Abstract
Context
Urban green spaces (UGS) are not evenly distributed within cities, and some neighborhoods with high socio-environmental demands require more UGS than others. This raises two challenges: green inequality and demand-based inequity. However, comprehensive assessments of UGS inequality and inequity in cities worldwide are lacking.
Objectives
We aim to develop a multi-level approach and supply-demand concept to assess UGS inequality and inequity across neighborhoods in international cities with contrasting geographical and socio-political contexts.
Methods
We measured multi-level green accessibility and human demands based on Earth Observation and statistical data. UGS inequality and supply-demand mismatches were assessed by Gini coefficients, spatial cluster analysis, and statistical models.
Results
We found that: (1) UGS inequality is primarily reflected by the public park per capita in three cities. New York has larger UGS inequality than Beijing and Amsterdam. (2) Demand-based inequity in terms of low supply and high demand is mainly scattered around the city center in three cities. Tree coverage does not align with environmental pressures (LST/PM2.5) in New York and Beijing. (3) Relations between green supplies and human demands vary by cities and indicators. A shorter distance to the nearest large park is associated with a higher proportion of the elderly and children in New York and Amsterdam.
Conclusions
Our findings can inform UGS allocations to improve landscape sustainability in the neighborhoods with low green supply and high human demand, and to prioritize specific green metrics based on demand-oriented equity.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
Urban green space (UGS), an indispensable natural element of metropolises, provides environmental and social benefits for urban residents (Triguero-Mas et al. 2015; Veerkamp et al. 2021). The Sustainable Development Goal (SDG) 11.7 has been set to provide universal access to safe, inclusive and accessible, green and public spaces, in particular for women and children, older persons and persons with disabilities by 2030 (UN 2021). However, UGS is unevenly distributed within cities, which is increasingly recognized as an environmental justice issue worldwide (Xiao et al. 2017; Nelson et al. 2021). Common strategies are “just green enough” for every district, such as developing small parks in less green areas (Wolch et al. 2014). Since the potential space for new UGSs is very limited in metropolises (Lin et al. 2022), a better strategy is to optimize the UGS allocation based on the distribution of human demands (Menconi et al. 2021; Liu et al. 2022). To this end, accurately assessing the green supply and its relations to human demand is required for scientifically planning demand-oriented UGS (Lin et al. 2022).
Previous studies have observed the urban green inequality by measuring the total amount (e.g. NDVI, green cover) or accessibility (e.g. distance or travel time to parks) (Wüstemann et al. 2017; Meng et al. 2020; Spotswood et al. 2021). They revealed that the characteristics of UGS matter greatly for its environmental and social benefits. For example, public parks are formal UGS and freely accessible to every inhabitant, and large parks can support more diverse human activities such as jogging or biking than small ones (Shanahan et al. 2015; Hoover and Lim 2021). However, little research assessed UGS inequality in terms of vegetation types, public or private ownership, and park size. In particular, neighborhood-scale inequality was rarely compared among global cities with contrasting geographical, socio-political, and climatic contexts (Fletcher et al. 2021; Veerkamp et al. 2021).
Most studies explored the uneven distribution of UGS without sufficiently considering human demand (Wüstemann et al. 2017; Han et al. 2022), defined as “UGS inequality” in this study. Beyond UGS inequality, recognizing distinct human demands for UGS would reveal mismatches between green supply and human demand (Hunter et al. 2019; Lin et al. 2022), which is defined as “demand-based inequity” (Pham et al. 2012). Pham et al. (2012) illustrated how equity in terms of the urban heat island effect (i.e. demand for cooling) requires different amounts and types of vegetation across blocks. We attempt to evaluate a more comprehensive inequity concerning multiple socio-environmental benefits of UGS, including cooling, air clarification, and the well-being of vulnerable people. As such, green inequality and demand-based inequity are conceptualized as pursuing “Gini perfect equality” and “demand-oriented equity” in this study, respectively.
A growing body of studies started to consider various human demands for UGS in terms of mitigating environmental pressures and supporting human well-being (Luederitz et al. 2015; Baró et al. 2015; Feng et al. 2019). These studies primarily focus on the locations of UGS beneficiaries, without specifically addressing vulnerable groups that could particularly benefit from UGS due to their heightened risk of illness and lifelong impacts. Recent research considered the low-income group and integrated it with environmental pressures and total population to map the overall demand (Fletcher et al. 2021; Lin et al. 2022). Another study considered the elderly and children to examine UGS inequalities, particularly for local parks (Kim et al. 2023). However, those vulnerable groups were often left out of the comprehensive assessment of human demand (Chen et al. 2022; Veerkamp et al. 2021). Moreover, existing demand-based studies have not correlated various UGS metrics with environmental and social demands, respectively, even though the quantity and quality of UGS would support different demands (details can be found in the theoretical basis section).
To fill the gaps in assessing demand-based inequity, this study presented a paradigmatic framework for multi-level analysis of green supply and human demand, implemented at the neighborhood level in New York, Amsterdam, and Beijing by integrating satellite and social data. The specific goals were to (1) quantify multi-level UGS inequality by the Gini coefficient, (2) analyze UGS inequity based on supply-demand mismatches, and (3) explore relationships between green supplies and human demands. Our results can guide demand-oriented UGS allocation to achieve city goals of green equity.
Theoretical basis
Although UGS was initially promoted for food production, modern UGSs are mainly used for leisure, recreation, health, and ecology (van Leeuwen et al. 2011). In this study, we focus on two dimensions of ecosystem services provided by UGSs: (a) environmental benefits, including cooling effects and air purification; (b) socio-economic benefits, encompassing leisure, recreation, character-building, therapy, social interactions, aesthetic value, and market value (Bratman et al. 2019; Hunter et al. 2019; Veerkamp et al. 2021; Lin et al. 2022). Urban residents seek these UGS benefits, and such human demands vary based on environmental pressures and socio-economic vulnerability, as demonstrated in the context of our study cities (Huang et al. 2020; Paulin et al. 2020; Schrammeijer et al., 2022; Wu et al. 2020). Thus, we correlate green supply with human demand at three levels (Fig. 1):
At the first level, the green amount meets the demand for environmental livability (hereafter named ‘environmental livability demand’). UGS contributes to the maintenance of a healthy urban environment by adapting to climate change and improving air quality toward a resilient and livable city (Meerow and Newell 2017; Sera et al. 2019). The urban heat island effect is a widely concerning climate issue in cities, caused by hotter built environments than their neighboring rural counterparts (Oke 1982; Hsu et al. 2021). In particular in the future, when the urban heat island effect is likely to intensify due to urbanization and global warming, moderating urban heat and the risk of heat-related illnesses will be of utmost importance (van Leeuwen et al. 2011). UGSs, especially trees, can deliver cooling effects through shade and evapotranspiration (Zhou et al. 2021). Air pollution is another urgent issue in the urban environment, whereas trees can improve air quality by absorbing certain airborne pollutants from the atmosphere at leaf surfaces (Yan et al. 2016). Thus, higher levels of environmental pressures represented by urban heat and air pollution lead to greater environmental livability demand for urban trees.
At the second level, access to public parks meets social demands in terms of socio-economic vulnerability represented by low-income households. Public parks, being a predominant type of UGS, often serve as venues for physical activity accessible to all residents, thereby contributing to enhanced public health (Bedimo-Rung et al. 2005; Schipperijn et al. 2017). Compared to high-income households, low-income households are more dependent on local public parks considering the travel costs to distant UGSs in surrounding rural regions and user fees of private UGSs (Feng et al. 2019; Basu and Nagendra 2021; Lin et al. 2022). Low-income households desire more accessible public parks to improve their livability since their neighborhoods are normally overcrowded with dense buildings, which has been particularly observed in our study cities (Huang et al. 2020). As an attractive urban setting, public parks can enhance the real estate values of low-income communities (Wüstemann et al. 2017). This economic effect may be paradoxical since the resulting gentrification might make it unaffordable for low-income people, but our study cities are developed capital cities and improving the living quality is the first strategy for those relatively low-income neighborhoods (Paulin et al. 2020). Thus, higher numbers of people especially low-income people lead to greater social demand for public parks.
At the third level, distance to a high-quality park meets social demands in terms of life cycle represented by elderly people (aged 65 +) and children (aged 14 −). We chose the size of the park to represent the quality because it is a universal standard for a high-quality park (Kmail and Onyango 2020; Hoover and Lim 2021; Remme et al. 2021). Compared to small parks, large parks can provide more diverse outdoor activities such as jogging and cycling which could reduce the risk for many chronic diseases, and thus enhance the physical health and psychological well-being of urban residents (Wolch et al. 2014; Remme et al. 2021). Such benefits are particularly essential for the elderly and children, since access to parks may increase the longevity of elderly people (Takano et al. 2002), and diverse outdoor activities in large parks can reduce obesity, develop a strong physique, improve attention deficit disorder, build active character, and other life-long impacts on children (Taylor et al. 2001; Wolch et al. 2014; Salthammer et al. 2016). Large parks are usually well managed and include facilities such as high-quality paths and toilets, which are essential to support children’s and elderly’s outdoor activities (Kemperman and Timmermans 2014). Moreover, senior citizens particularly rely on greenspace within walking distance due to their reduced mobility and cognitive function (Alves et al. 2008), and nearby parks can motivate children to incorporate physical activity into daily life (Sallis et al. 2012). Thus, higher numbers of people especially the elderly and children lead to increased social demand for adjacent large parks.
Given that the above three supply-demand relations are not strictly one-to-one correspondence, the overall balance between green supply and human demand should be assessed by integrating the environmental and social benefits of UGS. According to the above theoretical basis, we define low-income households, elderly people, and children as vulnerable people. Cross-supporting services exist across three supply-demand levels, i.e., the total green amount generally supports social demands, and access to public parks or large parks benefits vulnerable people (Feng et al. 2019; Fletcher et al. 2021). Co-effects also exist between environmental livability and social demands, e.g., elderly people are vulnerable to heat pressure due to a high risk of illness (Hsu et al. 2021). Therefore, high levels of environmental pressure, total population, and vulnerable population would simultaneously lead to increased demand for UGS. Mismatches between green supply and human demand (e.g. low supply-high demand clusters) can be identified from this comprehensive perspective. Such overall mismatches can guide the optimization of urban green allocation toward demand-oriented equity.
Data and methods
Study area
We chose New York, Amsterdam, and Beijing as the study cities for the following reasons. First, they are known as the populous cities on different continents with 2963 inhabitants/km2 in New York, 3272 in Amsterdam, and 8502 in Beijing in 2015 (Florczyk A. 2019). Second, existing evidence has revealed that urban residents in dense and aging cities usually have scarce and dwindling access to nature and greatly desire public open space (Remme et al. 2021). Third, those developed capital cities have sufficient neighborhood-scale data (e.g. population of age groups) supporting our in-depth supply/demand assessments. Fourth, their municipalities have greatly focused on enhancing and equalizing UGS. The municipality of Amsterdam epitomizes the desire to improve the quality and accessibility of public greenspaces and set the standard of a minimum green provision of 24 m2 per household (Gemeente Amsterdam 2018). In New York, Million-Trees had aimed to equalize urban forests but failed to “prioritize low-canopy, low-income communities of color” (Garrison 2019). Chinese official UGS planning standards point out that UGS construction should be guided by the ecological civilization strategy and give full play to the multiple functions of UGS in ecology, recreation, landscape, and protection (Ministry of Housing and Urban-Rural Development 2019). Therefore, identifying the UGS inequality and diverse supply-demand mismatches could provide essential information for these official goals.
Our unit of analysis is the neighborhood. We investigated all the neighborhoods within the city boundary that is the metropolitan area of Amsterdam, the borough area in New York, and the area inside the fifth ring road in Beijing. The neighborhoods with less than five people were excluded as outliers.
Urban greenspace measurement and indicator review
Access to UGS has been widely measured by amount, proximity, distance, and quality indices (Wüstemann et al. 2017; Spotswood et al. 2021). We summarized their advantages and disadvantages as shown in Table 1. The amount/proximity metrics are the common measurement that can be obtained from public and timely remote sensing imagery, but they fail to reflect the functional use and quality of UGS (Wüstemann et al. 2017; Schwarz et al. 2018; Yan et al. 2020). The distance metrics can measure geographical accessibility. Many studies have used Geographic Information Systems (GIS) to measure the Euclidean distance or travel times from the centroid of a neighborhood to parks (Giuliani et al. 2021; Wang et al. 2021), but they did not consider the size of UGS. However, small parks are rather frequent and evenly distributed in cities which makes the distance from neighborhoods to parks similarly short. The quality metrics are the highest level of measurement while most existing indicators are subjective or over complicated to be used for multiple cities (Hoover and Lim 2021; Knobel et al. 2021). Trade-offs also exist in different indicators. For example, a study in Baltimore found that although Blacks were more likely than Whites to live within walking distance of a park, Whites had access to more park acres (Boone et al. 2021). Therefore, we combined amount, proximity, distance, and quality to comprehensively measure access to UGS.
Based on the indicator review, we proposed three metrics to assess the multi-level green supply (Fig. 1; Table 2). We selected trees and public parks that can efficiently provide ecosystem services for urban residents, i.e., trees deliver a higher cooling efficiency by creating shades and evapotranspiration than other vegetation types (Zhou et al. 2021). First, we used a remote sensing product - ESA World Cover at 10 m resolution in 2020 to compute the sum of tree pixels in each neighborhood and get the tree cover. Second, we combined neighborhood-level census data with Open Street Map (OSM) which provides function and ownership information of UGS to compute the public park area per capita. We extracted the public parks under the ‘leisure’ tag in OSM and compared the obtained maps to the official maps of city parks. A 400 m buffer was used in the calculation of tree cover and public park area per capita for each neighborhood because the UN-Habitat recommends using 400 m as an international walking distance for analyzing the access to public space at a neighborhood level (UN-Habitat 2020). Previous global research has tested the sensitivity of different buffer distances and found a consistent ranking of greenspace exposure (Chen et al. 2022). Third, we combined the distance with quality to measure the highest level of green supply: the distance to the nearest large park. We chose the size of a park to represent quality because larger parks can provide more services (Hoover and Lim 2021). Our threshold for a large park was set as 20 ha based on a widely used standard (Natural England 2010) and previous research for our study cities (Feng et al. 2019). We also analyzed the sensitivity by adjusting the threshold from 15 ha to 25 ha and found no significant impact on the distance to the nearest large park. Euclidean distance tool in ArcGIS pro was used to calculate the distance from each cell to the closest large park excluding the park areas, and the average distance of a neighborhood to the nearest large park was then obtained. Together, we calculated tree cover, public park area per capita, and the distance to the nearest large park (> 20 ha) to measure the multi-level green supply.
Estimating human demands for urban greenspace
The current definitions of human demand for UGS mainly include actual consumption and potential demand. The actual consumptions are based on empirical results, i.e., the sum of actual use or consumption of green services within a certain area over a certain period (Burkhard et al. 2012). The potential demands include the desire to mitigate environmental pressures (Lin et al. 2022), the amount of a service required by society (Luo and Li 2021), and subjective willingness for green services (Fletcher et al. 2021).
In this study, we integrated the metrics of environmental livability and social demand to estimate potential human demand (Table 2). Based on remote sensing products of LST and PM2.5 at 1 km resolution, we used the zonal statistics (mean) to calculate the average LST during summer daytime because heat exposure on summer days is likely to be at a maximum value (Hsu et al. 2021) and annual PM2.5 concentration in each neighborhood (van Donkelaar et al. 2021). Population data of different age groups and income data at the neighborhood level were mainly collected from national and local statistical websites (NHGIS 2020; Statistics 2020; CBS 2021). Since up-to-date statistical data were not available at the neighborhood level for some metrics in New York and China, we used Living Atlas, rent distribution (Lianjia website), and estimated population (Jing et al. 2020) as ancillary data. Specific data and methodology for each demand metric and city are listed in Table 2.
Quantification of inequality/inequity and supply-demand clusters
Although the Gini coefficient is popular in measuring income inequality (Hu and Wang 2005; Xie and Zhou 2014), it has also been used for environmental or social inequality such as water resources, flood drainage, technology, and population (Wu and Xu 2010; Yuan et al. 2017; Zhang et al. 2020). Recent research started to apply the Gini coefficient to UGS (Wüstemann et al. 2017; Feng et al. 2019). As such, we used the Gini coefficient to quantify the distributional “inequality” in green supplies and demands across neighborhoods in each city. The Gini coefficient was calculated as the ratio of the area between the perfect equality line and the Lorenz curve (A) divided by the total area under the perfect equality line (A + B) (Fig. S1). The Gini coefficient ranges between 0 and 1, with 0 representing perfect equality and 1 complete inequality (Wüstemann et al. 2017; Sitthiyot and Holasut 2020). To further explore green “inequity” considering that each neighborhood has different circumstances of demands, we drew the Lorenz lines of green supplies over the ranking of human demands (Fig. 2). Such Lorenz lines could present how UGS disproportionately concentrates on some neighborhoods over others.
We further conducted a cluster analysis to identify overall matches and mismatches between green supply and human demand. Before clustering, metrics of supply and demand were rescaled into values of 0–1 based on min-max normalization within a city. Overall supply and demand in each neighborhood were then calculated by the equally weighted mean of rescaled supply and demand metrics, respectively. Using the overall supply and demand, we applied two algorithms to obtain four clusters: low supply-low demand, low supply-high demand, high supply-low demand, and high supply-high demand (Fig. 6). The first algorithm is that we graded the levels of high and low by the median value of overall supply/demand (Lin et al. 2022). Since we have properly defined these four clusters to guide urban green design, unsupervised algorithms such as k-means were not adopted to segment clusters of similarity. The second algorithm is spatial analysis - bivariate Local Indicators of Spatial Association (LISA), which was applied to map spatial clusters of extremes of green supply and surrounding human demand. Bivariate LISA analysis maps how the value of one variable is surrounded by the values of a second variable and determines the statistical significance for each cluster (Anselin 1995; Tate et al. 2021). In this case, the identified high-high clusters are where high green supply is surrounded by high human demand with significantly positive spatial autocorrelation, and vice versa.
Models for supply-demand relations
The ordinary least square (OLS) model and spatial Durbin model (SDM) were adopted to analyze the current relationship between human demand and green supply. Since LST and air pollution are inappropriate to explain the distribution of green supply, social demands were set as independent variables and green supplies were dependent variables. To avoid the collinearity between the total population and age groups, we transformed the density of the elderly and children to the percentage of the elderly and children (\(EC\)). We also performed logarithmic processing on population density (\(PD\)) and income per capita (\(IC\)) to reduce estimation errors. The specification of the OLS model is:
where Y is the green supply metric, \(\beta\) is the vector of coefficients, and \(\epsilon\) is the random error. Since urban trees and parks are normally concentrated in several places (Tian et al. 2014; Haaland and van den Bosch 2015), a neighborhood with large tree cover or park areas is likely to have similarly green neighbors. We demonstrated such spatial autocorrelation in green supplies by Moran’s I test. Given that green space in a certain region may also be affected by the population in surrounding regions, we included the spatial lag of the dependent variable (WY) and population density (WPD) to yield the SDM (Durbin 1960):
where \(\rho\) is the spatial lag parameter of the dependent variable, and W is the spatial weight matrix computed by using the Queen contiguity which defines neighbors by a common boundary (Bivand 2008). The Lagrange Multiplier (LM) test was applied to choose the type of spatial model between the spatial lag model and the spatial error model and whether the spatial model is better than the OLS model (Anselin 1988; Park et al. 2021).
Results
Green inequality within cities
First, Gini coefficients show the inequality in green supplies and human demands across neighborhoods (Table 3). Among green supply metrics, amount metrics, especially the public park per capita, identify larger inequalities than the distance metric. Green inequality is greater in New York and Amsterdam than in Beijing. The Gini coefficients of the public park per capita in New York and Amsterdam are remarkably close to 1 (complete inequality). Inequality in human demand is generally smaller than that in green supply. Among human demand metrics, environmental livability demands are distributed more evenly than social demands, of which Gini coefficients are below 0.1.
Second, Lorenz curves illustrate how green supplies are disproportionately distributed in neighborhoods with uneven human demands (Fig. 2). At the first level (green amount-environmental livability demand), tree cover is more heavily allocated to lower-LST neighborhoods in Beijing (the Lorenz curve is fully located above the Gini perfect equality line) but to higher-PM2.5 neighborhoods in Amsterdam (below the Gini perfect equality line). At the second and third levels (green accessibility-social demands), lower-population-density blocks share more tree cover and lower-income blocks share less public park area per capita in New York. Neighborhoods with more children and elderly people share more distance to the nearest large park in Beijing but share less distance in Amsterdam. Notably, the jagged lines between public park area per capita and income are related to the great inequality in public park area per capita across neighborhoods (Table 3). Overall, Amsterdam performs better than the other cities, according to its close-to-Gini perfect equality lines and uneven green supply but generally follows the ranking of demands.
Demand-based inequity
The spatial inequity in urban green is revealed by comparing distributions of green supply to human demand, as well as integrated supply-demand clusters (Figs. 3, 4, 5 and 6). Overall, high peaks of green supplies are mostly concentrated on the fringe of the three cities, while the distribution of high peaks of human demands differs across cities.
In New York, green supplies are concentrated in southwestern and northern parts, but the demands are relatively high in central parts according to the high-density population, the elderly and children (Fig. 3), resulting in overall low supply-high demand clusters in central regions - i.e. Brooklyn district (Fig. 6). The opposite mismatching type (high supply-low demand cluster) is mainly observed in the fringe, especially in the Staten Island reflected as high tree cover but low environmental pressures as well as short distance to large parks but low population density. The LISA map also depicts a large area of high-low clusters in Staten Island representing high supply with surrounding low demand (Fig. 6). Across three supply-demand levels, notable mismatches are observed between tree cover and urban heat and between distance to the nearest large park and elderly/children density.
In Amsterdam, green supplies are low in the city center, where high peaks of demands in LST, population density, and elderly/children density are located (Fig. 4), resulting in the overall low supply-high demand cluster scattering around the central area (Fig. 6). The opposite mismatching type (high supply-low demand cluster) is mainly observed in the fringe, especially in southwestern areas reflected as high tree cover but low LST as well as short distance to large parks but low population density. The LISA map presents a large area of low-low clusters in the northeast representing a spatial match with low supply and surrounding low demand (Fig. 6). Across three supply-demand levels, the notable mismatches are similar to those in New York. Interestingly, the demand for improving air quality increases from the city center to the edge, which generally matches the corresponding green supply (tree cover) pattern.
In Beijing, overall low supply-high demand clusters cover a large area in both the north and south (Fig. 6), which are reflected at different supply-demand levels (Fig. 5). The low-high mismatches in the north are mainly reflected as low public park per capita but high population density, while mismatches in the south are mainly reflected as low tree coverage but high environmental pressures. Matches occur in the southeastern part where the large public park area per capita meets the high demand of low-income neighborhoods. The LISA clusters are mostly not significant indicating weak spatial autocorrelation between green supply and human demand (Fig. 6).
Relations between green supply and human demand
In addition to supply-demand maps, OLS and SDM models further unraveled diverse associations between green supply and social demand in cities (Table 4). Our models are quite robust since OLS and SDM estimates are generally similar. Overall, Amsterdam performs well since most social demands are positively associated with green supplies (p < 0.05), while most associations in Beijing and New York are negative (p < 0.05) or insignificant. Specifically, higher population density is related to less tree cover in New York and Beijing (p < 0.05). Neighborhoods with a higher proportion of the elderly and children tend to have a shorter distance to the nearest large park representing easier access to high-quality parks in New York and Amsterdam (p < 0.01). Notably, income per capita is not significantly connected with green supplies, except for the negative connections with distance to the nearest large park in New York (p < 0.01) which means that higher social demands of low-income neighborhoods tend to have less supply of high-quality parks.
When we look at spatial lag factors in SDM, most spatial lag effects of dependent variables (lag.Y) are significantly positive, suggesting that green supply in a neighborhood is significantly related to green supplies in its bordering neighborhoods. Population density also has spillover effects (lag.Population density) on tree cover and public park area per capita, which means that more green supplies in a certain neighborhood are associated with higher population density in bordering neighborhoods.
Discussion
Gini perfect equality or demand-oriented equity
Urban green inequality in study cities has been observed in previous research in terms of green coverage and urban park accessibility, which is consistent with our findings (Feng et al. 2019; Paulin et al. 2020; Pipitone and Jović 2021). We integrated tree cover, park accessibility, and green quality to assess multi-level UGS inequality. In contrast to a previous finding that UGS inequality is mainly reflected by coverage instead of distance in Germany (Wüstemann et al. 2017), our results show comparable inequalities in tree coverage and distance (Table 3). In addition to distinct cities, varying measurements of distance can be another reason: the previous study used the common distance metric – distance to any nearest UGS, while small and informal UGSs are rather frequent but have fewer benefits than large parks. To address such shortages, we considered green quality (i.e. park size) and obtained an improved distance metric – distance to the nearest large park (> 20 ha), which identified meaningful inequalities in three cities. The measurement strategy of green accessibility can therefore have a significant impact on the findings and makes it difficult to compare our findings to others.
City performance varies by the objective of environmental justice – “Gini perfect equality” or “demand-oriented equity”. According to the concept of perfect equality implied in the Gini coefficient, Beijing performs the best among study cities as its Gini coefficients of green supplies are the smallest indicating the most evenly distributed UGS (Table 3). However, when we involve the demand to get the concept of equity, Amsterdam has the best performance since UGS distribution follows the ranking of human demands (Fig. 2). For instance, neighborhoods with higher PM2.5 pressures share more trees and those with more elderly people and children share more nearby large parks in Amsterdam (Fig. 2). Those findings suggest that evenly distributed UGS is not necessarily concurrent with demand-oriented equity and vice versa – highlighting the necessity to concern demand variation towards environmental justice (Chen and Huang 2021).
Our study contributes to the scientific understanding of urban environmental justice, by highlighting the concept of demand-oriented equity. We should reflect on whether the Gini perfect equality or demand-oriented equity is the ideal goal for environmental justice. On the one hand, Gini perfect equality represents evenly distributed UGS across neighborhoods, which is the definition of environmental justice in most previous research (Wolch et al. 2014; Xiao et al. 2017). On the other hand, as UGS is serving people, UGS should be more allocated to neighborhoods with higher demand. This UGS inequality is consistent with the demand variation which can be regarded as demand-oriented equity. Nevertheless, demand-oriented equity has difficulty determining how much UGS should be prioritized in high-demand neighborhoods. Moreover, the distribution of human demands changes over time, and thus investigating how to achieve demand-oriented equity based on circular statistics is warranted (Li et al. 2022).
Supply-demand mismatches: calling for UGS allocation optimization
In addition to measurements of inequality and inequity, we presented spatial patterns of supply-demand mismatches, which are generally consistent with previous studies. Specifically, in New York, we found mismatches between tree cover and environmental pressures, and a recent study in US cities also revealed that tree planting is not prioritized by urban heat distribution (Zhou et al. 2021). In Amsterdam, previous research detected an increasing gradient of tree coverage from the center to the edge but an opposite gradient of population density and cooling demands (Paulin et al. 2020; Schrammeijer et al. 2022), which are also shown in our maps (Fig. 4). In Beijing, an earlier study observed a mismatch between the spatial distribution of urban parks and population in 2017, particularly for elderly residents (Feng et al. 2019), and our study found a similar mismatch for the year 2020 and more specific mismatch between distance to the nearest large park and vulnerable population (Fig. 5). Compared to existing findings, our results reflected more dimensions and spatial detail of up-to-date mismatches between green supplies and human demands.
Similar to most studies on supply-demand mismatches (Luo and Li 2021; Lin et al. 2022), we identified mismatches as low supply-high demand and high supply-low demand clusters. Beyond that, we distinguished multi-level mismatches that can separate the deficit for environmental livability demands and social demands. For example, Fig. 6 presents a large area of low supply-high demand mismatches in Beijing, and Fig. 5 further reveals that such mismatches are mainly caused by low park accessibility-high social demand in the North but by low tree cover-high environmental livability demand in the South. Accordingly, UGS development in Beijing should focus on developing public parks in the North and planting trees in the South, rather than making undifferentiated plans.
Relationships between green supplies and human demands vary by city and metric (Table 4). Overall, the distribution of green supplies meets the distribution of social demands in Amsterdam with positive relationships, while they are not positively related in New York and Beijing. Our models indicate that higher-demand neighborhoods with greater population density tend to have lower green supplies with less tree cover and longer distances to the nearest large park in New York and Beijing. Such mismatches can be explained by that high population density is usually related to high built-up density (Li et al. 2019) and are consistent with the previous studies (Feng et al. 2019; Chen and Huang 2021). Beyond the consistent findings, our spatial lag factors imply that green supply in a certain neighborhood tends to meet the demand of population density in its bordering neighborhoods. Moreover, existing studies have shown that low-income people have less access to green space than other people in our study cities (Wolch et al. 2014; Hughey et al. 2016; de Vries et al. 2020; Wu et al. 2020). Our models unravel more specific mechanisms that the low-income communities tend to have shorter distances to the nearest large park in New York and Amsterdam but have no significant association with tree cover or public park area per capita. In addition, our spatial models reveal that green supplies in a neighborhood are significantly influenced by green supplies in its bordering neighborhoods, which can be explained by that the distribution of parks is primarily determined by geographical locations and the development history (Tian et al. 2014; Haaland and van den Bosch 2015).
Our results on supply-demand mismatches at the neighborhood level can help guide local green planning toward meeting diverse human demands (Liu et al. 2020; Menconi et al. 2021). The low supply-high demand clusters we identified should be the focus of future urban greening projects. Negative relationships between green supply and human demand would cause a series of inequity issues, particularly warranting intervention. As such, our results provide concrete green advice for study cities: Green plans in New York (e.g. Million-Trees NYC) can plant more trees in the blocks with high population density, and expand the public parks around the low-income blocks. The municipality of Amsterdam should particularly develop large public parks around low-income neighborhoods to more efficiently achieve its target of increasing the public park area per capita and UGS quality. With the largest supply-demand mismatches, Beijing officials are supposed to make extra efforts to allocate more trees and new public parks to neighborhoods with higher levels of temperature, PM2.5, and vulnerable people. How to better meet human demand with limited UGS is an important challenge for environmental equity and sustainable development in these capital cities.
Uncertainties and limitations
Uncertainties exist in the data source which might influence the results. Given that our study involves a comparative analysis of three international cities, it is imperative to adopt globally remote sensing products with uniform definitions and methodology for estimating tree cover, LST, and PM2.5 concentration. However, the LST and PM2.5 datasets at 1 km resolution are relatively coarse to capture neighborhood-level environments and can omit localized or fine-grain patterns, which might result in underestimated inequalities across neighborhoods. Due to this limitation, we did not involve LST and PM2.5 concentration in the statistical models but only depicted spatial patterns. The user’s accuracy of the tree cover in the World Cover 2020 product is around 80%, also affecting the precise assessment of the distribution of green spaces (Tsendbazar et al. 2021). Therefore, we recommend the use of globally consistent products based on high-resolution data for future investigations into urban green spaces and environments in international cities. For instance, further research can produce high-resolution LST products from Landsat 8 and Sentinel-2 imagery to obtain more accurate LST at the neighborhood level.
Furthermore, the statistical data we gathered to assess social demands exhibit variations among cities regarding which data are publicly available, at what administrative resolution, and how up-to-date the datasets are. For instance, while Amsterdam and New York provide neighborhood-level income data for the year 2018, Beijing only offers district-level income information, which was then downscaled using rent distribution for every neighborhood. These disparities in data sources may influence the absolute value of demands, but we mainly look at the ranking of demands where the influence is diminished. Our regression models between green supply and social demand were built for each city, and thus the regression coefficients are comparable between cities.
Our study design and methodology also have some limitations. First, different divisions of neighborhoods and income concepts between cities can lead to differences in the Gini coefficient we used to measure inequality. Second, we directly compared green supply with human demand without converting them into the same evaluation system, i.e., ecosystem services (Lin et al. 2022; Liu et al. 2020). Environmental services provided by UGS can be calculated by cooling rates and PM2.5 removal amount per day of trees, woodland, shrubs, and grass, and then linked to environmental livability demands (Lin et al. 2022). Third, our assessment of ‘human demand’ is not sufficient. We assessed the potential demands without considering the actual use of parks. Given park space in low-income communities may be perceived as unsafe for children (Wolch et al. 2014). Although previous research in the Global North has revealed that UGS could decrease the number of total crimes and gun assaults by relieving stress, studies in the Global South found tree cover associated with higher sexual crimes (Hunter et al. 2019; Venter et al. 2022). Further research is needed to explore causal mechanisms behind crime-green space associations and complement the demand assessment in terms of safety. We also used environmental pressures such as LST to represent the demand but did not concern the various heat tolerance of residents. Future research can improve the assessment of human demands for UGS through more indicators and surveys, such as conducting surveys and interviews on human preferences. Fourth, we established a singular set of regression coefficients to obtain a general relationship between green supply and human demand within each city, disregarding spatially varying relationships. Although spatially varying relationships can be analyzed by models such as the Geographically Weighted Regression (GWR), it requires a unified distance or range used for spatial correlation which is difficult to set due to varying neighborhood size and density in three international cities. Nevertheless, we recommend future research to apply GWR on a large scale, creating a map of the green supply-demand relation.
Conclusions
Using a unique green supply-human demand concept at the neighborhood level, we proposed “demand-oriented equity” as a sustainable goal of urban green justice, rather than the common “Gini perfect equality”. This study has measured multi-level green inequality and analyzed supply-demand mismatches based on spatial clusters and models.
Overall, the public park per capita reflects the largest inequality in green supplies, and the mismatching neighborhoods with low supply and high demand are mainly scattered around urban centers. Among study cities, New York is the city with the greatest inequality in green supplies, and mismatches between tree cover and population density and between the access to public parks and low-income households are remarkable. In Beijing, the green supplies are more evenly distributed compared to other cities, but the distribution of green supplies does not match that of environmental livability demands and social demands, particularly of the elderly and children. Amsterdam performs well in demand-oriented equity with green supplies generally consistent with the demand distribution, while the green supply around low-income communities can be enhanced.
Our findings can guide urban greening projects by pinpointing areas with low supply and high demand, facilitating the development of neighborhood-specific strategies for landscape sustainability. A key contribution of this study lies in creating a pyramid of green supply indicators corresponding to human demand indicators. This framework is applicable across diverse cities and analysis scales. Future research can improve the assessment of human demand and debate how to achieve demand-oriented equity of urban green in cities, ultimately fostering urban sustainability.
References
Alves S, Aspinall PA, Ward Thompson C, Sugiyama T, Brice R, Vickers A (2008) Preferences of older people for environmental attributes of local parks. Facilities. https://doi.org/10.1108/02632770810895705
Anselin L (1988) Lagrange multiplier test diagnostics for spatial dependence and spatial heterogeneity. Geogr. Anal. 20(1):1–17. https://doi.org/10.1111/j.1538-4632.1988.tb00159.x
Anselin L (1995) Local Indicators of Spatial Association—LISA. Geogr Anal. 27(2):93–115. https://doi.org/10.1111/j.1538-4632.1995.tb00338.x
Baró F, Haase D, Gómez-Baggethun E, Frantzeskaki N (2015) Mismatches between ecosystem services supply and demand in urban areas: a quantitative assessment in five European cities. Ecol Indic 55:146–158
Basu S, Nagendra H (2021) Perceptions of park visitors on access to urban parks and benefits of green spaces. Urban For Urban Green. 57:126959
Bedimo-Rung AL, Mowen AJ, Cohen DA (2005) The significance of parks to physical activity and public health: a conceptual model. Am J Prev Med 28:159–168. https://doi.org/10.1016/j.amepre.2004.10.024
Bivand R (2008) Creating Neighbours. Applied Spatial Data Analysis with R 239–251
Boone CG, Buckley GL, Grove JM, Sister C (2021) Parks and people: an environmental justice inquiry in Baltimore, Maryland. Ann. Assoc. Am. Geogr. 99(4):767–787
Bratman GN, Anderson CB, Berman MG, Cochran B, De Vries S, Flanders J, Folke C, Frumkin H, Gross JJ, Hartig T, Kahn Jr PH (2019) Nature and mental health: an ecosystem service perspective. Sci. Adv. 5(7):eaax0903
Burkhard B, Kroll F, Nedkov S, Müller F (2012) Mapping ecosystem service supply, demand and budgets. Ecol Indic 21:17–29. https://doi.org/10.1016/j.ecolind.2011.06.019
CBS (2021) Wijk- en buurtkaart 2021. https://www.cbs.nl/nl-nl/dossier/nederland-regionaal/geografische-data/wijk-en-buurtkaart-2021
Chen Z, Huang G (2021) Greenspace to meet people’s demand: a case study of beijing in 2005 and 2015. Remote Sens. (Basel) 13(21):4310. https://doi.org/10.3390/rs13214310
Chen B, Wu S, Song Y, Webster C, Xu B, Gong P (2022) Global patterns in human exposure to greenspace: contrasting inequality between cities of the South and North. Nat Commun. 13:1–16. https://doi.org/10.1038/s41467-022-32258-4
de Vries S, Buijs AE, Snep RPH (2020) Environmental justice in the Netherlands: Presence and quality of greenspace differ by socioeconomic status of neighbourhoods. Sustain (Switzerland). https://doi.org/10.3390/SU12155889
Durbin J (1960) Estimation of parameters in time-series regression models. J R. Stat. Soc. Series B: Stat Methodol. 22(1):139–153. https://doi.org/10.1111/j.2517-6161.1960.tb00361.x
Feng S, Chen L, Sun R, Feng Z, Li J, Khan MS, Jing Y (2019) The distribution and accessibility of urban parks in Beijing, China: implications of social equity. Int J Environ Res Public Health 16(24):4894. https://doi.org/10.3390/ijerph16244894
Fletcher DH, Likongwe PJ, Chiotha SS, Nduwayezu G, Mallick D, Md NU, Rahman A, Golovátina-Mora P, Lotero L, Bricker S, Tsirizeni M (2021) Using demand mapping to assess the benefits of urban green and blue space in cities from four continents. Sci Total Environ. https://doi.org/10.1016/j.scitotenv.2021.147238. 785:
Florczyk A, Corbane C, Schiavina M, Pesaresi M, Maffenini L, Melchiorri M, Politis P, Sabo F, Freire S, Ehrlich D, Kemper T, Tommasi P, Airaghi D, Zanchetta L (2019) GHS Urban Centre Database 2015, multitemporal and multidimensional attributes, R2019A. European Commission, Joint Research Centre (JRC). https://data.jrc.ec.europa.eu/dataset/53473144-b88c-44bc-b4a3-4583ed1f547e
Garrison JD (2019) Seeing the park for the trees: New York’s million trees campaign vs. the deep roots of environmental inequality. Environ Plan B Urban Anal City Sci. https://doi.org/10.1177/2399808317737071
Gemeente A (2018) Amsterdamse referentienorm
Giuliani G, Petri E, Interwies E, Vysna V, Guigoz Y, Ray N, Dickie I (2021) Modelling accessibility to urban green areas using open earth observations data: a novel approach to support the urban SDG in four European cities. Remote Sens. (Basel) 13(3):422. https://doi.org/10.3390/rs13030422
Haaland C, van den Bosch CK (2015) Challenges and strategies for urban green-space planning in cities undergoing densification: a review. Urban For Urban Green
Han Y, He J, Liu D, Zhao H, Huang J (2022) Inequality in urban green provision: a comparative study of large cities throughout the world. Sustain Cities Soc. 89:104229
Hoover F-A, Lim TC (2021) Examining privilege and power in US urban parks and open space during the double crises of antiblack racism and COVID-19. Socioeco Prac Res 3:55–70. https://doi.org/10.1007/s42532-020-00070-3
Hsu A, Sheriff G, Chakraborty T, Manya D (2021) Disproportionate exposure to urban heat island intensity across major US cities. Nat Commun 12:1–11. https://doi.org/10.1038/s41467-021-22799-5
Hu H, Wang L (2005) The gini coefficient’s application to general complex networks. Adv Complex Syst. https://doi.org/10.1142/S0219525905000385
Huang JH, Hipp JA, Marquet O, Alberico C, Fry D, Mazak E, Lovasi GS, Robinson WR, Floyd MF (2020) Neighborhood characteristics associated with park use and park-based physical activity among children in low-income diverse neighborhoods in New York City. Prev Med (Baltim) 131:105948
Hughey SM, Walsemann KM, Child S, Powers A, Reed JA, Kaczynski AT (2016) Using an environmental justice approach to examine the relationships between park availability and quality indicators, neighborhood disadvantage, and racial/ethnic composition. Landsc Urban Plan 148:159–169. https://doi.org/10.1016/j.landurbplan.2015.12.016
Hunter RF, Cleland C, Cleary A, Droomers M, Wheeler BW, Sinnett D, Nieuwenhuijsen MJ, Braubach M (2019) Environmental, health, wellbeing, social and equity effects of urban green space interventions: a meta-narrative evidence synthesis. Environ Int 130:104923. https://doi.org/10.1016/j.envint.2019.104923
Jing C, Zhou W, Qian Y, Yan J (2020) Mapping the urban population in residential neighborhoods by integrating remote sensing and crowdsourcing data. Remote Sens (Basel). https://doi.org/10.3390/rs12193235
Kemperman A, Timmermans H (2014) Green spaces in the direct living environment and social contacts of the aging population. Landsc Urban Plan 129:44–54. https://doi.org/10.1016/j.landurbplan.2014.05.003
Kim Y, Corley EA, Won Y, Kim J (2023) Green space access and visitation disparities in the phoenix metropolitan area. Landsc Urban Plan 237:104805. https://doi.org/10.1016/j.landurbplan.2023.104805
Kmail AB, Onyango V (2020) A GIS-based assessment of green space accessibility: case study of Dundee. Appl Geomatics. https://doi.org/10.1007/s12518-020-00314-7
Knobel P, Dadvand P, Alonso L, Costa L, Español M, Maneja R (2021) Development of the urban green space quality assessment tool (RECITAL). Urban For Urban Green 57:126895. https://doi.org/10.1016/j.ufug.2020.126895
Li X, Zhou Y, Eom J, Yu S, Asrar GR (2019) Projecting Global Urban Area Growth through 2100 based on historical Time Series Data and Future Shared Socioeconomic pathways. Earths Future 7:351–362. https://doi.org/10.1029/2019EF001152
Li Q, Thapa S, Hu X, Luo Z, Gibson DJ (2022) The relationship between Urban Green Space and Urban Expansion based on gravity methods. Sustainability 14(9):5396. https://doi.org/10.3390/su14095396
Lin Y, Zhang M, Gan M, Huang L, Zhu C, Zheng Q, You S, Ye Z, Shahtahmassebi A, Li Y, Deng J (2022) Fine identification of the supply–demand mismatches and matches of urban green space ecosystem services with a spatial filtering tool. J Clean Prod 336:130404. https://doi.org/10.1016/j.jclepro.2022.130404
Liu H, Remme RP, Hamel P, Nong H, Ren H (2020) Supply and demand assessment of urban recreation service and its implication for greenspace planning-A case study on Guangzhou. Landsc Urban Plan 203:103898. https://doi.org/10.1016/J.LANDURBPLAN.2020.103898
Liu K, Li X, Wang S, Gao X (2022) Assessing the effects of urban green landscape on urban thermal environment dynamic in a semiarid city by integrated use of airborne data, satellite imagery and land surface model. Int J Appl Earth Obs Geoinf 107:102674. https://doi.org/10.1016/J.JAG.2021.102674
Luederitz C, Brink E, Gralla F, Hermelingmeier V, Meyer M, Niven L, Panzer L, Partelow S, Rau AL, Sasaki R, Abson DJ (2015) A review of urban ecosystem services: six key challenges for future research. Ecosyst Serv 14:98–112. https://doi.org/10.1016/j.ecoser.2015.05.001
Luo C, Li X (2021) Assessment of ecosystem service supply, demand, and balance of urban green spaces in a typical mountainous city: a case study on chongqing, China. Int J Environ Res Public Health. https://doi.org/10.3390/ijerph182011002
Meerow S, Newell JP (2017) Spatial planning for multifunctional green infrastructure: growing resilience in Detroit. Landsc Urban Plan 159:62–75. https://doi.org/10.1016/j.landurbplan.2016.10.005
Menconi ME, Sipone A, Grohmann D (2021) Complex systems thinking approach to urban greenery to provide community-tailored solutions and enhance the provision of cultural ecosystem services. Sustain (Switzerland) 13(21):11787. https://doi.org/10.3390/su132111787
Meng Y, Xing H, Yuan Y, Wong MS, Fan K (2020) Sensing urban poverty: from the perspective of human perception-based greenery and open-space landscapes. Comput Environ Urban Syst 84:101544. https://doi.org/10.1016/j.compenvurbsys.2020.101544
Ministry of Housing and Urban-Rural Development (2019) Urban Green Space Planning standards. GB/T 51346 – 2019
Natural England (2010) ‘Nature Nearby’ Accessible Natural Greenspace Guidance. https://www.gov.uk/government/organisations/natural--england
Nelson JR, Grubesic TH, Miller JA, Chamberlain AW (2021) The equity of tree distribution in the most ruthlessly hot city in the United States: Phoenix. Arizona. Urban For Urban Green 59:127016
NHGIS (2020) U.S. CENSUS DATA. https://www.nhgis.org/
Oke TR (1982) The energetic basis of the urban heat island. Q J R Meteorol Soc. https://doi.org/10.1002/qj.49710845502
Park Y, Guldmann JM, Liu D (2021) Impacts of tree and building shades on the urban heat island: combining remote sensing, 3D digital city and spatial regression approaches. Comput Environ Urban Syst 88:101655. https://doi.org/10.1016/j.compenvurbsys.2021.101655
Paulin MJ, Remme RP, de Nijs T, Rutgers M, Koopman KR, de Knegt B, van der Hoek DC, Breure AM (2020) Application of the Natural Capital Model to assess changes in ecosystem services from changes in green infrastructure in Amsterdam. Ecosyst Serv 43:101114. https://doi.org/10.1016/j.ecoser.2020.101114
Pham TTH, Apparicio P, Séguin AM, Landry S, Gagnon M (2012) Spatial distribution of vegetation in Montreal: an uneven distribution or environmental inequity? Landsc Urban Plan 107:214–224. https://doi.org/10.1016/j.landurbplan.2012.06.002
Pipitone JM, Jović S (2021) Urban green equity and COVID-19: Effects on park use and sense of belonging in New York City. Urban For Urban Green 65:127338
Remme RP, Frumkin H, Guerry AD, King AC, Mandle L, Sarabu C, Bratman GN, Giles-Corti B, Hamel P, Han B, Hicks JL (2021) An ecosystem service perspective on urban nature, physical activity, and health. Proc Natl Acad Sci U S A 118:1–10. https://doi.org/10.1073/PNAS.2018472118
Sallis JF, Floyd MF, Rodríguez DA, Saelens BE (2012) Role of built environments in physical activity, obesity, and cardiovascular disease. Circulation 125:729–737. https://doi.org/10.1161/CIRCULATIONAHA.110.969022
Salthammer T, Uhde E, Schripp T, Schieweck A, Morawska L, Mazaheri M, Clifford S, He C, Buonanno G, Querol X, Viana M (2016) Children’s well-being at schools: impact of climatic conditions and air pollution. Environ Int. 94:196–210
Schipperijn J, Cerin E, Adams MA, Reis R, Smith G, Cain K, Christiansen LB, Van Dyck D, Gidlow C, Frank LD, Mitáš J (2017) Access to parks and physical activity: an eight country comparison. Urban Urban Green 27:253–263. https://doi.org/10.1016/j.ufug.2017.08.010
Schrammeijer EA, Malek Verburg PH (2022) Mapping demand and supply of functional niches of urban green space. Ecol. Indic. 140:109031
Schwarz K, Berland A, Herrmann DL (2018) Green, but not just? Rethinking environmental justice indicators in shrinking cities. Sustain Cities Soc. https://doi.org/10.1016/j.scs.2018.06.026
Sera F, Armstrong B, Tobias A, Vicedo-Cabrera AM, Åström C, Bell ML, Chen BY, de Sousa Zanotti Stagliorio Coelho M, Matus Correa P, Cruz JC, Dang TN (2019) How urban characteristics affect vulnerability to heat and cold: a multi-country analysis. Int J Epidemiol. https://doi.org/10.1093/ije/dyz008
Shanahan DF, Lin BB, Gaston KJ, Bush R, Fuller RA (2015) What is the role of trees and remnant vegetation in attracting people to urban parks? Landsc Ecol 30:153–165. https://doi.org/10.1007/s10980-014-0113-0
Sitthiyot T, Holasut K (2020) A simple method for measuring inequality. Palgrave Commun. https://doi.org/10.1057/s41599-020-0484-6
Spotswood EN, Benjamin M, Stoneburner L, Wheeler MM, Beller EE, Balk D, McPhearson T, Kuo M, McDonald RI (2021) Nature inequity and higher COVID-19 case rates in less-green neighbourhoods in the United States. Nat Sustain. https://doi.org/10.1038/s41893-021-00781-9
Statistics BMB (2020) of Beijing Statistical Yearbook. http://tjj.beijing.gov.cn/tjsj_31433/
Takano T, Nakamura K (1978) Watanabe M (2002) Urban residential environments and senior citizens’ longevity in megacity areas: The importance of walkable green spaces. J Epidemiol Community Health 56:913–918. https://doi.org/10.1136/jech.56.12.913
Tate E, Rahman MA, Emrich CT, Sampson CC (2021) Flood exposure and social vulnerability in the United States. Natural Hazards. 106:435–457. https://doi.org/10.1007/s11069-020-04470-2
Taylor AF, Kuo FE, Sullivan WC (2001) Coping with ADD. The surprising connection to green play settings. Environ Behav. https://doi.org/10.1177/00139160121972864
Tian Y, Jim CY, Wang H (2014) Assessing the landscape and ecological quality of urban green spaces in a compact city. Landsc Urban Plan 121:97–108. https://doi.org/10.1016/j.landurbplan.2013.10.001
Triguero-Mas M, Dadvand P, Cirach M, Martínez D, Medina A, Mompart A, Basagaña X, Gražulevičienė R, Nieuwenhuijsen MJ (2015) Natural outdoor environments and mental and physical health: relationships and mechanisms. Environ Int. https://doi.org/10.1016/j.envint.2015.01.012
Tsendbazar N, Li L, Koopman M, Carter S, Herold M, Georgieva I, Lesiv M (2021) WorldCover Product Validation Report
UN (2021) SDG indicator metadata: Indicator 11.7.1. 1–6
UN-Habitat (2020) Public Space Site-Specific Assessment. Guidelines to Achieve Quality Public Spaces at Neighbourhood Level
van Donkelaar A, Hammer MS, Bindle L, Brauer M, Brook JR, Garay MJ, Hsu NC, Kalashnikova OV, Kahn RA, Lee C, Levy RC (2021) Monthly global estimates of fine particulate matter and their uncertainty. Environ Sci Technol 55:15287–15300. https://doi.org/10.1021/acs.est.1c05309
van Leeuwen E, Nijkamp P, de Noronha Vaz T (2011) The multifunctional use of urban greenspace. Urban Agriculture: Diverse Activities and Benefits for City Society 20–25
Veerkamp CJ, Schipper AM, Hedlund K, Lazarova T, Nordin A, Hanson HI (2021) A review of studies assessing ecosystem services provided by urban green and blue infrastructure. Ecosyst Serv 52:101367. https://doi.org/10.1016/j.ecoser.2021.101367
Venter ZS, Shackleton C, Faull A, Lancaster L, Breetzke G, Edelstein I (2022) Is green space associated with reduced crime? A national-scale study from the Global South. Sci Total Environ. https://doi.org/10.1016/j.scitotenv.2022.154005
Wang S, Wang M, Liu Y (2021) Access to urban parks: comparing spatial accessibility measures using three GIS-based approaches. Comput Environ Urban Syst 90:101713. https://doi.org/10.1016/j.compenvurbsys.2021.101713
Wolch JR, Byrne J, Newell JP (2014) Urban green space, public health, and environmental justice: the challenge of making cities just green enough. Landsc Urban Plan 125:234–244. https://doi.org/10.1016/j.landurbplan.2014.01.017
Wu C, Xu Z (2010) Spatial distribution of the environmental resource consumption in the Heihe River Basin of Northwestern China. Reg Environ Change 10:55–63. https://doi.org/10.1007/s10113-009-0091-y
Wu J, He Q, Chen Y, Lin J, Wang S (2020) Dismantling the fence for social justice? Evidence based on the inequity of urban green space accessibility in the central urban area of Beijing. Environ Plan B Urban Anal City Sci 47:626–644. https://doi.org/10.1177/2399808318793139
Wüstemann H, Kalisch D, Kolbe J (2017) Access to urban green space and environmental inequalities in Germany. Landsc Urban Plan 164:124–131. https://doi.org/10.1016/j.landurbplan.2017.04.002
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. Landsc Urban Plan 157:383–393. https://doi.org/10.1016/j.landurbplan.2016.08.007
Xie Y, Zhou X (2014) Income inequality in today’s China. Proc Natl Acad Sci U S A. https://doi.org/10.1073/pnas.1403158111
Yan J, Lin L, Zhou W, Ma K, Pickett ST (2016) A novel approach for quantifying particulate matter distribution on leaf surface by combining SEM and object-based image analysis. Remote Sens Environ. https://doi.org/10.1016/j.rse.2015.11.033
Yan J, Zhou W, Zheng Z, Wang J, Tian Y (2020) Characterizing variations of greenspace landscapes in relation to neighborhood characteristics in urban residential area of Beijing, China. Landsc Ecol 35:203–222. https://doi.org/10.1007/s10980-019-00943-3
You H (2016) Characterizing the inequalities in urban public green space provision in Shenzhen, China. Habitat Int 56:176–180. https://doi.org/10.1016/j.habitatint.2016.05.006
Yuan Q, McIntyre N, Wu Y, Liu Y, Liu Y (2017) Towards greater socio-economic equality in allocation of wastewater discharge permits in China based on the weighted Gini coefficient. Resour Conserv Recycl. 127:196–205
Zhang D, Shen J, Liu P, Zhang Q, Sun F (2020) Use of fuzzy analytic hierarchy process and environmental gini coefficient for allocation of regional flood drainage rights. Int J Environ Res Public Health. https://doi.org/10.3390/ijerph17062063
Zhou W, Huang G, Pickett STA, Wang J, Cadenasso ML, McPhearson T, Grove JM, Wang J (2021) Urban tree canopy has greater cooling effects in socially vulnerable communities in the US. One Earth 4:1764–1775. https://doi.org/10.1016/j.oneear.2021.11.010
Acknowledgements
This research was funded by the National Natural Science Foundation of China (42271096), China Scholarship Council (NO. 201904910503), and Open Foundation of the State Key Laboratory of Urban and Regional Ecology of China (SKLURE2023-2-7). Special thanks to Dr. Jingli Yan for his contribution to the data and manuscript.
Author information
Authors and Affiliations
Contributions
Yunyu Tian: Conceptualization, Methodology, Software, Investigation, Visualization, Writing. Eveline van Leeuwen: Conceptualization, Writing—Review and Editing, Supervision. Nandin-erdene Tsendbazar: Conceptualization, Methodology, Writing—Review and Editing, Supervision. Chuanbao Jing: Data resources. Martin Herold: Conceptualization, Methodology, Writing—Review and Editing, Supervision.
Corresponding author
Ethics declarations
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Tian, Y., van Leeuwen, E., Tsendbazar, Ne. et al. Urban green inequality and its mismatches with human demand across neighborhoods in New York, Amsterdam, and Beijing. Landsc Ecol 39, 60 (2024). https://doi.org/10.1007/s10980-024-01874-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10980-024-01874-4