographic
Accessibility to Emergency Stroke and Myocardial Infarction Care in East Tennessee
Using Geographic Information Systems and Network Analysis." Annals of epidemiology
20 (12): 924-930. doi:10.1016/j.annepidem.2010.06.013.
Rigolon, Alessandro, and Travis L Flohr. 2014. "Access to Parks for Youth as an Environmental
Justice Issue:." Buildings 4: 69-94. doi:10.3390/buildings4020069.

65
Sadahiro, Yukio. 2005. "Spatiotemporal analysis of the distribution of urban facilities in terms of
accessibility." Papers in regional science 84 (1): 61-84. doi:10.1111/j.1435-
5957.2005.00004.x.
Schlossberg, Marc, Jessica Greene, Page Paulsen Phillips, Bethany Johnson, and Bob Parker.
2006. "School Trips: Effects of Urban Form and Distance on Travel Mode." Journal of
hte American Planning Association 72 (3): 334-346. doi:10.1080/01944360608976755.
Schoeppe, Stephanie, Mitch J Duncan, Hannah M Badland, Amanda L Rebar, and Corneel
Vandelanotte. 2016. "Too far from home? Adult attitudes on children's independent
mobility range." Children's Geographies 14 (4): 482-489.
doi:10.1080/14733285.2015.1116685.
Seeland, Klaus, Sabine Dübendorfer, and Ralf Hansmann. 2009. "Making friends in Zurich's
urban forests and parks: The role of public green space for social inclusion of youths
from different cultures." Forest Policy and Economics 11 (1): 10-17.
doi:10.1016/j.forpol.2008.07.005.
The Trust for Public Land. n.d. ParkScore. Accessed 02 22, 2018.
http://parkscore.tpl.org/methodology.php.
United States Geological Survey. n.d. Split by Attribute Tool. Accessed October 5, 2017.
https://www.umesc.usgs.gov/management/dss/split_by_attribute_tool.html.
US Census Bureau. 2017. rl_2016_51_tract. US Census Bureau. Washington, DC, July 2.
Accessed July 2, 2017. https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml.
van Loon, Joshua, and Lawrence Frank. 2011. "Urban Form Relationships with Youth Physical
Activity: Implications for Research and Practice." Journal of Planning Literature 26 (3):
280-308. doi:10.1177/0885412211400978.
Wang, Dong, Gregory Brown, Guoping Zhong, Yan Liu, and Iderlina Mateo-Babiano. 2015.
"Factors influencing perceived access to urban parks: A comparative study of Brisbane
(Australia) and Zhongshan (China)." Habitat international 50: 335-346.
doi:10.1016/j.habitatint.2015.08.032.
Wolch, Jennifer, John P Wilson, and Jed Fehrenbach. 2005. "Parks and Park Funding in Los
Angeles: An Equity-Mapping Analysis." Urban Geography 26 (1): 4-35.
doi:10.2747/0272-3638.26.1.4.
Wolch, Jennifer, Michael Jerrett, Kim Reynolds, Rob McConnel, Roger Chang, Nicholas
Dahmann, Kirby Brady, Frank Gilliland, Jason G Sue, and Kiros Berhane. 2011.
"Childhood obesity and proximity to urban parks and recreational resources: A
longitudinal cohort study ." Health & Place 17 (1): 207-214.
doi:10.1016/j.healthplace.2010.10.001.

66
Yeager, Charles D, and Jay D Gatrell. 2014. "Rural food accessibility: An analysis of travel
impedance and the risk of potential grocery closures." Applied geography 53: 1-10.
doi:10.1016/j.apgeog.2014.05.018.
Zhang, Xingyou, Hua Lu, and James B Holt. 2011. "Modeling spatial accessibility to parks: a
national study." International Journal of Health Geographics 10:31. doi:10.1186/1476-
072X-10-31.

67
Appendix A
This appendix shows the residential parcels with no access to age-appropriate active play parks
within each age group.
Figure 21 No Active Park Access for Youth Aged Five to Nine

68
Figure 22 No Active Park Access for Youth Aged Ten to Seventeen

69
Figure 23 Park Congestion for Youth Ages Five to Nine

70
Figure 24 Park Congestion for Youth Ages Ten to Seventeen

Abstract (if available)

Abstract Park accessibility is important for city planners because the accessibility of parks can impact people throughout the community. Youth park accessibility is especially important, as parks positively impact physical, emotional, and social development. This study uses dasymetric mapping of census block group population data to estimate segments of youth population at each residential parcel, and then associates those segments with age-appropriate active play features at each park. Network analysis connects parcels to parks and their amenities, providing a more precise accessibility rating at the city-level than studies based solely on geodesic buffers from park centroids. ❧ This study shows that while Alexandria, Virginia has many parks throughout the city, the distribution of age-appropriate active play features is not uniform. Most children in Alexandria have access to at least one active-play park. Only 132 parcels have zero access to age-appropriate, active-play parks, a rate of less than one-hundredth of a percent. There are areas for improvement, but the City of Alexandria has done an excellent job ensuring children have access to active play parks. For other cities, this sort of accessibility analysis could help planners to target areas to increase funding for fitness amenities and programs within parks, establish new parks, or add pedestrian paths to improve walkability to existing park resources.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Analysis of park accessibility in Redan, Georgia Web GIS application

PDF

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

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

Mapping uniformity of park access using cadastral data within Network Analyst in Wake County, NC

PDF

Providing a new low-cost primary care facility for under-served communities: a site suitability analysis for Service Planning Area 6 in Los Angeles County, California

PDF

Preparing for earthquakes in Dallas-Fort Worth: applying HAZUS and network analysis to assess shelter accessibility

Asset Metadata

Creator Fox, Alexander (author)

Core Title Access to active play parks for youth segments in Alexandria, Virginia

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 07/26/2018

Defense Date 03/09/2018

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

Tag accessibility,active play park features,age appropriate active play features,Alexandria, Virginia,city planning,dasymetric mapping,geospatial analysis,GIS,greenspace,network analysis,OAI-PMH Harvest,park access,Parks,pedestrian access

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bernstein, Jennifer (committee member), Tau, Ran (committee member), Vos, Robert (committee member)

Creator Email alexdfox06@gmail.com,foxa@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-31623

Unique identifier UC11672489

Identifier etd-FoxAlexand-6514.pdf (filename),usctheses-c89-31623 (legacy record id)

Legacy Identifier etd-FoxAlexand-6514.pdf

Dmrecord 31623

Document Type Thesis

Format application/pdf (imt)

Rights Fox, Alexander

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