ACCESSIBILITY AND EQ UITY ASSESSMENT O F PUBLIC PARKS AND RECREATIONAL FACILIT IES IN ALACHUA COUNT Y By KEXIN CHEN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE RE QUIREMENTS FOR THE DEGREE OF MASTER OF URBAN AND REGIONAL PLANNING UNIVERSITY OF FLORIDA 2018
2018 Kexin Chen
To my loving family, boyfriend and Alachua County
4 ACKNOWLEDGMENTS I w ould like to thank Dr. Paul D. Zwick for making this study possible and for the guidance and encouragement through my entire process I would also like to thank M r. Stanley Latimer for h is contribution to my study as well as for providing me inspiration an d good advice when I run into bottleneck They give generously of their time whenever possible to discuss this paper with me and push me to break through the boundaries and limitations of my personal thinking Last of all, I would like to thank my family my boyfriend and my friends for their unconditional love and for supporting me when I am lost, helpless and hesitated as my forever harbor. Without the companionship of these important people in my life, I can't spend the lonely years abroad.
5 TABLE OF CON TENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 13 1.1 Research Background ................................ ................................ ...................... 13 1.2 Research Objectives ................................ ................................ ......................... 14 2 LITERATURE REVIEW ................................ ................................ .......................... 16 2.1 Definition of Local Parks and Recreational Facilities ................................ ........ 16 2.2 The Function of Public Parks ................................ ................................ ............ 17 2.3 Accessibility of Public Parks and Recreational Facilities ................................ ... 19 2.3.1 Concept of Accessibility ................................ ................................ ........... 19 2.3.2 Relationship between Accessibility and Equity of Public Facilities .......... 2 1 2.3.3 The Method of Measuring the Accessibility ................................ ............. 22 184.108.40.206 Nearest neighbor distance method ................................ ................ 23 220.127.116.11 Buffer analysis method ................................ ................................ ... 24 18.104.22.168 Network analysis method ................................ ............................... 24 22.214.171.124 Cost distance model method ................................ ......................... 24 126.96.36.199 Potential model method ................................ ................................ 25 188.8.131.52 Two step Floating Catchment Area method (2SFCA) .................... 25 2.3. 3.7 Selection of method to measure the accessibility .......................... 26 2.4 Social Equity associated with Access to Local P arks ................................ ........ 27 2.4.1 The Conce pt and Characteristics of Equity in Urban Geography ............ 27 184.108.40.206 Territorial equality ................................ ................................ ........... 27 220.127.116.11 Locational equity ................................ ................................ ............ 28 18.104.22.168 Spatial equity ................................ ................................ ................. 28 22.214.171.124 Social equity ................................ ................................ ................... 29 126.96.36.199 Social equity of publi c parks distribution ................................ ........ 31 2.4.2 Environmental Justice and Social Equity in Urban Green Space ............ 32 3 STUDY AREA AND DATA SOURCE ................................ ................................ ...... 36 3.1 Study Area ................................ ................................ ................................ ........ 36
6 3.2 Data Source ................................ ................................ ................................ ...... 36 3.2.1 Public Parks ................................ ................................ ............................ 36 3.2.2 Classification of Public Parks into Active Parks and Passive Parks ........ 37 188.8.131.52 Active parks ................................ ................................ ................... 38 184.108.40.206 Passive parks ................................ ................................ ................. 38 3.2.3 Population Data at Census Tract ................................ ............................. 39 3.2.4 The Representative Point of Census Tr acts and Green Space ............... 39 3.2.5 Road Network Data ................................ ................................ ................. 40 4 MEASURING ACCESSIBIL ITY OF PARKS ................................ ........................... 46 4.1 Analysis of Area Ratio of Public Parks within Different Walking Distance Thresholds ................................ ................................ ................................ ........... 46 4.2 Evaluation of Public Parks Accessibility based on 2SFCA Method ................... 47 4.2.1 Accessibility of Public Parks ................................ ................................ .... 49 4.2.2 Accessibility of Active Parks ................................ ................................ .... 49 4.2.3 Accessibility of Passive Parks ................................ ................................ 50 5 EQUITY ANALYSIS ................................ ................................ ................................ 55 5.1 Spatial Distribution of Vulnerable Groups ................................ ......................... 55 5.2 Socioeconomic Factors Regarding High Access Versus Low Access Census Tracts ................................ ................................ ................................ ..... 56 5.2.1 Socioeconomic Factors Regarding High Access Versus Low Acces s to Active Parks ................................ ................................ .............................. 59 5.2.2 Socioeconomic Factors Regarding High Access Versus Low Access to Passive Parks ................................ ................................ ........................... 59 5.3 Spatial Clusterin ........ 61 ...... 63 .... 66 6 CONCLUSION ................................ ................................ ................................ ........ 78 LIST OF REFERENCES ................................ ................................ ............................... 81 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 89
7 LIST OF TABLES Table page 4 1 Service area ratio of public parks within differe nt walking distance thresholds ... 51 5 1 Comparison of Socioeconomic Factors of Vulnerable Groups Regarding High Access Versus Low Access to Active parks ................................ ............... 69 5 2 Equity Analysis on Active Parks ................................ ................................ .......... 69 5 3 Comparison of Socioeconomic Factors of Vulnerable Groups Regarding High Access Versus Low Access to Passive parks ................................ ............ 70 5 4 Equity Analysis on Passive Parks ................................ ................................ ....... 70 5 5 Summary table about Equity Analysis by census tract with Low park accessibility and High proportion of vulnerable groups ................................ ....... 71
8 LIST OF FIGURES Figure page 2 1 Th e diagrams of the equality and equity ................................ ............................. 35 3 1 Study area and its spatial distribution of Black, Hispanic and Asian population. ................................ ................................ ................................ .......... 42 3 2 Public P arks in Alachua County ................................ ................................ .......... 43 3 3 Active Pa rks in Alachua County ................................ ................................ .......... 44 3 4 Passi ve P arks in Alachua County ................................ ................................ ....... 45 4 1 Service area ratio of Public parks within 400m buffer ................................ ......... 52 4 2 Service area ratio of Public parks within 800m buffer ................................ ......... 52 4 3 Service area ratio of Public parks within 1200m buffer ................................ ....... 53 4 4 Total accessibility of public parks in Alachua County. ................................ ......... 53 4 5 Spatial accessibility of active parks in Alachua County ................................ ....... 54 4 6 Spatial accessibility of passive parks in Alachua County ................................ .... 54 5 1 Guide to understanding the results of the Mann Whitney U test ......................... 72 5 2 Bivariate LISA Cluster Map of access to active parks and percent under age 18 ................................ ................................ ................................ ....................... 73 5 3 Bivariate LISA Cluster Map of access to active parks and percent over age 64 73 5 4 Bivariate LISA Cluster Map of access to active parks and percent below the poverty line ................................ ................................ ................................ ......... 74 5 5 Bivariate LISA Cluster Map of access to active parks and percent below ninth grade ................................ ................................ ................................ .................. 74 5 6 Bivariate LISA Cluster Map of access to active parks and percentage of non Whites ................................ ................................ ................................ ................ 75 5 7 Bivariate LISA Cluster Map of access to passive parks and percent under age 18 ................................ ................................ ................................ ....................... 75 5 8 Bivariate LISA Cluster Map of access to passi ve parks and percent over age 64 ................................ ................................ ................................ ....................... 76
9 5 9 Bivariate LISA Cluster Map of access to passive parks and percent below the poverty line ................................ ................................ ................................ ......... 76 5 10 Bivariate LISA Cluster Map of access to passive parks and percent below ninth grade ................................ ................................ ................................ .......... 77 5 11 Bivariate LISA Cluster Map of access to passive parks and percentage of non Whites ................................ ................................ ................................ ......... 77
10 LIST OF ABBREVIATIONS LISA: L ocal I ndicators of S patial A ssociation SES: Socioeconomic Status 2SFCA : Two step F loating C atchment A rea
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Urban and Regional Planning ACCESSIBILITY AND EQ UITY ASSESSMENT O F PUBLIC PARKS AND RECREATIONAL FACILIT IES IN ALACHUA COU NTY By Kexin Chen December 2018 Chair: Paul D. Zwick Cochair: Stanley Latimer Major: Urban and Regional Planning conveniently and equitably enjoy the accessibility favors vulnerable groups are key considerations of environmental justice, urban sustainable development, and social equity This paper examines accessibility of active parks and passive parks and then evaluates their social functions through the equity assessment of public parks in Alachua County by means of GIS, SPSS, and GeoDA. The major vulnerable groups are youth, the el derly, non W hites, low educated residents, and low income residents. The Mann Whitney U test is adapted to evaluate the ethnic and socioeconomic differences between high access and low access areas. The Bivariate LISA method is also used to analyze the spa tial clustering patterns between the distribution of vulnerable groups and the accessibility of public parks. Results indicate that the service supply of public parks has not been skewed toward vulnerable groups. The results of the assessment provide an im portant foundation and reference that urban planners can use to design a more
12 rational layout of urban parks and recreational facilities, which would help residents in Alachua County have better and more equitable access to such areas.
13 CHAPTER 1 INTRODU CTION 1.1 Research Background Urban public parks are an indispensable element of urban ecological environments and are also important for realizing that sustainable urban development has significant functions such as protecting biodiversity; easing urban pressures; and providing leisure and entertainment (Wolf, 2003). Parks are part of the natural service system and ensuring that residents can easily enjoy parks and recreational facilities is one of the primary norms of urban modernizati on. However, the construction of urban public parks has been guided by core indicators such as the per capita public green area and green coverage ratio. Although grasping the quantitative characteristics of urban green space is generally helpful, merely u nderstanding the per capita area of public green space in isolation is not enough (Van Herzele & Wiedemann, 2003). Whether urban residents can conveniently and equitably enjoy the services of urban parks and recreational facilities is an important indicato r of environmental justice, urban sustainable development, and social equity (Dominski, 1992; Roseland, 1997). Accessibility of parks and recreational facilities is the best embodiment of the fairness and equity of sharing such resources. Policymakers need to learn more about the division and use of public parks at all levels of society to see whether those who enjoy maximum use of public green space include those who need it most. All citizens should be equally able to reach public places, and public insti tutions should provide services fairly to community citizens. The supply of public parks and recreational facilities as public goods should reflect the ideas of equity and efficiency; equity refers to the provision of services for various social groups, an d efficiency refers to the provision
14 of services for those who really need them. Social equity is an important social responsibility for every professional who operates parks and recreational facilities. Therefore, when evaluating the equity and efficiency of public parks, I need to consider spatial layouts; accessibility to residents; actual populations covered by park services; and socioeconomic attributes of population coverage such as economic, race, religion, sex, and age factors (Talen, 1997) 1.2 Res earch Objectives This paper has three main objectives: 1. To analyze the accessibility of public parks in Alachua County, and the more concrete accessibility assessment would be based on the classification of public parks. Solving the first problem can figur e out the number s of public parks that residents living in different spatial locations can enjoy, explaining their spatial interactions with the green space. 2. To analyze whether census tracts with high proportion of vulnerable groups like minorities and lo w socioeconomic status groups have high accessibility of public parks, that is whether social equity exists. Addressing the second problem can help to understand the racial and socio economic differences between high access and low access areas to obtain w hether the service supply of public parks has been skewed toward vulnerable groups to achieve social equity. 3. To analyze the spatial pattern s between the accessibility of public parks and the spatial distribution of vulnerable groups, and whether the spatia l layout of public parks is consistent with the demand distribution of vulnerable groups. Solving the third problem would help to identify areas with low park accessibility and a high proportion of vulnerable groups from the spatial level which should be t argeted as priority areas for optimization of park layout This paper is organized into six chapters. Chapter 1 presents the research background and research objectives. Chapter 2 presents a literature review based on current studies about public parks, ac cessibility measurements of public parks, and social equity assessment of public parks. Chapter 3 present s the study area and data sources. Chapter 4 measure s the accessibility of active parks and passive parks
15 Chapter 5 analyzes the equity of spatial dis tribution of public parks and Chapter 6 presents the conclusion
16 CHAPTER 2 LITERATURE REVIEW 2.1 Definition of L ocal P arks and R ecreational F acilities Webber believed that the urban green space has the value of protection of natural resources. It reflects the historical and cultural value and embodies the comprehensive value that integrates landscape appreciation with leisure and entertainment (Webber, 1963) The Polish scholar Ostrowski defined open space as not only space that is open, closed, and lack ing constraints, but also publicly accessible space that serve the public, including parks and green space" ( Ostrowski, 1975 ). Cordell et al. define d and accessible to the public, even if there may be temporary use restrictions, including various publicly available space such as parks, gardens, sports venues, etc. ( Cordell & Green, 2002 ). Byrne et al. define d public parks as "a part of urban open space with a large number of plants that directly or indirectly provid e health, safety, comfort, happiness and aesthetics for urban residents ( Byrne & Wolch, 2009 ) In summary urban parks all emphasize the natural features within the urban background, that is, they are all space for maintaining, restoring or creating natur al landscapes. Many scholars further differentiate urban parks into public parks and private parks according to their specific ownership and usage rights. Private parks are attributed to individuals, whose use is restricted unless they are allowed by the o wner, such as courtyards and private gardens in private houses. Correspondingly, public parks as public infrastructure can be freely used by the public. They mainly include natural green space with dense vegetation, such as forests, botanic gardens etc a nd artificially modified green space ( Wolch & Byrne, 2014; Yao & Liu, 2014)
17 2.2 The F unction of P ublic Parks Valuing equity and universal accessibility of public parks and recreational facilities has many benefits, including promotion of community cohesio n ; improvement of mental and physical health ; reduction in crime ; and growth of economic profit. For promotion of community cohesion, where parks and recreational services are abundant, residents have strong place attachment to the communities. Based on th e community stud ies across the United States, the physical beauty, networking opportunities and openness of the community affect ed greatly the creation of emotional bonds within the community and the communities with greatest cohesion had the highest GDP ( Kresl, 2013). The term human network created by urbanist Jane Jacobs are reinforced in some communities by parks and recreational facilities which offer chances for people across all age groups and from all walks of life to interact, learn, and understand. Through the interaction in the safe open space, community cohesion would your & Crompto n, 2014; Jacobs, 1961). Numerous studies have shown that the more social networks of neighborhoods have, the more secure the community is. Any activit ies and open places that c ould promote community ties w ould add value to the neighborhood and extend to th e city (Harnik & Welle, 2009). For improvement of mental and physical health, the recreational environment like parks could support physical activity in the free time and help to reduce chronic diseases rate in such vulnerable groups, especially youth and the elderly (Sallis & Kerr, 2007). The physical factor living near parks was consistent with high level of physical activity (Humpe & Leslie, 2002). The less access to recreation facilities for some lower
18 income groups would likely result in decreased phys ical activities and increased obesity (Gordon Larsen, 2006). Communities with more parks and organized park program would raise the usage of parks and increase physical activity (Mowen, 2010). Based on a study conducted on residents of Wilmington, over 15, 498 Wilmingtonians t ook part in parks actively to upgrade their physical quality in 2008. The Parks Health Benefits Calculator compute d the medical care cost savings due to their park usage which change d into increased physical activity and result ed in eco nomic savings and saved $4,322,000 in total (Harnik & Welle, 2009). Another research show ed that about 16,500 adult residents in the City of Plano using its park s and recreation facilities to participate in physical activity acquire d a number of health ben efits, bringing an annual medical cost savings of $21.2 million (Bowen & Parry2015). For reduction in crime, well organized parks with recreation services and healthy program would help to create an active and healthy atmosphere in communities which would bright about a reduction in crimes and injurious activities. The Summer Night Lights program, undertaken by GRYD, which focus on increasing community resiliency and reducing gang violence, ke pt parks open at night and provide d free food distribution. The p rogram in its sixth year has already cut down 73% gang related crimes near those parks and 85% assaults (Dunworth, 2010). For growth of economic profit, parks with trails and open space can raise the value of nearby house since people who want to live near those facilities willing to pay more. And the rise in property value would bring about the increased property tax parks can also attract visitors to generate direct vis itor spending and sales tax receipts
19 from spending by tourists (Harnik & Crompton, 2014; Fausold & Lilieholm, 1999). A research conducted in Long Island and New York showed that parks in that region bring about the economic benefit of $2.74 billion annuall y (Crompton, 2001). A study conducted in Plano indicated that the parks increase d the value of nearby homes by $337 million and br ought about property tax revenues by $6.08 million a year (Crompton, 2001). 2.3 Accessibility of Public P arks and R ecreational F acilities 2.3.1 C oncept of A ccessibility Accessibility, as an index reflecting the cost of transportation, first appeared in the field of transportation research. In 19th century, Von Thnen regarded the transportation cost as an indispensable aspect of the determinants of costs of location model reflected on agricultural location theory (Von Thnen, 1966). Until 1959, the connotation of accessibility was formally proposed for the first time. Hansen defined it as the size of the opportunity for each node to interact with each other in the transportation network. Since that beginning, the research es about accessibility of public resources gradually deepened into the stud ies of urban planning, urban transportation planning and other regional and spatial rese arch es Dalvi and Martin further added the objects involved in accessibility and the ways a certain facility from a certain place based on a given mode of transporta in overcoming spatial barriers, that is, the smaller spatial barriers the better accessibility, 96). Morris et al. also believe d that accessibility reflect ed the amount of opportunities to get close to target within a unit
20 of time (Morris & Dumble, 1979). Shen define d the potential of spatial interaction as accessibility, that is, the magnitude of mu tual interaction between two point s in space is characterized as accessibility (Shen, 1998). Briefly, accessibility reflects the amount of resistance that users need to overcome to reach the location of a given facility from their location. The actual res earch and application analyze d the source of resistance and measure d the magnitude of resistance from four perspectives such as land use, traffic, time, and individual. These four aspects interact ed and jointly determine d the accessibility ( Geurs & Van, 20 04). The perspective of land use mainly includes three aspects: 1) the quantity, quality and spatial layout of facilities ; 2 ) the location of the user s (such as the distribution of residential area ); 3 ) the series of competitions caused by the unbalanced p attern between facilit supply and user demand. The traffic aspect reflects all the transportation costs that users spend on a given transportation system between the two places, including transportation time and expenses. The time perspective reflect s the time constraints, such as opening hours of facilit ies Personal perspectives reflect the individual's need for access to facilities (depending on age, income, education level, etc.) and the physical and economic capabilities they possess. In summary, accessibility is a complex concept of three dimensions covering time, space, and social factors It includes not only the objective level, which measures the degree of convenience in transportation between points, but also implies the subjective (psycholo gical) level, that and ability.
21 As for researches on the accessibility of green space, Nicholls (2001) conducted comparative analysis of the accessibility and equ ity of parks based on different ways of travel. The combination of GIS technology and economic data was extended to the evaluation of urban public service facilities accessibility. Subsequently, studies about the impact of public parks accessibility on r esidents' health, quality of life and lifestyle has gradually increased (Fors, 2015; Wolch, 2011; Khotdee, 2012; Neutens, 2015). So, the accessibility of public parks is suggest ed as the degree of difficulty from any point in the space to public parks Gre ater spatial resistance to a certain point indicate poorer accessibility, and vice versa. From the perspective of time, the interaction of two points is mainly achieved through the transport system. Time is the most important resistance factor, so time cos t is an important aspect that reflects accessibility. From the perspective of social value, higher accessibility will help improve the satisfaction of use of public facilities and the attractiveness of urban landscapes. The evaluation of acce ssibility must have three elements, namely, the starting point (residential area), the transportation system used to access the end point, and the destination (public parks). 2.3.2 Relationship between A ccessibility and E quity of P ublic F acilities The acce ssibility of public parks is concerned with the social equity of public resource allocation. Applying accessibility of public facilities visually describe d the geographical differences in public service allocation, reflecting the equity associated with acc ess to public parks, thereby identifying areas that should be taken care of in the absence of appropriate public facilities (Gu & Yi, 2010). Therefore, among methods of evaluating spatial equity, accessibility was the most widely used and effective tool fo r
22 revealing the inequity of the spatial layout of public facilities (Talen, 1998; Tsou & Chang, 2005). Delmelle and Casas found that the bus rapid transit based accessibility was associated with the socioeconomic status of the residents in Cali, Colombia. The residents of the middle income class were the largest beneficiary of bus rapid transit based accessibility while the residence of highest income and lowest income level ha d relatively low accessibility (Delmelle & Casas, 2012). Talen use d the data of P ueblo and Macon, Colorado to comprehensively analyze the difference s of social equity of urban park distribution in these two cities through the correlation between accessibility of public parks and selected socioeconomic factors (Talen,1997). Lotfi and Ko ohsari used buffer analysis and Boolean operations to study the accessibility of parks, schools and shops and then found that there was an inconsistent spatial relationship between distribution of facility and beneficiary populations (Lotfi & Koohsari, 200 9). Through questionnaire survey, Erkip comprehensively evaluated the accessibility and equity of Ankara Park by using the number of parks, population distribution, travel time and accessibility as indicators. Results show ed the equity of park distribution should depend on the needs of different social groups (Erkip, 1997). 2.3.3 The M ethod of M easuring the A ccessibility Studies about accessibility of public parks is mainly divided into qualitative analysis and quantitative analysis. Qualitative analysis ma inly analyzes the factors that hinder accessibility in certain types of parks, but qualitative description cannot obtain quantitative indicators. While quantitative analysis is mainly based on statisti cal analysis and GIS spatial analysis. The statistical analysis refers to obtaining data through sociological methods such as questionnaires, interviews, and on site observations, and
23 then uses mathematical statistics to summarize resistance factors and evaluation criteria for accessibility (Nicholls & Shafer, 2001). GIS spatial analysis is a comprehensive information technology system supported by computer network system, which collects, stores, retrieves, analyzes and displays geographic information, and uses its analysis module to collect and edit spatial in formation about public parks to measure their usage level. The GIS based analysis is relatively simple and accurate compared with the statistical analysis. Currently the GIS based methods applied to accessibility assessment mainly include: Nearest neighbo r distance method (Kessel, 2009), Buffer analysis method (Potestio, 2009; Richardson, 2010), Network analysis method (Comber, 2008), Cost distance model method (Jobe, 2009), Potential model method (Geertman & Ritsema, 1995), and Two step Floating Catchment Area method (2SFCA) (Luo & Qi, 2009; Dai, 2011) 220.127.116.11 Nearest neighbor distance method The n earest n eighbor d istance method calculates the shortest distance between the location of residents and the nearby green space to get the accessibility, where th e residential area and the green space are abstracted as points. This evaluation method intuitively reflect ed the behavior of residents choosing the nearest park or recreational facility for recreational activities (Talen & Anselin, 1998). The nearest neig hbor distance method mainly relie d on data analysis for the calculation of accessibility, which is not an intuitive spatial analysis and then relatively difficult to handle the optimization of the spatial layout of parks and recreational facilities. Althou gh it is possible to find a residential area that lacks services of public parks it is difficult to answer the question of where the newly park can be built to achieve the maximum efficiency of service and the evaluation error is large (Apparicio, 2008)
24 18.104.22.168 Buffer analysis method The buffer analysis model creates the service radius of public parks, which is the equal distance from the edge of the nearest parks. Urban residents within the buffer zone can easily and conveniently access parks and recreat ional facilities, while urban residents outside the buffer zone are considered to be unable to access. The principle and nature of calculation are consistent with the nearest neighbor distance method. The advantage of this method is that the calculation is simple and easy to operate in the planning, but it gives no considerations to the influence of road network and other influencing factors different from the actual situation. Both the buffer analysis method and the nearest neighbor distance method ignore obstacles in the process of accessing to destination, and it is easy to overestimate the accessibility of public facilities (Nicholls & Shafer, 2001). 22.214.171.124 N etwork analysis method The network analysis model employs the actual road network and construct s a networked model in the GIS. The advantage of such method is the use of actual traffic network which can calculate vector data fully reflect ing the accessibility of public facilities. The network analysis method can simulate user's actual usage and actua l paths. It can not only analyze the actual traffic conditions and capacity, but also can identify spatial obstacles, which is also the fundamental difference s compared with other methods (Comber, 2008). 126.96.36.199 C ost distance model method The cost resistan ce model mainly studies the impact of roads and other different types of land use on individual movement speeds. The model reflects the obstacles that residents need to overcome in the process of horizontal movement of residents to reach
25 urban park s Gener ally, distance, time and cost are used as evaluation indicators. This method can more intuitively reflect the transportation cost of residents in the process of movement. However, this method does not take into account factors such as the population distri bution, the attractiveness of different categorie s of green space and the different modes of transportation used in the movement. Therefore, it cannot fully reflect the efficiency and equity in the use of urban green space (Jobe, 2009). 188.8.131.52 P otential m odel metho d The potential model means that the potential between two points of travel is proportional to the traffic potential of the starting point and to traffic attraction potential of arrival point and is inversely proportional to nth power of the traf fic distance or time between points. The advantage of this method is to consider attractive interaction between arrival point and starting point in the analysis of accessibility which incorporates the attraction into the calculation (Ma & Cao, 2006). Howev er, the overall model is much complex with different modeling methods, that is, under different conditions, their calculation results have different meanings, and most of them are dimensionless constants. Therefore, the meaning of the calculated constant c annot be judged correctly, effectively and intuitively (Geertman & Ritsema, 1995). 184.108.40.206 T wo step Floating Catchment Area method (2SFCA) The 2SFCA method was first proposed by Luo et al. (2003) for the measure ment of accessibility of medical facilities which has become an important paradigm for research es about health care access ibility and has been applied by many scholars. For example, Cervigni et al (2008) calculated the spatial accessibility of pediatric services by using the 2S FCA method Since thi s method overcomes the limitation of the boundary of administrative region and is intuitive and easy to operate, it is widely used
26 in the research es about accessibility of public facilities. It is assumed that the accessibility is equal within a certain se arch radius and does not conform to the law that accessibility decreases with distance actually. Later, researchers included Luo (2006) have proposed the improvement on this method. Based on the same theoretical framework, the gravity model and the 2SFCA method all take into consideration the influence of the supply scale, demand scale and resistance between supply and demand (Luo et al., 2003). While t heir difference s lie in the treatment of distance factors : The former uses the continuous distance attenu ation function, which takes into account the characteristics of attenuation of facilities service capacity with distance, but does not limit the effective search radius of the facility; and the 2SFCA method uses the dichotomy method to deal with distance attenuation, for example the accessibility within the search radius threshold is the same, but it is completely unreachable beyond the search radius. 220.127.116.11 Selection of method to measur e the accessibility Each method of calculation of accessibility has i ts own advantages and disadvantages. For example, the buffer analysis method and nearest neighbor distance method are relatively simple since both methods adopt Euclidean distance which is quite differen t from residents' actual travel routes and ultimately overestimates the accessibility of green space. The potential model method calculates the mutual attraction between the green space and the starting point, but still uses the Euclidean distance to represent the distance. In addition, the modeling method a nd parameter setting of this method are complex which result in difficult ies in interpret ing the results The network analysis method on the basis of actual road network and actual park entrance can relatively accurately assess the accessibility of green s pace. The
27 calculation results are relatively true and reliable, but the detailed traffic network data is not easy to obtain. Among all accessibility evaluation methods, 2SFCA is the most widely used with comprehensive factors and solid theoretical foundati on 2SFCA benefits from the idea s of two step floating catchment, which is easy to understand and have stronger operability and provide s a viable basic framework for various possible extensions (McGrail et al, 2009). Based on the above advantages of 2SFCA, this paper uses this method to evaluate the accessibility of public parks 2. 4 Social E quity associated with A ccess to L ocal P arks 2.4.1 The C oncept and C haracteristic s of E quity in U rban G eography The term equity was first seen in the social sciences. Ho wever, in the study of urban geography, the idea s of equity are associated with the rationality of spatial layout of public service s or services or facilities has institutional and systemic discrimination against certain special and evaluation methods of equity of public resource allocation are endless and diverse, and generally show the continuous improvement over time with the improvement of people's level of knowledge and technology and the development of society. It has 18.104.22.168 T errit orial equality The social movements of the 1960s and 1970s triggered concerns about urban rights and social justice. Th ese social movement s inspired the research boom of social justice. At this stage, western governments have established a sound social wel fare system to achieve social security for all. The issue of equity was based on equal
28 distribution, emphasizing that the per capita public services enjoyed by residents in different regions should be equal (Rich, 1979). Bleddyn Davies proposed the concept of geospatial space (Davies, 1970). Territorial justice and spatial equality are the core concepts at this stage, focusing on whether the per capita public service s are equal, regardless of human needs, the actual spatial layout of facilities and service benefits. 22.214.171.124 L ocational equity In the 1980s, since the welfare models was too costly and inefficient, western countries focused on efficiency of public service Correspondingly, at this stage the locational equity emphasize d the service benefits and focuse d on quantity of facilities allocation. Locational equity means fairness and equality in public resources allocation. Wicks and Crompton propose d four sub conc e d the locational equity of public services. The provision of public services should meet the minimum demand level with the same opportuni ty and without sacrificing public service benefits (Wicks & Crompton, 1986). The concept of locational equity basically follows the idea s of equal distribution but proposes the minimum demand standard and the idea s of properly considering the efficiency of public services 126.96.36.199 S patial equity the size of residents' access ibility to public resources, emphasizing the equal opportunity to access to public park s regardless of where residents live. It considered
29 placed on equally spatial separation or spatial proximity of residents from public facilities (Talen, 1998; Smith,1994; Kinman, 19 99; Ogryczak, 2000). Compared with the previous stage, spatial equity research mainly shows two major advances: (1) Evaluation method: GIS technology is used to quantitatively calculate the actual distance between the user's geographic unit and the public facility and the equity assessment is represented by the accessibility hierarchical map; (2) The evaluation scale: the evaluation unit is reduced from larger administrative district to neighborhood level, making the equity research more micro and accurate. However, this space and differentiation of social groups. The actual needs of different target groups such as children and ethnic minorities, will be neglected because of the homogenization process, which corresponds to the long standing equality. 188.8.131.52 S ocial equity From the 20 century, western academic circles questioned the los s of equity on account of paying too much attention to market and efficiency and raised the importance of civil rights and humanism in public service. Therefore, the social equity is proposed in the context of social spatial differentiation, which take pub lic facilities and user characteristics into consideration that means the distribution of public facilities should consider different regions, different economic, ethnic and political groups (Omer, 2006). Thompson propose d that allocation should consider t he characteristics of various social groups, that is, the equity issue of public services should consider the characteristics of users, and the results of evaluating equity should be based on who gets what and not where can get what (Thompson, 2002). The s ocial equity shifts from
30 equality and fairness among different social groups, emphasizing the reverse inequality of opportunities for different economic, ethnic and po litical groups to achieve equal access to public resources (Lineberry, 1977). Social equity combines the spatial distribution characteristics of public resources and the demand of users, emphasizes on demand distribution, and gives certain care and prefere ntial treatment to vulnerable groups. It is a relative fairness -"Equity". The issue of social equity is based on the assumption that social unfair ness exits: there are racial polarization and injustice in the public services allocation, and this inequity tends to be distributed in the lower classes (Smith, 1994). For example, park s cluster in proximity to high income neighborhoods, dominant racial groups (Abercrombie et al., 2008) and people with high socioeconomic status (Wolch et al., 2005; Lindsey et al ., 2001). Therefore, Talen proposed that the degree of rationality of facility layout should be based on the different needs of the public. The fundamental point is to reduce the inequality of facilities supply caused by hierarchical differentiation and ma ximize the consideration of needs of low income people (Talen & Anselin, 1998). The Figure 2 1 illustrates the relationship between e quality equity and social to social equity. It is unreasonable to regard citizens of different social groups as having equal needs for parks because citizens do not have the same starting line with the same characteristics and attributes. Therefore, equal opportunity has nothing to do with social equity. Even if the distribution of public resources themselves is equal, it can
31 make those who have higher incomes, higher education, and higher socioeconomic status win over vulnerable groups (Wicks & Crompton, 1986). 184.108.40.206 S ocial equity of publ ic parks distribution In its early stages, the equity associated with urban green space is affected by equality policies. Generally, equity assessments of public parks are evaluated by per capita public green space; this is the so called territorial equal ity stage. The evaluation method reflects quantity equity but ignores spatial equity. With the advent of the accessibility model, the limitations of equity assessment of urban green space were broken, and equity research about public green space entered th e stage of spatial equity. Currently, based on spatial equity and differences among social groups, we have entered the social equity stage; the ultimate goal is to narrow the gap caused by social differentiation. Therefore, we should pay attention to the f ourth type of equity: social vulnerable groups so that social resources can be enjoyed equally by people of all walks of life. Defining social equity as giving more oppo rtunities to disadvantaged groups so that vulnerable groups can have better access to public park s may cause controversy among the wealthy; they pay more taxes and contribute more to the construction of public facilities. However, it is precisely because t he wealthy occupy more social resources and enjoy a higher socioeconomic status than disadvantaged groups do that more urgent work should focus on improving conditions for vulnerable populations in order to reduce such disparities. Furthermore, because the y have higher socioeconomic statuses, wealthier social groups are better able to enjoy the services of private parks and golf courses ; wealthier groups can spend more time and money on visiting passive
32 parks in rural areas. Therefore, vulnerable groups who do not have the conditions to enjoy the services of private parks should have better access to public parks In summary, the concept of social equity considers the capabilities and needs among social groups and advocates that when such capabilities and ne eds differ, the distribution of public park s should be skewed toward disadvantaged groups. The explicit aim of ensuring equality of outcomes in a particular sphere can justify inequality and unequal treatment. For example, unequal inputs of public service can compensate for existing low levels of health, particular local hazards, or other criteria of need 2.4.2 Environmental J ustice and S ocial E quity in U rban G reen S pac e For a long time environmental justice focused on the inequity problem that people exp osed to harmful natural elements like air pollution, water safety, garbage disposal sites and so on. Currently environmental health has incorporated the living environments or artificial built environments (Sallis et al., 2006). Scholars have expanded the ir research on environmental justice to explore the distribution of architectural environmental elements, including urban public green space. Boone believe d that equity is the high level of environmental justice considering socioeconomic factors such as de mand, choice and value (Boone et al., 2009). So, the core concern of public parks is their social equity of spatial distribution. The spatial distribution of public parks should meet the different demands of different social groups in consideration of thei r socioeconomic status, ethnicity, age and gender (Byrne, 2009 Lineberry, 1977). V ulnerable residents need better accessibility of public services (Harvey, 1973). Due to different research objects, analysis methods, research areas and evaluation indicators in different studies the results of equity assessment of urban green space are not the same. It is believed that the distribution of public parks is not
33 always equal, and their accessibility is often associated with economic income, ethnicity, age, gende r, political rights, and other factors (Byrne et al., 2009; McConnachie & Shackleton, 2010). Earlier studies have shown that areas with the high percentage of marginalized groups have no disadvantage in the spatial distribution of public products like urba n parks. For instance, Lineberry found poor communities were actually favored in the park allocation (Lineberry, 1977). Mladenka and Hill f ound that public park s were configured with no special discrimination for low income communities (Mladenka et al., 19 77). However, recent researches do not agree with this "equitable" distribution. Some research cases have found that the accessibility of parks or other recreational facilities wa s relatively low in areas with low socioeconomic groups, high proportion of f oreign races or minorities (Estabrooks & Gyurcsik, 2003; Harris et al., 2015; Powell & Chaloupka, 2004). Erkip pointed out that parks and recreational facilities e ntering Ankara depend ed mainly on the level of income of individuals (Erkip, 1997). Sister an d so on found Hispanics, non W hite, or low income groups had lower park accessibility in the United States (Sister et al., 2010). And accessibility of green space is even worse where the proportion of African Americans is higher in Atlanta (Dai, 2011). Pha m et al. found that in Montreal, accessibility of urban parks was often less in the neighborhoods where low income resident s and minorities concentrated (Apparicio et al., 2012). Meanwhile, the equity assessment of access to public parks by vulnerable gro ups began to emerge. For example, some studies have investigated the differences in distribution of recreational facilities and its relationship with the physical and mental health of adolescents (Cutts et al., 2009; Reyes & Morency, 2014; Kabisch & Haase,
34 2014). Maas (2006) pointed out that in large cities, the elderly, adolescents and low educat ed populations we re more likely to benefit from the supply of public parks in the living environment than general populations. Therefore, this paper analyzes acce ssibility of public parks and recreational facilities in Alachua County concerning socioeconomic indicators like economic income, ethnicity, age, and gender to indicate whether the distribution of public parks in Alachua County has specific discrimination against d isadvantaged groups and whether accessibility of parks and recreational facilities gives certain care and preferential treatment to vulnerable groups to achieve s ocial e quit y
35 Figure 2 1 The diagrams of the equalit y and equity Adapted from All of Our Children Deserve a Chance to Succeed, In Georgia Education Equity Coalition, n.d., Retrieved September 29, 2018, from http://www.gastandardsequity.com/about/
36 CHAPTER 3 STUDY AREA AND DATA SOURCE 3. 1 Study Area The study area is A lachua County, a county in the U.S. state of Florida. Based on the 2010 Census, the estimated 2017 population was 266,944 (US Census Bureau, n.p., n.d.). The county seat is Gainesville (National Association of Counties, n.p., n.d.). The maj ority of its population is W hite, making up almost 70.0% of the total population. 20.6% of the population is B lack ; 6.1% of the population is Asian and 9.8% of the population is Hispanic or Latino (US Census Bureau, n.p., n.d.). The selection of study area is appropriate because of the racial and ethnic structure of Alachua County shown in Figure 3 1 and the guiding principle of environmental justice. race, color, national ori gin or income is entitled to equal protection from environmental n.p., n.d.). According to environmental injustices, minorities were exposed to higher unequal environmental risks than other s egments of society (Hartley, 1995). For many years, ethnic differences have had the bad influence on the lives of the people in important to improve minorit conditions by providing them more opportunities to reduce these d isparities and achieve environmental justice ( "Understanding Racial Inequity in Alachua County", 2018 ) 3.2 Data Source 3.2.1 Public Parks The Data about public parks (total: 160) was acquired from the Alachua County Growth Management Office, combined with the GeoPlan Parks and Recreational
37 Facilities Boundaries in Alachua County in 2017 (Figure 3 2). Consistent with previous studies, public parks in this paper includ e natural park s ( e.g., boat ramp multi use trail preserve and water space ), nationhood pa rks ( e.g., community garden mixed use recreation playground and walking path), state parks, recreational facilities ( e.g., bike trails child recreation museum and sports recreation), and historical preservations (e.g., battlefields), because they are m ain open space of the city where the public can access with freedom (Comber et al., 2008; Richardson et al., 2010). i, 2011). The size in acres of each public parks is of great significance to the subsequent analysis since it is used to represent its capacities (Boone et al., 2009; Harnik, 2004). Total area for public parks is 37596.700369 acres and the total population in Alachua County is 266,944 So, the per capita park area is 140.84 acres per 1000 residents. Compared with the basic standard proposed by the NRPA that the provision of urban parks should be 10 acres per 1000 residents, Gainesville has adequate supply l evel of public parks 3.2.2 Classification of Public Parks into Active P arks and P assive P arks Since the equity assessment of public parks is to get a better understanding of relationship between the supply of public parks and recreational facilities and t he demand of vulnerable groups and then to meet the needs of vulnerable group s understanding the different needs of the public for different uses of parks and recreational facilities can better adjust the layout of public parks based on different demand s Therefore, to delve deeper into accessibility of parks and recreational
38 facilities for different uses and their links to socio economic characteristics, I need to make a simple classification of public parks and recreational facilities. Public parks that provide passive and active recreation opportunities are important contributions to the A lachua C ounty area general plan, n.p., n.d.). Therefore, the classification of public parks needs to consider the active and p assive recreation opportunities they provide. 220.127.116.11 Active p arks Active parks being closer to our daily lives ha ve the urban features and require intensive development. In densely populated urban areas, public parks that provide active entertainment oft en appear as neighborhood parks and are important community assets for people. They usually involve team activities, such as playgrounds, stadiums, swimming pools gymnasiums and skate parks. Due to the need of provid ing more recreational facilities to mee t the needs of nearby residents, such as child recreation, leisure activities, sports recreation, active parks usually involve more management and n.d.). So, the first category of pub lic parks in this paper would be a total of 106 active parks, including neighborhood park and recreational facilities and so on The Data about neighborhood parks and recreational facilities was selected from the A lachua County Growth Management Office, c ombined with the GeoPlan Parks and Recreational Facilities Boundaries in Alachua County in 2017 ( F igure 3 3). 18.104.22.168 Passive parks Then the other classification of public park s in this paper is what provides passive recreation. Passive parks with low inte nsity entertainment usually require very little management and can be offered at very low cost. These parks are more of the
39 natural park, emphasizing the open space aspect and allowing the conservation of natural habitats. They are usually on the outskirts of densely populated urban areas and involve low levels of development, not necessarily close to people's daily li fe as residents may choose to spend more t ime and money on visit ing passive parks during a longer interval. Passive park officials usually p rovide only rural picnic areas, boat ramps and trails, so visitors can perform such as horse riding and mountain biking; or sedentary activities such as observing nature, b ird watching painting, photography Is Palisades Park Passi ve or Active n.d.) So, the second category of public parks in this paper would be a total of 54 passive parks, including natural parks and conservation area, and so on The Data about natural parks and conservation area was selected from the A lachua County Growth Management Office, combined with the GeoPlan Parks and Recreational Facilities Boundaries in Alachua County in 2017 ( F igure 3 4) 3.2.3 Population D ata at C ensus T ract The related data about population at census tract level was obtained from the census 2010 Summary File 1 (US Bureau of Census, 2010a). The total number of census tracts in Alachua county is 56 for population distribution, because it is the lowest areal unit used in the current pr actice of S hortage A 869, 2003). 3.2.4 The R epresentative P oint of C ensus T racts and G reen S pace When calculating the accessibility of public parks using GIS, I need to figure out the centroids that can represent the census tracts and public parks. T he related s patial distribution information of census tract and block group s were collected from the 20 10 Census TIGER/Line files (US Bureau of Census, 20 10 b).
40 e for some open green space, any point along the perimeter can arguably serve as the entry point (Dai, pg 236, 2011) In addition, the population weighted centroid was used to represent the census tract. As the population is seldom distributed homogeneou sly within a census tract, the population weighted centroid instead of the simple geographic centroid of a census tract represents the location of population more accurately. The population centroid of a tract may be distant from its geographic centroid, p articularly in rural or peripheral suburban areas where tracts are large, and population tends to concentrate in limited space. Weighted centroids are computed based on block level population data (Luo et al., pg. 869, 2003) ( 3 1) (Luo et al., pg. 869, 2003) where Xc and Yc are the x and y coordinates of the weighted centroid of a census tract c; X i and Yi are the x and y coordinates of the ith block centroid within that census tract; Pi is the population at the ith census block within that census tract; and nc is the total number of blocks within that census tract (Luo et al., pg. 869, 2003) 3.2.5 Road N etwork D ata This paper use d network analysis in ArcGIS to calculate the catchment of 10 min driving time for population weighted centroids of each census tracts and geographic t he ArcGIS Network An alyst extension makes it possible to model potential travel paths along diverse types of d the road network data (TIGER/Line files) from the US Census Bureau to create network dataset for the Road shapefiles. Sin ce the field of drive time was missing in the attribute table of the shapefiles of roads
41 network and it was appropriate in this paper to choose driving time to measure the catchment size, the field of driving time (DT) was created by using the formula belo w. (3 2 )
42 Figure 3 1 Study area and its spatial distribution of Black, Hispanic and Asian population
43 Figure 3 2 Public P arks in Alachua County
44 Figure 3 3 Active P a rks in Alachua County
45 Figure 3 4 Passive P arks in Alachua County
46 CHAPTER 4 MEASURING ACCESSIBIL ITY OF PARK S 4.1 Analysis of A rea R atio of P ublic P arks within D ifferent W alking D istance T hr esholds T he walking speed of ordinary people is about 5km/h and it takes about 5 minutes to walk 400 meters. Residents' willingness to visit public parks will decrease as travel distance or time cost increase. It was believe d that 15 minutes is the longest walking time that most people can accept (McCormack, et al., 2010; Willemse, 2013) In this paper 400 meters (5 minutes), 800 meters (10 minutes) and 1200 meters (15 minutes) are used as walking distance thresholds, and the buffer analysis is used to mea sure the area ratio of public parks within different walking buffer t o examine the level of coverage of public parks Specifically, the buffer tool in the ArcGIS is use d to generate buffers of different radii for the public parks and then this paper analy zes and counts the proportion of the residential area that falls into the public buffer to the total area of residential area. The formula for calculating the service area ratio of public parks is: ( 4 1) This ratio uncovers the opportunities of access to public parks for residences in each of cen sus tract s T he residential parcels were used in this paper to represent the actual living area. The more accurate service area ratio of public green space should just consider the proportion of residential area w ithin the public parks service buffer to th e total residential area of each census tract.
47 As shown in Figure 4 1, Figure 4 2 and Figure 4 3, the service area ratio of public parks under different walking distance thresholds, reflect s the coverage of public parks. And the service area ratio of publi c parks increases with the increase of walking distance threshold. The service area ratio of public parks within the walking distance of 1200 meters reaches 52.95%, which covers about 52.95% of residential areas. While the service area ratio of public park s within the 400 meter walking distance is only 19.14%, covering only about 19.14% of residential area s shown in Table 4 1. In addition, under the three walking distance thresholds, the service area ratio of public parks is gradually decreasing from the ce nter of Gainesville, and the service area ratio of public parks within Gainesville downtown is the highest to a first approximation, followed by the southern region, followed by the eastern region and the outermost area in the western side, while the north ern area and the west of Gainesville have the lowest area ratio of public parks 4.2 Evaluation of P ublic P arks A ccessibility based on 2SFCA M ethod The service area ratio of public parks calculated by the buffer analysis can just reflect the coverage level of public parks and the opportunit ies for residents to approach public parks but it does not consider the impact of such factors as the supply of public parks the demand of population and travel resistance (road network constraints) on the accessibility of public parks. Based on the previous literature review, 2SFCA is most widely used with comprehensive factors and solid theoretical foundation among the accessibility evaluation methods. Therefore, this paper calculates the accessibility using the 2SFCA method green space beyond this catchment w as assumed to be inaccessible to the residents pg. 237, 2011).
48 At the first step, for each green space location j, search all populati on locations (k) that are within a threshold travel time d0 from j, thus formulating the catchment for green space location j. S um up the total populations within the catchment for j as the potential users for the green space at j (Dai, pg. 237, 2011) Equ ation 4 2 below outlines the provider to population ratio (supply ratio) (Rj) of public parks. The size of each public parks is divided by population living within that d0 drive time catchment. ( 4 2 ) Where Pk is the population at location k whose centroid falls into the catchment (i.e., dkj d0) from green space location j; dkj is the travel time between population location k and green space location j; Sj is the capacity (i.e., size in acres) of gre en space at j (Dai, pg. 23 7 2011) At the second step, for each population location i, search all green space l within the threshold time d0 from i, thus formulating the catchment for the population at i. Sum up total R within the catchment area i to obtain the spatial accessibility at population location i as follows (Dai, pg. 237, 2011): ( 4 3 ) Where l denotes all green spaces within the catchment of population location i, and all other notations are the same as equation above. The accessibility score (Ai) suggests the amount of green spaces (in acres) for every 1000 res idents in a neighborhood (Dai, pg. 237, 2011). pg. 237, 2011). The spatial query above about service area ratio of public parks within different walking distance thresholds showed that 92.27% census tracts are beyond the walking buffer of 400 meters. Nearly 83.97% quarter mile but are likely to drive pg. 237, 2011; Boone et al., pg. 772, 2009). Therefore, when
49 deciding the catchment size (d0), it is appropriate in this paper to choose driving time. The longer drive time would produce less var iation. With reference to Dai's previous research, I can find a 10 min catchment can reflect the spatial variation in the accessibility of public parks. Catchments less than 10 min will result in zero accessibility in many census tracts, making it difficu lt to reveal disparities. Catchments larger than 30 min will over smooth the accessibility, thus concealing the variation in accessibility (Dai, pg. 237, 2011). Therefore, the following accessibility assessment would choose the 10 min driving time as thre shold 4.2.1 Accessibility of Public Parks Using the 2SFCA method, Figure 4 4 shows the accessibility of all public parks. Darker colors indicate higher accessibility of public parks within the census tract. The eastern area and some areas in downtown Gai nesville have the highest accessibility levels. The northern and southern regions have the lowest accessibility. When public park s are not classified, its accessibility may not have practical significance. People may spend more time in specific passive par ks such as Paynes Prairie Preserve State Park for camping, fishing, or horseback riding. Active parks may be more important for people's daily lives because people can easily play basketball, walk, and enjoy other leisure activities in neighborhood parks o r nearby recreational facilities. Therefore, public parks should be classified into active parks and passive parks, and then accessibility of each should be considered 4.2.2 Accessibility of Active Parks Using the 2SFCA method based on the 10 min catchmen t, Figure 4 5 presents spatial accessibility of active parks such as neighborhood parks and recreational facilities. Darker colors indicate higher accessibility within the census tract. Accessibility
50 of active parks is highest within the downtown and weste rn areas of Gainesville, followed by the area around downtown Gainesville, followed by the eastern and southwestern regions of Gainesville. The northern, southern, and western regions of Gainesville have the lowest accessibility levels 4.2.3 Accessibility of Passive Parks Using the 2SFCA method based on the 10 min catchment, Figure 4 6 presents spatial accessibility of passive parks such as natural parks and conservation areas. Darker colors indicate higher accessibility of passive parks within the census tract. The spatial accessibility of passive parks is similar to the accessibility of all public parks. The eastern area and some other areas within Gainesville have the highest accessibility levels. The northern, southern, and western regions all have the lowest accessibility.
51 Table 4 1 Service area ratio of public parks within different walking distance thresholds Buffer (walking distance) Area ratio (total) 400m 19.14% 800m 25.62% 1200m 52.95%
52 Figure 4 1 Service area ratio of Public parks within 400m buffer Figure 4 2 Service area ratio of Public parks within 800m buffer
53 Figure 4 3 Service area ratio of Public parks within 1200m bu ffer Figure 4 4 Total accessibility of public parks in Alachua County.
54 Figure 4 5 S patial accessibility of a ctive parks in Alachua County Figure 4 6 S patial access ibility of p assive parks in Alachua County
55 CHAPTER 5 EQUITY ANALYSIS 5.1 Spatial D istribution of V ulnerable G roups To optimize the layout of public parks, it is imperative to indentify priority area s that need to be optimized. According to the definition o f society equity in this paper, the accessibility of public parks should be skewed toward the disadvantaged groups. Therefore, some census tracts with low park accessibility but a high proportion of vulnerable groups should be set to priority areas for opt imization. In some research es of environmental justice, indicators such as income, living conditions, and education level are used to reflect the socio economic situation of residents. People with social and economic disadvantages often face low income ( e .g., no vehicles available and below the poverty line), low level of education and lack of knowledge and skills required for employment, poor living conditions ( e.g., rental housing) and other disadvantages. The low access to recreation facilities for some lower income groups would likely result in decreased physical activities and increased obesity (Gordon Larsen, 2006). The study by Maas et al. pointed out that in large cities, low educated populations seem ed to be able to acquire more benefits from the s upply of public parks (Maas et al., 2006). At the same time, some social justice seminars believe d that youth, the elderly and no W hites are potentially vulnerable groups. The elderly is more vulnerable to health problems due to their declining body functi ons. Children with better access to parks tend to participate in more physical activity which would reduce the risk of obesity (Sanders et al., 2015; Wolch et al., 2011). The elderly living in communities with better access to parks tend to participate in more outdoor activities and are less prone to
56 physical and mental illness (Bertram & Rehdanz, 2015; Gong et al., 2014). Boone et al. found that W hites c ould enjoy more public parks within walking distances (Boone et al., frican Americans and Asian Pacific islanders have W hite dominated areas in LA (Wolch et al., 2005). Therefore, the major vulnerable groups in the equity analysis are youth, the elderly, non whites, low educated residents, and low income residents represented by the following seven variables: percent under age 18 percent over age 64 percentage of non White s p ercent below nin th grade p ercentage of carless housing units percent of rental housing units and percent below the poverty line Housing tenure as well as the vehicle ownership are used as an indicator for low income residents People who are considered to need high accessibility of public parks were non Whites, low income residents (presumably those who rent their hou se and ha ve no available vehicle s and those below the poverty line), youth and the elderly, and low educated residents 5.2 Socioeconomic Factors Regarding High Access Versus Low Access C ensus T racts C onsidering the combination of accessibility and equity can be achieved using Mann Whitney U test, a two sample statistical test, which involves comparing the socioeconomic characteristics of those residents who are considered to have high accessibility, with those for whom accessibility is low (Talen, 1997; Ni cholls, 2001; Xiao d the socioeconomic characteristics of two groups Nicholls (2001) carried out the equity analysis in Bryan using t he Mann Whitney U test.
57 Whitney U test. Therefore, the equity analysis can figure out whether the spatial layout of public parks is equitable or not across vulnerable groups using Mann Whitney U test. The rationality of u sing the nonparametric Mann Whitney U test as research method is based on the independence of two samples, equal variance and lack of normal distribution. The Mann Whitney U test is utilized to make a comparison between the median value of certain socioeconomic variable in census blocks with high accessibility and that in census blocks with low accessibility to see if park a ccessibility favor s a particular socio economic group or not. The measuring formula is listed below : ( 5 1) Where, n 1 and n 2 U U is standard deviation of U. U 1 and U 2 are calculated value for sample1 and sample2. R 1 and R 2 are sum of adjusted rankings for sample1 and sample2. To calculate the test separate groups. Both groups are ranked together but ordered in separate groups. This test can be used with interval or ratio data when the sample data are not normally distributed. The test automatically converts the interval or ratio data to ordinal ranks and s of both samples are the same, which means they are not significantly different.
58 The equity analysis between the socioeconomic characteristics of vulnerable groups and the accessibility of public parks is assessed by the Mann Whitney U test. For example, the test can detect whether the census tract with high aging level enjoys better access to public parks than that with low aging level. Setting the average of park accessibility value of all census tracts in the study area as the dividing line, all census tract was divided into high accessibility groups and low accessibility groups, and then the Mann Whitney U test i s used to identify differences in the socioeconomic characteristics of vulnerable population between the high access census tracts and low acce ss census tracts. Table 5 1 and Table 5 3 contain results of the equity analysis using Mann Whitney U test in SPSS. Figure 5 1 describes how I can understand the results of the Mann Whitney U test. First, when a significant difference between two groups, o ne with high park accessibility and the other one with low park accessibility was found (p<0.05), it was essential to compare the median value of certain socioeconomic variable with high accessibility to that with low accessibility to see if park accessibi lity favor ed a particular socio economic group or not, that was equity or inequity. When there wa s no significant difference (p>0.05), results were interpreted as implying that those disadvantaged groups had equal opportunities compared to the rest of the community. Therefore, equality rather than demand based equity is obvious. Because of the equity definition adopted here, an equitable distribution was put forward when the proportion of youth or the elderly, the proportion of minorities, or the proportion of low income residents is significantly higher within the area with high park accessibility than those with low accessibility.
59 5.2.1 Socioeconomic Factors Regarding High Access Versus Low Access to Active Parks As shown in Table 5 1 and Table 5 2, when considering percentage of carless households as an indicator, I found a significant difference between two groups, one with high park accessibility and the other one with low park accessibility (p < 0.05). The median value of that variable within the high access area was compared to the same value within the low access area; I determined that census tracts with larger percentages of carless housing units are more likely to have higher accessibility of active parks. Except for carless housing units, I saw no significant difference among Therefore, the spatial distribution of active parks takes into account the low income groups represented by carless households; to some extent, this reflects so cial equity because carless households have better access to active parks. However, accessibility of active parks does not favor youth, elderly, non W hite, low educated, or low income populations. Compared to general populations, these vulnerable groups do not enjoy greater access to active parks, which does not reflect social equity 5.2.2 Socioeconomic Factors Regarding High Access Versus Low Access to Passive Parks As shown in Table 5 3 and Table 5 4, when considering population under age 18 and below ni nth grade as an indicator, I found a significant difference between two groups, one with high park accessibility and the other with low park accessibility (p < 0.05). The median value of these two variables within high access areas was compared to the valu e within low access areas; I determined that census tracts with larger percentages of population under age 18 and below ninth grade are more likely to have lower accessibility of passive parks. Because these two populations show a higher level
60 of inequity regarding accessibility of passive parks, I must pay greater attention to younger and less educated social groups. Except in the case of youth and less educated populations, the results of equity analysis based on accessibility of passive parks show no si gnificant difference among social groups. Such results also reveal that accessibility of passive parks is not in favor of any vulnerable groups. The spatial distribution of passive parks in Alachua County appears to show equality rather than need based equ ity. One possible reason that passive parks may not be significantly in favor of any that is, the open space aspect and the preservation of natural habitat. The spatial distri bution of such passive parks cannot be controlled or changed by human factors. cannot take the distribution of social groups or vulnerable groups into consideration. Ther efore, Table 5 1 to Table 5 4 show that comparing socioeconomic characteristics of vulnerable groups with high or low accessibility of active parks or passive parks reveals very similar results. The spatial distribution of active parks and passive parks in Alachua County is biased toward equality rather than demand based equity. Regardless of whether the active park or the passive park is taken into consideration, individuals who are elderly, minorities, below the poverty line, or living in rental housing u nits are not particularly well served by public products: urban parks. Compared to the rest of the community, the disadvantaged groups receive equal opportunities. However, this reveals equality and inequity rather than demand based equity.
61 5.3 Spatial Clu stering Pattern of V ulnerable G roups to P ublic P ark s Since the Mann Whitney U test can only evaluate whether the vulnerable groups s between the distribution of pu blic parks and vulnerable groups. It is impossible to locate the unreasonable part of the spatial layout of public parks just through the Mann Whitney U test. Therefore, this paper draws on the method of Li et al (2015), using the Bivariate LISA (Local Ind icators of Spatial Association) to identify the spatial correlation between the distribution of vulnerable groups and the distribution of public parks LISA (Local indicators of spatial association) is utilized to figure out whether spatial clusters of biv ariate variables, accessibility and related socioeconomic characteristics is statistically significant or not (Anselin, 1995). The measuring formula is shown as follows: ( 5 2 ) Where, and is the deviation from average. is spatial weight. A uthentically t he advantage of LISA would be to detect the important spatial patterns of association at each location (Anselin, 1995). Only a few cases propose d a spatial correlation model of accessibility and socio economic characteristics (Talen & Anselin, 1998). Therefore, they proposed LISA wa s the most suitable method for the research goal (Anselin, 199 5). T he statistically spatial association pattern s of accessibility and socioeconomic characteristics can be established with the help of a software tool, Geoda. To reflect the degree of relevance, a LISA Cluster Map was constructed using accessibility and related demographic characteristic s as variables.
62 The main advantage of the LISA method is the ability to detect significant correlation patterns in local regions, including hotspots and outliers. Therefore, this paper uses the Bivariate LISA to analyze t he spatial correlation pattern s between vulnerable groups youth (represented by percent under age 18), the elderly (represented by percent over age 64), low income population (represented by percent below the poverty line), low educat ed population (represe nted by percent below ninth grade) and minorities (represented by percentage of non White s ) and accessibility of active parks or passive parks. The Bivariate LISA Cluster m ap s shown from Figu re 5 2 to Figure 5 1 1 respectively display the spatial cor relatio n between the six socioeconomic indicato rs above and the accessibility of active parks or passive parks through the Bivariate LISA enable dividing the study area into four sub regions corresponding to the clusters and high and low low locations (positive local spatial autocorrelation) represent spatial clusters, while the high low and low high locations (negative local spatial autocorrelation) represent p g. 10, 2013). The high high sub region (census tract in red) indicates the census tract have high accessibility of active parks or passive parks which surrounded by areas with a high percentage of vulnerable groups. The l ow low sub region (census tract in dark blue) indicates the census tract have low accessibility of active parks or passive parks surrounded by areas with a low proportion of vulnerable groups. These two kinds of spatial clusters (high high sub region s and low low sub region s ) all represent that the proportion of vulnerable groups in certain census tract and corresponding public park accessibility are at a
63 relatively consistent level, whether they are higher or lower. However, the high low sub region (cens us tract in pink) indicates the census tract have high accessibility of active parks or passive parks surrounded by areas with a low proportion of vulnerable groups. The low high sub region (census tract in light blue) indicates the census tract have low a ccessibility of active parks or passive parks surrounded by areas with a high proportion of vulnerable groups. Unlike the high high and low low spatial clusters, the spatial outliers (high low sub region s and low high sub region s ) show that the proportion of vulnerable groups in the census tract is inconsistent with public park accessibility. Therefore, in order to achieve social equity, that is, to make vulnerable groups enjoy more public resources, greater attention should be paid to the low high sub re gion s the census tract s with relatively high proportion of vulnerable people and relatively low public park accessibility. Meanwhile, only the statistically significant census tract s ha ve brightly painted shadows and most of the census tract s within the n ot s category are indicated by light gray 5.3.1 The Bivariate LISA maps in Figure 5 2 to Figure 5 6 show the spatial clustering patterns between vulnerable groups youth, the elderly, non W hite s low educated residents, and low income residents and accessibility of active parks. Figure 5 2 shows the spatial clustering pattern between youth and the accessibility of active parks. The figure also reveals that two census tracts wi th high percentages of youth and high accessibility of active parks are located in the southwestern and eastern regions of Gainesville. Above all, attention should focus on the low high subregions census tracts in light blue. The census tracts have a relat ively high proportion of youth and a relatively low accessibility of active parks. Census tracts
64 with low accessibility of active parks but a high proportion of the youth population are located in the western region of Alachua County; such tracts need to p rovide a greater number of active parks to achieve social equity for adolescents. Figure 5 3 shows the spatial clustering pattern between the elderly and accessibility of active parks. This LISA map demonstrates that a majority of the census tracts the light gray sections indicate no significant relationship between the elderly population and accessibility of active parks. The map also reveals that the urban planning in Alachua County did not consider proximity to urban parks as a necessary living standa rd for elderly groups. Only two census tracts shaded in light blue in the northeast of Gainesville have a high presence of elderly groups and low accessibility of active parks. In order to achieve social equity for the elderly, these two tracts must priori tize improvements in accessibility levels. Figure 5 4 shows the spatial clustering pattern between population below the poverty line and accessibility of active parks. About eight census tracts shaded in light blue are located in southwestern areas of Gai nesville city. These regions have a high percentage of low income residents and low accessibility of active parks; therefore, the region is lacking in parks and dominated by a high percentage of population below the poverty line. In order to achieve social equity for low income groups, the tracts in light blue must prioritize improvements in accessibility of active parks. Additionally, many of the census tracts in the western region of Alachua County represent the low low subregion census tracts in dark blu e. The region indicates that census tracts with low accessibility of active parks are surrounded by low proportions of low income groups.
65 Figure 5 5 shows the spatial clustering pattern between population with low education level and accessibility of acti ve parks. Only two census tracts shaded in light blue have a high presence of less educated groups and low accessibility of active parks; the areas are located in the northern region of Gainesville and the northeastern region of Alachua County, respectivel y. In order to achieve social equity for the less educated, the two tracts must prioritize improvements in accessibility of active parks. Figure 5 6 shows the spatial clustering pattern between the non W hite social group and accessibility of active parks. According to this LISA map, most of the census tracts in bright shades are the high high subregions census tracts in red and the low low sub regions census tracts in dark blue. These two kinds of spatial clusters high high and low low subregions indicate that in certain censuses, the relationships between non W hite populations and accessibility of active parks remain at relatively consistent levels, whether high or low. Table 5 5 summarizes the census tract findings and shows where social equity is lacking Census tracts with a high proportion of vulnerable groups but low accessibility should be the priority areas for optimization of park layout. For active parks, the census tracts with UAID numbers 449409, 449437, 449440, 449406, 449442, 449439, 448614, 44 9407, 449405, 448525, 448574, 448573, 448524, 448572, 448570, 448569, 448523, 448528, and 449331 all have a high proportion of certain vulnerable groups but low accessibility. Among them, the census tracts with UAID 449437, 449440, 449442, and 449439 shoul d receive our special attention because they demonstrate low accessibility of both active parks and passive parks as well as a high percentage of youth population. Additionally, the census tract with UAID 449407 has a high percentage of
66 the elderly populat ion but low accessibility of active parks and passive parks. The census tract with UAID 448525 has a high percentage of the low educated population but low accessibility of active parks and passive parks. The census tract with UAID 449331 has a high percen tage of minorities but low accessibility of active parks and passive parks. Therefore, urban planning should pay special attention to the demands of vulnerable groups living in these census tracts. 5.3.2 to Passive Parks The Bivariate LISA maps presented from Figure 5 7 to Figure 5 11 show the spatial clustering pattern between vulnerable groups youth, the elderly, non W hites, low educated residents, and low income residents and accessibility of passive pa rks. Figure 5 7 shows the spatial clustering pattern between youth and accessibility of passive parks. Many of the census tracts in the western area of Alachua County and one in the eastern area of Gainesville are low high subregions; these tracts are in light blue. They indicate that such tracts have low accessibility of passive parks surrounded by areas with a high proportion of youth. In order to achieve social equity for adolescents, tracts with a high proportion of youth but low accessibility of passi ve parks need to provide more passive parks. Figure 5 8 shows the spatial clustering pattern between the elderly and accessibility of passive parks. The LISA map indicates that most of the census tracts colored in light gray show no significant relation ship between the elderly and accessibility of passive parks; this is the same result that is seen in the elderly parks in Alachua County is not considered a necessary living standard for the elderly.
67 The connection between the spatial distribution of the elderly and the accessibility of passive parks is not very significant. Figure 5 9 shows the spatial clustering pattern between population below the poverty line and accessibility of passive parks. About seven census tracts shaded in light blue in the southern area of Gainesville show a high percentage of low income residents and low accessibility of passive parks; these areas also house a high percentage of population below the poverty line and few nearby passive parks. In order to achieve social equity for low income groups, the tracts in light blue must prioritize improvements for accessibility of passive parks. Figure 5 10 shows the spatial clustering pattern betwe en population with low education level and accessibility of passive parks. Only two census tracts shaded in light blue show a high presence of the less educated and the low accessibility of passive parks; these areas are located on the eastern side of Alac hua County. In order to achieve social equity for the less educated, the two tracts must prioritize improvements for accessibility of passive parks. I also found that almost all the census tracts are shaded in light gray, indicating no significant relation ship between less educated social groups and accessibility of passive parks. Figure 5 11 shows the spatial clustering pattern between non W hites and accessibility of passive parks. About four census tracts shaded in light blue are located on the eastern s ide of Gainesville city and have a high percentage of minorities and low accessibility of passive parks. The areas are dominated by a high percentage of non W hites and have few nearby passive parks. In order to achieve social equity for
68 minorities, the tra cts in light blue must prioritize improvements for accessibility of passive parks. Table 5 5 summarizes the findings by census tract to show where social equity is lacking. Census tracts with a high proportion of vulnerable groups but low accessibility sho uld be the priority areas for optimization. For passive parks, the census tracts with UAID numbered 449331, 449411, 449438, 448616, 449332, 448618, 449330, 449435, 449328, 449329, 449437, 449440, 449442, 449439, 449407, and 448525 all have a high proportio n of certain vulnerable groups but low accessibility. The census tract with UAID 449330 should receive special attention because it has not only a high proportion of youth, minority, and low educated populations but also low accessibility of passive parks.
69 Table 5 1 Comparison of Socioeconomic Factor s of Vulnerable Groups Regarding H igh Access Versus L ow Access to A ctive parks V ariable Median Value of Variable High Access (22) Low Access (34) Mann Whitney U p value P ercenta ge of non White s 3 1.2 2 2.90 2 99.50 0 .211 P ercent over age 64 8 .88 1 2.72 3 26.00 0 .421 P ercent under age 18 1 6.3 2 0.22 2 95.00 0 .185 Percentage of carless housing units 1 0.16 5 .32 2 04.00 0 .004 P ercent of rental housing units 5 5.78 2 9.56 2 98.00 0 .202 P erc ent below the poverty line 2 5.365 1 5.36 2 99.00 0 .208 Percent below ninth grade 2 .015 1 .69 3 60.50 0 .821 Table 5 2 Equity A nalysis on Active Parks Variable Indicates Equity or Inequity? P ercent age of non White s Inequity P er cent over age 64 Inequity P ercent under age 18 Inequity Percent age of carless housing units E quity P ercent of rental housing units Inequity P ercent below the poverty line Inequity Percent below ninth grade Inequity
70 Table 5 3 Comparison of Socioeconomic Factors of Vulnerable Groups Regarding High Access Versus Low Access to Passive parks V ariable Median Value of Variable High Access (2 3 ) Low Access (3 3 ) Mann Whitney U p value P ercentage of non White s 28 23.7 359.50 0 .739 P ercent over age 64 8.92 13.56 320.00 0.322 P ercent under age 18 16.62 20.3 259.00 0.045 Percentage of carless housing units 7.54 5.57 300.00 0.185 P ercent of rental housing units 54.8 26.86 295.00 0.159 P ercent below the poverty line 25.73 16.29 318.00 0.306 Percent below ninth grade 1.27 2.59 244.50 0.024 Table 5 4 Equity A nalysis on Passive Parks Variable Indicates Equity or Inequity? P ercent age of non White s Inequity P ercent over age 64 Inequity P ercent under age 18 Inequity Percent age of carless housing units Inequity P ercent of rental housing units Inequity P ercent below the poverty line Inequity Percent below ninth grade Inequity
71 Table 5 5 Summary table about Equity Anal ysis by census tract with L ow park accessibility and H igh proportion of vulnerable groups C ensus Tract UAID Access Active Parks Access Passive Parks Ratio of youth Ratio of the elderly Ratio of the low income Ratio of the low educated Ratio of the minori ties 449409 L ow H igh 449437 L ow L ow H igh 449440 L ow L ow H igh 449406 L ow H igh 449442 L ow L ow H igh 449439 L ow L ow H igh 448614 L ow H igh 449407 L ow L ow H igh 449405 L ow H igh 448525 L ow L ow H igh 448574 L ow H igh 448573 L ow H igh 448524 L ow H igh 448572 L ow H igh 448570 L ow H igh 448569 L ow H igh 448523 L ow H igh 448528 L ow H igh 449331 L ow L ow H igh 449411 L ow H igh 449438 L ow H igh 448616 L ow H igh 4 49332 L ow H igh 448618 L ow H igh 449330 L ow H igh H igh H igh 449435 L ow H igh 449328 L ow H igh 449329 L ow H igh
72 Figure 5 1 Guide to understanding the results of the Mann Whitney U test
73 Figure 5 2 Bivariate LISA Cluster Map of access to active parks and percent under age 18 Figure 5 3 Bivariate LISA Cluster Map of access to active parks and percent over age 64
74 Figure 5 4 Bivariate LISA Cluster Map of access to active parks and percent below the poverty line Figure 5 5 Bivariate LISA Cluster Map of access to active parks and percent below ninth grade
75 Figure 5 6 Bivariate LISA Cluster Map of access to active parks and percentage of non White s Figure 5 7 Bivariate LISA Cluster Map of access to passive parks and percent under age 18
76 Figure 5 8 Bivariate LISA Cluster Map of access to passive parks and percent over age 64 Figure 5 9 Bivariate LISA Cluster Map of access to passive parks and percent below the poverty line
77 Figure 5 10 Bivariate LISA Cluster Map of access to passive parks and percent below ninth grade Figure 5 11 Bivariate LISA Cluster Map of access to passive parks and percentage of non White s
78 CHAPTER 6 CONCLUSION Valuing equity and u niversal accessibility of public parks and recreational facilities has many benefits, including promotion of community cohesion; improvement in mental and physical health; reduction in crime; and growth of economic profit. Therefore, people are increasingl y worried about whether the unequal distribution of social groups will influence their access to public parks. People are also concerned about whether allocation of public parks takes social equity into consideration that is, whether the provision of publi c parks is skewed toward vulnerable groups. To examine the issue, this paper assessed accessibility of public parks in Alachua County and discussed the equity of public parks by exploring socioeconomic disparities in high access and low access census trac ts. Finally, by using the Bivariate LISA method to analyze the spatial correlation of various vulnerable groups and their accessibility of public parks, the shortage areas with low accessibility and a high percentage of vulnerable groups were identified an d can be prioritized for improvement. First, when considering park accessibility, this paper classified public parks based on their characteristics and use, thereby making subsequent equity assessments more realistic and operational. Due to the natural pro perties of passive parks and their lack of alteration by humans, it is more effective to identify and consider development and improvement to areas with low accessibility of active parks so that their distribution can be consistent with the spatial distrib ution of vulnerable groups. Second, this paper used the 2SFCA method to assess spatial accessibility of active parks and passive parks. Each census tract can receive an accessibility score based on the number of active parks or passive parks that 1,000 res idents can
79 potentially access. Using 2SFCA to calculate park accessibility has many advantages such as considering the interaction between the park and the population and using the road network to calculate travel resistance rather than Euclidean distance. Accessibility of passive parks in the southern, northern, and western areas of Alachua County is low. Additionally, accessibility of active parks in the southern and northern areas of Alachua County and the western area of Gainesville is low. Therefore, a lthough Alachua County has a very rich range of parks and recreational facilities, its distribution is not balanced relative to the demands of the population. Third, this paper selected youth, the elderly, non W hites, low educated residents, and low income residents as vulnerable groups; the paper then used the Mann Whitney U test to evaluate the ethnic and socioeconomic differences between high access and low access areas. Except for census tracts with high proportions of carless households, which have hig her accessibility of active parks, socially disadvantaged communities do not have better park accessibility. Therefore, the degree of social equity is very low. Less advantaged groups especially youth, the elderly, non W hites, low educated residents, and l ow income residents do not tend to have better access to active parks or passive parks; this indicates the absence of demand based equity. Fourth, the Bivariate LISA method was used to analyze the spatial clustering pattern between distributions of vulner able groups and active parks or passive parks. In order to achieve social equity that is, to allow vulnerable groups to enjoy more public resources urban planners should pay more attention to the low high subregions shown in light blue on the LISA maps and to the census tracts with high proportions of
80 vulnerable people but low accessibility of public parks. The results can help urban planners identify shortage areas and ensure equity of accessibility. Therefore, low accessibility of public parks in Alachua County is widening the equity gap. The service supply of public parks has not been skewed toward vulnerable groups such as youth, the elderly, non W hites, low educated residents, and low income residents. Urban planning should pay special attention to the demands of vulnerable groups. Planning methods that focus on social equity should ensure that vulnerable residents have better access to public parks. When setting up new parks or expanding existing ones, urban planners should focus on low accessibility ar eas and prioritize socioeconomically disadvantaged areas.
81 LIST OF REFERENCES Abercrombie, L. C., Sallis, J. F., Conway, T. L., Frank, L. D., Saelens, B. E., & Chapman, J. E. (2008). Income and racial disparities in access to public parks and private recrea tion facilities. American journal of preventive medicine 34 (1), 9 15. Allen, W. B., Liu, D., & Singer, S. (1993). Acce s sibility measures of US metropolitan areas. Transportation Research Part B: Methodological 27 (6), 439 449. Anselin, L. (1995). Local in dicators of spatial association LISA. Geographical analysis 27 (2), 93 115. Apparicio, P., Abdelmajid, M., Riva, M., & Shearmur, R. (2008). Comparing alternative approaches to measuring the geographical accessibility of urban health services: Distance type s and aggregation error issues. International journal of health geographics 7 (1), 7. Apparicio, P., Sguin, A. M., Landry, S., & Gagnon, M. (2012). Spatial distribution of vegetation in Montreal: an uneven distribution or environmental inequity? Landscape and urban planning 107 (3), 214 224. Bertram, C., & Rehdanz, K. (2015). The role of urban green space for human well being. Ecological Economics 120 139 152. Boone, C. G., Buckley, G. L., Grove, J. M., & Sister, C. (2009). Parks and people: An environme ntal justice inquiry in Baltimore, Maryland. Annals of the Association of American Geographers 99 (4), 767 787. Bowen, K. J., & Parry, M. A. R. I. S. S. A. (2015). The evidence base for linkages between green infrastructure, public health and economic bene fit. project Assessing the Economic Value of Green Infrastructure Byrne, J., & Wolch, J. (2009). Nature, race, and parks: past research and future directions for geographic research. Progress in Human Geography 33 (6), 743 765. Byrne, J., Wolch, J., & Zha ng, J. (2009). Planning for environmental justice in an urban national park. Journal of Environmental Planning and Management 52 (3), 365 392. Cervigni, F., Suzuki, Y., Ishii, T., & Hata, A. (2008). Spatial accessibility to pediatric services. Journal of C ommunity Health 33 (6), 444 448. Comber, A., Brunsdon, C., Hardy, J., & Radburn, R. (2009). Using a GIS Based network analysis and optimi z ation routines to evaluate service provision: A case study of the UK post office. Applied Spatial Analysis and Policy 2 (1), 47 64.
82 Cordell, H. K., Green, G. T., & Betz, C. J. (2002). Recreation and the environment as cultural dimensions in contemporary American society. Leisure sciences 24 (1), 13 41. C ounty of A lameda E den Area General Plan (n.d.). P arks and Recreation Element Retrieved September 4, 2018, from https://www.acgov.org/cda/planning/generalplans/documents/05_ParksandRec.p df Crompton, J. L. (2001). The impact of park s on property values: A review of the empirical evidence. Journal of leisure research 33 (1), 1. Cutts, B. B., Darby, K. J., Boone, C. G., & Brewis, A. (2009). City structure, obesity, and environmental justice: an integrated analysis of physical and socia l barriers to walkable streets and park access. Social science & medicine 69 (9), 1314 1322. Dai, D. (2011). Racial/ethnic and socioeconomic disparities in urban green space accessibility: Where to intervene? Landscape and Urban Planning 102 (4), 234 244. Dalvi, M. Q., & Martin, K. M. (1976). The measurement of accessibility: some preliminary results. Transportation 5 (1), 17 42. Davies, B. (1970). Social needs and resources in local services. Public Administration 48 (2), 214 215. Delmelle, E. C., & Casas, I. (2012). Evaluating the spatial equity of bus rapid transit based accessibility patterns in a developing country: The case of Cali, Colombia. Transport Policy 20 36 46. Dominski, T. (1992). The three stage evolution of eco cities: Reduce, reuse, recyc le. Sustainable cities 16 18. Dunworth, T., Hayeslip, D., Lyons, M., & Denver, M. (2010). Evaluation of the Los Angeles gang reduction and youth development program: final Y1 report. Evaluation Erkip, F. B. (1997). The distribution of urban public servic es: the case of parks and recreational services in Ankara. Cities 14 (6), 353 361. ESRI. (n.d.). What is the ArcGIS Network Analyst extension? Retrieved September 5, 2018, from http://desktop.arcgis.com/en/arcmap/latest/extensions/network analyst/what is network analyst .htm Estabrooks, P. A., Lee, R. E., & Gyurcsik, N. C. (2003). Resources for physical activity participation: does availability and accessibility differ by neighborhood socioeconomic status? Annals of behavioral medicine 25 (2), 100 104.
83 Fausold, C. J., & Lilieholm, R. J. (1999). The economic value of open space: A review and synthesis. Environmental Management 23 (3), 307 320. Fo rs, H., Molin, J. F., Murphy, M. A., & van den Bosch, C. K. (2015). User participation in urban green spaces For the people or the parks? Urban Forestry & Urban Greening 14 (3), 722 734. Geertman, S. C., & Ritsema Van Eck, J. R. (1995). GIS and models of a ccessibility potential: an application in planning. International journal of geographical information systems 9 (1), 67 80. Gong, Y., Gallacher, J., Palmer, S., & Fone, D. (2014). Neighborhood green space, physical function and participation in physical ac tivities among elderly men: the Caerphilly Prospective study. International Journal of Behavioral Nutrition and Physical Activity 11 (1), 40. Gordon Larsen, P., Nelson, M. C., Page, P., & Popkin, B. M. (2006). Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics 117 (2), 417 424. Gu, Mingdong & Yi, Haiwei. (2010). Review on spatial accessibility and equity of public facilities. Urban Problems 5 27 31. Hansen, W. G. (1959). How accessibility sha pes land use. Journal of the American Institute of planners 25 (2), 73 76. Harnik, P., & Crompton, J. L. (2014). Measuring the total economic value of a park system to a community. Managing Leisure 19 (3), 188 211. Harnik, P., & Simms, J. (2004). Parks: Ho w far is too far. Planning 70 (11), 8 11. Harnik, P., & Welle, B. J. (2009). Measuring the economic value of a city park system Trust for Public Land. Harris, C. D., Paul, P., Zhang, X., & Fulton, J. E. (2015). Park access among school age youth in the Un ited States. Journal of Physical Activity and Health 12 (6 Suppl 1), S94 S101. Hartley, T. W. (1995). Environmental justice: An environmental civil rights value acceptable to all world views. Environmental Ethics 17 (3), 277 289. Harvey, D. (1973). Social Justice and the City. Johns Hopkins Studies in Urban Affairs. Center for metropolitan Planning and Research, The Johns Hopkins University, The Johns Hopkins University Press, Baltimore, Maryland, USA (s 30).
84 Humpel, N., Owen, N., & Leslie, E. (2002). Envi ronmental factors associated with American journal of preventive medicine 22 (3), 188 199. Is Palisades Park "Passive" or "Active"? (n.d.). Retrieved September 4, 2018, from https://www.friendsofpalisadespark.net/is palisades park active or passive/ Jacobs, J. (1961). The death and life of great American cities. Vintage, New York Jobe, R. T., & White, P. S. (2009 ). A new cost distance model for human accessibility and an evaluation of accessibility bias in permanent vegetation plots in Great Smoky Mountains National Park, USA. Journal of vegetation science 20 (6), 1099 1109. Kabisch, N., & Haase, D. (2014). Green justice or just green? Provision of urban green spaces in Berlin, Germany. Landscape and Urban Planning 122 129 139. Kessel, A., Green, J., Pinder, R., Wilkinson, P., Grundy, C., & Lachowycz, K. (2009). Multidisciplinary research in public health: a case study of research on access to green space. Public health 123 (1), 32 38. Khotdee, M., Singhirunnusorn, W., & Sahachaisaeree, N. (2012). Effects of green open space on social health and behavior of urban residents: A case study of communities in Bangkok. Procedia Social and Behavioral Sciences 36 449 455. Kinman, E. L. (1999). Evaluating health service equity at a primary care clinic in Chilimarca, Bolivia. Social Science & Medicine 49 (5), 663 678. Kresl, P. K. (2013). Empirical approaches to urban comp etitiveness analysis. P. Kresl y J. Sobrino (coords.), Handbook of Research Methods and Applications in Urban Economies, Cheltenham, Edward Elgar 32 59. Li, H., Wang, Q., Shi, W., Deng, Z., & Wang, H. (2015). Residential clustering and spatial access to p ublic services in Shanghai. Habitat International 46 119 129. Lindsey, G., Maraj, M., & Kuan, S. (2001). Access, equity, and urban greenways: An exploratory investigation. The Professional Geographer 53 (3), 332 346. Lineberry, R. L. (1977). Equity and u rban policy. Beverly Hills: Sage Publications Lotfi, S., & Koohsari, M. J. (2009). Measuring objective accessibility to neighborhood facilities in the city (A case study: Zone 6 in Tehran, Iran). Cities 26 (3), 133 140. Luo, W., & Qi, Y. (2009). An enhanc ed two step floating catchment area (E2SFCA) method for measuring spatial accessibility to primary care physicians. Health & place 15 (4), 1100 1107.
85 Luo, W., & Wang, F. (2003). Measures of spatial accessibility to health care in a GIS environment: synthes is and a case study in the Chicago region. Environment and Planning B: Planning and Design 30 (6), 865 884. Ma, J. (2016). Green Infrastructure and Urban Livability: Measuring Accessibility and Equity (Doctoral dissertation, ResearchSpace@ Auckland). Ma, L B., & Cao, X. S. (2006). A GIS based evaluation method for accessibility of urban public green landscape. Acta Scientiarum Naturalium Universitatis Sunyatseni 45 (6), 111 115. Maas, J., Verheij, R. A., Groenewegen, P. P., De Vries, S., & Spreeuwenberg, P (2006). Green space, urbanity, and health: how strong is the relation? Journal of Epidemiology & Community Health 60 (7), 587 592. accessibility. Transportation Research Part B: Methodological 30 (1), 47 79. LISA to explore the crash risky locations in urban areas. In Proceedings of the Conference of Network Association of European Researchers on Urbani z ation in the South, Enschede, The Netherlands (pp. 1 12). McConnachie, M. M., & Shackleton, C. M. (2010). Public green spac e inequality in small towns in South Africa. Habitat International 34 (2), 244 248. McCormack, G. R., Rock, M., Toohey, A. M., & Hignell, D. (2010). Characteristics of urban parks associated with park use and physical activity: A review of qualitative rese arch. Health & place 16 (4), 712 726. McGrail, M. R., & Humphreys, J. S. (2009). Measuring spatial accessibility to primary care in rural areas: improving the effectiveness of the two step floating catchment area method. Applied Geography 29 (4), 533 541. Mladenka, K. R., & Hill, K. Q. (1977). The distribution of benefits in an urban environment: Parks and libraries in Houston. Urban Affairs Quarterly 13 (1), 73 94. Morris, J. M., Dumble, P. L., & Wigan, M. R. (1979). Accessibility indicators for transport planning. Transportation Research Part A: General 13 (2), 91 109. Mowen, A. (2010). Parks, playgrounds and active living. A Research Synthesis. Princeton, NJ: Active Living Research, a National Program of the Robert Wood Johnson Foundation National Associ ation of Counties. (n.d.). Alachua County, Florida. Retrieved September 3, 2018, from
86 https://web.archive.org/web/20110531210815/http://www.naco.o rg/Counties/Pag es/FindACounty.aspx Neutens, T. (2015). Accessibility, equity and health care: review and research directions for transport geographers. Journal of Transport Geography 43 14 27. Nicholls, S. (2001). Measuring the accessibility and equity of public parks: A case study using GIS. Managing leisure 6 (4), 201 219. Nicholls, S., & Shafer, C. S. (2001). Measuring Accessibility and Equity in a Local Park System: The Utility of Geospatial Technologies to Park and Recreation Professionals. Journal of Park & Recreation Administration 19 (4). Ogryczak, W. (2000). Inequality measures and equitable approaches to location problems. European Journal of Operational Research 122 (2), 374 391. Omer, I. (2006). Evaluating accessibility using house level data: A spatial equity perspective. Computers, Environment and Urban Systems 30 (3), 254 274. Ostrowski, W. (1975). Potestio, M. L., Patel, A. B., Powell, C. D., McNeil, D. A., Jacobson, R. D., & McLaren, L. (2009). Is there an association between spatial access to parks/green space and childhood overweight/obesity in Calgary, Canada? Internatio nal Journal of Behavioral Nutrition and Physical Activity 6 (1), 77. Powell, L. M., Slater, S., & Chaloupka, F. J. (2004). The relationship between community physical activity settings and race, ethnicity and socioeconomic status. Evidence Based Preventive Medicine 1 (2), 135 44. Reyes, M., Pez, A., & Morency, C. (2014). Walking accessibility to urban parks by children: a case study of Montreal. Landscape and Urban Planning 125 38 47. Rich, R. C. (1979). Neglected issues in the study of urban service dis tributions: a research agenda. Urban Studies 16 (2), 143 156. Richardson, E., Pearce, J., Mitchell, R., Day, P., & Kingham, S. (2010). The association between green space and cause specific mortality in urban New Zealand: an ecological analysis of green sp ace utility. BMC public health 10 (1), 240. Roseland, M. (1997). Dimensions of the eco city. Cities 14 (4), 197 202. Sallis, J. F., & Kerr, J. (2007). Physical activity and the built environment. Sallis, J. F., Cervero, R. B., Ascher, W., Henderson, K. A., Kraft, M. K., & Kerr, J. (2006). An ecological approach to creating active living communities. Annu. Rev. Public Health 27 297 322.
87 Sanders, T., Feng, X., Fahey, P. P., Lonsdale, C., & Astell Burt, T. (2015). Greener neighborhoods, slimmer children? Evi dence from 4423 participants aged 6 to 13 years in the longitudinal study of Australian children. International Journal of Obesity 39 (8), 1224. Shen, Q. (1998). Location characteristics of inner city neighborhoods and employment accessibility of low wage workers. Environment and planning B: Planning and Design 25 (3), 345 365. Sister, C., Wolch, J., & Wilson, J. (2010). Got green? Addressing environmental justice in park provision. GeoJournal 75 (3), 229 248. Smith, D. M. (1994). Geography and social justi ce. Talen, E. (1997). The social equity of urban service distribution: An exploration of park access in Pueblo, Colorado, and Macon, Georgia. Urban geography 18 (6), 521 541. Talen, E. (1998). Visualizing fairness: Equity maps for planners. Journal of the American Planning Association 64 (1), 22 38. Talen, E., & Anselin, L. (1998). Assessing spatial equity: an evaluation of measures of accessibility to public playgrounds. Environment and planning A 30 (4), 595 613. Thompson, C. W. (2002). Urban open space i n the 21st century. Landscape and urban planning 60 (2), 59 72. Tsou, K. W., Hung, Y. T., & Chang, Y. L. (2005). An accessibility based integrated measure of relative spatial equity in urban public facilities. Cities 22 (6), 424 435. United States Environmental Protection Agency. (n.d.). Environmental Justice. Retrieved September 3, 2018, from https://www.epa.gov/environmentaljustice/environmen tal justice your community US Census Bureau. (n.d.). Alachua County, Florida. Retrieved September 3, 2018, from https://factfinder.census.gov/faces/tablese rvices/jsf/pages/productview.xhtml?src =CF US Census Bureau. (n.d.). Alachua County, Florida. Retrieved September 3, 2018, from https://www.census.gov/quickfacts/fac t/table/alachuacountyflorida/PST045217 Understanding Racial Inequity in Alachua County. (2018, January 12). Retrieved from https://www.bebr.ufl.edu/economics/racial inequity Van Herzele, A ., & Wiedemann, T. (2003). A monitoring tool for the provision of accessible and attractive urban green spaces. Landscape and urban planning 63 (2), 109 126.
88 Von Thnen, J. H. (1966). Isolated state Pergamon Press. Webber, M. M. (1963). Order in diversit y: community without propinquity. Cities and space: The future use of urban land 23 54. Wicks, B. E., & Crompton, J. L. (1986). Citizen and administrator perspectives of equity in the delivery of park services. Leisure Sciences 8 (4), 341 365. Willemse, L (2013). A Flowmap geographic information systems approach to determine community neighborhood park proximity in Cape Town. South African Geographical Journal 95 (2), 149 164. Wolch, J. R., Byrne, J., & Newell, J. P. (2014). Urban green space, public heal th, and Landscape and Urban Planning 125 234 244. Wolch, J., Jerrett, M., Reynolds, K., McConnell, R., Chang, R., Dahmann, N., ... & Berhane, K. (2011). Childhood obesity and prox imity to urban parks and recreational resources: a longitudinal cohort study. Health & place 17 (1), 207 214. Wolch, J., Wilson, J. P., & Fehrenbach, J. (2005). Parks and park funding in Los Angeles: An equity mapping analysis. Urban geography 26 (1), 4 35 Wolf, K. L. (2003). Ergonomics of the city: Green infrastructure and social benefits. In Engineering Green: Proceedings of the 11th National Urban Forest Conference. Washington DC: American Forests (Vol. 115). Xiao, Y., Wang, Z., Li, Z., & Tang, Z. (2017 ). An assessment of urban park access in Shanghai Implications for the social equity in urban China. Landscape and Urban Planning 157 383 393. Yao, L., Liu, J., Wang, R., Yin, K., & Han, B. (2014). Effective green equivalent A measure of public green spa ces for cities. Ecological indicators 47 123 127.
89 BIOGRAPHICAL SKETCH Kexin Chen was born in Guangzhou, l ocated in southern China She always wants to explore the world more, so she earned her Bachelor of Science degree from the Wuhan University in Resou rce Management and Urban and Rural Planning and a graduate minor in International Economics in 2016 After college graduation Kexin came to the United States and graduated again from the University of Florida with the degree of Urban and Regional Planning, concentrating on spatial analysis urban redevelopment, and GIS application. Her dream originally was to study economic geography and got a chance to work for enterprise location selection Through her two years of graduate study she also develo ped great attention to environmental justice, giving her the inspiration to focus on social justice issues for her thesis topi c Kexin is a dedicated traveler and ha s traveled to major cities and national parks in the United States and some important coun tries in South America during these two years abroad explore the unknown world in the future