Urban Park Mobility Analysis

Parks and green spaces are highly valued urban destinations, enhancing the quality of life for city residents (El Murr, Boisjoly, & Waygood, 2023). These spaces contribute significantly to well-being (El Murr, Boisjoly, & Waygood, 2023), environmental health (Haase et al., 2014), social cohesion (Peters, Elands, & Buijs, 2010), and economic prosperity (Hoshino & Kuriyama, 2010). Consequently, accessibility to parks and green spaces is considered a crucial component of life quality (Cracu, Schvab, Prefac, Popescu, & Sîrodoev, 2024).

Ensuring equitable access to these areas is essential for fostering inclusive urban environments. By prioritizing accessibility, urban planners can ensure that all residents, regardless of their socio-economic status or background, benefit from the physical and mental health advantages provided by green spaces. This focus supports a more sustainable and livable urban environment, contributing to overall urban resilience and community well-being.

However, the accessibility of urban greenspaces is an increasing concern (Comber, Brunsdon, & Green, 2008). Addressing such disparities is crucial for achieving equitable urban development.

International literature (Comber, Brunsdon, & Green, 2008) identifies a 2-hectare minimum for urban park areas. Sadeghian and Vardanyan (2015) categorize urban parks into three types, setting specific area and distance standards for each, notably city parks and community parks with a 2-hectare minimum and a 2.4 km maximum from residential areas. Romanian law classifies parks as "public green spaces with unlimited access" and requires a minimum of 1 hectare, incorporating both natural and built environments to facilitate cultural, educational, and recreational activities. In this project, hexagon grids measuring 900 by 900 square feet are utilized to conduct hotspot analysis. Given that one hectare would encompass nearly 2.5 hexagon grids, this area may be excessively large for detailed analysis. Therefore, one acre, which is approximately 0.4 hectares, has been selected as a more suitable scale. This choice ensures a more manageable and precise spatial resolution for our analytical purposes.

Both the distance and the available area within that distance were emphasized as two elements of accessibility in the previous research. The analysis of accessibility examines the daily mobility patterns of travelers, focusing on their journeys and preferred methods of transportation, especially walking. The statistical analysis of trip lengths is used to establish an adaptive threshold, which in turn aids in calculating a cumulative potential measure for assessing accessibility. (Rojas, Páez, Barbosa, & Carrasco, 2016). The distance is typically measured either as the Euclidean distance, which is a straight line between two points, or as the network distance, which represents the shortest route along an actual path from origin to destination. The network distance generally provides a more accurate reflection of actual travel conditions compared to the Euclidean distance (Reyes, Páez, & Morency, 2014).

Synthesizing previous studies, it is observed that the standard walking distance for accessing urban green spaces is commonly set within a 5-10 minute threshold. Conversely, the cycling distance to access these areas varies more widely, ranging from approximately 1300 meters up to 5 kilometers (Richardson, Pearce, Mitchell, Day, & Kingham, 2010).

NYC Open Data (2023) depicts that "Walking distance" refers to a maximum of a 1/4-mile for smaller sites like playgrounds and sitting areas, while for larger parks that cater to a broader area, such as those over 8 acres or near water, the distance extends to up to a 1/2-mile.

Based on the walking distance and time threshold for park accessibility, the project references the article by Curran (2024) and sets a 4 km (about 2.48 miles) threshold for cycling distance.

The selection of New York City, Boston, and Washington DC for this study on urban park accessibility is informed by their park access metrics, which are exemplary in the U.S. Washington DC, with 99% of its residents living near a park and the highest park-related public spending, sets the benchmark for accessibility. New York City and Boston, both boasting high accessibility with 99% and 100% of residents respectively living within a 10-minute walk of a park, offer robust comparative cases (Trust for Public Land, n.d.). These cities provide diverse yet successful models of urban park accessibility, making them ideal for examining how public green spaces can enhance urban living.

The Geographic Information System (GIS) is a software designed to collect, store, and display diverse types of data, including physical, biological, environmental, ecological, and geological aspects. This integration facilitates comprehensive spatial and analytical evaluations (Gharehbaghi & Scott-Young, 2018).

GIS is essential for addressing urban park accessibility by integrating complex spatial and non-spatial data for comprehensive analysis and decision-making. It manages diverse variables like park area, distances, and demographic profiles, enhancing the understanding of green space accessibility across populations. Capabilities of GIS in modeling network distances and conducting hotspot analyses are crucial. Moreover, the use of Model Builder in GIS allows for the automation of these processes, enabling consistent application of analysis criteria across different urban areas. This systematic approach helps urban planners in strategic planning and connectivity optimization between neighborhoods and parks, thus improving the quality of life through better access to green spaces.

• Densely populated blocks are predominantly located near parks.

• The majority of residents live within walking and cycling distance to parks.

The project will employ the GIS tools described previously to evaluate park accessibility in New York City, Boston, and Washington DC. Building on the hypothesis that densely populated blocks tend to cluster near parks, the analysis will define service areas for each city. This approach will facilitate a detailed comparison of the effectiveness with which these urban environments provide their residents access to green spaces. This comparative study aims to highlight variations in park accessibility and its correlation with population density, thereby providing insights into urban planning and the distribution of recreational areas.

This section will outline the methodologies utilized within ArcGIS Pro.

From three swipe maps above, it could be concluded that population distribution has no strong relationship with where parks locates. Except for DC, NYC and Boston has a high population density in central areas, whereas parks, especially big-flagship parks are not necessarily sorrounded with high density populations. DC, however, although sorrounded by high density populations, not only in big-flagship parks, but also small-community parks, due to the small area size of DC's region, this result has no enough extenaility. As a result, no conclusion could be achieved of people choosing residential areas according to park cluster areas from this analysis.

D.C. and Boston satisfy the criteria of 10.8 acres of all parkland per 1,000 residents by NRPA. D.C. even has over 10.8 acres of large parkland per 1,000 residents.

NYC's park strategy seems focused on accessibility and high population coverage with smaller parks spread throughout the city, which is typical for its dense urban environment. In contrast, Washington D.C. emphasizes larger parks, possibly benefiting from more available space and a city layout that can supports expansive green areas. Boston appears to balance both approaches, with a decent proportion of large parks and good overall accessibility.

From the chart, NYC has similar walk-accessible and bike-accessible populations. It also offers the most extreme difference in park acreage per 1,000 residents. Boston provides a moderate level of metrics, only percentage of population within cycling accessibility is small. D.C. shows the most generous park allocation.

In the project, numerous parks are filtered out based on their categories and area. This selection criterion could potentially lead to discrepancies in the correlation between population and park density. It is important to consider that people may frequently visit smaller parks as well, despite their exclusion from the analysis due to size constraints. Ignoring these smaller parks might distort the actual accessibility and usage patterns within the urban landscape, which could, in turn, affect the reliability of the study's findings regarding the distribution and utilization of green spaces by the local population.

Moreover, when residents consider where to live, many factors come into play beyond just access to green spaces. Housing prices are a significant consideration, as affordability often dictates the neighborhoods people can choose. Commute times are another critical factor, as longer commutes can negatively impact work-life balance and overall quality of life. Proximity to schools, healthcare facilities, and shopping centers also play crucial roles in these decisions. Safety, community amenities, and the overall reputation of a neighborhood are additional factors that influence residential choices. All these elements combined shape the preferences and decisions of residents when selecting their living environments.

Morran's Index offen comes with Hot Spot Analysis (Getis-Ord G*) when doing spatial analysis and statistics. Morran's Index is a tool analyze the global spatial pattern, whereas Hot Spot Analysis focus on local spatial clusters. In this report, an analysis on how parks distribute and cluster within different cities is more likely to be a local spatial pattern, considering of the purpose of picking out big-flagship parks' cluster for further analysis. In this case, Morran'I could be one reference, but not a necessarity.

Differences in walking and biking accessibility are largely dictated by urban area restrictions on the construction of large parks. Walking accessibility in this study is based on parks larger than 1 acre in the green space category (excluding purely hard surfaces), while biking accessibility is based on large parks that can be linked together to form hotspot clusters in the above parks. Large, contiguous parks require sufficient space, and thus the dense residential neighborhoods or narrow landmasses in the three cities rarely have such conditions, leaving several hard-to-reach areas within cycling accessibility. The distribution of walking accessibility is relatively even in all three cities, demonstrating excellent community-level urban park planning.

Ideally, the selection of facilities should be based on the generation of real entrances and exits of the park; but in this project, due to the lack of entrance and exit points or park route feature class, only all junctions within 30m from the park polygons can be selected as facility points. Such a method may cause errors for large parks in dense road areas, resulting in more “entrances and exits” than in reality. For example, in New York's Central Park, the selected junctions are more than the actual number of entrances and exits due to Manhattan's dense road network around the park.

This study assumes an even distribution of population within each block and selects only blocks that are entirely within the accessible area for analysis. There are several limitations to this approach. First, assuming an even distribution of population within each block may not accurately reflect true population and residential patterns, which can vary widely, especially in urban areas with mixed land uses. In addition, excluding blocks that are only partially within accessible areas may result in an underestimation of the population that may benefit from proximity to parks and amenities. This approach ignores the portion of the population that lives near the defined boundary but not entirely within the accessibility definition. As a result, the study may not fully reflect the actual extent of the population that is accessible to park facilities, thus affecting the accuracy of the accessibility impact assessment.

The analysis assumes that individuals predominantly cycle to parks with significant GI Bin Scores identified through hotspot analysis. This assumption restricts the estimated cycling service areas, resulting in noticeably lower coverage compared to walking service areas. However, this may not accurately reflect real-world behaviors, as individuals might cycle to a broader range of parks, not solely those rated highly in the GI Bin Scores. Furthermore, the study lacks empirical data on cycling patterns to parks, which constrains the ability to validate and refine the assumptions made about cycling accessibility. This limitation highlights the necessity for comprehensive data collection on urban cycling behavior to enhance the accuracy of service area modeling and ensure that urban planning decisions are informed by representative activity patterns.

Urban planners and policymakers should prioritize the development of new green spaces in densely populated urban areas where such amenities are currently insufficient. Simultaneously, incentives or zoning adjustments could be implemented to guide residential developments closer to existing large park clusters. This dual strategy will enhance the overall accessibility to green spaces, fostering healthier urban environments and improving quality of life for city residents. Furthermore, regular assessments should be conducted to ensure that the distribution of green spaces continues to meet the evolving needs of the urban population.

This study highlights the critical role of parks in enhancing the quality of urban life, corroborating the findings of previous research. The position is that densely populated blocks tend to be situated near parks, with most residents residing within walking and cycling distances of these green spaces. To investigate this hypothesis, GIS tools were employed to construct heatmaps, perform hotspot and network analyse to assess park accessibility in New York City, Boston, and Washington D.C.

Results demonstrate that Washington D.C. and Boston adhere to, and sometimes surpass, the National Recreation and Park Association's (NRPA) recommended standard of 10.8 acres of parkland per 1,000 residents. New York City, with its dense urban structure, showcases significant variance in park acreage per capita, indicative of a strategic focus on maximizing accessibility through numerous smaller parks. Conversely, Washington D.C.'s generous parkland allocation likely reflects urban planning strategies that leverage available space for larger parks. Boston balances accessibility and size, with a considerable proportion of large parks and favorable accessibility statistics.

Nevertheless, the study encounters limitations concerning the Facilities Selection and Block-level Population Selection, which may skew the correlation between population density and park accessibility. Despite these constraints, the research contributes valuable insights into the spatial distribution and accessibility of urban parks, highlighting crucial areas for policy enhancement.

Feiyang Ren: Network analysis of biking mode and related part in story map.

Ruixin Gan：Network analysis of walking mode and related part in storymap.

Xueliang Yang: Hotspot analysis for park areas; heatmap for population density distribution.

City of Boston. (2016). Parks and recreation.  https://www.cityofboston.gov/parks 

Comber, A., Brunsdon, C., & Green, E. (2008). Using a GIS-based network analysis to determine urban greenspace accessibility for different ethnic and religious groups. Landscape and Urban Planning, 86(1), 103-114.  https://doi.org/10.1016/j.landurbplan.2008.01.002 

Cracu, G.-M., Schvab, A., Prefac, Z., Popescu, M., & Sîrodoev, I. (2024). A GIS-based assessment of pedestrian accessibility to urban parks in the city of Constanța, Romania. Applied Geography, 103(229).  https://doi.org/10.1016/j.apgeog.2024.103229 

Curran, D. (n.d.). How long to cycle 4 km (By age gender & speed). Condition and Nutrition.  https://conditionandnutrition.com/how-long-to-cycle-4-km-by-age-gender-speed/ 

El Murr, K., Boisjoly, G., & Waygood, E. O. D. (2023). Measuring accessibility to parks: Analyzing the relationship between self-reported and calculated measures. Journal of Transport Geography, 103(550).  https://doi.org/10.1016/j.jtrangeo.2023.103550 

Gharehbaghi, K., & Scott-Young, C. (2018). GIS as a vital tool for Environmental Impact Assessment and Mitigation. IOP Conference Series: Earth and Environmental Science, 127(1), 012009.  https://doi.org/10.1088/1755-1315/127/1/012009 

Haase, D., et al. (2014). A quantitative review of urban ecosystem service assessments: Concepts, models, and implementation. AMBIO, 43(4), 413-433.  https://doi.org/10.1007/s13280-014-0509-5 

Hoshino, T., & Kuriyama, K. (2010). Measuring the benefits of neighbourhood park amenities: Application and comparison of spatial hedonic approaches. Environmental and Resource Economics, 45(3), 429-444.  https://doi.org/10.1007/s10640-009-9321-5 

NCPC Home Page. (n.d.).  https://www.ncpc.gov 

NYC OpenData. (2023). Walk to a Park Service area.  https://data.cityofnewyork.us/Recreation/Walk-to-a-Park-Service-area/5vb5-y6cv 

Peters, K., Elands, B., & Buijs, A. (2010). Social interactions in urban parks: Stimulating social cohesion? Urban Forestry & Urban Greening, 9(1), 93-100.  https://doi.org/10.1016/j.ufug.2009.11.003 

Reyes, M., Páez, A., & Morency, C. (2014). Walking accessibility to urban parks by children: A case study of Montreal. Landscape and Urban Planning, 125, 38-47.  https://doi.org/10.1016/j.landurbplan.2014.02.002 

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 space utility. BMC Public Health, 10, Article 240.  https://doi.org/10.1186/1471-2458-10-240 

Rojas, C., Páez, A., Barbosa, O., & Carrasco, J. (2016). Accessibility to urban green spaces in Chilean cities using adaptive thresholds. Journal of Transport Geography, 58, 12-22.  https://doi.org/10.1016/j.jtrangeo.2016.10.012 

Rossi, F., & Becker, G. (2019). Creating forest management units with Hot Spot Analysis (Getis-Ord Gi*) over a forest affected by mixed-severity fires. Australian Forestry, 82(4), 166–175.  https://doi.org/10.1080/00049158.2019.1678714 

‌Rozhkov, Anton (2024). Apatial Analyst 1[Lecture PowerPoint Slide]. Available:   https://brightspace.nyu.edu/d2l/le/lessons/341690/topics/9961922  

‌