This study examines the pedestrian accessibility of urban green spaces (UGS) in New York City, focusing on disparities across different boroughs and neighborhoods. Leveraging the H3 hexagonal grid, pedestrian network data, and GIS-based analyses, the research identifies significant gaps in UGS quality and quantity, particularly for the Brooklyn borough. The study also explores correlations between UGS accessibility, population demographics. Policy recommendations include the creation and expansion of UGS, improved pedestrian infrastructure, and relocation of residents. The findings underscore the need for equitable green space distribution to enhance public health and social equity.

Introduction

Urban green spaces (UGS) are critical for public health, environmental quality, and social equity in urban settings (NYC, 2011). In dense cities like New York, the distribution and accessibility of these spaces are vital for ensuring that all residents have equal access. Moreover, disparities in UGS accessibility can lead to inequitable outcomes, especially for vulnerable populations.

Previous research highlighted the importance of UGS in mitigating urban heat islands, improving air quality, and providing recreational opportunities. Studies by Roy et al. (2012) and Morpurgo et al. (2023) explored various definitions and classifications of urban greenery, emphasizing the need for accessible and high-quality green spaces in cities. However, less attention has been paid to the specific spatial disparities in UGS accessibility within New York City.

Within this framework, this study examines the accessibility of UGS in New York City, with a focus on identifying underserved areas and analyzing the relationship between access disparities, and population demographics.

Enabling Data and Methodology

This study’s methodology is structured into three main parts: Data Collection, Data Processing, and Network Construction and Analysis. All three parts focus on three categories of UGS: Forest Areas and Nature Reserves (FANR), Open Grass and Scrublands (OGS), and Recreational Green Spaces (RGS). These categories were defined to represent different types of green spaces based on their primary functions, respectively: (i) natural conservation areas, (ii) transitional green zones, and (iii) green spaces predominately designed for recreational use.

Data Collection

The data collection phase consisted in the acquisition of all the datasets required to perform the analysis. All the datasets used for developing the following analysis are open-source and were acquired from OpenStreetMap (OSM). A comprehensive collection of New York City’s green spaces, validated against official open data (NYC, n.d.) was collected using OSMnx in Python and QuickOSM in QGIS. The pedestrian network was retrieved from QuickOSM, and demographic data were sourced from the US Census.

Data Processing

The data processing phase was structured around the implementation of  Uber’s Hexagonal Hierarchical Spatial Index, at resolution level 9, as starting grid for all the following analyses. Specifically, the grid was clipped to the NYC administrative boundary, with centroid points generated from each hexagon cell and later used for the network analyses.

The input layers were cleaned and merged. Then, the final UGS dataset was produced by combining the three green space categories, with further refinements applied to handle duplicate geometries.

 Population data were integrated by intersecting grids with block-level data, allowing for the proportional allocation of population counts. This enabled the calculation of age group percentages within each grid, highlighting three demographic groups, namely: under 5 (<5), school-aged (5-17), and senior (65+) populations. All layers were projected to NAD 1983 StatePlane New York Long Island FIPS 3104 (meters).

Network Construction and Analysis

The last phase, network construction and analysis, involved building the pedestrian network using OSM data, used as starting point for the following processes, developed in ArcGIS Pro. Specifically, the accessibility was evaluated by calculating a service area from each grid centroid, representing a 10-minute walking distance. These accessibility sheds were used to estimate the intersection with UGS layers, and green space areas were aggregated.

Moreover, a Closest Facility Network Analysis determined travel times to the nearest UGS, which was later used to estimate clusters of high and low travel times across the city through the evaluation of the Getis-Ord Gi* statistic (Getis & Ord, 1992). The distance band for the Hot Spot Analysis was determined using Incremental Spatial Autocorrelation (ESRI, n.d.), a tool that finds the distance at which spatial patterns are most statistically significant, showing where clustering of high or low values is strongest. To reflect the travel in NYC’s pedestrian network, Manhattan distance was used, which measures travel along streets rather than straight-line distance, offering a more accurate representation of how people move through the city.

Results

The results of this study collectively provide a detailed examination of the spatial inequities in UGS accessibility across the city, highlighting the areas accessible within a 10-minute walk, and the distribution of accessible UGS by different categories. Additionally, the article presents a combined analysis showing the target populations counts in relation to UGS.

Figure 1 shows how much of total UGS areas can be reached within 10 minutes from the centroids of each grid after combining data on all three UGS categories, namely Forest Areas and Nature Reserves, Open Grass and Scrublands, Recreational Green Spaces,. The results show that nearly all areas in New York City are within a 10-minute walk of some form of UGS. However, Brooklyn and Queens have less UGS coverage than other boroughs.

Subsequently, Figure 2 presents the results of a more detailed classification, analyzing the accessible areas within a 10-minute walking distance from each grid’s centroid, categorized by OGS, RGS, and FANR. The findings indicate that most areas classified as OGS exhibit low values, except for the Lower East Side of NYC and parts of Staten Island. The areas with high values are mainly due to the presence of parks such as Freshkills Park and Brookfield Park in Staten Island, Marine Park in the Lower East, and Jacob Riis Park in Lower Queens, all depicted in orange or red. The distribution of RGS differs, with a greater number of high values concentrated in central Manhattan, parts of Queens and the Bronx, and Prospect Park in Brooklyn. The FANR map reveals that Staten Island has significantly higher values compared to other boroughs.

Figure 2 Accessible areas for each UGS category within a 10-Minute Walk (Square Kilometers)

To visualize the simultaneous representation of different factors, this study employs a bivariate mapping technique, building upon the principles of spectrally encoded two-variable maps developed by Olson (1981). Figure 3 illustrates the number of each targeted age group – children, pre-college students, and seniors – alongside the accessible UGS area for each grid. This approach identifies areas with high population but low accessible green space, represented in pink, and areas with low population but high accessible urban green spaces, represented in light blue.

For population counts in relation to accessible urban green spaces (UGS), differences are more pronounced at the neighborhood level than at the borough level. Generally, lower Brooklyn and parts of Queens exhibit limited UGS despite high population counts, whereas Staten Island shows the opposite trend, with a relatively low population count but extensive UGS.

Figure 3 Bivariate maps for each population group in comparison with accessible Urban Green Spaces (UGS)

Travel Time and Accessibility Analysis

In addition to examining accessibility within a 10-minute walk, this study also analyzes the nearest UGS for each grid cell, and explores how areas with longer and shorter travel times are spatially clustered in different parts of NYC. Figure4 shows that travel times to UGS are generally shorter in Staten Island and Manhattan, while Brooklyn and outer Queens experience longer travel times.

Figure 5 shows the result of the hot spot analysis, identifying areas with consistent spatial distribution of travel time. In particular, it emerges how longer travel times are found in Brooklyn, Queens, and northern part in Staten Island, while Manhattan and most part in Staten Island show shorter travel times, indicating better UGS accessibility.

Accessible Metrics

Several calculations were performed to assess the extent of uncovered accessible areas within New York City. It was determined that only 4.02% of the total NYC area remains uncovered by accessible urban green spaces. Notably, 72.24% of these uncovered areas are designated as UGS. However, these areas exhibit a significantly lower pedestrian density of 0.067 km per square kilometer, compared to the NYC average of 37.256 km per square kilometer.

Discussion

The results reveal that New York City offers broad access to Urban Green Spaces (UGS), with nearly all areas located within a 10-minute walk of some form of green space, including Forest Areas and Nature Reserves (FANR), Open Grass and Scrublands (OGS), and Recreational Green Spaces (RGS). However, accessibility disparities are evident.

Brooklyn and Queens, in particular, exhibit lower UGS coverage compared to other boroughs, which impacts high-need areas within these communities. For instance, neighborhoods like Bensonhurst, lower Brooklyn, and areas around Park Slope near Prospect Park reflect substantial UGS demand among children and pre-college students, who have limited access to these spaces. In Queens, there is a pronounced need for UGS among seniors in neighborhoods surrounding Whitestone, Flushing, and Auburndale, where green space access is low. Meanwhile, Douglaston, which has ample UGS, has a relatively low population of children who might otherwise benefit from these resources, underscoring an imbalance between UGS availability and local demographic needs. Similar patterns emerge in the Bronx, particularly in the West Bronx and Pelham Parkway, which have high population counts but limited UGS resources. These three boroughs also have a longer travel time, representing reduced accessibility.

Conversely, Staten Island and areas surrounding Central Park in Manhattan generally demonstrate better UGS accessibility and quality, coupled with shorter travel times to green spaces, which enhances access for residents. However, other parts of Manhattan, such as the Lower East Side, face UGS limitations despite higher populations of children, pre-college students, and seniors. Midtown also exhibits this trend, where children and seniors encounter restricted green space access.

Policy Recommendations

Based on the findings, it is essential to enhance both the accessibility and equity of UGS, particularly in areas of Brooklyn and Queens, where the accessible UGS is limited. In Brooklyn, expanding green spaces and improving access, particularly around Prospect Park, would significantly improve UGS availability for children, pre-college students and seniors. Ensuring equitable UGS distribution across these groups is critical. In Queens, addressing the gaps in UGS accessibility for seniors by expanding or improving green spaces in underserved areas is necessary to achieve a more balanced distribution of UGS. In Manhattan, priority should be given to increasing UGS accessibility in Midtown and lower east side, especially for children and seniors, through the development of new green spaces or the enhancement of existing ones. More broadly, across all boroughs, improving UGS access by upgrading the pedestrian network, and potentially residents’ relocation to ensure ease of access to these spaces, will contribute to a more equitable and inclusive distribution of UGS throughout the city. We hope that the results included in this article may contribute to the ongoing green-infrastructure improvement (The Nature Conservancy, 2021) and expansion plans (NYLCV, 2022).

Conclusions and Future Works

This study highlights the accessibility and equity to Urban Green Spaces (UGS) across New York City, with Brooklyn and Queens emerging as particularly underserved. The analysis also reveals longer travel times to UGS in Brooklyn and outer Queens, further emphasizing accessibility challenges. To address these issues, targeted policy interventions are recommended, including the creation and expansion of UGS. While limitations such as the use of grid centroids and the lack of considering socio-economic factors, the study provides valuable insights into the need for more equitable distribution of UGS.

Improving UGS accessibility is essential for promoting public health, environmental quality, and social equity. By addressing the disparities identified in this research, New York City can create a more inclusive and sustainable urban environment. In future analysis, incorporating data including privately owned green spaces and rooftop green spaces, as well as examining the progress or changes following policy implementation, could provide a more comprehensive understanding of UGS accessibility, further enhancing the analysis and informing more effective policy interventions.

Acknowledgments

The research work has been carried out in collaboration with Hongying Wu, a Master’s student in Applied Urban Science and Informatics at New York University, Class of 2025, as part of her internship at Systematica US Inc. The analyzed data were treated according to the GDPR-General Data Protection Regulation (EU, 2016/679). This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Esri (n.d.). Incremental Spatial Autocorrelation (Spatial Statistics). ArcGIS Pro documentation. Accessed September 4, 2024. Retrieved from: https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/incremental-spatial-autocorrelation.htm

Getis, A., & Ord, J. K. (1992). The analysis of spatial association by use of distance statistics. Geographical analysis, 24(3), 189-206. https://doi.org/10.1111/j.1538-4632.1992.tb00261.x

Morpurgo, J., Remme, R. P., & Van Bodegom, P. M. (2023). CUGIC: The consolidated urban green infrastructure classification for assessing ecosystem services and biodiversity. Landscape and Urban Planning, 234, 104726. https://doi.org/10.1016/j.landurbplan.2023.104726

NYC (n.d.). NYC Parks and Open Data. Department of Parks & Recreation. Accessed February 25, 2025. Retrieved from: https://www.nycgovparks.org/about/data

NYC (2011). PlaNYC update April 2011: A Greener, Greater New York. City of New York. Accessed February 25, 2025. Retrieved from: https://www.nyc.gov/html/planyc/downloads/pdf/publications/planyc_2011_planyc_full_report.pdf

NYLCV (2022). New York City Policy Agenda 2022. New York League of Conservation Voters & New York League of Conservation Voters Education Fund. Accessed February 25, 2025. Retrieved from: https://nylcv.org/wp-content/uploads/NYLCV_2022_NYCPolicyAgenda_webpgs.pdf

Olson, J. M. (1981). Spectrally encoded two-variable maps. Annals of the Association of American Geographers, 71(2), 259-276. https://doi.org/10.1111/j.1467-8306.1981.tb01352.x

Roy, S., Byrne, J., & Pickering, C. (2012). A systematic quantitative review of urban tree benefits, costs, and assessment methods across cities in different climatic zones. Urban Forestry & Urban Greening, 11(4), 351–363. https://doi.org/10.1016/j.ufug.2012.06.006

The Nature Conservancy (2021). NYC Urban Forest Agenda: Toward A Healthy, Resilient, Equitable, and Just New York City. The Nature Conservancy. Accessed February 25, 2025. Retrieved from: https://forestforall.nyc/wp-content/uploads/2021/06/NYC-Urban-Forest-Agenda-.pdf