University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

The Role of Amenities in Measuring Park Accessibility:
A Case Study of Downey, California

by

Edgar H. Jimenez

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)

May 2016

Copyright ® 2016 by Edgar H. Jimenez

I would like to dedicate this manuscript to my wife Hilda for her endless support and my Aunt
Trini who has always encouraged me in my educational pursuits.

iv
Table of Contents
List of Figures ................................................................................................................................ vi
List of Tables ............................................................................................................................... viii
List of Equations ............................................................................................................................ ix
Acknowledgements ......................................................................................................................... x
List of Abbreviations ..................................................................................................................... xi
Abstract ......................................................................................................................................... xii
Chapter 1 Introduction .................................................................................................................... 1
1.1 Motivation ............................................................................................................................2
1.2 Theoretical Background .......................................................................................................5
1.3 Study Site .............................................................................................................................7
1.4 Research Question ...............................................................................................................9
1.5 Manuscript Navigation .........................................................................................................9
Chapter 2 Literature Review ......................................................................................................... 11
2.1 Brief histories of parks .......................................................................................................11
2.2 Parks and Environmental Justice .......................................................................................13
2.3 Measuring Population ........................................................................................................13
2.4 Past Methods for Measuring Park Accessibility ................................................................15
2.5 Network Analysis and Park Accessibility ..........................................................................17
2.6 Measuring Park Amenities .................................................................................................19
Chapter 3 Methods ........................................................................................................................ 23
3.1 Dasymetric Mapping ..........................................................................................................24
3.2 Collection of Amenity Data ...............................................................................................30
v
3.3 Park Amenity Scoring ........................................................................................................33
3.4 Service Area Data Preparation ...........................................................................................40
3.5 Service Area Analysis ........................................................................................................43
3.6 Conclusion .........................................................................................................................46
Chapter 4 Results .......................................................................................................................... 47
4.1 Quarter Mile Service Area .................................................................................................48
4.2 Half Mile Service Area ......................................................................................................52
4.3 One Mile Service Area .......................................................................................................58
4.4 Two Mile Service Area ......................................................................................................63
Chapter 5 Discussion and Conclusion .......................................................................................... 71
5.1 Population Mapping ...........................................................................................................72
5.2 Service Area Analysis ........................................................................................................74
5.3 Amenity Scoring ................................................................................................................76
5.4 Limitations and Improvements ..........................................................................................77
5.5 Future Directions ...............................................................................................................80
5.6 Final Conclusions ...............................................................................................................81
References ..................................................................................................................................... 83
Appendix A: PARA Operational Definitions and Protocols ........................................................ 88
Appendix B: Adjusted PARA Auditing Instrument ..................................................................... 96

vi
List of Figures
Figure 1 City of Downey .............................................................................................................. 20
Figure 2 Downey Population per Census Block ........................................................................... 37
Figure 3 Downey Residential Land Parcels .................................................................................. 39
Figure 4 Downey Population Per Parcel ....................................................................................... 41
Figure 5 Histogram of Park Acreage Values ................................................................................ 44
Figure 6 Furman and Independence Park Playgrounds ................................................................ 48
Figure 7 Improvised and Designed Physical Activity Amenities ................................................. 49
Figure 8 Half Mile Service Area Results ...................................................................................... 56
Figure 9 Histogram of 0.25-mile SA Parcel Values ..................................................................... 60
Figure 10 Histogram of 0.25-mile Play Acre SA Parcel Values .................................................. 61
Figure 11 Parcel Values for 0.25-mile SA .................................................................................... 63
Figure 12 Histogram of 0.50-mile SA Parcel Values ................................................................... 65
Figure 13 Histogram of 0.50-mile Disamenity Adjusted Acre SA Parcel Values ........................ 66
Figure 14 Parcel Values for 0.50-mile SA .................................................................................... 68
Figure 15 Detail of Byron Zinn Play Acre 0.25-mile SA ............................................................. 69
Figure 16 Histogram of 1-mile SA Parcel Values ........................................................................ 71
Figure 17 Histogram of 1-mile PA Acre SA Parcel Values ......................................................... 72
Figure 18 Parcel Values for 1-mile SA ......................................................................................... 74
Figure 19 Histogram of 2-mile SA Parcel Values ........................................................................ 76
Figure 20 Histogram of 1-mile NA Acre SA Parcel Values ......................................................... 77
Figure 21 Parcel Values for 2-mile SA ......................................................................................... 79
Figure 22 Detail of Wilderness Park NA 2-mile SA .................................................................... 80

vii

Figure 23 Service Areas for Discovery Sports Complex and Independence Park ....................... 84
Figure 24 Downey Restricted Watering Signs .............................................................................. 90
Figure 25 Emergency Call Boxes ................................................................................................. 91

viii

List of Tables
Table 1 List of Parks in Study Area .............................................................................................. 43
Table 2 List of Category Scores for Parks .................................................................................... 47
Table 3 Tree Amenity Values ....................................................................................................... 51
Table 4 Parcel Values for Each Service Area ............................................................................... 53
Table 5 Percent Downey Parcels for Each SA .............................................................................. 59

ix
List of Equations
Equation 1 Calculation for People per Residential Unit ............................................................... 40
Equation 2 Calculation for Population per Land Parcel ............................................................... 40
Equation 3 Calculation for Total Category Acres per Park .......................................................... 46
Equation 4 Calculation for Disamenity Adjusted Acres per Park ................................................ 46
Equation 5 Calculation for Acres per Person ................................................................................ 57
Equation 6 Calculation for Category Acres per Person ................................................................ 57

x
Acknowledgements
I am very grateful for the guidance and patience given to me by my thesis advisor Dr. Robert
Vos. I would also like to thank my professors at the USC Spatial Sciences Institute who
consistently worked with me as I faced the many challenges of completing my graduate course
work.

xi
List of Abbreviations
CPAT Community Park Audit Tool
EAPRS Environmental Assessment of Public Recreation Spaces
GIS Geographic information system
L.A. Los Angeles
LAeGIS Los Angeles County Enterprise GIS

NA Nature access amenity category

NRPA National Recreation and Park Association

PA Physical activity amenity category

PARA Physical Activity Resource Assessment

PRORAGIS Park and Recreation Operating Ratio and Geographic Information System

SA Service area

SAGE Systematic Audit of Greenspace Environments

U.S. United States of America

xii
Abstract
Previous studies of park accessibility have utilized network analysis and dasymetric
mapping to investigate pedestrian accessibility to park resources measured in acres per capita.
Through a case study of Downey, California, this study extends on previous work in this area by
combining network analysis and dasymetric mapping with robust park amenity auditing. The
intention of this study is to provide a more detailed examination of how accessibility is affected
by park condition and the types of facilities provided to park users. The study uses a method for
dasymetrically mapping population data to land parcels, Esri’s ArcGIS 10.1 Service Area
Network Analyst tool, and a park amenity scoring system based on the Physical Activity
Resource Assessment (PARA) instrument. The results of this research reveal that park
accessibility in Downey is limited at multiple Service Area (SA) distance levels due to the
presence of parks with high pedestrian accessibility but low amenities in the geographic center of
the city and parks with low pedestrian accessibility but high amenities on the city’s periphery.
The results of this case study inform policy suggestions for future park developments. These
policy suggestions include planning strategies for increasing pedestrian access to parks with
developed amenities, which are distant from residential areas. Also, the study indicates which
parks to nominate for development in highly accessible areas with few amenities.
1
Chapter 1 Introduction
Public parks are a common service offered by governmental bodies at many levels including
nation, state, county, and city. The following study investigated access to public parks of all
types at the scale of the city. Accessibility will be the main focus of the research discussed in
this paper. More specifically, a quantifiable method for measuring accessibility to park amenities
will be utilized in a case study of the Los Angeles (L.A.) County City of Downey.
The site chosen for this study is the roughly 12 square mile city of Downey, California.
Downey belongs to a conglomerate of southeast L.A. County Governments known as the
Gateway Cities. The area in general has a higher average population of Hispanic and Latino
residents than the rest of L.A. County. Downey was chosen as a study area because of its
surrounding geographical features and its independent city services. As will be explained later in
this paper, both of these factors play a large role in making Downey a site of interest.
Network analysis was the method utilized to investigate park accessibility in the city of
Downey. Large inspirations for this study were taken from methods utilized by both Parsons
(2015) and Ibes (2015) in their studies on park accessibility. In his 2015 thesis on park
accessibility in Wake County, North Carolina, Parsons established a method of using both
dasymetric mapping and service area analysis to determine population accessibility based on
residential land parcel locations. These methods allowed for a more realistic model of
accessibility than alternative geodesic distance models that have been utilized.
Ibes (2015) also provided a unique perspective on park accessibility in her work on park
equity in Phoenix, Arizona. Utilizing a classification system based on park amenities, Ibes
(2015) created a typology for Phoenix parks. Comparing park type to demographic data, Ibes
(2015) explained how different demographic groups in Phoenix had differing access to parks
2
with certain amenities. Using ideas form both Ibes (2015) and Parsons (2015), the following
study utilizes network analysis, dasymetric mapping, and a park amenity scoring system to
assess how accessible Downey parks are to its residents.
Wolch and Byrne (2009) state five main areas of park research: history and ideology,
access and use, fostering sustainable urban life, ecological benefits, and benefits to health and
well being of residents. The present study focuses on the area of park access and use. In a
review of past accessibility studies, Kabish et al. (2015) explained the different types of
statistically observable phenomenon in urban green space research. Kabish et al. (2015) found
that in general, city-planning experts in many cities throughout the world agree that there are
benefits to increasing accessibility to urban green space. It is also suggested by Kabish et al.
(2015) that future research should focus on more generalizable and quantifiable methods such as
those that utilize Geographic Information systems (GIS) as opposed to traditional survey
methods which can be locally contingent.
The remaining portions of this chapter are divided into four sections that explore the basis
of this research in more depth. Section 1.1 looks further at the motivation for this research and
what benefits it can provide for urban green space planning and the spatial sciences. In section
1.2, the theoretical background of network analysis in park research is discussed in more detail.
The study site of Downey is described in depth in section 1.3. Finally, section 1.4 presents an
overview of the entire manuscript.
1.1 Motivation
The impetus for this project was to contribute to the areas of urban planning, public
health, and spatial sciences. Urban planning research has often made the argument that public
green space is needed in an urban environment (Mehta 2013). Benefits are often associated with
3
areas of public health. By studying accessibility to parks in a small city like Downey, this
research has contributed to an increased understanding of public green space in areas outside of
major metropolitan centers. Working with data at the land parcel scale has also helped contribute
to the need for more detailed maps to help further spatial science studies. Finally, this case study
will provide an additional option for small governments to perform GIS studies at a community-
based level.
Studies of public space vary greatly and often use different variables to exemplify the
role they play in an urban environment. Mehta (2013) suggests that public space contains a mix
of safety, comfort, pleasure-ability, inclusiveness, and meaningful activity. Parks or public green
space is one of the most common examples used when discussing public space and the benefits it
provides to urban residents (Mehta 2013). Parks have also been the focus of many studies that
explore issues of environmental justice and public health (Suminski et al. 2012; Wolch et al.
2014; Parsons et al. 2015)
Environmental Justice involves the understanding of how productive and destructive
features of a physical environment are spatially distributed among a population based on social
demographics (Parsons et al. 2015; Parsons 2015). In this example, public parks would be an
example of a productive feature of the physical environment due to its positive effects on the
population (Mehta 2015). An example of a destructive feature would be a landfill or sewage
processing plant (Bullard 1996). Studies have shown that areas with lower income and where
minority populations are high, often have differential access to quality and quantities of public
green space (Taylor et al. 2007; Byrne and Wolch 2009; Marcelli 2010; Suminski et al. 2012;
Parsons et al. 2015).
4
Access to public parks has been associated with different types of benefits for a
community. Studies have shown that the presence of public green space promotes social
interaction that is often lacking in low-income areas and amongst neighborhoods where income
levels are mixed (Coley 1997; Taylor et al. 2007; Krellenberg et al. 2014). Among the many
benefits of parks, improved physical health has been linked to public park access, especially in
low-income communities (Cohen et al. 2007). Correlations have also been found between access
to public green space and better mental health (Richardson et al. 2013). By investigating the
accessibility to public green space in Downey the following research has contributed to a better
understanding of how smaller communities have differential access to productive environmental
features.
In a review of past public green space studies, Kabish et al. (2015) found that GIS
methods were most often used in studies of environmental justice and urban green space
planning. Considering the importance of GIS analysis in this area of study, detailed methods for
determining accessibility should be explored. The use of dasymetric mapping and network
analysis in this study contributed to establishing a more accurate and detailed model of
accessibility. This benefits the spatial sciences at large because it helps exemplify the ability to
widely improve such methods, a direction that has been encouraged to improve future studies
(Kabish et al. 2015).
The benefit of using Downey as a case study for this research also helps establish a model
for other small cities to engage in their own accessibility analysis. All of the data used in this
study was readily available over the Internet. Increasingly, open source GIS software is also
becoming more readily accessible to people with less experience in the spatial sciences.
5
Considering these two aspects, this study hopes to aid in establishing a method for local
governments to engage in their own accessibility studies to improve urban planning.
1.2 Theoretical Background
The use of park amenities and network analysis in this study have been informed by
previous research which have proven their benefits to urban green space studies. Two studies in
particular have been used as primary inspirations for the methodology used in this study.
Dorothy C. Ibes (2015) work on park type access in the Phoenix Arizona area served as an
inspiration for looking at park amenities. Jonathan Parsons’ (2015) thesis on park accessibility in
Wake County North Carolina served as a model for the use of dasymetric mapping and network
analysis.
Studying park amenities has often been a focus of environmental justice studies of public
green space. Bruton and Floyd (2014) found that parks in North Carolina showed a differing
quality of amenities based on their proximity to minority neighborhoods. While this study lends
to the common findings of previous environmental justice research, other studies have also found
that minority populations use parks in different ways and certain amenities are favored over
others (Stodolska et al. 2011; Li 2014). When creating the methodology for this study, it was
apparent that general park accessibility needed to be qualified by considering the type of
amenities that were accessible.
Ibes (2015) provides a useful model for considering park amenities in urban green space
studies. Ibes used a combination of amenity types for determining a park typology. Using this
process, it is possible to analyze park accessibility in a way that considers the differing and
similar characteristics of parks. By creating a typology, the results of park accessibility have an
added layer of context, which can speak to the quality of the type of park that is accessible to city
6
residents. In this study park amenity and disamenity types were used in a similar fashion to give
context to Downey park accessibility. The idea of disamenities has been utilized in past park
accessibility research (Weiss et al. 2001; Byrne et al. 2005; Lee et al. 2005; Bruton et al. 2014;
Parsons et al. 2015). Disamenities, in this study, are those characteristics of parks that deter or
prevent attendees from utilizing its resources. Some examples of disamenities are broken glass,
evidence of drug use, graffiti, and dead or overgrown grass.
Prior to network analysis the most common form of studying park accessibility was to
use some variant of distance buffers using geodesic measures. (Boone et al. 2009; Sister 2010).
Buffering methods of investigation often neglected street networks, walking paths, and other
forms of urban travel. Network analysis has been increasingly used as a more acceptable method
for more accurately modeling travel distances in an urban environment (Heckert 2013; Morar
2014; Parsons 2015). Using network analysis, models can be built that consider park access
points, travel paths that restrict certain transportation methods or constraints that would make a
short geodesic distance unavailable.
Recent studies, which have used network analysis to study park accessibility, have used
large-scale population measures and have not considered park amenities in analyzing pedestrian
accessibility along road networks (Heckert 2013; Morar 2014). Parsons’ (2015) methodology of
utilizing dasymetric mapping and network analysis improves upon previous studies by creating a
more detailed and accurate model of accessibility. The choice to use network analysis along with
a measure for park amenities in this study has expanded upon both the previous methods of Ibes
(2015) and Parsons (2015) by providing a more realistic model of accessibility that considers the
quality of park facilities.

7
1.3 Study Site
The City of Downey California is located in southeast L.A. County and is estimated to
have a population of approximately 114,000 people according to the 2014 United States (U.S.)
Census. Spatially the city only covers 12.4 square miles and is mostly bordered on all sides by
smaller municipalities. Demographically Downey differs from the California average in that it is
70% percent Hispanic or Latino (Census 2010). Economically Downey is similar to the 2010
California Median Income, coming in at just over $60,000. Compared to its neighboring cities
Downey is a city of relatively wealthy predominantly Latino residents.
The City of Downey has never been previously studied for park accessibility. In terms of
geography, Downey is bordered on both the east and west sides by two riverbeds, the Rio Hondo
and Rio San Gabriel respectively. To the north, the Interstate 5 freeway borders Downey. These
three features form natural travel barriers for those navigating into and out of Downey. Though
small parts of the city are on the outside of these natural borders, all of the cities public parks and
recreation spaces are within them. Figure 1 depicts Downey’s location in the southeast of L.A.
County.

8

Figure 1 City of Downey
Downey, unlike most of its bordering cities also maintains its own school district, parks,
police, and fire services. This is uncharacteristic for the area of L.A. County that Downey is
located in. The cities surrounding Downey are more likely to contract city services to L.A.
County, contain additional county services with in their borders (such as L.A. County owned
parks), or cooperate with neighboring city services (such as the Norwalk-La Mirada unified
9
school district). These factors combined make Downey a much more isolated city in terms of city
services that are available to its citizens, but also give the residents a greater opportunity for local
control of park services.
1.4 Research Question
The intention of the following study was to determine if there was equity in land parcel
access to public parks and their amenities. For the purpose of this study accessibility was seen as
the amount of access each land parcel’s population has to the nearest park in acres per person. In
previous studies, accessibility has been viewed as a ratio that looked at the spatial distribution of
public facilities compared to population totals or demographics of administrative units such as
census blocks. (Sister et al. 2010; Ibes 2015) Using a model of accessibility that focused on land
parcels allowed this study to look at what park accessibility is like for a resident of Downey. This
study also went further to investigate how accessibility differs when looking at specific park
amenities and how it is affected when factoring in park disamenities.
1.5 Manuscript Navigation
The remaining chapters of this manuscript will contain more detail on how the study was
informed, planned, and executed. In an attempt to ease the navigation of the manuscript, quick
summaries for each chapter have been provided.
In chapter 2 of this paper, there is a detailed explanation of the past studies on urban
green space accessibility. Past methods and models for investigating environmental justice and
park amenities is explored. Closer examination is also given to the use of network analysis in
public park accessibility research.
Chapter 3 discusses the methods that were used to perform the research. There will be
clearer descriptions on how methods from Ibes (2015) and Parsons (2015) have been combined
10
in a more meaningful way. Data sources will be revealed and all data processing will be detailed.
There is also an explanation of the software and analytical tools that were used to accomplish
this study. Finally, the details of the workflow for the dasymetric and network analysis are
outlined.
Chapter 4 contains the detailed results of service area analysis done on the parks of
Downey. There is an exploration of what was learned about park accessibility and the
effectiveness of the methodology used. There are detailed explanations on how each method
used provided its own insight into the dynamics of park accessibility in Downey.
Chapter 5 contains a comprehensive discussion about the implications of the results from
this research. Potential improvements and limitations of the research are explored in detail. There
are also suggestions for future research and potential policy directions for the City of Downey.

11
Chapter 2 Literature Review
The use of GIS methods to analyze park accessibility is not in itself a novel concept.
Advancements in methodology have helped improve models of park accessibility by embracing
network analysis, making finer scale population estimates, and quantifying amenity
characteristics. Though research exists that combines any two of these aspects in one study, no
study has yet established a methodology that combines all three. In this chapter, previous
research on parks and accessibility will be explored as a means for justifying the need for a
methodology that utilizes recent advancements in population estimates, network analysis, and
park amenity classifications.
This chapter will first give background on the history of urban parks and their use in
section 2.1. The next section (2.2) then provides some further clarification on environmental
justice and the motivations behind contemporary accessibility studies. Section 2.3 briefly
summarizes advances in spatial population modeling for parks using dasymetric mapping.
Section 2.4 is a comprehensive look at past accessibility methods which utilize geodesic
measures both with and with out dasymetric population models. Section 2.5 provides a detailed
exploration of the current state of network analysis in park accessibility studies. In the final
Section 2.6 the study of park amenities will be carefully examined.
2.1 Brief histories of parks
The term public space itself is not an easy term to define. In her creation of a public space
index, Mehta (2013) explains that public space does not just pertain to who owns the land.
Instead Mehta (2013) explains that public space is effective only when it satisfies five
components. These components require the space to be inclusive, support meaningful activity,
12
pleasurable, safe, and comfortable. Though parks can be seen as an archetypal example of this
definition of public space, this has not always been true.
The meaning in an urban landscape has evolved over time. Parks were originally seen as
areas where nature could be kept and put on display only for those in higher socioeconomic
classes (Byrne et al. 2009). Over time this meaning changed, and parks were viewed as having
more of a democratic purpose where people of different backgrounds could interact (Byrne et al.
2009). As recently as the late 1990’s, Coley et al. (1997) found that the inclusion of trees and
other elements of natural landscaping attracted people and promoted social interaction in an
urban environment. These areas of natural landscaping are generally referred to as public green
space and have been found to also benefit urban ecology and public health (Wolch et al. 2014).
It is because of these benefits that there has been much concern about the idea of public green
space in urban environments.
Though the benefits of public green spaces have been recognized for some time, it has
only been more recently that the differential access to these spaces have been studied using
detailed spatial methods (Wolch et al. 2005; Cohen et al. 2007; Taylor et al. 2007; Byrne et al.
2009; Kabisch et al. 2014). Though there is a general consensus on the desire to increase public
green space in cities, methods have not always been effective for those most in need (Wolch et
al. 2005; Joassart-Marcelli 2010; Wolch et al. 2014). Efforts to simply increase the acreage of
public green space by appropriating unused industrial sites such as rail yards or concrete
riverbeds have not always lead to equitable access (Wolch et al. 2014). Studies have shown that
programs such as park bond funding and relying on federal or state investment can actually
exacerbate park accessibility issues (Wolch et al. 2005; Joassart-Marcelli 2010).
13
Parks have historically been areas of minority exclusion (Byrne et al. 2009). In light of
this, contemporary park planners increasingly look at how ethnically and socioeconomically
homogenous areas can be better served by strategic park location and design (Wolch et al. 2005;
Byrne et al. 2009; Li 2014; Wolch et al. 2014). Contemporary studies concerning park
accessibility often stem from the differing accessibility of demographically homogenous
neighborhoods and the ideas of environmental justice.
2.2 Parks and Environmental Justice
Initially environmental justice was seen as the idea that ecological and spatial
characteristics of a persons surrounding should not be unduly burdened by the presence of
unhealthy locations such as landfills and refineries (Bullard 1994). As research in the field grew,
environmental justice advocates also started looking at the lack of access to neighborhood
improving services (Taylor et al. 2007).
Parks have been correlated with many benefits for neighborhoods including improved
health, improved community engagement, improved property value, and improved
environmental conditions (Coley et al. 1997; Cohen et al. 2007; Taylor et al. 2007). The
prevalence of parks in neighborhoods of homogenous ethnic and socioeconomic demographics
has often been found to be lacking (Cohen et al. 2007; Taylor et al. 2007). It is because of this
finding that quantifiable methods for measuring accessibility have been increasingly important
for both environmental justice advocates and city planning researchers (Kabisch et al. 2014).
2.3 Measuring Population
When studying the accessibility to an urban amenity such as a public park, it is important
to use a scale of reference that appropriately models park accessibility issues. Zhang et al. (2011)
provides an example of how traditional methods for measuring accessibility have been
14
approached. In their study, accessibility was measured at the national level using census tract
level data (Zhang et al. 2011). The challenges in using this method are that local accessibility
issues can be lost at this scale of analysis. In a review of methodology for studying parks,
Kabisch et al. (2014) explains that the current direction of research is moving towards more
localized case studies. In order to increase the effectiveness of more localized studies, new
methods for analyzing population have been explored.
Dasymetric mapping is the disaggregation of data into more meaningful spatial
distributions (Eicher et al. 2001). Eicher et al. (2001) describes two basic methods for dasymetric
mapping, the first is a binary raster based method and the second is a vector-based method,
which uses limiting variables. The raster method converts data into a field of cells which cover
the entire spatial extant of the study area and contain a binary value (Eicher et al. 2001). In terms
of studying population, this would mean each cell in the field would either carry a value
representing the presence or absence of population. This binary method helps distribute data
more accurately across an area that may not be uniformly populated (Eicher et al. 2001).
The limiting variable (vector based) method disaggregates population data in one set of
polygons to another more meaningfully distributed set of polygons (Eicher et al. 2001). The
population value for each new polygon is proportional to a limiting variable (Eicher et al. 2001).
In the study by Eicher et al. (2001), population amongst an entire country was disaggregated
according to land type and each land type was given a maximum allowable population density
(people per square kilometer) to determine its population. The conclusion of the study found that
the limiting variable (vector based) method of dasymetric mapping was a more accurate
representation of a population’s spatial distribution.

15
2.4 Past Methods for Measuring Park Accessibility
In their review of public space research Kabisch et al. (2014) suggests that research into
the accessibility of parks should be scalable and utilize methods which are comparable to those
used by park planners. The use of a GIS to analyze park accessibility satisfies both of these
concerns and it is for this reason that the following section will focus its attention on GIS
methodology. In terms of GIS methods there are two general areas that vary in their execution
amongst past studies. First, some studies use of geodesic measures of distance while others use
network distance. Second, some studies use dasymetric-mapping techniques for modeling
population while others incorporate whole census tracts or take simple proportional percentages
of buffers that intersect tracts.
Of the four methodological combinations possible based on the above variations, the use
of Euclidian distance measures and non-dasymetric mapping is the least complex. The most
complex method for GIS workflow is network distances with dasymetric mapping. The
advantages and disadvantages of each combination of methods will be discussed further in the
remainder of this section.
The use of a physical survey of park guests is often associated with contemporary studies
of park accessibility (Cohen et al. 2014; Rossi et al. 2015). In studies by both Cohen et al. (2014)
and Rossi et al. (2015) surveys of park visitors were used to examine park accessibility. In the
study by Cohen et al. (2014), a linear half-mile buffer was used to create the survey area for
examining community accessibility to parks. The research by Rossi et al. (2015) surveyed
attendees on site and self disclosed attendee origins were aggregated into neighborhoods. In both
cases, GIS methods were only lightly utilized in conjunction with surveying methods and both
were not easily scalable.
16
Lara-Valencia et al.’s (2013) study of community access to public spaces used more
common GIS methods for accessibility. In the study, Lara-Valencia et al. (2013) used centroids
for parks and neighborhoods to determine accessibility by using linear buffers around park
centroids to determine service areas. These methods oversimplify park access by using centroids
as destinations and risk misrepresenting access by aggregating populations by colonias, which
are Mexican national population measuring units generally larger than census block groups in
U.S. terms (Lara-Valencia et al. 2013). In a 2012 study, Hewko et al. found that disaggregating
population data helps increase accuracy of service area estimates for amenities such as
playgrounds.
Dasymetric mapping is a commonly used way to provide a more accurate model of
population when studying accessibility (Langford et al. 2006; Oh et al. 2007; Boone et al. 2009;
Maroko et al. 2009; Sister et al. 2009; Maantay et al. 2013; Morar et al. 2014; Parsons 2015).
Maroko et al. (2009) is uncommon in relation to most contemporary studies in that they used a
raster data format to model population and public park space. The kernel density method, which
was utilized, transformed census block group population data, in the form of centroids and park
spaces into two fields of cells (Maroko et al. 2009). However, using this method did not allow
for the utilization of the distinct boundaries to travel that exist in an urban landscape. Using a
similar model for population, Sister et al. (2009) utilized a combination of the Landscan
Population Grid and theissian polygons around park centroids to determine the “PSA (Park
Service Area)”, to study accessibility. In this instance the model determining service areas could
also be improved because they utilized linear distances to determine the “PSA” for each park.
Maantay et al. (2013) and Boone et al. (2009) both provide a method for modeling
population that utilizes both dasymetric mapping and the natural boundaries of an urban
17
environment. In both studies dasymetric methods were used to disaggregate census data to the
scale of land parcels (Boone et al. 2009; Maantay et al. 2013). The use of parcels as a unit for
measuring population is ideal because it represents the actual location of were people live. In the
case of the Maantay et al. (2013) study, dasymetric mapping of the population was used along
with a linear buffer around roads to model exposure to air toxins from car traffic. This was a
more appropriate model for this study because air toxins can travel linearly and are not limited to
network travel. The dasymetric mapping served to model the population while the linear buffers
modeled the reach of possible air borne toxins. This type of linear measurement was also used by
Boone et al. (2009) to determine park service areas. Linear geodesic measures are not as
appropriate for accessibility models because the nature of the urban landscape means that people
cannot navigate directly to a destination, instead they must move within predetermined routes of
travel.
2.5 Network Analysis and Park Accessibility
When studying accessibility in an urban environment it is difficult to ignore how streets
and other routes affect travel. Network analysis is a tool that allows one to study accessibility
while also recognizing the travel limitations inherent in a city. In separate studies Nicholls
(2001) and Sarah et al. (2001) compared methods of studying accessibility to public parks, both
found that network analysis provided a more accurate estimate then traditional geodesic buffers.
One example of network analysis being used to study accessibility was done in 2006. In
this study Apparicio et al. (2006) used network analysis to determine the access to various
services from public housing locations. Each service (destination) and public housing complex
(origin) was visualized as a point to determine what the true distance is between each when
accessing a service via the street network. Ultimately these methods would not be directly
18
applicable to park studies because each park was visualized with only a single access point and
each public housing complex had no population value so service density could not be estimated
(Apparicio et al. 2006).
Heckert (2012) provides us with an example of an accessibility study that utilized
network analysis and population estimates. He used the percentage of census block groups that
fell within a 0.5-mile of street access to a park. The difficulty in utilizing this method is that
census blocks do not account for differing population densities. The assumption that block
groups have uniform population density, even in urban environments, does not account for areas
that have been zoned for higher density habitation (i.e. apartments) or lower density habitation
(i.e. industrial zones). These concerns also plague similar methods, which utilize postal zip codes
to help model population distributions (Comber et al. 2008).
The use of network analysis in conjunction with dasymetric mapping methods has shown
much promise in recent research on accessibility. Langford et al. (2006) used network analysis
and dasymetric mapping in a raster data format to study accessibility to healthcare services. In
their research, they rasterized both the street networks and population in their study area
(Langford et al. 2006). As was used in past studies, population was modeled in a binary fashion
which meant that each cell in the field had a value which meant it was either populated or not
(Langford et al. 2006). Though this method was found to be more effective then population
estimates that assumed uniform population across a polygon (Langford et al. 2006), it still did
not provide specific population densities for the study area.
Using similar methods, Morar et al. (2014) and Oh et al. (2007) used a more detailed
dasymetric mapping method for modeling population. In both studies, the square footage of
buildings was used to help estimate population densities (Oh et al. 2007; Morar et al. 2014). In
19
both cases, the method provided a more detailed way to model population, but the assumption
that population density was the same from house to house also leaves room for errors. In the case
of Morar et al. (2014), these concerns were even more pronounced because aerial imagery was
used to estimate building square footage. This method is also not easily scalable since different
urban areas have unique building methods and detailed data on square footage is not always
easily accessible.
The final example of network analysis and dasymetric mapping that will be discussed is
by Jonathan Parsons (2015). In this study Parsons used census block data and land parcel data to
create a more accurate model of population in Wake County, North Carolina (Parsons 2015).
Using this method Parsons (2015) was able to create a more accurate spatial distribution of the
population that considered both the areas of residential zoning and the approved habitation for
each land parcel (i.e. single family home). In addition to this Parsons (2015) also visualized the
parks with multiple access points for the network analysis. Parsons (2015) methods currently
provide one of the best options for studying park accessibility in a city. The one area which has
been given little to no attention in these previously reviewed studies is the idea of accessibility to
specific park amenities.
2.6 Measuring Park Amenities
The National Recreation and Park Association (NRPA) is the long cited authority on park
planning in the United States. According to the NRPA, cities or communities should set local
standards for which park amenities are essential. Since this means there is no nationally set
standard for park amenities there have been various methods devised to study this area. In the
following section there will be a deeper investigation into the different ways park amenities
shave been studied.
20
The Trust for Public Land maintains an assessment instrument for comparing park
systems between cities (Trust for Public Land 2015). The assessment instrument is known as
ParkScore (Trust for Public Land 2015) and it uses a series of measures based on the areas of
acreage, accessibility, investment, and facilities to determine a cities park score. Parks systems
are scored out of a total of “100” and the assessment outputs a map that indicates areas serviced
by parks and those that are in need (Trust for Public Land 2015).
The advantages of this type of assessment are that city park systems can be compared to
each other, and there is associated age and income demographics with the results (Trust for
Public Land 2015). The disadvantage to this system is that park amenity data is aggregated into a
single score (Trust for Public Land 2015), so accessibility to certain types of park amenities
cannot be investigated. ParkScore (Trust for Public Land 2015) also only uses a single 0.5-mile
service area distance to determine resident accessibility to parks. In general, ParkScore ‘s (Trust
for Public Land 2015) purpose is to investigate an entire park systems quality and give a
generalized result of accessibility needs.
Previous work by Moore et al. (2008) and Weiss et al. (2011) utilized raster data formats
to create density fields of park features. In the study by Weiss et al. (2011), the focus was
specifically on features referred to as “disamenities.” These were characteristics such as
incidences of murder, traffic fatalities, and other characteristics that would discourage park
attendance (Weiss et al. 2011). In both cases the raster data format generalized the data in such a
way that dynamics in specific neighborhoods were not discernible.
Dorothy Ibes (2015) provides a more detailed view of park amenities. She measured
various amenities that were available in parks and performed a principal components analysis to
form a park typology (Ibes 2015). Once each park was categorized into a type, GIS methods
21
using linear buffers were used to determine population accessibility to park types (Ibes 2015).
While this method is effective at maintaining the detail of a vector data format, it places each
park in a typology and does not allow for the consideration of parks serving multiple purposes
for a community.
There are many different auditing instruments that exist to systematically survey public
parks (Byrne et al. 2005; Lee et al. 2005; Saelens et al. 2006; Greer et al. 2014). The goal of each
of these audit instruments is to evaluate the many different amenities and in some cases
disamenities of parks. Disamenities in this case are aspects such as graffiti or presence of litter,
which may keep people from utilizing parks. Many of these instruments use Likert scales to not
only record the presence of features but also the condition or severity of each feature.
Some auditing instruments also focus on park amenities like recreational equipment or
landscapes rather than the general quality of the parks environment. The Environmental
Assessment of Public Recreation spaces (EAPRS) and Systematic Audit of Greenspace
Environments (SAGE) are both good examples of audit surveys that seek detailed information on
the amenities at parks but have less focus on any observational disamenities (Byrne et al. 2005;
Saelens et al. 2006). In the case of SAGE the recorded disamenities (i.e. Litter) are binary (i.e.
present or not present) and do not allow for a more multi dimensional examination (Byrne et al.
2005). EAPRS however does provide a scale by which to rate disamenities but the audit tool uses
vague indicators (i.e. perceived degree of safeness) as opposed to specific occurrences of
disamenities (Saelens et al. 2006). Though both tools have been shown to be effective in
previous studies (Sister et al. 2009; Bruton et al. 2014) they do not provide as comprehensive a
view as other audit tools.
22
The Community Park Audit Tool (CPAT) and Physical Activity Resource Assessment
(PARA) are two audit tools that provide more detail on disamenities (Greer et al. 2014; Lee et al.
2005). While CPAT provides more areas to highlight potential disamenities in public parks,
PARA provides a more detailed audit instrument because it allows for a degree of severity for
each disamenity listed (Greer et al. 2014; Lee et al. 2005). PARA also provides a more thematic
recording of amenities that focuses on both the physical landscape of the park and general
features that are present (Suminski et al. 2012).
This study draws on the strengths of the various audit instruments discussed above. The
next chapter explains the auditing instrument developed for this study and also how this study
has combined previous methods in park amenity studies, network analysis, and dasymetric
mapping to create model for examining park accessibility in Downey, California.
23
Chapter 3 Methods
Creating a quantifiable method for measuring park amenity accessibility in Downey, CA
involved three general steps: dasymetric mapping of the population, collecting and scoring park
amenity data, and performing service area analysis. The overarching objective was to create park
accessibility metrics at the level of residential parcels for the entire city. Each step of the
methodology was done to specifically address a challenge faced in previous studies. Dasymetric
mapping was used to create a more detailed spatial distribution of population density. The
collection of amenity data and scoring procedure provided a quantifiable method for comparing
park amenity quality. Network service area analysis was used to provide the most accurate model
of accessibility in an urban environment.
The use of dasymetric mapping involved redistributing census block population data to
the scale of land parcels. This step was needed because the results of the study were visualized as
accessibility values for each individual parcel. After the dasymetric mapping was done, amenity
data for the parks in the study area was collected and each park scored on the basis of four
categories. Three categories represented the types of amenities available for park users, and one
scored for disamenities like graffiti or litter that would make a park less attractive to park users.
Available park acres were adjusted based on these scores.
The third and final step in the methodology involved the use of network service area
analysis to determine accessibility values for each parcel in a park’s service area. The service
area analysis provided a means for determining, which parcels parks serviced, and which parcels
parks did not service. Using service area analysis allowed for the assignment of park amenity
scores to parcels in Downey. The result of the service area analysis allowed for a visualization of
how parcels in Downey differed in their access to parks. All of the analysis, unless otherwise
24
stated, was done using Esri’s ArcGIS 10.1 software. All data sets for the analysis were on the
North American 1983 Geographic Coordinate System. For the purposes of spatial measurements
and mapping all data for this study was projected using the State Plan Coordinate System
California zone V FIPS 0405. The following chapter will provide a further details of the
methodology used in this study.
3.1 Dasymetric Mapping
There were two main data sets that were imported into ArcGIS for the dasymetric
mapping: U.S. Census population data and land parcel data provided by the L.A. County Office
of the Chief Information Officer and L.A. County Enterprise GIS (LAeGIS). A third data set of
political boundaries for L.A. County was also procured from LAeGIS and was used to clip all
datasets to the boundaries of the study area.
The U.S. Census data contained population estimates for census block groups according
to the 2010 U.S. Census. Population data was downloaded as a data table and joined to a polygon
data set of census block polygons based on the GEO.id field. Downey is roughly divided into
approximately 83 census blocks. Figure 2 depicts the population of Downey and all census
blocks that intersect within a 2-mile buffer of the city (study area). The figure shows a rough
estimate of population for the city and all census blocks within the greater study area. In general,
it can be seen that more people live in Downey’s southern than northern areas. The census block
lacking data in the southwest section of the city contained no residential parcels and so this study
was unaffected by its absence.
25

Figure 2 Downey Population per Census Block
Residential land parcel data for the city of Downey was obtained from a larger land
parcel data set for all of L.A. County. The land parcel data set was downloaded from the LAeGIS
data portal and was compiled by the L.A. County Office of the Assessor. The data set was a
polygon data set of 2014 land parcels along with a vast amount of information on each parcel’s
zoning designation and attributes. Initially, the parcel data was queried to include only those
26
parcels that were classified as residential. In past dasymetric mapping efforts, parcels were then
further reclassified depending on the general zone classification for each parcel (Parsons 2015).
While the parcel data set had general classifications for each parcel (single family, duplex, multi-
unit, etc.) these were not needed for this study because the data set contained a field specifying
the exact number of residential units for each parcel. Due to the authoritative source of the
parcel data, no further testing of the data set was done.
In order to create a smaller land parcel data set for the study, the census block polygons
of the study area were dissolved into a single polygon and used to select a subset of parcels. Any
parcel whose center was contained within a census block was retained for the analysis. Since the
study area expanded 2-miles beyond the boundary of Downey, all residential parcels in the city
were retained. The 2-mile buffer used to determine the study area was chosen based on the
maximum length that would be used in the service area analysis. By clipping all data sets
according to the 2-mile buffer, edge effects for Downey parcels were modeled. On the one hand,
some of Downey’s parcels are serviced by neighboring city parks in adjacent cities. On the other
hand, some Downey parks may serve residential parcels in neighboring cities, potentially
contributing to overcrowding at these parks.
In Figure 3 the distribution of residential parcels for the study area is shown. The
distribution of the parcels shows how the census block data had aggregated data across sections
of the city that were not populated. Areas of Downey in the southeast and southwest show large
areas where there are no residential parcels. However, if they are visualized solely on the basis of
the census block data, they appear to be heavily populated areas. Figure 3 also depicts the parks
of Downey and those in neighboring cities that were in the study area. All parks outside of
27
Downey were given a 2-mile service area (as defined in Section 3.5) and only those parks whose
buffer entered or immediately bordered Downey parcels were included.

Figure 3 Downey Residential Land Parcels
The calculations for the dasymetric mapping were done using the field calculator,
summary statistic and spatial join tools in ArcGIS. In order to do the calculations each residential
28
parcel needed to be assigned to a census block. To do this, a spatial join was done with the
census block polygons and the land parcel polygons. Land parcels were assigned to census
blocks based on the location of the land parcel centroids. To verify the parcel, count the number
of records in the spatial join data set was compared to the original parcel data set and both were
found to be equal.
The ultimate goal of the calculations was to determine the Population Per Parcel (Pp). In
order to determine this value, there were two main formulas that were used. The first was the
calculation for People Per Unit (Pu) for each parcel and the second for Pp. To determine the Pu
for each parcel the total population for each block group (B) was divided by the sum of all units
(Bu) for every parcel whose centroid resided in it. Spatially joining the parcels to a census block
group and running a summary statistic to add all units for parcels that belonged to a common
census block, achieved this. The second calculation to determine the Pp for each parcel was done
using the total housing units field (U) provided in the land parcel data set. Each parcels’ total
units (U) was multiplied by the Pu to determine the total population (Pp) for each residential land
parcel. The formulas below summarize the calculations done to determine the Pu and Pp for each
parcel.
Pu
(i…j)
= B /Bu

(1)
Pp
(i…j)
= Pu (U
(i…j)
) (2)
The results of dasymetric mapping calculations allowed for a finer scale spatial
distribution of the population in the study area. Figure 4 shows the results of the dasymetric
mapping that was done on the study area. According to the dasymetric analysis, the vast majority
of Downey consists of land parcels that are mostly within a common range, with some larger
multi-unit parcels in the central northern and southern parts of the city. The area void of
29
residential parcels toward the geometric center of the city is the area known as Downtown
Downey and it contains one of the largest clusters of retail businesses in the city. The areas in the
western part of the city with no population represent a mix of commercial, industrial, and newly
developed retail areas. This parcel population data was eventually used in conjunction with the
service area and amenity data to determine accessibility.

Figure 4 Downey Population Per Parcel
30
3.2 Collection of Amenity Data
Amenity data was collected for a total of 20 parks that were both in and immediately
surrounding Downey. A polygon data set of land type data was downloaded from the LAeGIS
data portal. The land type data set is maintained by LAeGIS but its sources are varied. The data
itself contained many different types of polygons including but not limited to schools,
businesses, and various government services. Within the attribute data was a classification
scheme that was queried to isolate public parks from the data set. Other then the fields used for
land type classifications and city, no other attribute data was needed for the analysis.
In order to account for any inconsistencies in park location or size, I verified all parks and
park locations by referencing the City of Downey Parks and Recreation site and visually
comparing the land type data to aerial imagery. Three parks (Temple, Discovery, and Treasure
Island) were within Downey and were added to the dataset by referencing aerial imagery. As
mentioned above, the study includes all parks outside of Downey’s municipal boundary that fell
within service areas for Downey parcels. These non-Downey parks whose service areas extended
into the city were scored for amenities. Details on the service area analysis will be explained
further in Section 3.4.
Table 1 lists all the parks, which were scored for amenities in the study along with the
corresponding acreage and the city in which it resides. There were a total of 12 parks in Downey
and 8 parks that resided in neighboring cities. The largest park in Downey is just over 24 acres,
and the largest two parks are located just outside of Downey’s boundaries in neighboring cities.
Figure 5 depicts the distribution of park acreages listed in Table 1. The mean value for the parks
was 13.2 acres with a standard deviation of 13.1. In Figure 5, the histogram shows a strong skew
to the right with most acreage values falling below 20.
31

Table 1 List of Parks in Study Area
Name Acres City
John Anson Ford Park 52 Bell Gardens
Hollydale Park 39.3 South Gate
Wilderness Park 24.4 Downey
Rio San Gabriel Park 17.5 Downey
Furman Park 14.4 Downey
Independence Park 14.2 Downey
Thompson Park 13.9 Bellflower
Santa Fe Springs Park 12.8 Sante Fe Springs
Apollo Park 11.9 Downey
Discovery Sports Complex 10.7 Downey
Veterans Memorial Park 9.7 Bell Gardens
Golden Park 8.1 Downey
All American Park 6.5 Paramount
Dennis the Menace Park 4.7 Downey
Lakeside Park 4.0 Norwalk
Crawford Park 3.7 Downey
Treasure Island Park 3.2 Downey
Byron Zinn Park 2.8 Bellflower
Brookshire Children's Park 1.2 Downey
Temple Park 0.4 Downey

32

Figure 5 Histogram of Park Acreage Values
Before the collection of the amenity data could begin, an auditing instrument was
selected to base observations on. The Physical Activity Resource Assessment (PARA)
instrument was chosen as starting point for collecting data on park amenities (Lee et al. 2005;
Suminski et al. 2012). PARA is an audit instrument that was designed to measure multiple types
of facilities that provide physical activity opportunities for people. PARA’s operational
definitions and protocols (see Appendix A) provided a simple established framework for
gathering data. The areas measured by PARA are generally divided into three different areas:
33
features, amenities, and incivilities. With the exception of the incivilities, these areas did not
remain as operational categories for the analysis. Each survey item on the PARA instrument is
presented as a four point Likert scale that ranges from 0-3
Preliminary data collection was done from October 23-24, 2015 between the hours of
11am-5pm. During this time, all 20 parks were surveyed for all the aspects listed in the original
PARA instrument. Supplementary visits to audit additional aspects not included in the original
PARA survey were done from November 1-2, 2015. The weather was between 65-75 degrees
Fahrenheit with no rain during all collection times and there were no special community-wide
events that occurred at any one particular park during the audit process that might have affected
park use. In addition to the 49 categories listed on the PARA instrument, there were an additional
seven survey items added for the purposes of this study (see Appendix B). The seven items
added to the survey included physical activity areas, nature access amenities, and various types
of disamenities that were not listed in the original PARA instrument. For the purposes of this
study, the operational definitions of PARA were augmented to accommodate the park amenity
scoring method used.
3.3 Park Amenity Scoring
The method used for scoring the amenities for each park necessitated a grouping of the
individual PARA survey items and supplementary items. All the data surveyed from the parks
was aggregated into four general categories: play, disamenity, physical activity (PA), and access
to nature (NA). The Children’s play (play) category covered park amenities such as playgrounds
that would enhance park use for young children. The disamenity category measured incivilities
such as broken glass and graffiti, which would discourage park use. The physical activity (PA)
category covered those amenities, which would be utilized by older park patrons such as adult
34
size basketball courts and exercise equipment. The final category, nature access (NA), measured
park amenities such as nature guide signage and amount of natural landscaping, which would
satisfy patron’s desires to connect with nature. Each park contained some mixture of
characteristics from each of these categories and none of them belonged exclusively to any one
of them. The distance bands for service areas for each category were also different to account for
the different types of access expected for each. A comprehensive look at the entire list of 56
items that were scored for each park is included in Appendix B.
Each category was scored based on a percentage of possible points. For example, the play
category had a total of 12 possible points, which could be earned if a park scored a “3” for each
of the 4 items that were measured. A park’s initial score for each category was determined by
dividing the point’s earned in each category by the total possible points. This initial score (Sx)
was then multiplied by the total acres (Pa) of the park to determine the number of category acres
provided by a given park (Ax). These calculations are summarized in the formulas listed below.
Ax
(i…j)
= Sx (Pa) (3)
In this formula, Sx is alternatively the play Score, disamenity Score, PA Score, or NA Score.
Calculating the disamenity-adjusted acres (Ad) required a further calculation that involved
subtracting the disamenity acres (Ax) from the total acres (Pa) for each park.
Ad = Pa – Ax (4)
Table 2 lists the category scores for each park in the study area along with the total
possible points for each category. Generally, every park had a score for each category with the
exception of Byron Zinn Park, which had no children’s play amenities. Because scores are
relative within each amenity category, it is not appropriate to compare accessibility values across
amenity categories.
35
Table 2 List of Category Scores for Parks
Park
Nature
Points
Nature
Score
(Sx)
Play
Points
Play
Score
(Sx)
PA
Points
PA
Score
(Sx)
Disamenity
Points
Disamenity
Score (Sx)
All American
Park
6 0.4 3 0.25 12 0.36 31 0.74
Apollo Park 1 0.07 9 0.75 25 0.76 23 0.55
Brookshire
Children’s Park
5 0.33 2 0.17 8 0.24 31 0.74
Byron Zinn Park 2 0.13 0 0 3 0.09 35 0.83
Crawford Park 2 0.13 2 0.17 6 0.18 33 0.79
Dennis the
Menace Park
4 0.27 7 0.58 7 0.21 32 0.76
Discovery Sports
Complex
5 0.33 3 0.25 13 0.39 38 0.9
Furman Park 6 0.4 7 0.58 26 0.79 33 0.79
Golden Park 3 0.2 6 0.5 15 0.45 34 0.81
Hollydale Park 4 0.27 4 0.33 15 0.45 28 0.67
Independence
Park
2 0.13 5 0.42 15 0.45 26 0.62
John Anson Ford
Park
7 0.47 10 0.83 25 0.76 30 0.71
Lakeside Park 1 0.07 2 0.17 8 0.24 35 0.83
Rio San Gabriel
Park
4 0.27 5 0.42 18 0.55 34 0.81
Sante Fe Springs
Park
4 0.27 7 0.58 14 0.42 24 0.57
Temple Park 4 0.27 3 0.25 3 0.09 42 1
Thompson Park 3 0.2 6 0.5 20 0.61 38 0.9
Treasure Island 7 0.47 4 0.33 7 0.21 36 0.86
Veterans
Memorial Park
4 0.27 7 0.58 19 0.58 34 0.81
Wilderness Park 11 0.73 6 0.5 13 0.39 29 0.69
Total Possible 15 1 12 1 33 1 42 1

The play category consisted of items that would be predominantly utilized by children.
This category consisted of items such as playgrounds, sand boxes, community centers, and
wading pools. Community centers were the only item that was supplemental to the original
PARA instrument. According to the PARA protocol, each item on the instrument was rated on a
scale from 0-3, with “0” indicating complete lack of the amenity and “3” being an indication that
the amenity is in excellent operating order. In addition to scoring whether equipment was in
working order, this also considered size and/or number of the amenity. When considering
community centers, both the size and components of the center that served children were
36
considered. An example of a community center with the value of “3” was one where there were
advertised structured activities for young children along with facilities to house them. Minimal
community centers, which were scored as “1,” advertised activity using only the existing open
park area and any constructed facilities were only for the park staff.
Playground equipment was central to the play score. If playgrounds were in perfect
working order but were very small relative to the rest of the park or had limited capacity, they
were given a lower score. In Figure 6 there is an example of two different playgrounds, which
would have both received a score of “3” on the original PARA instrument but were scored
differently according to the adjusted standards for this study.

Figure 6 Furman Park Playground vs. Independence Park Playground

On the left in Figure 6 is an image of a playground at Furman Park and on the right is an image
of a playground at Independence Park. Both parks are about the same size, but Furman Park’s
playground is substantially bigger and can accommodate more children. In this case Furman Park
received a score of “3” and Independence received a score of “2”. In general, the goal was to
determine the relative amount of playground acres for each park by auditing the prominence of
the playground in the park.
37
Due to the nature of the PARA instrument, the PA category was more detailed than the
NA and play categories. In short, the PA category contained items that accounted for structured
physical activities such as basketball courts, soccer fields, handball courts, and baseball
diamonds (for complete list see Appendix B). As with the play category, the PARA protocols
were augmented so that the survey considered both the condition and the prominence of the PA
amenity within the park. The two items added to the PA category were horseshoe stations and
handball courts. When surveying the PA items, special attention was paid to distinguishing areas
designed for a specific activity from areas adapted for physical activities by the public. In Figure
7 there are two examples of PA use at different park facilities. The picture on the left is an empty
field that is being utilized for a soccer game and on the right is a tennis court that has specifically
been erected for this activity.

Figure 7 Improvised and Specifically Designed PA Amenities

In this example, the park with the empty field that was being utilized effectively as a soccer field
received a score of “0” for that amenity since the people at the park are the ones who supplied
the resource to utilize the space as a soccer field. However, the park depicted on the right did
receive a score for the presence of a tennis court, even though the score was lowered to “1”
because the net was broken and the court was in general poorly maintained.
38
The NA category largely consisted of additional items that were added to those used in
the PARA audit. The intended purpose of the NA category was to quantify the accessibility to
nature that is provided by public parks. Only two of the items in the PARA audit were used for
this category: landscaping efforts and fountains. In each case the original protocols for the PARA
audit were used. The supplemental items that were added to this category were access to the
riverbed, nature guide signage, and tree coverage.
Due to value of the San Gabriel and Rio Hondo riverbeds as a natural resource, park
access to river adjacent walking and bike trails was surveyed as an added value to park attendees
who use parks to access nature. Since all the parks surveyed were not directly adjacent to
revitalized areas of the riverbed, this type of nature access was simply scored a “1” if present and
a “0” if absent. Interpretive signs for nature were only surveyed in a couple of parks in the study
area. However, these warranted their own score because of the important role it played in
assisting attendees in accessing nature. The scores for the nature guides ranged from a single
informational sign (1) to a series of signs or activities through out the park (3). Nature guides,
which were vandalized or not maintained, received reduced scores no lower than “1.”
In order to score the tree cover for each park, the iTree Canopy v6.1 web-GIS application
was utilized (United States Forest Service 2015). The iTree Canopy software allowed for a
convenient method for auditing tree canopies using aerial imagery. iTree Canopy requires an
auditor to draw a closed polygon over an area of interest in an aerial image. The auditor is then
asked to confirm or deny the presence of a tree on a specific location of an image that is
randomly sampled by the program. Over the course of the sampling, iTree Canopy can reduce its
error rate at predicting the percentage of tree cover in the area of interest.
39
Table 3 lists the iTree derived tree cover values and the corresponding standard error for
each park in the study area. Each park was sampled in iTree Canopy 150 times or until the
standard error fell below 4%. To determine the scores for each park, the mean of percent tree
canopy coverage for all parks (Mean = 20.53%) was used as the mid point. All parks with tree
canopy percent coverage lower than half a standard deviation (SD = 9.89) below the mean
received a score of “1,” and all those with values at least half a standard deviation above the
mean were given a “3.”
Table 3 Tree Amenity Values
Park %Tree %Non-Tree SE Score
All American Park 42.7 57.3 4.04 3
Apollo Park 11.7 88.3 3.9 1
Brookshire Children’s Park 22.5 77.5 3.96 3
Byron Zinn Park 14.6 85.4 3.9 1
Discovery Sports Complex 12.8 87.2 3.79 1
Crawford Park 8.6 91.4 3.86 1
Dennis the Menace Park 46.7 53.3 4.07 3
Furman Park 27.6 72.4 3.96 3
Golden Park 36.7 63.3 3.98 3
Hollydale Park 16.3 83.7 3.98 1
Independence Park 20.4 79.6 3.97 2
John Anson Ford Park 24.1 75.9 3.97 3
Lakeside Park 12.2 87.8 3.9 1
Rio San Gabriel Park 23.7 76.3 3.98 3
Sante Fe Springs Park 28.8 71.2 3.94 3
Temple Park 11.7 88.3 3.9 1
Thompson Park 12.2 87.8 3.61 1
Treasure Island 16.3 83.7 3.98 1
Veterans Memorial Park 22.6 77.4 3.9 3
Wilderness Park 19.4 80.6 3.9 2

When surveying the items for the disamenity category, the PARA protocols were
followed very closely. Many of the disamenities were scored based on the number of times the
items appeared. During each survey, the auditor spent between 20-40 minutes at each park
depending on the size. The parks were surveyed over as large a span as was possible. Generally,
40
the disamenity category recorded items such as litter, evidence of drug use, graffiti, and noise
pollution (for complete list see Appendix B). The only item that was added to this category was
the presence of power lines, since power lines are obstructive to the enjoyment of the natural
landscape of the park and limit the ability to use areas of the park. The prevalence of power lines
on park grounds was scored depending on the severity of their intrusion.
3.4 Service Area Data Preparation
Service areas for this paper are defined as distance-based buffers calculated on publically
accessible paths around a park. These depict the location of people who are more likely to utilize
the resources of the park. Past studies have most commonly used 0.5 mile (Heckert 2012; Cohen
et al. 2014; Greer et al. 2014; Parsons, A. A. et al. 2015) and 0.25-mile measures (Boone et al.
2009) to study either geodesic buffers or network-based service areas for parks. Often these
distances were used as a general distance measure for park access that did not consider specific
park amenity types. For this study, four different network-based service area distances were used
to represent an intuitive sense of appropriate distances for residents to different types of
amenities. Since parks could not be classified as anyone particular type of park each park was
subjected to all the service area analysis and amenity scoring.
Each amenity category was given a specific service area distance that would best suit the
parks intended use. The 0.25- mile service area distance was used for the play category as this
was the shortest distance seen in past studies, and it is likely that parents need short travel
distances for daily play activity for children to access park facilities. The 0.5-mile service area
was used to measure the effects of disamenities on park access. Since the 0.5-mile service area is
the most common used in park accessibility studies, this was the service area chosen to see how
disamenities affect park access in a way that is most comparable to past studies. The 1.0-mile
41
service area was used for the PA category. This was chosen because the amenities in the PA
category were more likely to be utilized by people willing to travel longer distances by foot,
bike, public transit, or automobile. Finally, the 2.0-mile service area was utilized for the NA
category. Unlike other L.A. County cities that may have deserts, mountains, or beaches close to
them, Downey is in the midst of contiguous urban landscape and it reasonable to assume
accessing natural landscape features of significant size might require some travel.
Based on the logic above, each parcel in Downey had scores for the different amenity
categories calculated at different distances. Each service area distance had a baseline calculation
done for each park to determine the acres per person (Ap). This provided a comparison data set
for the amenity category acres per person (Ac) that was calculated. Table 4 lists the amenity type
and the service area distance for the 8 calculations that were run for each parcel.
Table 4 Parcel Values for Each Service Area
Amenity Type Service Area Distance
Acres/Person (Ap) 0.25-mile
Play Acres/Person (Ac) 0.25-mile
Acres/Person (Ap) 0.5-mile
Disamenity Acres/Person
(Ac)
0.5-mile
Acres/Person (Ap) 1-mile
PA Acres/Person (Ac) 1-mile
Acres/Person (Ap) 2-mile
Nature Acres/Person
(Ac)
2-mile

Before beginning the service area analysis for each park, the amenity data was added to
ArcGIS. To do this, initial scores (Sx) based on the audit for each category for each park were
calculated using Microsoft Excel and were exported into a .CSV file. The initial scores (Sx) were
then joined to the park data set in ArcGIS based on the park names.
42
To represent access points on the network, the park data was then converted to point data
using the vertex to point tool in ArcGIS. The park point data set was edited to only include
points that were located or moved to points of park access. To determine access points for each
park, a TIGER/Line street centerline data set was used. Using an aerial image as a base map for
reference, each park access point was placed on the nearest street line feature. All parks within
the 2-mile Downey buffer service area were included in the service area analysis. An initial
service area analysis of all parks within the study area revealed that only 8 non-Downey parks
contained Downey parcels within their service areas. All other non-Downey parks were included
in the service area analysis to allow for more accurate service area population estimates but
amenity scores were not needed.
The street data set was downloaded from the LAeGIS GIS data portal and had been
previously checked for accuracy by LAeGIS. The data set was clipped so that only the data in the
study area was used. The U.S. Census MTCC classification was used to query the TIGER/Line
data to separate it into local streets, highways, interstates, railways, and private roads. Of the 5
MTCC classifications, only 2 had limited or no access for park attendees. Railways were not
considered navigable routes for determining park access. Interstates were only considered
navigable for the 2-mile service area since automobiles are required to navigate them. Though
highways can be classified as limited access in come cities, in Downey all state highway routes
were lined with sidewalks and were easily accessed. Each one of the street data classifications
was then exported into separate shape files and combined into a feature data set. Local roads
consisted of all other publically accessible roads smaller then state highway routes. Finally a
network data set was built from the feature dataset with points at the end and intersections.

43
3.5 Service Area Analysis
The service area analysis tool on the ArcGIS 10.1 Network Analyst extension was run for
each service area distance. In each use of the tool, the park access points were loaded in as
facilities and the analysis settings were set “towards the facility.” Service area polygons were set
not to overlap, instead each polygon would go to the nearest facility. This process is similar to
the Theissian polygon method used to generate service areas in previous studies (Sister et al.
2009). Service areas for the 0.25-, 0.5-, and 1.0-mile distances were only based on local street
and highway network access. The 2.0-mile distance included interstate routes of travel since the
two-mile travel distance makes it more likely that public transit or car travel would be used.
Figure 8 shows the results for the 0.5-mile service area distance used for scoring disamenities. In
the figure both the highways and local streets are depicted as streets and interstates are visualized
for reference. Service areas were created for all parks within a 2-mile buffer of Downey. Only
parks with service areas that entered or immediately bordered Downey were retained for the
study and were scored for amenities.
44

Figure 8 Half-Mile Service Area Results
The next step was to associate parcels with each park’s service area. Once the service
area tool was run, the resulting polygon data set was exported to its own shape file. Each service
area data set was run through the dissolve tool in ArcGIS with the name of the park used as a
dissolve field. The dissolved polygons where then used to select land parcels through the select
by location tool. This allowed the creation of separate parcel data sets for each service area
45
buffer. In each case where service area polygons cover the same parcel, the parcel is allocated
based on the location of its center. Only at the distance of two miles are all the parcels in
Downey mostly covered by at least one service area.
Calculations for the four service area distances were done in an identical fashion. In each
case a spatial join was done to assign each parcel in a service area to a park service area. A
summary statistic was then performed to sum the total population of each residential land parcel
(Pp) with in a service area, using the park name as a case field. The resulting table provided the
total population per service area (Ps) for each park. This data was then joined to the polygon park
data set to perform the final calculations.
Using the Ps and the acreage (A) of each park, the acreage per person (Ar) was found for
each park. Equation 5 below explains how park acres (A) were divided by the total population in
its service area (Ps) to determine a park’s acres per person (Ar). The same calculation was done
for each category by replacing the acreage of the park with the category acreage (Ax), or
disamenity adjusted acreage (Ad) to find the category acreage per person (Ac). The formulas
below summarize the calculations used to derive both the Ar and the Ac.
Ar
(i…j)
= A/Ps (5)
Ac
(i…j)
= Ax/Ps or Ad/Ps (6)
Since all of the amenity scoring data was previously attached to the park data set, all calculations
were done using the field calculator. Once the acres per person (Ar) and category acres per
person (Ac) were calculated for each service area that data was spatially joined to the service
area parcels and visualized for analysis.
46

3.6 Conclusion
The methods used for this study involved dasymetric mapping, amenity scoring, and
service area analysis. The results of the dasymetric mapping and amenity scoring were both
needed to complete the final service area analysis. Dasymetric mapping provided detailed
population values at the parcel scale that helped calculate accurate service area populations.
Amenity scoring provided category values that allowed for the adjusting of a parks standard acre
per person calculation. In the next chapter, the results of each of the service area analyses
outlined in Table 4 will be discussed in greater detail.

47

Chapter 4 Results
When examining park accessibility in Downey from the perspective of residential parcels, it
becomes very apparent that park accessibility is not geographically uniform. Generally,
Downey’s parks are located on the city’s periphery. The larger parks maintained by Downey
such as Wilderness, Golden, and Furman parks are located at the edges of the city. Generally,
this places the service areas of the large parks away from the densely populated areas of the city
that are located in the geographic center. Further, some of the parks that serve Downey’s
residents that are located closer to the city’s border or belong to adjacent cities are less accessible
to Downey’s residents because of restrictions posed by the Interstates and waterways.
The following chapter explores the results of the service area analysis in more detail. The
chapter looks closely at what information was obtained from each service area (SA) analysis
distance (0.25-mile, 0.5-mile, 1-mile and 2-mile). Each distance is discussed in terms of how
effective it is in covering the residential parcels of Downey. There is also an exploration of how
park amenity scoring affects parcel accessibility for each SA distance.
As can be expected, it was found that the shorter the distance for the service area
threshold, the more parcels were left out of a service area. Table 5 lists the total number of
residential parcels in Downey, along with the percentage of parcels that were covered by a
service area for each distance. In total there are 22,029 residential parcels in Downey.
48

Table 5 Percent Downey Parcels for each Service Area
Service
Areas
%Accessible
Parcels
% Non-Accessible
Parcels
0.25- mile 13% 87%
0.5- mile 39% 61%
1-mile 84% 16%
2-mile 97% 3%

When looking at the parcel coverage for each service area it became apparent that the greatest
gains in coverage occurred when increasing the service areas to 0.5-mile and 1.0- mile.
Increasing the service area to 0.5-mile raised the total parcel coverage 26%, and increasing the
service area to 1.0-mile further raised total coverage by 45%, such that at 1.0-mile 84% of the
city’s residential parcels had park access of some kind. Doubling the service area threshold to
2.0-miles only further raised total coverage by only 13%, but was sufficient to grant park access
to nearly 100% of Downey’s residential parcels. The parcel coverage is only a brief insight into
park accessibility in Downey; the following chapter will reveal the results of the four service
area analyses in more detail.
4.1 Quarter Mile Service Area
The quarter mile service area had the lowest parcel coverage and had no overlapping
service areas. There were 2,597 residential parcels covered by the 14 Downey parks and 2 parks
from a neighboring city. Byron Zinn Park and All American Park in neighboring cities
respectively provided 4.1% and 3% of Downey parcel coverage for the 0.25-mile service area.
Figures 9 and 10 show the distribution of acre per person and play acre per person values
for the 0.25-mile service area. In both distributions the values were skewed to the left. The mean
49
value for the 0.25-mile acre per person service area parcels was 0.037 acres per person with a
standard deviation of 0.13. The mean for the play acre per person values was 0.018 play acres
per person and the standard deviation was 0.059. As expected, the factoring of the play score
reduced accessibility across all parks and reduced the mean. The difference in the variance
between the two data sets can be explained by the reduction of scores for larger parks. Using the
standard acre per person scoring methods, large parks received high scores based solely on their
size. When adjusting large park acreage scores for play amenities, scores were reduced. An
example of this can be seen with Wilderness Park, which had 0.38 acres per person but only 0.19
play acres per person.

Figure 9 Histogram of 0.25-Mile SA Parcel Values
50

Figure 10 Histogram of 0.25-Mile Play Acre SA Parcel Values
Figure 11 depicts the parcel scores for acres per person for the 0.25-mile park service
area on the left and parcel scores for the play acres per person on the right. The maps are
visualized with seven interval classes determined by natural breaks and a separate class for all
parcel values with “0.” Looking closer at the parcel coverage for the acre per person map,
clustering can be seen in the southern part of the city with the service areas of Golden,
Brookshire, and Apollo Parks. These three parks serviced just under 40% (1,013) of the total
residential parcels covered with the 0.25-mile service areas. Most of the other parcel coverage is
distributed in the periphery in the northern and eastern parts of the city. Independence and
51
Wilderness are two parks with relatively high acre per person values (0.38 Wilderness and 2.78
Independence). In both cases the high value can be attributed to the lack of access by
surrounding residential parcels. The Rio San Gabriel River and Interstate 605 restrict Wilderness
Park’s access, while Independence Park is less accessible because non-residential land parcels
mostly surround it.
Most all of the parcels covered by the 0.25-mile service area had access to 0.00029 – 0.23
acres per person. The scoring of parks for 0.25-mile children’s play amenities (play) accessibility
can be seen in Figure 11 on the right. Though the majority of parcels still fall within the same
range as the acre per person map, there is a noticeable difference in parcel coverage. Brookshire,
Temple, and Byron Zinn Park all have play acre per person scores, which were below 0.0001
acres per person (1 acre per 1,000 people). When looking at contributing factors, Bryon Zinn
Park’s score was “0” because it had no children’s play amenities. In the cases of Brookshire and
Temple parks, the play acres per person score were so low because of the size and condition of
the amenities in the park. This result is particularly interesting because Brookshire Park is one of
the few parks located toward the center of Downey and it serves a densely populated area of the
city. All other parcels, which were serviced by parks with in a 0.25-mile service area, had access
to at least 0.0018 play acres per person.
52

Figure 11 Parcel Values for 0.25-Mile Service Areas
4.2 Half Mile Service Area
As was previously listed in Table 5, the 0.5-mile service area distance saw a large
increase in parcel coverage compared to the 0.25-mile service area. At 0.5 miles, park service
areas began to overlap with each other; this meant that parcels were assigned to the nearest park.
The 0.5-mile service area distance is one of the most commonly used in park accessibility studies
(Heckert 2012; Cohen et al. 2014; Greer et al. 2014: Parsons et al. 2015). For this reason, the 0.5-
mile service area distance was used to measure the changes disamenities have on accessibility.
As expected, the increased service area distance caused parcel coverage to increase but parcel
acres per capita values to decrease. There were a total of 8,322 residential parcels covered by the
53
0.5-mile park SA; non-Downey parks in adjacent cities serviced just fewer than 13% of those
parcels.
Figure 12 depicts the distribution of parcel values for 0.5-mile park service areas. The
same left skewing data distribution is seen again, though the skew relative to the normal curve is
not as dramatic as for 0.25-mile access. The mean acre per person parcel value was 0.0091 and
the standard deviation was 0.015. Comparing this to the distribution of disamenity-adjusted acres
in Figure 13, it can be seen that the leftward skew of the data remains but values have generally
decreased. The mean value for the disamenity-adjusted acres was 0.0068 acres per person and
the standard deviation was 0.01.
54

Figure 12 Histogram of 0.5-Mile SA Parcel Values
55

Figure 13 Histogram of 0.5-Mile Disamenity Adjusted SA Parcel Values
Figure 14 depicts both the mapped parcel values for 0.5-mile acre per person SA and the
0.5-mile disamenity acre per person SA. For each map, the parcels visualized with seven
interval classes determined by natural breaks and a separate class for all parcel values with “0”
value. Unlike the 0.25-mile analysis, there were no parks whose disamenities were so great that
they provided “0” disamenity adjusted acres per person. The clustering of park service areas in
the south of Downey is more prominent in the 0.5-mile SA analysis. There are three parks in
Downey that service more than a third of all parcels covered in the 0.5-mile SA analysis.
56
Brookshire and Apollo Park in southern Downey both service about 13% of the parcels each
while Furman Park in the north services 14%. This is important to note because Furman Park has
a larger area than Brookshire and Apollo Park combined. When comparing parcel values, it is
clear that accessibility is unequal as those parcels serviced by Brookshire Park have access to
0.0007 acres per person and those by Furman Park have 0.0096 acres per person.
Apollo, Independence, and Wilderness Parks were amongst the lowest scoring parks
when it came to adjusting for disamenities. The parcels in Apollo Park’s SA had a reduction
from 0.0065 acres per person to 0.0035 disamenity adjusted acres per person. This drop in score
was mostly due to the large amounts of graffiti and litter observed at the park. Both Wilderness
and Independence Park also had high observations of disamenities, however the lack of parcel
accessibility at the 0.5-mile distance meant that their scores were not as visibly noticeable in
Figure 14. In the case of Independence Park, there are a low number of residential parcels
located in its SA this causes high acre per person values for those parcels that are in its SA.
Wilderness Park is a different situation because access is limited due to the presence of the San
Gabriel Riverbed and Interstate 605, as both present obstacles to easily accessing the park. On
average there was a 77% reduction in acres per person after adjusting for disamenities.

57

Figure 14 Parcel Values for 0.5-Mile Service Areas
The results of the 0.5-mile SA analysis also revealed details about how Downey parcels
are serviced by facilities of neighboring cities. Bryon Zinn and Thompson Park provided a
combined 4.9% of the Downey parcel coverage for the standard 0.5-mile SA. This is an
interesting occurrence because Byron Zinn and Thompson Parks are maintained by Bellflower
but their 0.5-mile service areas extend into an area that is not serviced by a Downey park. Figure
15 provides a more detailed look into the service areas of Byron Zinn and Thompson Park. The
specific area of Downey serviced by these two Bellflower parks is particularly susceptible to
accessibility problems. The building of Interstate 105 in the 1990’s interrupted many local streets
that would have normally connected this part of the city to Independence Park.
58

Figure 15 Detail of Byron Zinn Play Acre 0.50-mile SA
4.3 One Mile Service Area
The 1-mile SA encompasses nearly all the parcels in the city of Downey. Though the 1-
mile SA distance is not as commonly used in park accessibility studies, the increased distance
was thought to account for those people in Downey who would be willing to commute longer
distances to participate in communal physical activities, like baseball or soccer, or to utilize
publically available resources for exercise and recreation (e.g., swimming pools). As before, the
increase in SA size caused a shift in the data distribution and mean parcel value. Of the 18,305
parcels, about 9% was serviced by non-Downey parks.
Figure 16 and 17 show both the data distributions for the 1-mile acre per person and PA
acre per person parcel values. The mean parcel value for the 1-mile SA is 0.0045 acres per
59
person with a standard deviation of 0.0062. When the parcel values are adjusted for the physical
activity (PA) category, the values for the data set generally decrease. The mean parcel value for
the PA acres 1-mile SA is 0.0019 acres per person with a standard deviation of 0.0026. Both data
sets show a strong leftward skew but there is substantially more variance in the acre per person
values compared to the PA acre per person values.
The lack of variance in the PA acre per person values can be explained by the specificity
of physical activity amenities located at some parks. An example of this can be seen when
comparing Rio San Gabriel and Apollo Park. Rio San Gabriel Park saw a large reduction in its
PA service area values compared with its general 1-mile SA values because it had only a
baseball and basketball amenity. Though each amenity was in good shape there were few
amenities compared to Apollo Park, which had some PA amenities in disarray, but there were a
wide variety of facilities. This reduced the SA value differences between the two parks from
0.0081 acres per person to 0.0038 PA acres per person.
60

Figure 16 Histogram of 1-Mile SA Parcel Values

61

Figure 17 Histogram of 1-Mile Physical Activity Acre SA Parcel Values
Figure 18 depicts both the spatial distribution of parcel values for the 1-mile acre per
person and PA acre per person data set. Both data sets were visualized with seven interval
classes determined by natural breaks and a separate class for all parcel values with a “0” score.
Compared to the prior SA distances, the clustering of coverage in the south of Downey was not
as prominent compared to coverage in the north. When reviewing the map on the left of Figure
18, there is a clear clustering of high parcel values in the east of Downey. This is because the
parcels on the east side of Downey are serviced by three of the four largest parks in Downey:
62
Wilderness, Rio San Gabriel, and Independence Park. It is also clear that the service areas for
Temple, Dennis the Menace and Brookshire Park are much lower then the surrounding parks.
This is directly attributable to the fact that all 3 parks are highly accessible at the 1-mile SA
distance but they are relatively small compared to other Downey Parks.
When the maps are compared, two immediate differences are noticed. First, most parks
saw a sharp reduction in the acre per person score. Brookshire, Temple, and Treasure Island
Parks provided the least amount of PA amenities compared to other Downey Parks. The SA of
just Brookshire and Temple Parks accounted for 15.6% (2,852) of total overall parcel coverage.
In the case of Brookshire Park, this was a significant reduction in PA acres values because it
occurs in a densely populated area of the city. The second difference noticed when the maps
were compared, was the continued clustering of parcels with higher values on the east side of
Downey. The presence of the high values in this area of Downey is attributed to the large size of
the 2 parks and the small amount of residential parcels in their SA.

63

Figure 18 Parcel Values for 1-Mile Service Areas
4.4 Two Mile Service Area
The final SA distance encompassed nearly all the parcels in the entire study area. The 2-
mile SA distance was assumed to reflect the amount a reasonable person would commute to
access large naturalized landscapes in an urban environment. As a result of the 2-mile distance
threshold, this final SA analysis contained the largest data set. Of the 21,167 parcels in Downey
that were covered by the 2-mile SA, 9.5% (2,016 parcels) were serviced by non-Downey parks.
Figure 19 shows the distribution of the parcel values for the 2-mile SA analysis. The
mean value was 0.0038 acres per person and the standard deviation was 0.0056. Following the
trend of the previous data sets the distribution of the parcel values is skewed to the left, although
for the 2-mile SA very few parcels (728) have no access at all (i.e., “0” scores). Figure 20 depicts
64
the distribution of the parcel values for the NA acres per person SA analysis. After adjusting for
the NA category, the values skewed even further to the left when compared with the basic 2-mile
SA. Furthermore, the mean parcel value decreased to 0.0014 NA acres per person and the
standard deviation was 0.004. Compared to the 0.5-mile acre per person and 0.5-mile PA acre
per person SA data sets, the variance change between the 2-mile acre per person and 2-mile NA
acre per person was not as pronounced.
When closely inspecting the park auditing data, it is clear that parks in Downey either
thrived or were severely lacking in this amenity category. This can be seen easily when looking
at Treasure Island and Wilderness Parks. Both parks were created with the idea of accessing
nature by including native vegetation, nature guides, and integration with a near by riverbed.
Since these were the only two parks in Downey that seemed to be designed in this way, their
high scores made surrounding parcels high outliers for nature access.
65

Figure 19 Histogram of 2-Mile SA Parcel Values
66

Figure 20 Histogram of 2-Mile Nature Acre SA Parcel Values
Figure 21 depicts the spatial distributions of parcel values for the 2-mile SA analysis. An
immediate observation from the standard acre per person map on the right of Figure 21 is that
there is still a clustering of high parcel values on the east side of Downey. It is also worth noting
that at the 2-mile SA distance all the residential parcels in the geometric center of Downey are
within a SA of a park. Another valuable observation is the lack of coverage in the small area of
Downey north of Interstate 5. Outside of the eastside of Downey, Golden Park and Crawford
Park are the only parks that provide relatively high acre per person values in surrounding parcels.
67
As was seen in the previous amenity SA analysis, Brookshire and Temple Park lack any
NA amenities. The result of this is a drastically reduced NA value for Downey parcels in the
south and the east. In this instance, both parks accounted for 16.3% (3,450) of the parcel
coverage for the 2-mile SA analysis. Another point of interest when comparing the two maps is
the change in the scores for Rio San Gabriel Park. In the initial 2-mile SA analysis, parcels
surrounding the Rio San Gabriel Park and Wilderness Park were able to maintain a relatively
higher acre per person value compared to the rest of Downey. After adjusting for the NA
amenities Rio San Gabriel Park’s SA became more in line with the values for the rest of the city
while Wilderness Park maintained its relative high value. This change can be attributed to Rio
San Gabriel Park’s low NA amenity score (0.20) compared to Wilderness Park (0.60). Both
Crawford and Golden Parks NA acre per person values also fell to values for in line with other
Downey parks.

68

Figure 21 Parcel Values for 2-Mile Service Areas
One final observation was made for the 2-mile SA analysis. Throughout the analysis,
there were many observable cases in which non-Downey parks assisted in SA coverage for
Downey land parcels. Also, during the final 2-mile SA analysis there was a clear case of a
Downey park servicing parcels outside of the city. Figure 22 shows a detailed spatial distribution
of the Wilderness park 2-mile SA.
69

Figure 22 Detail of Wilderness Park Nature Access 2-Mile SA

Wilderness Park was the only Downey Park that had relatively high parcel values for the 2-mile
SA analysis. Consequently, it turned out that it was one of the few Downey parks that serviced
parcels outside of the city. The parcels with high scores to the south of Wilderness Park are
actually in the City of Norwalk. This is notable because in the case of Wilderness Park, it was
shown that access was limited at all of the SA distances. Wilderness Park is the largest park in
Downey and to find that it is less accessible to Downey residents then those in a neighboring city
is quite interesting.
In general the results of the multiple distance SA analysis revealed that areas closest to
the geometric center of the city are more likely to face lower park accessibility. After visual
inspections of the spatial distribution of the park service areas, it was also found that the south
side of Downey has a tighter cluster of park service areas. However, when amenity scores were
70
integrated into the analysis it was further revealed that parks which provided coverage for dense
settlement areas in Downey often did not have sufficient amenities to serve all types of park use.
In the next chapter the implications of these results for Downey and the approach taken here for
studying park accessibility is discussed in more detail.

71

Chapter 5 Discussion and Conclusion
The intention of this study was to develop a method to measure the difference in individual land
parcel accessibility to parks and park amenities. In order to achieve this, a combination of
dasymetric mapping, amenity scoring, and service area (SA) analysis was used. Each individual
step in the methodology of the study discussed in this paper was inspired by previous work and
in some cases was altered to account for previously unaccounted for aspects of accessibility.
Parsons (2015) served as the inspiration for the dasymetric mapping method used for this
study. Of the three steps previously mentioned, dasymetric mapping was the method that was
most closely replicated from previous research. Parsons (2015) previously documented the
effectiveness of this same type of limiting variable dasymetric mapping. Eicher et al. (2001)
provided a well-tested methodology for redistributing census block data population. The built in
Service Area Analysis tool in the ArcGIS 10.1 Network Analyst extension also provided a
trustworthy method for determining parcels which were serviced by the parks of Downey.
The combined use of the Service Area Tool and Parsons (2015) methods for dasymetric
mapping was further developed in this study by incorporating multiple distance thresholds for
the service areas and a park amenity scoring system. Using multiple service area thresholds
allowed for a closer analysis of how accessibility is affected by different types of park use. The
use of the augmented PARA audit instrument allowed for the measurement of park
characteristics, which satisfied portions of Mehta’s (2013), five benefits of public space. As
previously mentioned in chapter 1, Mehta’s (2013) five benefits of public space include:
inclusivity, support of meaningful activity, pleasurable, safe, and comfortable. The use of the
PARA instrument allowed for the measurement of park facilities that support meaningful
72
activity, pleasure-ability, and safety. The method by which the auditing items were categorized
also provides a methodology that can easily be augmented for diverse local needs (Kabish et al.
2015).
The following chapter discusses in more detail the effectiveness of the methods utilized,
the implications of the results, and suggests new directions for future studies. Sections 5.1-5.3
are each dedicated to discussing the results for the population mapping, SA analysis, and amenity
scoring done for this study. Each of these sections investigates the effectiveness of the method in
the context of the study and any augmentations, which proved to be improvements over past
research. Section 5.4 contains an introspective critique on the limitations of the study and
potential improvements. Finally, Section 5.5 discusses suggestions for future research on park
accessibility.
5.1 Population Mapping
The dasymetric mapping methodology used for this study had various benefits that have
not been utilized generally in previous research on park accessibility. The research done by Lara-
Valencia et al. (2013) provides the most common example of traditional population estimates
that use large neighborhood scale aggregations. This type of population mapping could be
deceiving in a small analysis scale study such as this. Figure 23 depicts a detailed map of the
residential parcels surrounding the Discovery Sports Complex and Independence Park. As can be
seen in the map, large portions of the potentially serviceable areas of the parks contain no
residential parcels.
73

Figure 23 0.50-Mile Service Areas for Discovery Sports Complex and Independence Park
Aggregating population data to census block group or larger would not properly represent
the service area for these 2 parks. If census block group data was used the population with in the
SA of Discovery Sports Complex and Independence Park would have been artificially inflated.
This is due to the fact that census block polygons overlap both residential and non-residential
parcels. The aggregation of the population numbers within a census block would have given
population values even for large areas that have no residential parcels. Other Downey parks that
are adjacent to similar non-residential areas are Rio San Gabriel Park, which has a large
commercial area to its south, and Apollo Park, which has a non-residential area to the west.
When considering future park developments, the City of Downey should consider
carefully the surrounding residential area. In the case of the parks mentioned above, accessibility
74
might be improved by increasing pedestrian access through non-residential zones to public green
space. One way that this can be done is through the development of more bike paths. The 4 parks
mentioned above are all easily accessible from either Imperial highway or Firestone Boulevard.
Both of these streets are main transportation routes through the city that do not have dedicated
bike paths. Investing in bike paths along these routes could help increase accessibility to parks
that are not immediately surrounded by residential areas.
The methods used in this study also improve upon previous dasymetric mapping used in
previous studies. Some studies of green space and other urban phenomena have previously used
building square footage as means to disaggregate population data (Oh et al. 2007; Morar et al.
2014). This approach has the potential for error because the disaggregation of population data
based on building square footage cannot account for a buildings use or homes that have large
land plots. By using the housing unit values of each land parcel, the dasymetric mapping
methods used in this study provide a more accurate assessment of potential population
distributions based only on plots that are designated for residential use.
5.2 Service Area Analysis
The way in which SA analysis was utilized for this study also expanded upon previous
work. The network analysis approach used in ArcGIS 10.1 allowed for a more realistic model of
park accessibility than prior research that used geodesic distance buffers (Boone et al. 2009).
Even in cases where past studies used network analysis to determine accessibility, data
aggregated to the level of a census tract inhibited the ability to look at neighborhood aspects of
park accessibility (Heckert 2012). Since specific population values existed for each land parcel,
the SA analysis methods used in this study provided more detail on how a given parks serves its
citizens. This is an improvement compared to prior research that uses only simple binary values
75
for population (Langford et al. 2006) or measures distances from parks to housing units without
considering population (Apparicio et al. 2006).
Though large parts of this study were inspired by the methods of Parsons (2015) research
in Wake County, NC, augmentations to the final accessibility analysis helped prevent the
processing challenges posed by Origin Destination Matrix analysis. By utilizing multiple
distances in the SA analysis tool, this study allowed for a more efficient use of computer
processing power. Using variable distances for the service areas also allowed for an efficient
examination of where there were gaps in park coverage depending on differential use. Using this
method allowed for the comparison of how citizens of Downey with differing thresholds of
acceptable park distance might be serviced. An example of this can be seen when comparing the
0.25-mile and 0.50-mile service areas of Downey parks.
The difference in parcel coverage between the 0.25-mile and 0.50-mile service areas
shows how parks may not be serving communities as well as intended. When considering that
the 0.25-mile SA covered only a third (5,598) of the parcels covered by the 0.50-mile SA
(14,972), it becomes clear that accessibility is highly dependent on the intended use of the park.
Parks such as Furman, Rio San Gabriel, Crawford, Dennis the Menace, Brookshire, and Golden
parks represent the best locations for 0.25-mile access because mostly residential areas surround
them. Consequently, a large portion of these easily accessible parks is located in the northern
part of Downey. It is only at the 0.50-mile level that the SA for the Downey parks in the south
starts to show better coverage. This lack of coverage can be explained by the increased amount
of commercial and industrial areas in southern Downey. Future park developments, however,
should make it a priority to create smaller parks, similar in size to Crawford Park or Brookshire
Park with in highly populated residential areas in south Downey.
76
5.3 Amenity Scoring
One aspect shared by most of the previous research on park accessibility is a lack of
attention to how park amenities and disamenities affect accessibility. One of the key
enhancements used in this study was its integration of park amenity scoring. In a previous study
by Weiss et al. (2011), disamenities were mapped as a density field across the study area using
kernel density cluster. This method used incidences of ‘incivilities’ such as traffic accidents and
murder to model the negative forces that would discourage people from using a particular park.
The methods used for this study were much more direct in that disamenities were recorded at and
attributed to specific park locations.
The use of the PARA audit instrument greatly assisted in the amenity scoring because of
its simple structure and its ability to record the quality or severity of an amenity or disamenity
(Lee et al. 2005; Greer et al. 2014). The way that each category of the audit was scored also
allowed for easy expansion to add supplementary items and categories. The addition of the NA
category also improved upon previous research with the PARA instrument, which focused
mainly on constructed park amenities (Lee et al. 2005; Greer et al. 2014).
Ibes (2015) research on park accessibility to park types in Arizona was a starting point
for creating the methodology for this study. However, where Ibes (2015) created a typology for
all the parks in the study area, this study embraced the idea that each park in Downey shared
similar aspects of 4 main categories. This resulted in the ability to compare all parks based on an
amenity category. This insight allowed this study to reveal parks such as Brookshire Park, which
provides no amenities for PA or NA to any of the parcels in its SA. This methodology also
revealed that parks such as the Discovery Sports Complex and Crawford had relatively higher
SA capacities then neighboring parks, because of their lack of disamenities.
77
Future park development in Downey should consider how park facility improvements
affect accessibility. One example of this is the clustering of natural access amenities in northern
Downey parks. Both Wilderness and Treasure Island parks scored high in NA and both are
located in the northern part of the city. While other Downey parks may have had a large amount
of tree canopy coverage, it was only Wilderness and Treasure Island Park, which had nature
guides and the most notable signs of purposeful landscaping.
Another aspect of amenity scoring, which could be considered by the City of Downey, is
the way in which public schools service residential communities. Downey’s schools, though not
open for general public use, often make their fields available on weekends and after school for
organized youth sports. Since some of these youth athletic organizations cost money to
participate, it may be interesting to investigate how public school facilities might be made
available for free to supplement the lack of park accessibility.
5.4 Limitations and Improvements
Areas in which this study is limited or can be improved involve, quality, quantity, and the
collection of the data used for the analysis. When considering the road network data used for this
study, it is important to note that not all routes of travel were included in the network analysis.
Though the study included all modes of transportation used by cars it did not consider bike lanes.
Though Downey does not have an extensive bike lane network, most of the Rio Hondo
and San Gabriel riverbeds maintain a biking and running path on either side of their banks.
Adding these routes of travel proved to be difficult because complete bike route data for the city
was unavailable. Time and resource restraints also did not allow for the creation of such a data
set. Other road network limitations include the lack of data on roads without sidewalks. This data
78
would have been helpful in determining potential barriers to pedestrian travel for shorter distance
service areas.
It is worth noting that bike lanes also were not scored as part of the PA amenities
category. However, the access to riverbed bike lanes at particular parks was scored as a NA
amenity because it improved access to naturally redeveloped portions of the riverbed. While
many of the parks in this study did not have bike lines within them, some were near bikes lanes
that traverse the riverbeds. Not scoring the access to bike lanes as a PA amenity is a further
limitation of this study that can be improved.
Other limitations to this study involve the generalized items that were included on the
auditing instrument that measured amenities. Some examples of this include unique facilities
such as batting cages and skate parks. Facilities such as these were only noted in single locations;
because of this, it was difficult to consider the value of such a unique park amenity in the study
area. Another aspect of amenity scoring which can be improved is the method in which amenities
were scored. Specifically, the methods for scoring play grounds and other spatially variable
amenities could be improved by using actual measurements. For example, a similar method as
was used for tree canopy coverage could be used to determine the percent of playground
coverage. Future studies should investigate methods such as these.
When considering disamenities, it is also important to mention that not all factors were
equal. An example of this would be the disamenity scores for brown grass and overgrown grass.
In general Downey’s parks had very little overgrown grass, but areas of dead grass were much
more prevalent. It was, however, hard to determine if this was truly a disamenity because
Downey had recently posted signs explaining the rationing of water mandated by the California
Governor in 2015 due to a record-setting drought.
79
Figure 24 shows the clearly posted sign that was present at most Downey parks
explaining the occurrence of dead grass. The dead grass disamenity accounted for only 3 of the
possible 42 disamenity points for the category. Though it affected all the parks equally, it did
cause a general reduction of disamenity-adjusted acres for all parks. These same signs were not
posted in parks scored outside of the city, so non-Downey parks did not necessarily suffer from
the same disamenity.

Figure 24 Downey Restricted Watering Sign
Other factors, which were not considered in the disamenity scoring methodology,
included the lack of acknowledgement of amenities specifically aimed to counteract park
incivilities. Figure 25 depicts a call box that was located at Hollydale Park in the neighboring
city of Bell Gardens. The effect of an emergency call box’s presence on park attendees may
counteract disamenities but was not considered in this study.

80

Figure 25 Emergency Call Boxes
5.5 Future Directions
When looking ahead to future studies on park amenities and its effects on accessibility,
data on demographics, service rates, and gravity analysis should be explored. By studying park
accessibility, this study hoped to discover more about unrecorded inequalities in park access. In
order to advance the ideas of this study, future studies should look at the possibility of
integrating demographic data along with population data. The importance of integrating
demographic data in areas such as income and race will help explore potential unequal
accessibility patterns, which can greatly contribute to the growing research on environmental
justice.
Another area that should be explored is the differing service rates of park amenities. In a
recent publication, NRPA (2015) discussed the potential for their Park and Recreation Operating
Ratio and Geographic Information System (PRORAGIS) to calculate national sourced county
81
benchmarks for park amenities. Using this data could provide possibility to compare service
areas for specific park amenities against national statistics. Since PRORAGIS is still in
development this research did not consider it when studying park amenities. Future research
should consider integrating this data as a comparative metric for measuring park amenities.
Finally, future research should explore new GIS methods for measuring accessibility at
larger distance thresholds, where service areas are likely to overlap. For example, gravity
analysis provides a way to model the potential of a person in one park SA to go to another.
Gravity analysis works from the premise that possible interactions between parcels and facilities
decrease as the distance between them increases (Ttth and Kincses 2015). However, this
potential interaction also increases depending on the size and quality of the facilities (Ttth and
Kincses 2015). In terms of park access this means that larger parks and parks with more
attractive amenities have a higher likelihood of servicing people at farther distances then smaller
parks. The advantage to using gravity analysis is that service or ‘market’ areas have fuzzier
boundaries, as opposed to models that assign people to the closest service area (Eck and Jong
1999). Gravity analysis is most often used in determining service areas for retail establishments
and often they involve the modeling of ‘resistance’ factors that would help explain why someone
would not participate in the nearest SA (Eck and Jong 1999; Ttth and Kincses 2015). For parks,
such “resistance” factors might include the sorts of disamenities measured in this study.
5.6 Final Conclusions
The results of the combined dasymetric mapping, service area analysis, and amenity
scoring have shown that accessibility can be affected by park quality. Though there are still
adjustments that can be made to the exact methods used in this study, the importance of park
amenity scoring has been exemplified. As more detailed methods are developed for modeling
82
park accessibility in an urban environment, this study has shown that it is important to consider
the quality of parks that are accessible.
Enhancing or acquiring more detailed data sets can alter difficulties that were
encountered in this study. Road networks with information on bike trails and pedestrian
accommodations should be an immediate goal for future studies built upon this work. More data
on park amenity use from larger organizations such as NRPA is also needed to help improve
methods of amenity scoring. Gravity analysis and other advanced GIS methods should also be
considered for future studies to help model complex interactions between parks and those who
live in its service area.
Overall the methods utilized in this study provide a relatively simple and efficient way to
study park accessibility given current GIS technologies. Such analysis should be attainable for
any city with a competent, professional GIS staff, meaning the study should be beneficial to
many small municipalities, which are seeking to understand accessibility as a small scale of
analysis.
83

References
Apparicio, P., and A.M. Séguin. 2006. “Measuring the Accessibility of Services and Facilities for
Residents of Public housing in Montréal.” Urban Studies 43 (11): 187–211.

Bedimo-Rung, A. L., A. J. Mowen, and D. A. Cohen. 2005. “The Significance of Parks to Physical
Activity and Public Health.” American Journal of Preventive Medicine 28 (2S2): 159–168.
Accessed April 2015.

Boone, C. G., G. L. Buckley, J. M. Grove, and C. Sister. 2009. “Parks and People: An Environmental
Justice Inquiry in Baltimore, Maryland.” Annals of the Association of American
Geographers 99 (4): 767–787. Accessed April 2015.

Bruton, C. M., and M. F. Floyd. 2014. “Disparities in Built and Natural Features of Urban Parks:
Comparisons by Neighborhood Level Race/Ethnicity and Income.” Journal of Urban
Health: 894–907.

Bullard, R. D., ed. 1994. Unequal protection: environmental justice and communities of color. San
Francisco, California: Sierra Club Books.

Byrne, J., J. Wolch, J. Swift, and C. Ryan. 2005. SAGE (Systematic Audit of Green-Space
Environments): Audit Form and Instructions. Los Angeles, California: University of Southern
California Center for Sustainable Cities.

Byrne, J., and J. Wolch. 2009. “Nature, Race, and Parks: Past Research and Future Directions for
Geographic Research.” Progress in Human Geography 33 (6): 743–765. Accessed April 2015).

Cohen, D., S. Lapham, K. Evenson, S. Williamson, D. Golinelli, P. Ward, A. Hillier, and T. Mckenzie.
2014. “Use of Neighborhood Parks: Does Socio-Economic Status Matter? A Four-City
Study.” Public Health 127 (4): 325–332.

Cohen, D. A., T. L. Mckenzie, A. Sehgal, S. Williamson, D. Golinelli, and N. Lurie. 2007.
“Contribution of Public Parks to Physical Activity.” American Journal of Public Health: 509–
514.

Coley, R. L., W. C. Sullivan, and F. E. Kuo. 1997. “Where Does Community Grow? The Social
Context Created by Nature in Urban Public Housing.” Environment and Behavior 29 (4):468–
494. Accessed April 2015).

84

Comber, A., C. Brunsdon, and E. Green. 2008. “Using a GIS-based Network Analysis to Determine
Urban Greenspace Accessibility for Different Ethnic and Religious Groups.” Landscape and
Urban Planning 86:103–114. Accessed November 2015.

U.S. Census. 2015. “State and County Quick Facts.” Last modified December 2. Accessed December
15, 2015. http://quickfacts.census.gov/qfd/states/06/0619766.html

Eck, J. R. V., and T. D. Jong. 1999. “Accessibility analysis and spatial competition effects in the
context of GIS-supported service location planning.” Computers, Environment and Urban
Systems 23:75–89. Accessed July 2015.

Eicher, C. L., and C. A. Brewer. 2001. “Dasymetric Mapping and Areal Interpolation: Implementation
and Evaluation.” Cartography and Geographic Information Science 28 (2): 125–138.

Greer, A. E., R. Marcello, and R. Graveline. 2014. “Community Members' Assessment of the Physical
Activity Environments in Their Neighborhood Parks: Utility of the Community Stakeholder
Park Audit Tool.” Health Promotion Practice 16 (2): 202–209. Accessed April 2015.

Heckert, M. 2012. “Access and Equity in Greenspace Provision: A Comparison of Methods to Assess
the Impacts of Greening Vacant Land.” Transactions in GIS 17 (6):808–827.

Hewko, J., K. E. Smoyer-Tomic, and M. J. Hodgson. 2002. “Measuring Neighborhood Spatial
Accessibility to Urban Amenities: Does Aggregation Error Matter?” Environment and
Planning A 34:1185–1206.

Higgs, G., R. Fry, and M. Langford. 2012. “Investigating the Implications of Using Alternative GIS-
based Techniques to Measure Accessibility to green space. Environment and Planning B:
Planning and Design 39:326–343.

Ibes, D. C. 2015. “A Multi-Dimensional Classification and Equity Analysis of an Urban Park System:
A Novel Methodology and Case Study Application.” Landscape and Urban Planning137: 122–
137. Accessed April 2015.

Joassart-Marcelli, P. 2010. “Leveling the playing field? Urban Disparities in Funding for Local Parks
and Recreation in the Los Angeles Region.” Environment and Planning A 42: 1174–1192.

Kabisch, N., S. Qureshi, and D. Haase. 2014. “Human–Environment Interactions in Urban Green
Spaces — A Systematic Review of Contemporary Issues and Prospects for Future
Research.” Environmental Impact Assessment Review50: 25–34. Accessed April 2015.

Krellenberg, K., J. Welz, and S. Reyes-Päcke. 2014. “Urban Green Areas and Their Potential for
Social Interaction – A Case Study of a Socio-Economically Mixed Neighborhood in Santiago
de Chile.” Habitat International 44:11–21. Accessed December 2015.
85

Langford, M., and G. Higgs. 2006. “Measuring Potential Access to Primary Healthcare Services: The
Influence of Alternative Spatial Representations of Population.” The Professional Geographer
58 (3): 294–306. Accessed May 2015.

Lara-Valencia, F., and H. García-Pérez. 2013. “Space for Equity: Socioeconomic Variations in the
Provision of Public Parks in Hermosillo, Mexico.” Local Environment 20 (3): 350–368.
http://www.tandfonline.com (last accessed April 2015).

Li, C. 2014. “Ethnicity, Culture and Park Design: Case Studies of Urban Parks in American
Chinatowns.” Journal of Urban Design 19 (2): 230–254. Accessed April 2015.

Maantay, J. A., A. R. Maroko, and C. Herrmann. 2007. “Mapping Population Distribution in the Urban
Environment: The Cadastral-based Expert Dasymetric System (CEDS).” Cartography and
Geographic Information Science 34 (2): 77–102.

Maas, J. 2006. “Green Space, Urbanity, and Health: How Strong is the Relation?” Journal of
Epidemiology and Community Health 15 (1): 587–592.

Maroko, A. R., J. A. Maantay, N. L. Sohler, K. L. Grady, and P. S. Arno. 2009. “The Complexities of
Measuring Access to Parks and Physical Activity Sites in New York City: a Quantitative and
Qualitative Approach.” International Journal of Health Geographics 8:34–34.

Mehta, V. 2013. “Evaluating Public Space.” Journal of Urban Design 19 (1): 53–88. Accessed April
2015.

Melcher, K. 2013. “Equity, Empowerment, or Participation: Prioritizing Goals in Community
Design.” Landscape Journal 32 (2): 167–182.

Moore, L. V., A. V. D. Roux, K. R. Evenson, A. P. Mcginn, and S. J. Brines. 2008. “Availability of
Recreational Resources in Minority and Low Socioeconomic Status Areas.” American Journal
of Preventive Medicine 34 (1): 16–22.

Morar, T., R. Radoslav, L. C. Spiridon, and L. Pacurar. 2014. “Assessing Pedestrian Accessibility to
Green Space using GIS.” Transylvania Review of Administrative Sciences 42:116–139.

Nicholls, S. 2001. “Measuring the Accessibility and Equity of Public Parks: A Case Study using GIS.”
Managing Leisure 6 (4): 201–219. Accessed May 2015.

NRPA. 2015. NRPA PRORAGIS: Database Report Counties. Ashburn, VA: National Recreation and
Parks Association. Accessed November 30, 2015.
http://www.nrpa.org/uploadedFiles/PageBuilder_Proragis/Content/common_elelments/PRORA
GIS-Database-Report-Counties.pdf
86

Oh, K., and S. Jeong. 2007. “Assessing the Spatial Distribution of Urban Parks using GIS.” Landscape
and Urban Planning 82:25–32. Accessed May 2015.

Trust for Public Land, 2015. “Parkscore Methodology.” Accessed December 15, 2015.
http://parkscore.tpl.org/methodology.php

Parsons, A. A., G. M. Besenyl, A. T. Kaczynski, S. A. Wilhelm Stanis, C. E. Blake, and D. J. Barr-
Anderson. 2015. “Investigating Issues of Environmental Injustice in Neighborhoods
Surrounding Parks.” Journal of Leisure Research 47 (2): 285–303.

Parsons, J. 2015. “Mapping Uniformity of Park Access using Cadastral Data within Network Analyst
in Wake County, NC.” Masters Thesis, Institute of Spatial Sciences, University of Southern
California.

Rossi, S. D., J. A. Byrne, and C. M. Pickering. 2015. “The Role of Distance in Peri-urban National
Park use: Who visits them and how far do they travel?” Applied Geography 63:77–88.
Accessed April 2015.

Richardson, E., J. Pearce, R. Mitchell, and S. Kingham. 2013. “Role of Physical Activity in the
Relationship Between Urban Green Space and Health.” Public Health 127:318–324. Accessed
April 2015).

Saelens, B. E., L. D. Frank, C. Auffrey, R. C. Whitaker, H. L. Burdette, and N. Colabianch. 2006.
“Measuring Physical Environments of Parks and Playgrounds: EAPRS Instrument
Development and Inter-Rater Reliability.” Journal of Physical Activity and Health 3: S190–
S207.

Sarah, N., and C. S. Shafer. 2001. “Measuring Accessibility and Equity in a Local Park System: The
Utility of Geospatial Technologies to Park and Recreation Professionals.” Journal of Park and
Recreation Administration 19 (4): 102–124.

Sister, C., J. Wolch, and J. Wilson. 2009. “Got green? Addressing Environmental Justice in Park
Provision.” GeoJournal 75:229–248.

Stodolska, M., K. J. Shinew, J. C. Acevedo, and D. Izenstark. 2011. “Perceptions of Urban Parks as
Havens and Contested Terrains by Mexican-Americans in Chicago Neighborhoods.” Leisure
Sciences 33 (2): 103–126. Accessed April 2015.

Suminski, R. R., E. K. Connolly, L. E. May, J. Wasserman, N. Olvera, and R. E. Lee. 2012. “Park
Quality in Racial/Ethnic Minority Neighborhoods.” Environmental Justice 5 (6): 271–278.

87

Taylor, W. C., M. F. Floyd, M. C. Whitt-Glover, and J. Brooks. 2007. “Environmental Justice: A
Framework for Collaboration Between the Public Health and Parks and Recreation Fields to
Study Disparities in Physical Activity.” Journal of Physical Activity & Health 4:S50–S60.

Ttth, G., and R. Kincses. 2015. “Accessibility Models Based on the Gravity Analogy: In Theory and
Practice.” SSRN Electronic Journal SSRN Journal 5 (1): 137–158.

United States Forest Service. 2015. “i-Tree Canopyv6.1 (i-Tree Canopy)” Accessed December 15,
2015. https://www.itreetools.org/canopy/index.php

Weiss, C. C., M. Purciel, M. Bader, J. W. Quinn, G. Lovasi, K. M. Neckerman, and A. G. Rundle.
2011. “Reconsidering Access: Park Facilities and Neighborhood Disamenities in New York
City.” Journal of Urban Health 88 (2): 297–310.

Wolch, J., J. P. Wilson, and J. Fehrenbach. 2005. “Parks and Park Funding in Los Angeles: An Equity-
Mapping Analysis.” Urban Geography 26 (1): 4–35. Accessed April 2015.

Wolch, J. R., J. Byrne, and J. P. Newell. 2014. “Urban Green Space, Public Health, and Environmental
Justice: The Challenge of Making Cities ‘Just Green Enough.’” Landscape and Urban
Planning 125:234–244. Accessed April 2015).

Zhang, X., H. Lu, and J. B. Holt. 2011. “Modeling Spatial Accessibility to Parks: A National Study.”
International Journal of Health Geographics 10:31–31.

88

Appendix A: PARA Operational Definitions and Protocols
Protocol

General Directions

At an indoor facility, stop at the reception area and introduce yourself to desk staff
and/or management. Briefly describe the project, and explain the purpose of your
visit.

If an outdoor location, drive around the resource perimeter to assess the safety before
getting out of the car. If anything looks dangerous or suspicious, write a note on the
assessment form and report to Project Manager. Move onto the next physical activity
resource to be assessed. If at any time conditions become unsafe, return to the car and
continue to the next assessment.

If there is a physical activity resource that is not on the list, collect data for it in a
blank Physical Activity Resource Assessment form. Include resource name and
street address.

The outlying boundary for a physical activity resource(s) will be as follows:

If a gate is surrounding the physical activity resource, then the physical activity
resource will be assessed from the gate in.

If there is no gate, but there is a sidewalk, then the physical activity resource will
be assessed from the outer edge of the sidewalk in.

If there are no consecutive posts that signify a boundary, then the physical
activity resource will be assessed from those posts in.

If there is no clear indicating boundary for the physical activity resource, then
the physical activity resource will be assessed from the end of the adjacent
street(s) in.

If there is an outlying ditch that signifies a boundary and there is no sidewalk,
gate, or posts, then the physical activity resource will still be assessed from the
adjacent street(s) in.

If there is an activity resource that starts inside the 1 mile diameter boundary and
extends beyond the boundary, then that activity resource should be fully surveyed
and assessed.

At top of form:

If the form is to assess a pre-identified physical activity resource, there will be a sticker
produced from the Excel record of businesses and physical activity resources for each Housing
development. The sticker should have the correct street address for the physical activity
resource. Please verify that the address on the label is correct. If it is not, please write a note in
the comments section of the PARA form.

Complete each field as specified:

1) Date = Date of data collection
2) Data Collector = Person on research team who collects the data
3) PA Resource ID= Unique physical activity resource identifying number

4) Time = The starting and ending time of data collection

89

5) Phone Call = Call the project manager at departure from the office and arrival back to
the office, and write in a time of when the phone calls were made

6) Type of Resource (circle ONE)
1 fitness club – aka health clubs
2 park – city park
3 sport facility – baseball fields, basketball and tennis courts, soccer fields

4 trail – walking or biking trail (other than sidewalk that is part of a street curb)
5 community center – public building, may include outdoor space
6 church or other religious organization

7 school
8 combination of 2 or more resources: describe in detail
7) Approximate Size (Circle one)

1 sm. = Small = ½ square block,

2 med. = Medium = > ½ square block up to 1 square block,
3 lg. = Large = > 1 sq. block

8) Capacity = (for an indoor facility) The maximum capacity number which should be
posted (in the US), or ask the management

9) Cost = cost for use of facility
1 Free, no charge to use

2 Pay at the door (You must pay to gain entry in the facility)
3 Pay only for certain program (You can use the facility for free, but certain
program/classes have a
fee)
4 Other (List any other type of cost or payment fee)

10) Hours of Operation = The hour that the resource opens and closes (write in US Military
Time; 5am = 0500, 6am = 0600, 7am = 0700, 8am = 0800, 9am = 0900, 10am = 1000,
11am = 1100, 12pm = 1200, 1pm = 1300, 2pm = 1400, 3pm = 1500, 4pm = 1600, 5pm
= 1700, 6pm = 1800, 7pm = 1900, 8pm = 2000, 9pm = 2100, 10pm = 2200, 11pm =
2300, 12am = 2400)
11) Signage – Hours of Operation = place a check on the appropriate box
12) Signage – Rules of Use = place a check on the appropriate box

Features -- Numbers 13 – 25

Rate each item by circling a number. Operational definitions describing each are found
below, in the section on Operational Definitions.

0 = Not Present 1 = Poor 2 = Mediocre 3 = Good

Special note on item 16) Play Equipment. If it is ‘typical’ equipment such as a slide,
swings, horizontal bar; no description is necessary. When the equipment is unusual,
please describe and use the Comments space as necessary.

Amenities -- Numbers 26 - 37

Rate each item by circling a number. Operational definitions describing each are found
below, in the section
on Operational Definitions.
0 = Not Present 1 = Poor 2 = Mediocre 3 = Good

For Incivilities
90

Numbers 38 - 49

Rate each item by circling a number. Operational definitions describing each are found
below, in the section on Operational Definitions.
Feature Poor Mediocre Good
Baseball field – Count
Surface of fields is
uneven, unsafe, no
overhead lighting, no
benches for players,
fencing in poor
condition or
nonexistent
Surface of fields is
uneven, slightly
unsafe, no overhead
lighting, + benches
for dugouts. Some
fencing existent, but
not 100% intact
Surface of fields is
uniform, no
rocks/barriers to
running bases, have
overhead lighting, +
benches for dugouts.
Have bleachers for
spectators, intact
backstop fencing
Basketball courts –
Count (BB courts)
Court of hoop is in
very bad condition,
almost unstable
Hoop is missing a net,
rim is bent, court has
cracks or weeds
Hoop is straight and
has a net or chain,
court is playable
Soccer fields – Count Grass coverage may
be poor in 50% or >
of the field, rough
surface, hazards
and/or trash on the
field
Grass coverage may
be sparse in a few
places, grass may be
too high, some trash
or debris on field
Field has uniform
grass coverage and is
well- mowed, no trash
or debris on field;
nets, if furnished, are
intact
Bike Rack Rack is in poor
condition, almost
unstable or has poor
access
Rack is bent, or
missing paint, but
otherwise usable
Rack is sturdy, usable,
may have a few
cosmetic blemishes
Exercise Stations with
Signage (Exer.
Station)

4 or > stations need
major repair – are not
safe to use. Signage
may be missing or in
poor condition for
several stations. Path
between stations is
unsafe.
3 or < stations may
need minor repair or
maintenance, path
between stations need
minor improvement
Stations themselves
are in good condition
and safe. 5 or >
stations with safe path
between them
Play equipment
(describe if different
than traditional play
equipment – slide,
swings, monkey bars)

Several pieces are in
need of major repair
and is almost or
unstable, there is a lot
of trash, and the
ground is overgrown
Some equipment is in
need of minor repair,
there is some trash,
and the ground needs
some improvement
In good condition,
variety of pieces,
ground in good
condition, well-kept
and clean
91

or barren
Pool > 3 ft deep Swimming pool has
very discolored water
or too little water,
surrounding surface is
in need of repair, trash
in or around pool –
not safe for use
Swimming pool or
deck needs minor
cleaning or treatment
Swimming pool is
clean, well-lit.
surrounding surface is
safe as well as
exit/entry points
Sandbox
Sandbox is < or 1⁄2
full, and/or needs
cleaning (replacement
sand). Box itself
needs major repair,
and is almost
unusable
Sandbox is only 3⁄4
full, and is mostly
clean; the box or
edging could use
minor
Sandbox has adequate
clean sand, all
sides/edging are
sturdy
Sidewalk

Sidewalk has major
damage and needs
repair, almost
unusable
Sidewalk has some
debris, cracks or
uneven surfaces, but
otherwise usable
Sidewalk is smooth,
clear of debris

Tennis courts –
Counts

Courts have cracked
surface, nets are in
major need of repair,
debris is evident;
almost unusable
Court surface and nets
are in need of some
repair, but otherwise
usable
Tennis court surface
and nets are in fairly
good condition

Trails –
running/biking
Surface is unsafe in
many places, there is
a lot of debris, no
signage about
appropriate use

Surface is in places
uneven or in need of
minor repair, may be
a few hazards or
avoidable debris

Surface is smooth,
without unmarked
hazards or debris , has
signage re:
appropriate users

VB courts – Count

Playing surface has
debris or cracks or
bumps all over, net is
almost unusable or
missing

Playing surface has
some debris or cracks
or has 1 – 5 bumps,
net is sagging or has
holes

Playing surface is free
of debris and smooth,
net is in good
condition

92

Wading Pool < 3 ft.

Wading pool has
discolored water, or
no water, trash in or
around pool – not safe
for use

Wading pool needs
minor cleaning or
repair

Wading pool is clean
and well-kept

Amenities Poor Mediocre Good
Access Points

Some appear as
potentially unsafe
areas, unkempt, not
well-marked

Not all access points
are clearly marked.
Some may have trash
or overgrown grass.

Clearly visible, safe,
free of debris or
overgrown grass. If
gated, works properly.

Bathrooms

Bathroom is not clean,
not well-stocked.
More than 50% of
fixtures are in
disrepair

Bathroom is fairly
clean, stocked, and
most sinks’ and
toilets’ plumbing is in
good working order.

Bathroom is clean,
well-lit, stocked, all
plumbing is
functioning well.

Benches – all types of
affixed seating. Count

Benches are in bad
condition, unusable

Benches are missing
some paint or boards,
may be crooked, but
otherwise usable

In good condition but
could have minor
cosmetic flaws

Drinking fountains –
Count

Either all or most
(50%) are broken

At least 1 of the total
fountains not in
working operation

Working, clean
fountains with clean
surrounding area

Fountains (decorative)

Water is unclean or
not flowing. Fountain
itself is in disrepair.
Area at base is in poor
shape

Water is clean;
fountain itself is in
adequate repair. Area
at base could use a
little improvement

Water is clean;
fountain is in good
condition (working).
Area at base of
fountain is well-kept

Landscaping efforts
(this does not include
grass)

Shrubs or flowering
plants appear dead or
more than 50%
overgrown with
weeds. (Does not
Shrubs or flowering
plants in ground, but
do not appear healthy
and/or colorful.
Attractive live shrubs
and/or flowering
plants, perhaps
decorative material
93

include grass)

Existing weeds.

such as rock or mulch

Lighting – Count For
an outdoor resource
such as a park, this is
within the boundaries

Area has limited
lighting, inadequate
for safety

They are usable, but
need minor repair,
partially clean

Area or building has
effective overhead
lighting which
sufficient for safety

Picnic tables shaded
Count
Tables are in need of
major repair, unclean,
almost unusable

Tables are usable, but
need minor repair,
partially clean

Tables are sturdy and
in good condition,
clean

Picnic tables no-shade
Count
Same as above Same as above Same as above
Shelters – Count

Structures are not
intact – so rain would
get into area. If
seating/tables are
present, they are in
major need of repair
or are missing

Structures are in need
of some repair,
provide protection
from weather. If
seating/tables are
present they are
usable but need minor
repair

Structures are intact,
provide protection
from weather. If
seating/tables are
present they are clean.

Shower/Locker room

Unclean, may not be
well-lit, inadequate
dressing space or
receptacles provided,
plumbing is almost
unusable

Most areas are clean,
lockers and/or
dressing space
provided (but is
inadequate), plumbing
could be improved,
but works

clean, well-lit, lockers
and/or dressing space
provided, plumbing
works well

Trash containers –
Count

Unclean and/or in
poor condition, more
care needed, Full with
trash or overflowing.

Partially unclean or in
< perfect condition,
but scattered, and
unstable

Clean on exterior,
scattered throughout,
not overflowing with
trash

Incivilities 1 2 3
Auditory annoyance Sound is not irritating, Sound(s) is (are) Noticeable sounds
94

but is (hardly)
noticeable

noticeable and
interfere(s) with
enjoyment of
resources

which are unpleasant.
Reaction is to leave
area.

Broken glass

A few pieces of
broken glass (the
equivalent of 1 bottle)

Several pieces of
broken glass (the
equivalent of 2 – 4
bottles)

Many pieces of
broken glass (5+
bottles)

Dog refuse

1 refuse pile from dog

2 – 4 dogs refuse piles
from dogs

5 or > refuse piles
from dogs

Dogs Unattended

1 dog unattended

2 – 4 dogs unattended;
may be associated
noise

5 or > dogs
unattended, definitely
unsafe, may be
associated noise

Evidence of alcohol
use

1 bottles, cans, or
bottle caps visible

2 – 4 bottles, cans, or
bottle caps visible

5 or > bottles, cans, or
bottle caps visible

Evidence of substance
use

1 piece: syringes,
paint cans, rags,
baggies, rolling
papers

2 – 4 pieces: syringes,
paint cans, rags,
baggies, rolling
papers

5 or > pieces:
syringes, paint cans,
rags, baggies, rolling
papers

Graffiti/tagging

1-3 small

4+ small or 1 large

2 + large

Litter

A few items (<5) are
on the ground

Several items (5-10)
are on the ground

Many items are on the
ground (11+)

No grass

A small area without
grass

A moderate portion of
the area without grass
A large area without
grass (more than with
grass)

95

Source: Lee et al. 2005

Overgrown grass

A little bit, hardly
noticeable

A moderate amount,
noticeable

A lot, very noticeable,
may be obstructing
some equipment

Sex paraphernalia

1 used or unused
contraceptive devices
and/or 1 pieces of
pornographic reading
material visible

2 - 4 used or unused
contraceptive devices
and/or 2 - 4 pieces of
pornographic reading
material visible

5 or > used or unused
contraceptive devices
and/or 5 or > pieces of
pornographic reading
material visible

Vandalism

Hardly noticeable, but
it appears up to a few
pieces of equipment
or an area of indoor
space has been
defaced

Noticeable, more than
a few pieces of
equipment are
vandalized, or < 50 %
of the space has been
rendered unusable by
vandalism

Very noticeable, more
equipment in disrepair
than in good order,
between 50%-100%,
because of vandalism.
Signs of vandalism
are obvious

96

Appendix B: Adjusted PARA Auditing Instrument
Adjusted PARA instrument as it was used for this research. Additional items, not part of the
original PARA instrument, are marked with “*”.
1. Date
2. Time Started
3. Time Ended
4. Cost: 1 Free, 2 Pay at door, 3 Pay for Certain Programs, 4 Other_____
5. Hours of Operation
6. Signage Hours: Yes, No
7. Signage Rules: Yes, No

General Features
8. Benches: 1,2,3
9. Picnic Tables in shade: 1,2,3
10. Picnic Tables no shade: 1,2,3
11. Shelter: 1,2,3
12. Sidewalk: 1,2,3
13. Restrooms: 1,2,3
14. Drinking Fountains: 1,2,3
15. Lighting: 1,2,3

Children’s Play Amenities
16. Play Equipment: 1,2,3
17. Sandbox: 1,2,3
18. Pool <3ft: 1,2,3
19. Community Center: 1,2,3*

Disamenities
20. Auditory Announce: 1,2,3
21. Broken Glass: 1,2,3
22. Dog Refuse: 1,2,3
23. Dog Unattended: 1,2,3
24. Alcohol Use: 1,2,3
25. Substance Use: 1,2,3
26. Graffiti: 1,2,3
27. Litter: 1,2,3
28. Dead Grass: 1,2,3
29. Overgrown Grass: 1,2,3
30. Sex Paraphernalia: 1,2,3
31. Vandalism: 1,2,3
32. Power Lines: 1,2,3*

Physical Activity Amenities
33. Baseball Fields: 1,2,3
34. Basketball Courts: 1,2,3
97

35. Soccer Fields: 1,2,3
36. Bike Rack: 1,2,3
37. Exercise Equipment: 1,2,3
38. Tennis Court: 1,2,3
39. Volleyball Court: 1,2,3
40. Pool >3ft: 1,2,3
41. Running/Walking Trails: 1,2,3
42. Handball Court: 1,2,3*
43. Horseshoe Pit: 1,2,3*

Nature Access Amenities
44. Fountains (water features): 1,2,3
45. Intentional Landscaping: 1,2,3
46. Riverbed Access: 1,2,3*
47. Tree Coverage: 1,2,3*
48. Nature Guides: 1,2,3*

Source: Lee et al. 2005

Abstract (if available)

Abstract Previous studies of park accessibility have utilized network analysis and dasymetric mapping to investigate pedestrian accessibility to park resources measured in acres per capita. Through a case study of Downey, California, this study extends on previous work in this area by combining network analysis and dasymetric mapping with robust park amenity auditing. The intention of this study is to provide a more detailed examination of how accessibility is affected by park condition and the types of facilities provided to park users. The study uses a method for dasymetrically mapping population data to land parcels, Esri’s ArcGIS 10.1 Service Area Network Analyst tool, and a park amenity scoring system based on the Physical Activity Resource Assessment (PARA) instrument. The results of this research reveal that park accessibility in Downey is limited at multiple Service Area (SA) distance levels due to the presence of parks with high pedestrian accessibility but low amenities in the geographic center of the city and parks with low pedestrian accessibility but high amenities on the city’s periphery. The results of this case study inform policy suggestions for future park developments. These policy suggestions include planning strategies for increasing pedestrian access to parks with developed amenities, which are distant from residential areas. Also, the study indicates which parks to nominate for development in highly accessible areas with few amenities.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

Measuring seasonal variation in food access: a case study of Everett, Washington

PDF

Analysis of park accessibility in Redan, Georgia Web GIS application

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

Applying least cost path analysis to search and rescue data: a case study in Yosemite National Park

PDF

Spatial and temporal patterns of long-term temperature change in Southern California from 1935 to 2014

PDF

Residential housing code violation prediction: a study in Victorville, CA using geographically weighted logistic regression

PDF

Calculating solar photovoltaic potential on residential rooftops in Kailua Kona, Hawaii

PDF

Estimating populations at risk in data-poor environments: a geographically disaggregated analysis of Boko Haram terrorism 2009-2014

PDF

A spatial analysis of veteran healthcare accessibility

PDF

Assessing the impact of a web-based GIS application to promote earthquake preparation on the University of Southern California University Park Campus

PDF

A spatial analysis of violent crime cold spots: testing the capable guardian component of the routine activity theory

PDF

Does the Bay Area have a social center? Delimiting the postmodern urban center of the San Francisco Bay Area

PDF

Finding the green in greenspace: an examination of geospatial measures of greenspace for use in exposure studies

PDF

Effect of spatial patterns on sampling design performance in a vegetation map accuracy assessment

PDF

Network accessibility and population change: historical analysis of transportation in Tennessee, 1830-2010

PDF

Using census data, urban-land cover classification, and dasymetric mapping to measure urban growth of the lower Rio Grande Valley, Texas

PDF

Urban green space accessibility and environmental justice: a GIS-based analysis in the city of Phoenix, Arizona

Asset Metadata

Creator Jimenez, Edgar H. (author)

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

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 02/22/2016

Defense Date 01/15/2016

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

Tag accessibility,audit,dasymetric mapping,GIS,OAI-PMH Harvest,Parks,service area analysis

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Vos, Robert (committee chair), Lee, Su Jin (committee member), Loyola, Laura (committee member)

Creator Email edgarhjimenez@yahoo.com,ehjimene@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-211267

Unique identifier UC11278371

Identifier etd-JimenezEdg-4131.pdf (filename),usctheses-c40-211267 (legacy record id)

Legacy Identifier etd-JimenezEdg-4131.pdf

Dmrecord 211267

Document Type Thesis

Format application/pdf (imt)

Rights Jimenez, Edgar H.

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