Next Article in Journal
An Eye-Tracking Study on Text Comprehension While Listening to Music: Preliminary Results
Previous Article in Journal
Comparable Analysis of Natural and Modified Starches from Kazakhstan: Physicochemical Properties, Applications, and Insights on Biodegradable Films
Previous Article in Special Issue
Design of Urban Indicators to Optimize the Implementation of Low-Impact Techniques in Semi-Arid Cities
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 15
Issue 7
10.3390/app15073940
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Urban Green Space Inequity, Socioeconomic Disparities, and Potential Health Implications in Metropolitan Melbourne

by
Parian Hoseini
1,*,
Pooriya Mohseni
2,
Piret Veeroja
3 and
Greg Foliente
4,5
1
School of Built Environment, University of New South Wales, Kensington 2033, Australia
2
School of Civil & Environmental Engineering, University of New South Wales, Kensington 2033, Australia
3
Centre for Urban Transition, Swinburne University of Technology, Hawthorn 3122, Australia
4
Department of Infrastructure Engineering, The University of Melbourne, Parkville 3010, Australia
5
Advanced Research Institute for Informatics, Computing and Networking (AdRIC), College of Computer Studies, De La Salle University, Manila 1004, Philippines
*
Author to whom correspondence should be addressed.
Appl. Sci. 2025, 15(7), 3940; https://doi.org/10.3390/app15073940
Submission received: 7 February 2025 / Revised: 7 March 2025 / Accepted: 19 March 2025 / Published: 3 April 2025
(This article belongs to the Special Issue Sustainable Urban Green Infrastructure and Its Effects)
Browse Figures
Versions Notes

Abstract

:
Urban green spaces (UGSs) provide city residents contact with nature, offering mental and physical health benefits. However, residents’ access to green spaces in cities can be associated with their socioeconomic status (SESs). This study utilizes the Kernel Density tool as an innovative method to measure UGS inequities and their relationship with cardiovascular disease (CVD) rates. Next, the UGS patterns and their potential implications for CVD are examined across suburbs with a range of SES levels in Melbourne, Australia. The proposed method is tested in conjunction with two commonly used measures of accessibility (Network Analysis) and provision (UGS per capita). The results show that more advantaged suburbs have better access to UGS and lower CVD rates. Moreover, the analysis reveals that a more geographically dispersed UGS pattern, predominantly observed in higher SES suburbs, can be associated with lower CVD than a concentrated pattern, and the integration of the SES and UGS indicators through Kernel Density analysis reveals that inequitable access to green spaces disproportionately impacts the health incomes of socioeconomically disadvantaged communities. Finally, the Kernel Density and Network Analysis tools in ArcGIS can serve as effective supplementary methods for addressing similar considerations in UGS planning and policy.

1. Introduction

With rapid urbanization, it has been estimated that around 68 percent of the population will be living in cities by 2050. The concerns around human health issues regarding urban settings has shaped a large body of research. However, these attempts are yet to provide measurable standards for healthy cities’ policy targets [1], and more research is required to understand the full extent of the impacts of urban settings on various health dimensions [2]. Particularly, the influence of urban green space (UGS) on individuals’ overall well-being and quality of life has been highlighted by numerous studies [3,4,5,6]. Research has shown a stronger association of UGS with improved physical health metrics, while the mental health impacts remain less clear [7]. Similar to other urban amenities, UGS can be distributed inequitably and in favor of communities with higher socioeconomic status (SES) [8]. Similar to other OECD countries, Australian cities have arguably failed to provide equitable access to health-supportive amenities and infrastructure [9].
Numerous studies have shown a positive relationship between UGS, physical health, and well-being in general [5,10,11,12,13,14]. Alongside personal and behavioral factors, environmental factors significantly contribute to decreased cardiovascular disease (CVD) rates [15,16,17,18]. Research has found that living in areas with higher and more variable UGS forms can decrease the CVD hospitalization rate by 37% [19]. An extensive study on over 100,000 Australians shows that living within 1.6 km of areas with at least 30% tree canopy significantly reduces the risk of CVD [20]. A longitudinal population-based cohort study in Canada showed that a 1.2% increase in Normalized Difference Vegetation Index (NDVI) exposure is associated with a 3% reduction in stroke risk [20]. These studies suggest a strong theoretical basis for a positive association between higher vegetation density and lower CVD rates, but the type of vegetation (if it is a large natural reserve that is accessible to the public or a sidewalk tree line or a gated golf course) and exposure is not clear.
A systematic review of the UGS and CVD shows that a considerable body of literature is mainly focused on measures like the UGS area per inhabitant, or UGS area/count/percent within a certain buffer zone or urban area [21]. The findings have confirmed the positive relationship but have not further suggested practical tools for urban planning and UGS policy making purposes. Despite being essential for the primary stages of identifying potential associations, from an urban planning view, these approaches may disregard the spatial nature of UGS. For example, by drawing unrealistic boundaries like buffer zones or even official urban boundaries, some significant aspects can be neglected in the study, like accessibility through an existing network, and the impacts of neighboring UGSs.
From an environmental justice perspective, UGS should be equitably distributed throughout a city [5], but even in developed countries, UGS availability is limited or distributed unevenly [8,22]. The uneven UGS disparity can be associated with several factors, namely central versus suburban city areas. For instance, a comparative analysis shows that every kilometer farther from the city center contributes an average increase of 73 and 67 m to trips to the most frequently used UGS in Luxembourg and Brussels, respectively [23]. Similarly, Kabisch and Haase [24] revealed that although the majority of sub-districts in Berlin meet the 6 m2 per person UGS threshold, inner-city areas with high immigrant rates and population density face disproportionate UGS shortages, confirming distributive injustice.
Therefore, another factor associated with UGS distribution can be the socioeconomic status (SES) of the area, meaning that socioeconomically disadvantaged communities may be less exposed to UGS in their living environment than more affluent communities [25,26]. For example, Silva et al. [5] argue that in Tartu, Estonia, ethnic minority areas have less UGS m2 per individual and lower access rates to UGS compared to local or more affluent residents, which could make UGS a subject of environmental justice [4], as well as a health issue. Upon this theoretical basis, a dual exploration on the socioeconomic inequalities and access to green space that could both contribute to improved health could be beneficial for understanding the public health aspect of planning policies. In Australia, CVD is one of the leading causes of death, affecting more than 4 million people equivalent to nearly 17% of the population [27], and costs the Australian economy 5 billion dollars annually, which is the largest share of health expenditure compared to other diseases [28]. Moreover, several studies have indicated an unbalanced UGS availability relating to SES in Australian cities [22,29,30,31,32]. Additionally, research has shown that the residents of lower SES areas in Australia are more likely to experience higher morbidity and mortality rates for chronic diseases such as CVD [9]. Therefore, it is strongly recommended that planning bodies take affirmative action to address socioeconomic inequities in the supplying and distribution of UGS, what is referred to as UGS availability in this paper.
This study aims to explore potential associations between UGS availability, SES, and CVD in Metropolitan Melbourne, Australia, through an urban planning lens. In this context, UGS policy and urban planning can act as a public health intervention [14,17,18]. But for this to be possible, the assessment tools need to be more explicit about what type(s) of exposure to UGS can bring the most benefit to residents, and more importantly, how an equitable UGS availability can be measured. Building on the existing literature, this research aims to measure UGS availability with respect to its spatial nature. We examine three indicators, UGS provision, accessibility, and density, with three different methods. Exploring the relation between each of these three indicators with SES and CVD rates in Melbourne, we discuss the most reliable approach and provide suggestions for improving UGS equitability based on the proposed approaches.

2. Methods

2.1. Study Area

This study is carried out in the Metropolitan Melbourne area, and the unit of analysis is Statistical Area Level 2 (SA2), n = 294, within the 31 Local Government Areas (LGAs) that comprise the metropolitan area, obtained from the Australian Census of Population and Housing [33]. According to these data, there were 303 SA2s in Metropolitan Melbourne in 2016. The SA2 level units are “functional areas that represent a community that interacts together socially and economically, and often align with the suburb and locality boundaries” [34]. These areas are about 7.5 km2 on average, with a population of about 14,700. In this study, 9 SA2s were excluded due to data unavailability.

2.2. Urban Green Space Data

UGSs can include “urban parks and wetlands that comprise some vegetation” ([35], p. 32). Based on this definition, the UGS spatial data are obtained from the Victorian Planning Authority [36]. The map layer contains data on all public open spaces in Melbourne. As Taylor and Hochuli [35] suggest, it is essential to determine a proper definition of UGS that aligns with the research questions and analysis methods. It should be noted that not all open spaces can be considered as UGSs. Private backyards or gardens, for example, do not fit within the “public” nature of UGSs. Therefore, a set of criteria is formulated to eliminate those green spaces that lack one or more of the requirements, so the remaining green space features are known as UGSs and used as inputs for this research. The criteria for the UGS area are as follows:
  • Have vegetation or green space area to provide a chance for contact with nature;
  • Open for public use without charge or membership;
  • Devoted to, or amenable to, recreational purposes.
By intersecting this study’s criteria with the open space data attributes and their definitions [37], four types of green spaces in the Melbourne Metropolitan area are identified as eligible, as follows:
  • Conservation reserves;
  • Natural and semi-natural open space;
  • Parks and gardens;
  • Sports fields.
One of the major challenges in measuring UGS inequities is the selection of indicators [38,39,40], so the research results can go beyond theoretical and descriptive levels to some practical and measurable outputs, that can be used for future UGS policymaking and planning. Regarding the aim of this research to demonstrate a more elaborated methodology, and based on priority and relevance to the study objectives and availability of data, three indicators are used herein to evaluate UGS availability in the Melbourne Metropolitan SA2 areas, as follows:
  • UGS density;
  • Accessibility;
  • Provision.

2.3. UGS Density

Unlike the other two indicators, UGS density using Kernel Density Estimation is rarely used as a tool in UGS inequality studies. The Kernel Density tool in ArcMap 10.5 allows to select polygons (the four UGS types in this case) and to calculate the density based on the selected features. This tool “Calculates a magnitude-per-unit area from point features that fall within a neighborhood around each cell” [41]. In other words, the Kernel Density tool calculates the amount or density of UGS, in this case, that is distributed over the study area. The tool also applies differential weights to objects based on their proximity [42], meaning that areas with a higher density of UGS, whether of a large size or higher quantity, would get a higher density score. After the Kernel Density calculation, a unique density value can be assigned to each SA2 using the Zonal Statistic tool in ArcGIS 10.5. The combination of these two tools helps to address the issue of unrealistic boundaries, like buffer zones or urban compartments. For example, if an SA2 boundary is cutting a UGS polygon into two or more parts, it cannot affect the overall density value of SA2s, which is more aligned with reality.
The Kernel Density tool uses point data as the input, so the polygon UGS layer should convert to point data first. For this method, it is critical to consider the wide range of UGS sizes within the Melbourne Metropolitan area while choosing the method for converting the polygon features to points. For instance, there are very large UGSs in the northeast of the study area, like natural reserves, in comparison to the ones in the compact inner-city areas. Therefore, any direct analysis of the UGS data and neglecting the significant difference between the UGS sizes would mislead the results. To address this issue, a mesh layer with 400 * 400 cells was built using the ArcMap 10.5 fishnet tool. The fishnet layer was intersected with UGS polygons using the Feature to Point tool, and a point was drawn at the center of each output feature. A graphical sample of the procedure is represented in Figure 1.
The output point layer is the input layer for the Kernel Density analysis, and the result is represented in Figure 2.
To identify a quantitative output for UGS density in each SA2, a unique density value is assigned to each SA2 using the Zonal Statistics tool, that computes a single output value for every SA2 based on the Kernel raster, and the result for Melbourne SA2s is represented in Figure 3.

Accessibility and Provision

Various accessibility and provision indicators are used in different UGS studies ([5,39,42,43]). The most common approaches are buffer zones or Euclidean distance from the point of residence, or from UGS areas [21]. Network Analysis has been used in some studies [44]; however, parameters can vary like the distance, type of destination (UGS), travel mode, etc. In this study, UGS accessibility is evaluated using road network distance in ArcGIS 10.5. The Network Analysis calculates the real travelling distance using the existing routes to a certain destination [42], while other mentioned methods analyze accessibility based on a direct line distance that does not represent the actual distance that a person should take towards, either on foot or drive towards, the destination.
For this purpose, green space polygon features are obtained from a Melbourne open space map [37] and the road line features from the Victorian Government open data website [45]. As suggested by the VPP guideline [46], areas with 400 m, equal to 5 min walking distance, from green space areas are determined using the Network Analysis tool, in ArcGIS 10.5.
In such accessibility analysis, given the relatively small distance size (400 m) and the UGS polygons, not points, as the reference, it is crucial to determine destinations correctly. In other words, access should be measured from the boundary of the UGS, not the centroid, which can be a critique of some studies using this tool for measuring distance. This issue is addressed using the ‘feature vertices to point’ tool to draw points on the UGS boundaries, as represented in Figure 4.
The Network Analysis is conducted using Melbourne’s Road network layer and the UGS boundaries’ point layer. The result illustrated areas that are within the 400 m distance from the boundary of at least one UGS. Therefore, each SA2 is divided into two portions, one of which is within 400 m UGS access, and the other one is out of 400 m UGS access. Figure 5 demonstrates an example of the output.
Then, the accessibility ratio is given by the following:
( A c c e s s i b i l i t y   r a t i o ) i = ∑ ( 400   m   s e r v i c e   a r e a ) i ( S A 2   a r e a ) i
The last indicator used is UGS provision, which is assessed by per capita number, indicating the sum of green space per person in each SA2. The green space per capita is commonly given by the following:
G r e e n   S p a c e   P e r   C a p i t a = ∑ U G S i ( S A 2   P o p u l a t i o n ) i

2.4. SES Data

The Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD) from the Socioeconomic Index for Areas (SEIFAs) is the reference for the SES indicator. This index was developed by the Australian Bureau of Statistics in 2016, based on the five-year census data [47]. The IRSAD score is determined based on multiple measures including education, income, and employment status to find relative socioeconomic advantages and disadvantages in SA2 level boundaries. Higher scores represent more socioeconomically advantaged areas, and lower scores show the opposite. Based on the IRSAD score, the SA2s are categorized into five distinct SES classifications (quintile), from most disadvantaged to most advantaged. Figure 6 shows the spatial distribution of socioeconomically different areas based on IRSAD quintiles in Metropolitan Melbourne.

2.5. Health Data

The proportion of people with CVD (including heart- and blood vessel-related diseases) at the SA2 level is used as the health indicator. The last available health data are provided by the National Health Survey in 2018, demonstrating key findings in long-term health conditions and mental and physical risk factors [27]. It is important to highlight that the SES data and CVD data have been collected 2 years apart. The former was based on the regular (and comprehensive) national census survey, while the latter is from a special-purpose survey. We assumed herein that the 2-year temporal mismatch will not have a significant difference and influence on the results.

2.6. Data Analysis

In the first part of the analysis, SA2s are divided based on their UGS accessibility and UGS density into groups based on the median value of each indicator. This was followed by a series of Chi-square analyses to test whether the difference in the proportion of these indicators in different IRSAD quintiles is statistically significant and whether there is an association between the variables. Next, we examined if there is a meaningful change in the proportion of people with CVD in areas with different UGS accessibility, density, and provision. For this, a series of one-way analyses of variances (ANOVAs) helped us to find the difference.
In the second part of the analysis, different SA2s are grouped based on their UGS accessibility and density, called UGS availability. The groups are as follows:
  • High Access—High Density Areas;
  • High Access—Low Density Areas;
  • Low Access—High Density Areas;
  • Low Access—Low Density Areas.
A set of Welch ANOVA tests helped us to reveal if there is a statistically significant difference in mean IRSAD Scores and the proportion of people having CVD between these groups.

3. Results

Table 1 presents the descriptive statistics of different measures used in the analysis. There are 294 different SA2s in this analysis. The mean IRSAD score is 1028, ranging from 819 to 1160 in different SA2s. Within the Melbourne Metropolitan area, different SA2s have a UGS per capita ranging from 0.4 m2 to 29,036 m2 per person with a mean and median of 427 and 55.14 square meters per person, respectively. On average, 4.43% of the total population suffers from CVD.
The test results suggest associations between UGS accessibility and density with the SES of areas; however, no association between the UGS per capita of areas and their SES is detected, as demonstrated in Table 2, and Figure 6 shows its spatial disparities.
As Figure 7 and Figure 8 suggest, the majority (59% and 61%, respectively) of quintiles 4 and 5 SA2s have high access to green space, whereas, in the other 3 lower quintiles, the majority of SA2s have low access to green space. We see the same pattern for green space density, wherein in the bottom 3 IRSAD quintiles, the majority of SA2s have low UGS density.
Next, a series of one-way ANOVA showed a statistically significant change in the percentage of people having CVD in areas of different UGS access and different density separately F (1, 292) = 3.983, p = 0.047 and F (1, 292) = 7.462, p = 0.007. Areas with lower UGS accessibility have a higher proportion of CVD among residents. The same trend was observed with the UGS density indicator as well, represented in Table 3 and Table 4.
As discussed previously, different SA2s are grouped based on their UGS accessibility and density, named UGS availability in this paper. A set of Welch ANOVA tests revealed that there is a statistically significant difference in mean IRSAD scores and the proportion of people having heart conditions between these groups (F (3, 111.465) = 9.231, p < 0.001 and F (3, 126.767) = 5.135, p = 0.002). Table 5 shows the mean value of IRSAD in these different UGS availabilities. Areas with high access—high density of UGS have the highest SES among other areas. Also, Table 6 shows the average CVD rate in these groups, suggesting that areas with high access—high density of UGS have fewer people with CVD.

4. Discussion

This study examined accessibility, density, and provision indicators to assess UGS availability in SA2s with different SES levels, from the most disadvantaged to the most advantaged quintiles. The findings suggest that in the Melbourne Metropolitan area, SA2s with higher UGS density and better access to UGS, defined as being within a five-minute walking distance, have a lower proportion of people with CVD. Furthermore, the areas with higher SES tend to have UGS privileges in terms of density and accessibility. This pattern is unlike SA2s with lower SES levels, where the UGS inequities could have aggravated the resident’s physical health, in addition to all the disadvantages associated with the lower social and economic conditions. In contrast to the accessibility measure utilized in this analysis, the UGS per capita has not demonstrated a meaningful difference among areas with different SES levels. Indeed, personal characteristics are influential factors for CVD, but living environment factors play a role too [48]. Our findings show a correlation between higher UGS availability and decreased CVD rates, similar to the existing studies ([15,17,48,49]). Our findings are specifically consistent with the study of Pereira et al. [19], which used the NDVI method and a 1600 m service area. By focusing on four specific types of UGS and excluding smaller tree canopies and bushes that may lack public use or access, this study provides a more targeted analysis of the features of UGSs that could influence CVD rates.
The study found that the per capita measure is not associated with socioeconomic disparity and, therefore, might not be the most appropriate area-level measure to include in similar studies. This can also be the case for other area-related measures like the UGS percent ratio to the total study unit boundary, that is broadly used in similar studies [21]. However, the findings in other contexts might not be similar to this study, but the Melbourne case shows that a more cautious approach should be taken regarding UGS studies. Besides the statistical analysis results, a spatial viewpoint can highlight the underlying reasons. The per capita number in a particular area (e.g., SA2 boundary) does not capture sufficient attributes about UGS availability within the area. As represented in Figure 9, there might be other UGS areas next to the defined SA2 border that are unrealistically excluded from the analysis. The defined border does not limit people’s exposure and access to UGS areas in the neighboring SA2s.
Unlike per capita, the results about UGS accessibility and density correlate with the SES in Melbourne. SA2s with higher SES (quintiles 4 and 5) have significantly better access to UGS areas and have higher UGS density. The more disadvantaged SA2s (quintiles 1, 2, and 3) have much less access to UGS areas, while the UGS density remains almost unchanged here. This finding is aligned with Sharifi at al.’s [4] earlier study, utilizing the gravity model, which showed not only that UGS is more accessible in higher SES areas but also lower income Melbourne residents’ mobility is towards areas with lower UGS access. Furthermore, a recent study showed that in addition to UGS accessibility and density, areas with higher SES benefit from better UGS quality in terms of safety and maintenance [50].
Regarding the nature of UGS Kernel Density Estimation, which includes the size, count, and distribution of UGS areas, the two latter factors have played a more significant role in achieving a higher density rank than large-scale UGSs. While SA2s with higher SES showed a higher density score, the spatial UGS patterns in different areas become noteworthy. Consequently, as Figure 10 suggests, SA2s with numerous and dispersed UGSs achieve a higher density rank, which is the case for quintiles 4 and 5. In contrast, the lower quintiles have lower UGSs and less distribution accordingly. Therefore, existing large-scale UGSs do not result in a higher UGS density. This viewpoint again questions the reliability of per capita measures. For instance, the per capita number for Figure 10-2 might be higher than Figure 10-1, but it cannot address the inequity issue. However, it should be noted that although the findings reveal spatial patterns, a detailed spatial relationship analysis is beyond the scope of this research, as the current methodology primarily aimed at identifying general associations.
In terms of methodology, the Network Analysis tool provides the walking distance to certain destinations using the existing roads, as well as the distance from the UGS boundary rather than centroids. The analysis using this tool shows that higher SES SA2s in Melbourne have shorter access to UGSs. One reason behind the improved access score could be the existing road network. For instance, in the central SA2s, which are mainly among higher quintiles, there is a 400 m access to at least one UGS from almost every spot. It shows there are more access alternatives to green spaces, which is probably a result of the compact urban form in central Melbourne. This finding is particularly consistent with Giles-Corti et al.’s [9] study on livability in Australian cities, showing inner suburbs are more livable in various terms, namely walkability and open spaces. There has been a limitation to this approach due to the large-scale case study. The Network Analysis tool requires point features as the destination; therefore, we used the ‘feature to point’ tool to convert the UGS feature boundaries into points. The tool draws a point at each vertex, which demonstrates the precise shape of the feature for irregularly shaped UGS. So, if there is a straight line without a fraction, the tool draws a single point for it regardless of the length. This may cause some levels of inaccuracy in the analysis. Therefore, future studies can develop a method for more accurate accessibility assessment or a smaller scale case study.
On the other side of the analysis, assessments of the correlation between UGS availability measures and the CVD rates show a lower proportion of the affected population in areas with higher UGS density and accessibility. The CVD rates among people living in areas with higher-than-average UGS density and accessibility are minimal (about 4.1 percent of the population) compared to other areas. Furthermore, in areas with lower UGS accessibility, regardless of changes in density, this percentage reaches near 4.6. So, future studies may include more specific indicators like physical activity, hypertension, obesity, etc. [48], as contributors to CVD. Also, studies with more subjective approaches like surveys among residents in addition to UGS qualities, like existing facilities, green coverage, etc., in different SES areas can deepen the understanding of the impacts of UGS on decreasing CVD chances.
Beyond the limitations of the analytical method, it is important to acknowledge that 400 m access to UGS is just one of many factors that may contribute to variations in CVD risk. Additionally, certain areas in Melbourne, similar to any other city, may be predominantly populated by younger adults, potentially affecting the health indicators at the area level. These limitations highlight the need for a careful interpretation of findings in studies of this nature, as the results are not definitive.

5. Conclusions

Health outcomes are influenced by a combination of individual and environmental factors. This study aimed to explore the potential health implications of short-distance access to specific types of green spaces and examine patterns across socioeconomically diverse areas. It also sought to identify differences between areas of Melbourne with lower cardiovascular disease (CVD) prevalence and those with higher CVD rates, while assessing the applicability of a less commonly used analytical tool in this context.
We showed that Kernel Density and Network Analysis tools in ArcGIS could be more reliable approaches in UGS studies compared to the simpler and more common area per capita measure. The Kernel Density tool, which had not been used as broadly as NDVI or buffer zone assessments in UGS studies, provides a better understanding of the equitability of UGS policies because it captures the spatial characteristics of green spaces in built-up and typically complex urban areas.
In the case study application in this paper, the spatial distribution of UGS is associated with a lower CVD rate in Metropolitan Melbourne. Although area-related parameters (e.g., UGS area, ratio, per capita) are some of the more common indicators for UGS assessments, our findings showed that higher UGS density and accessibility demonstrate a more significant contribution to lower CVD levels than UGS size. Additionally, in terms of equity, we show a higher health benefit of UGS in higher SES urban areas in Melbourne. Despite having a lower total UGS area, statistical area SA2s with higher SES have better UGS density and accessibility as well as lower CVD rates.
Metropolitan Melbourne areas with the highest UGS density and accessibility have a more dispersed UGS pattern with a relatively higher UGS distributed evenly throughout the whole SA2. In contrast, areas with a spatially focused UGS pattern have a lower UGS density rank and a higher proportion of CVD among their residents. Although large-scale UGS brings advantages to cities, they might not be sufficient in terms of UGS equity and health benefits. The conventional UGS per capita indicator does not effectively capture the spatial characteristics of UGSs in cities. For instance, in smaller urban areas, like an SA2, a large and concentrated UGS pattern may result in a large UGS number (i.e., m2/person) per inhabitant on paper, but the distribution pattern may deprive a certain population of green spaces’ benefits that can be achieved by having easy access to more dispersed UGS. Therefore, it is crucial to consider the spatial nature of spaces and parks and choose more appropriate evaluation indicators and methods accordingly.
Finally, this study suggests that methods like those utilized in the research could be used for future UGS planning to improve UGS distribution, density, and accessibility. Furthermore, with more detailed neighborhood-level health data and more granular UGS typological classification—including relative compositions of natural and man-made park elements—there is a better potential to develop improved methods for understanding inequalities and health impacts associated with UGS spatial patterns

Author Contributions

Conceptualization, P.H., P.V. and G.F.; Methodology, P.H. and P.M.; Software, P.H. and P.M.; Validation, P.V.; Formal analysis, P.H. and P.M.; Resources, P.H.; Writing—original draft, P.H. and P.M.; Writing—review & editing, P.V. and G.F.; Visualization, P.H. and P.M.; Supervision, P.V. and G.F. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The data presented in this study are openly available in ABS website (https://data.aurin.org.au/dataset/au-govt-abs-seifa-ieo-aust-sa1-2011-sa1#:~:text=Socio-Economic%20Indexes%20for%20Areas%20%28SEIFA%29%20is%20a%20product,are%20based%20on%20information%20from%20the%20five-yearly%20Census, accessed on 18 March 2025).

Acknowledgments

ChatGPT was used in few occasions to edit certain sentences to improve readability and understanding.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Giles-Corti, B.; Saghapour, T.; Turrell, G.; Gunn, L.; Both, A.; Lowe, M.; Rozek, J.; Roberts, R.; Hooper, P.; Butt, A.; et al. Spatial and socioeconomic inequities in liveability in Australia’s 21 largest cities: Does city size matter? Health Place 2022, 78, 102899. [Google Scholar] [CrossRef]
  2. Leng, H.; Li, S.; Yan, S.; An, X. Exploring the relationship between green space in a neighbourhood and cardiovascular health in the winter city of China: A study using a health survey for harbin. Int. J. Environ. Res. Public Health 2020, 17, 513. [Google Scholar] [CrossRef]
  3. Browning, M.; Lee, K. Within what distance does “greenness” best predict physical health? A systematic review of articles with gis buffer analyses across the lifespan. Int. J. Environ. Res. Public Health 2017, 14, 675. [Google Scholar] [CrossRef]
  4. Schindler, M.; Le Texier, M.; Caruso, G. How far do people travel to use urban green space? A comparison of three European cities. Appl. Geogr. 2022, 141, 102673. [Google Scholar]
  5. Sullivan, W.C.; Kuo, F.E.; Depooter, S.F. The fruit of urban nature: Vital neighborhood spaces. Environ. Behav. 2004, 36, 678–700. [Google Scholar]
  6. Zhang, R.; Zhang, C.-Q.; Cheng, W.; Lai, P.C.; Schüz, B. The neighborhood socioeconomic inequalities in urban parks in a High-density City: An environmental justice perspective. Landsc. Urban Plan. 2021, 211, 104099. [Google Scholar] [CrossRef]
  7. Richardson, E.; Pearce, J.; Mitchell, R.; Kingham, S. Role of physical activity in the relationship between urban green space and health. Public Health 2013, 127, 318–324. [Google Scholar] [CrossRef]
  8. Wood, L.; Hooper, P.; Foster, S.; Bull, F. Public green spaces and positive mental health—Investigating the relationship between access, quantity and types of parks and mental wellbeing. Health Place 2017, 48, 63–71. [Google Scholar] [CrossRef]
  9. Hoffimann, E.; Barros, H.; Ribeiro, A.I. Socioeconomic inequalities in green space quality and Accessibility—Evidence from a Southern European city. Int. J. Environ. Res. Public Health 2017, 14, 916. [Google Scholar] [CrossRef]
  10. Feng, X.; Astell-Burt, T.; Standl, M.; Flexeder, C.; Heinrich, J.; Markevych, I. Green space quality and adolescent mental health: Do personality traits matter? Environ. Res. 2022, 206, 112591. [Google Scholar] [CrossRef]
  11. Ghanem, A.; Edirisinghe, R. Socio-economic disparities in greenspace quality: Insights from the city of Melbourne. Smart Sustain. Built Environ. 2024, 13, 309–329. [Google Scholar]
  12. Selmi, W.; Weber, C.; Rivière, E.; Blond, N.; Mehdi, L.; Nowak, D. Air pollution removal by trees in public green spaces in Strasbourg city, France. Urban For. Urban Green. 2016, 17, 192–201. [Google Scholar]
  13. Taylor, L.; Hochuli, D.F. Defining greenspace: Multiple uses across multiple disciplines. Landsc. Urban Plan. 2017, 158, 25–38. [Google Scholar] [CrossRef]
  14. Yu, W.; Ai, T.; Shao, S. The analysis and delimitation of Central Business District using network kernel density estimation. J. Transp. Geogr. 2015, 45, 32–47. [Google Scholar] [CrossRef]
  15. Astell-Burt, T.; Feng, X. Urban green space, tree canopy and prevention of cardiometabolic diseases: A multilevel longitudinal study of 46 786 Australians. Int. J. Epidemiol. 2020, 49, 926–933. [Google Scholar] [CrossRef]
  16. Park, Y.M.; Kwan, M.-P. Multi-contextual segregation and environmental justice research: Toward fine-scale spatiotemporal approaches. Int. J. Environ. Res. Public Health 2017, 14, 1205. [Google Scholar] [CrossRef]
  17. Sharifi, F.; Nygaard, A.; Stone, W.M. Heterogeneity in the Subjective Well-being Impact of Access to Urban Green Space. Sustain. Cities Soc. 2021, 74, 103244. [Google Scholar] [CrossRef]
  18. Zhang, L.; Tan, P.Y.; Richards, D. Relative importance of quantitative and qualitative aspects of urban green spaces in promoting health. Landsc. Urban Plan. 2021, 213, 104131. [Google Scholar] [CrossRef]
  19. Raza, W.; Bojke, L.; Coventry, P.A.; Murphy, P.J.; Fulbright, H.; White, P.C. A Systematic Review of the Impact of Changes to Urban Green Spaces on Health and Education Outcomes, and a Critique of Their Applicability to Inform Economic Evaluation. Int. J. Environ. Res. Public Health 2024, 21, 1452. [Google Scholar] [CrossRef]
  20. Astell-Burt, T.; Navakatikyan, M.A.; Feng, X. Urban green space, tree canopy and 11 year risk of dementia in a cohort of 109,688 Australians. Environ. Int. 2020, 145, 106102. [Google Scholar]
  21. Maroko, A.R.; Maantay, J.A.; Sohler, N.L.; Grady, K.L.; Arno, P.S. The complexities of measuring access to parks and physical activity sites in New York City: A quantitative and qualitative approach. Int. J. Health Geogr. 2009, 8, 34. [Google Scholar] [CrossRef]
  22. Astell-Burt, T.; Feng, X.; Mavoa, S.; Badland, H.M.; Giles-Corti, B. Do low-income neighbourhoods have the least green space? A cross-sectional study of Australia’s most populous cities. BMC Public Health 2014, 14, 292. [Google Scholar] [CrossRef] [PubMed]
  23. Shuvo, F.K.; Feng, X.; Akaraci, S.; Astell-Burt, T. Urban green space and health in low and middle-income countries: A critical review. Urban For. Urban Green. 2020, 52, 126662. [Google Scholar] [CrossRef]
  24. Kabisch, N.; van den Bosch, M.; Lafortezza, R. The health benefits of nature-based solutions to urbanization challenges for children and the elderly—A systematic review. Environ. Res. 2017, 159, 362–373. [Google Scholar] [CrossRef] [PubMed]
  25. Kabisch, N.; Haase, D. Green justice or just green? Provision of urban green spaces in Berlin, Germany. Landsc. Urban Plan. 2014, 122, 129–139. [Google Scholar] [CrossRef]
  26. Pereira, G.; Foster, S.; Martin, K.; Christian, H.; Boruff, B.J.; Knuiman, M.; Giles-Corti, B. The association between neighborhood greenness and cardiovascular disease: An observational study. BMC Public Health 2012, 12, 466. [Google Scholar] [CrossRef] [PubMed]
  27. National Health Survey. Available online: https://www.abs.gov.au/statistics/health/health-conditions-and-risks/national-health-survey-first-results/2017-18 (accessed on 20 July 2021).
  28. Australian Health Expenditure—Demographics and Diseases: Hospital Admitted Patient Expenditure 2004-05 to 2012-13. Available online: https://www.aihw.gov.au/reports/health-welfare-expenditure/health-expenditure-australia-2017-18/contents/summary (accessed on 18 March 2025).
  29. Liu, X.X.; Ma, X.L.; Huang, W.Z.; Luo, Y.N.; He, C.J.; Zhong, X.M.; Dadvand, P.; Browning, M.H.; Li, L.; Zou, X.G.; et al. Green space and cardiovascular disease: A systematic review with meta-analysis. Environ. Pollut. 2022, 15, 118990. [Google Scholar] [CrossRef]
  30. Sandifer, P.A.; Sutton-Grier, A.E.; Ward, B.P. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst. Serv. 2015, 12, 1–15. [Google Scholar] [CrossRef]
  31. Sharifi, F.; Nygaard, A.; Stone, W.M.; Levin, I. Green gentrification or gentrified greening: Metropolitan Melbourne. Land Use Policy 2021, 108, 105577. [Google Scholar] [CrossRef]
  32. Wolch, J.R.; Byrne, J.; Newell, J.P. Urban green space, public health, and environmental justice: The challenge of making cities “just green enough”. Landsc. Urban Plan. 2014, 125, 234–244. [Google Scholar] [CrossRef]
  33. Australian Statistical Geography Standard (ASGS). Available online: https://www.abs.gov.au/ausstats/abs@.nsf/lookup/by subject/1270.0.55.001~july 2016~main features~statistical area level 2 (sa2)~10014 (accessed on 20 July 2021).
  34. The 2016 Census of Population and Housing. Available online: https://www.abs.gov.au/%0Aausstats/abs@.nsf/Lookup/by (accessed on 15 July 2021).
  35. 68% of the World Population Projected to Live in Urban Areas by 2050. Available online: https://www.un.org/development/desa/en/news/population/2018-revision-of-world-urbanization-prospects.html (accessed on 18 March 2025).
  36. Wang, D.; Brown, G.; Liu, Y. The physical and non-physical factors that influence perceived access to urban parks. Landsc. Urban Plan. 2015, 133, 53–66. [Google Scholar] [CrossRef]
  37. VPA Open Data. Available online: https://data-planvic.opendata.arcgis.com/ (accessed on 18 March 2025).
  38. Boulton, C.; Dedekorkut-Howes, A.; Byrne, J. Factors shaping urban greenspace provision: A systematic review of the literature. Landsc. Urban Plan. 2018, 178, 82–101. [Google Scholar] [CrossRef]
  39. Chen, Z.; Sarkar, A.; Rahman, A.; Li, X.; Xia, X. Exploring the drivers of green agricultural development (GAD) in China: A spatial association network structure approaches. Land Use Policy 2022, 112, 105827. [Google Scholar]
  40. Kaplan, R. The nature of the view from home: Psychological benefits. Environ. Behav. 2001, 33, 507–542. [Google Scholar] [CrossRef]
  41. Point Density (Spatial Analysis). Available online: https://www.esri.com/en-us/arcgis/products/arcgis-pro/resources (accessed on 10 November 2021).
  42. Mavoa, S.; Koohsari, M.J.; Badland, H.M.; Davern, M.; Feng, X.; Astell-Burt, T.; Giles-Corti, B. Area-Level Disparities of Public Open Space: A Geographic Information Systems Analysis in Metropolitan Melbourne. Urban Policy Res. 2014, 33, 306–323. [Google Scholar] [CrossRef]
  43. de Sousa Silva, C.; Viegas, I.; Panagopoulos, T.; Bell, S. Environmental justice in accessibility to green infrastructure in two European Cities. Land 2018, 7, 134. [Google Scholar] [CrossRef]
  44. De Vries, S.; Verheij, R.A.; Groenewegen, P.P.; Spreeuwenberg, P. Natural environments—Healthy environments? An exploratory analysis of the relationship between greenspace and health. Environ. Plan. A 2003, 35, 1717–1731. [Google Scholar] [CrossRef]
  45. Victoria Planning Authority, Metropolitan Open Space Network, Map Journal. Available online: https://planvic.maps.arcgis.com/apps/MapJournal/index.html?appid=8965d9701575408e8ea7f59772e82e10 (accessed on 21 August 2022).
  46. Mytton, O.T.; Townsend, N.; Rutter, H.; Foster, C. Green space and physical activity: An observational study using Health Survey for England data. Health Place 2012, 18, 1034–1041. [Google Scholar] [CrossRef]
  47. Australian Bureau of Statistics, Socio-Economic Indexes for Areas. Census of Population and Housing 2016. Available online: https://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/2033.0.55.0012016?OpenDocument (accessed on 20 July 2021).
  48. Lin, B.; Meyers, J.; Barnett, G. Understanding the potential loss and inequities of green space distribution with urban densification. Urban For. Urban Green. 2015, 14, 952–958. [Google Scholar] [CrossRef]
  49. Bilal, U.; Auchincloss, A.H.; Diez-Roux, A.V. Neighborhood Environments and Diabetes Risk and Control. Curr. Diabetes Rep. 2018, 18, 62. [Google Scholar] [CrossRef]
  50. Giles-Corti, B.; Moudon, A.V.; Lowe, M.; Cerin, E.; Boeing, G.; Frumkin, H.; Salvo, D.; Foster, S.; Kleeman, A.; Bekessy, S.; et al. What next? Expanding our view of city planning and global health, and implementing and monitoring evidence-informed policy. Lancet Glob. Health 2022, 10, e919–e926. [Google Scholar] [PubMed]
Applsci 15 03940 g001
Figure 1. UGS features’ preparation for Kernel Density analysis.
Figure 1. UGS features’ preparation for Kernel Density analysis.
Applsci 15 03940 g001
Applsci 15 03940 g002
Figure 2. UGS Kernel Density map.
Figure 2. UGS Kernel Density map.
Applsci 15 03940 g002
Applsci 15 03940 g003
Figure 3. UGS density level in SA2 areas.
Figure 3. UGS density level in SA2 areas.
Applsci 15 03940 g003
Applsci 15 03940 g004
Figure 4. UGS boundaries vertices to point for Network Analysis.
Figure 4. UGS boundaries vertices to point for Network Analysis.
Applsci 15 03940 g004
Applsci 15 03940 g005
Figure 5. UGS accessibility in Caulfield North.
Figure 5. UGS accessibility in Caulfield North.
Applsci 15 03940 g005
Applsci 15 03940 g006
Figure 6. IRSAD quintile by SA2s and UGS areas.
Figure 6. IRSAD quintile by SA2s and UGS areas.
Applsci 15 03940 g006
Applsci 15 03940 g007
Figure 7. UGS access in different quintiles.
Figure 7. UGS access in different quintiles.
Applsci 15 03940 g007
Applsci 15 03940 g008
Figure 8. UGS density in different quintiles.
Figure 8. UGS density in different quintiles.
Applsci 15 03940 g008
Applsci 15 03940 g009
Figure 9. (1–3) Per capita flaws.
Figure 9. (1–3) Per capita flaws.
Applsci 15 03940 g009
Applsci 15 03940 g010
Figure 10. UGS distribution inequities in SA2s with different SES.
Figure 10. UGS distribution inequities in SA2s with different SES.
Applsci 15 03940 g010
Table 1. Descriptive statistics of measures by statistical area 2 (SA2).
Table 1. Descriptive statistics of measures by statistical area 2 (SA2).
VariableNMeanMinMax
UGS:
   Area (Ha)2943166148.8786,417.26
   Per capita (m2)294427.280.4029,036.15
Diseases (% of Total Pop)2944.431.407.80
IRSAD Score2941028.62819.001160.00
Table 2. Results of Chi-square test.
Table 2. Results of Chi-square test.
VariableNDFX2p-Value *
UGS:
   Access294417.3020.002
   Density294411.1050.025
   Per capita29445.2100.266
* p-value < 0.05 statistically significant.
Table 3. CVD and UGS access.
Table 3. CVD and UGS access.
Variable:UGS Access
CVD (%)LowMean4.56
Std. Dev.1.17
Min2.20
Max7.80
HighMean4.30
Std. Dev.1.03
Min1.40
Max6.90
Table 4. CVD and UGS density.
Table 4. CVD and UGS density.
Variable:UGS Density
CVD (%)LowMean4.60
Std. Dev.1.13
Min2.20
Max7.80
HighMean4.26
Std. Dev.1.06
Min1.40
Max7.00
Table 5. Comparing SES in different UGS availability *.
Table 5. Comparing SES in different UGS availability *.
UGS AvailabilityNIRSAD MeanStd. Dev.MinMax
High Access—High Density1041054.2958.61920.001160.00
High Access—Low Density431021.1983.02862.001147.00
Low Access—High Density431010.1967.92846.001115.00
Low Access—Low Density1041013.6466.91819.001144.00
* Games–Howell Test for multiple comparisons found statistically significant differences between High Access—High Density Areas and Low Access—High Density Areas, p = 0.002, and Low Access—Low Density Areas, p < 0.001.
Table 6. Comparing proportion of CVD in different UGS availability *.
Table 6. Comparing proportion of CVD in different UGS availability *.
UGS AvailabilityNDisease MeanStd. Dev.MinMax
High Access—High Density1044.130.1051.406.90
High Access—Low Density434.710.1172.806.60
Low Access—High Density434.560.1482.607.00
Low Access—Low Density1044.430.1222.207.80
* Games–Howell Test for multiple comparisons found statistically significant differences between High Access—High Density Areas and High Access—Low Density Areas, p = 0.002, as well as with Low Access—Low Density Areas, p = 0.043.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Hoseini, P.; Mohseni, P.; Veeroja, P.; Foliente, G. Urban Green Space Inequity, Socioeconomic Disparities, and Potential Health Implications in Metropolitan Melbourne. Appl. Sci. 2025, 15, 3940. https://doi.org/10.3390/app15073940

AMA Style

Hoseini P, Mohseni P, Veeroja P, Foliente G. Urban Green Space Inequity, Socioeconomic Disparities, and Potential Health Implications in Metropolitan Melbourne. Applied Sciences. 2025; 15(7):3940. https://doi.org/10.3390/app15073940

Chicago/Turabian Style

Hoseini, Parian, Pooriya Mohseni, Piret Veeroja, and Greg Foliente. 2025. "Urban Green Space Inequity, Socioeconomic Disparities, and Potential Health Implications in Metropolitan Melbourne" Applied Sciences 15, no. 7: 3940. https://doi.org/10.3390/app15073940

APA Style

Hoseini, P., Mohseni, P., Veeroja, P., & Foliente, G. (2025). Urban Green Space Inequity, Socioeconomic Disparities, and Potential Health Implications in Metropolitan Melbourne. Applied Sciences, 15(7), 3940. https://doi.org/10.3390/app15073940

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop