Urban green space exposure serves as a critical indicator for assessing the extent to which residents come into contact with and utilize green areas. It is shaped by a combination of socioeconomic and demographic factors (31). A broad consensus in the literature suggests that inequalities in green space exposure persist across income, education, and gender groups, reflecting deeper structural disparities in the distribution of green infrastructure.

At the individual level, higher levels of income, wealth, and educational attainment are associated with increased willingness and capacity to access quality environmental amenities. Consequently, middle- and high-income groups are more likely to prioritize green space access in their residential choices and daily mobility patterns (31–33). At the structural level, economic development in cities often leads to greater investment in public budgets, human resources, and land supply, all of which enhance the capacity for green space development and maintenance, thereby improving overall accessibility and usage opportunities for residents (34, 35). In terms of demographic differences, subjective perceptions of green space exposure vary across population subgroups. For instance, Sun et al. proposed a gender-based analytical framework to examine the relationship between green exposure and satisfaction. Their findings indicate that men’s satisfaction is more strongly influenced by green space accessibility indices, whereas women’s satisfaction is more affected by landscape features, suggesting divergent perceptual mechanisms between genders (36).

Green space quality is also recognized as a significant determinant of exposure. Early studies tended to operationalize quality through simplified indicators such as vegetation coverage (37, 38). Later, Gidlow et al. expanded the construct to include both natural elements (e.g., vegetation, water bodies) and built features (e.g., amenities, maintenance), emphasizing the holistic environmental condition of the green space (39). Building on this, Knobel et al. developed an 11-dimensional evaluation framework that integrates both subjective perceptions (e.g., aesthetic appeal, safety, signs of incivility) and objective attributes (e.g., land cover, biodiversity, accessibility), enabling a more comprehensive assessment of ecological function and usage potential (40).

Empirical studies highlight two primary mechanisms through which green space quality influences exposure: subjective and objective pathways. The subjective pathway emphasizes residents’ evaluations of aesthetic appeal, safety, and comfort, which in turn shape their willingness to use green spaces, thereby determining actual exposure levels (41). In contrast, the objective pathway focuses on how physical characteristics—such as canopy coverage, green space area, and amenity provision—enhance microclimatic conditions and reduce access costs, thereby increasing usage frequency. This effect is particularly pronounced in high-density urban environments (42).

Based on the above, the following hypotheses are proposed:

Perceived social benefits refer to individuals’ recognition of the positive impacts derived from social relationships and activities (43–45). Within the domain of urban green space and green infrastructure research, this concept is employed to assess residents’ evaluations of the social functions of green spaces, such as facilitating social interaction, strengthening community identity, alleviating psychological stress, and enhancing well-being and life satisfaction (46). This construct underscores the sociocultural value of green spaces and highlights the underlying mechanisms linking environmental attributes with interpersonal interactions (47–49).

Existing studies demonstrate that the quality of green environments significantly affects the perception of social benefits. On one hand, green space quality determines residents’ willingness to use these spaces, as well as the amount of time spent and frequency of activities undertaken there (50). High-quality urban green spaces typically offer safety, comfort, and aesthetic appeal, all of which enhance residents’ leisure experiences and promote physical and psychological well-being (51, 52). On the other hand, well-maintained green spaces foster place attachment and further promote mental relaxation and social interaction by offering visual enjoyment, recreational convenience, and spaces for social engagement (53–55). Thus, green space quality contributes to the enhancement of perceived social benefits through two pathways: by improving the environmental conditions for use, and by activating socio-psychological mechanisms that encourage engagement and emotional response.

Accordingly, the following hypothesis is proposed:

Urban green space justice primarily concerns the disparities among social groups in terms of accessibility, spatial distribution, and opportunities for green space use (56, 57). Early studies on environmental justice focused largely on the unequal distribution of environmental burdens such as pollution (58), whereas more recent research has expanded the scope to include spatial inequalities in green space allocation, usage, and related social processes within urban contexts (32, 59).

Empirical evidence reveals persistent deficiencies in the planning, management, and equitable distribution of green space exposure (59, 60). For example, Rao et al. used the Normalized Difference Vegetation Index (NDVI) and Green View Index (GVI) to quantify the extent and visibility of green space exposure at the activity-event level. Their study found that increased greenness enhances activity satisfaction, with a curvilinear (inverted U-shaped) relationship between green visibility and satisfaction. The results also indicated clear patterns of environmental injustice for marginalized groups (61). Similarly, Wang et al. constructed a measurement framework that captures both static and dynamic geographic contexts of green space exposure. Their findings reveal that disparities in accessibility to residential or workplace green spaces, as well as street-level greening, lead to unequal experiences of dynamic environmental exposure (62).

Beyond exposure, perceived social benefits also represent a key psychological mechanism influencing the perception of green space justice. Wolch et al. argued that individuals’ perceptions of the social functions provided by green spaces—such as safety and belonging—substantially shape access patterns among different groups, thereby affecting the realization of green space justice (57). Additional studies suggest that high-income groups are more likely to benefit from the social, recreational, and mental health advantages of high-quality green spaces, while low-income groups often face a “double disadvantage”: limited green space exposure and insufficient access to positive social experiences therein (50).

Based on this, the following hypotheses are proposed:

Subjective well-being refers to individuals’ self-evaluation of their overall life conditions, and is commonly used to assess life satisfaction, psychological health, and general welfare (14). In recent years, urban green space has been widely recognized as an essential strategy for enhancing residents’ well-being. Numerous studies have shown that green spaces not only improve residential environments and alleviate psychological stress, but also contribute to higher levels of overall well-being (6, 63, 64). However, the acquisition of well-being is not solely determined by the quantity or size of green spaces; fairness has increasingly emerged as a crucial dimension influencing residents’ welfare outcomes (14).

Specifically, whether green spaces are equitably distributed—and whether individuals perceive their access and use as respectful and equal—constitutes a fundamental precondition for deriving psychological benefits (64). Cross-national empirical studies have indicated that while the physical size of parks does not directly affect emotional or well-being outcomes, there are significant disparities in park use experiences across different income groups (65, 66). For instance, residents of low-income communities are more likely to encounter adverse environmental conditions, such as air pollution or crime, which may reduce the psychological benefits derived from green space use (67, 68). Furthermore, under rapid urbanization, Rui et al. conducted an integrated analysis using streetscape greenery (SG), neighborhood greenery (NG), and public green space (PGS), revealing the suppressive effects of green space inequality on the well-being of socially vulnerable populations. Their findings emphasize that green space must not only exist, but also be accessible, usable, and perceptible (14). This highlights that well-being is not solely associated with the presence or quality of green spaces, but is deeply embedded in the fairness of their distribution and the psychological responses it elicits.

Accordingly, the following hypothesis is proposed:

In summary, this study develops a comprehensive conceptual model to systematically examine the interrelationships among socio-demographic characteristics, green space quality, green space exposure, perceived social benefits, perceived green space justice, and subjective well-being (Figure 1).

Data collection for this study was conducted in Changsha, a major city in Hunan Province, China, located between 112°53′–113°10′E longitude and 27°51′–28°41′N latitude. Changsha lies in a humid subtropical monsoon climate zone, characterized by distinct seasons and synchrony between rainfall and temperature. With abundant green space resources and a strong ecological foundation, the city provides a suitable context for urban green space research.

This study, based on the Changsha City Master Plan (2003–2020) (69), employed a stratified random sampling approach and ultimately selected five parks as research samples. The sample encompasses major categories of urban green spaces, including comprehensive parks, specialized parks, linear parks, and community parks. Specifically, these include Houhu Park (comprehensive park), Martyrs Park and Yuelu Mountain Scenic Area (specialized parks), Jinjiang River Greenway (linear park), and Sunshine 100 Houhai Garden (community park). The site selection comprehensively considered factors such as service population size, functional representativeness, and spatial distribution diversity, thereby ensuring that the sample adequately reflects the characteristics of Changsha’s urban green space system. Figure 2 presents the spatial distribution of the study area, a statistical summary of park types, and photographs collected during on-site field surveys.

This study employed a mixed-methods approach combining offline field surveys and online questionnaires via the “Wenjuanxing” platform—a widely used digital survey tool in China. Surveys were conducted simultaneously across six districts in Changsha, targeting representative urban green parks. Selection criteria for survey sites included wide service coverage and the presence of typical green space functions, ensuring the inclusion of both well-equipped parks and areas with limited green infrastructure. This approach helped avoid biased perceptions of “green justice” that may arise from sampling only high-quality parks. The selected locations represented regions with both high green space density and insufficient green provision, thereby minimizing spatial sampling bias. Data were collected from September to December 2024.

To enhance the scientific validity and contextual appropriateness of the measurement tools, a pilot survey was conducted prior to full implementation. Items with unclear logic, ambiguous wording, or contextual irrelevance were removed. During the offline survey, trained researchers provided respondents with detailed explanations of technical terms and key concepts to ensure comprehension of the questionnaire items and intent (see Figure 1). For the online version, an introductory page with visual aids and glossary definitions was presented before the survey to facilitate accurate understanding and completion. Additionally, online responses containing abnormal geolocation data or inconsistencies with local behavioral patterns were excluded to ensure spatial representativeness and alignment with the offline survey areas. After rigorous data cleaning and screening, a total of 524 valid responses were obtained, providing an adequate sample size for subsequent empirical analyses.

The final questionnaire consisted of eight sections, covering both objective behaviors and subjective perceptions: Basic demographic and behavioral variables: Age, gender, marital status, household size, type of residence, and duration of residence, were used to analyze demographic influences on green space exposure. Socioeconomic status: Education level, occupation, household income, average monthly expenditure, household financial balance, and ownership of real estate or vehicles, to assess how economic background shapes exposure. Park visit behavior and spatial attributes: Frequently visited park types, the size of the nearest neighborhood park, and the main purposes for park visits, reflecting both behavioral patterns and spatial context.

To comprehensively evaluate residents’ subjective experiences and psychological responses to urban green spaces, the questionnaire also incorporated several validated scales: Perceived Green Space Quality Scale: Adapted from the RECITAL tool (40), it includes six dimensions—species diversity, plant color, seasonal features, architectural harmony, spatial layering, and aesthetic ambiance—to capture ecological and aesthetic appeal. Perceived Social Benefits Scale (70): Measures subjective evaluations of green space functions, including mental health support, recreation, social interaction, community cohesion, quality of life, and well-being. Subjective Well-Being Scale (71): Comprises five indicators—pleasure, relaxation, vitality, sleep recovery, and enjoyment of life—to explore links between green space exposure and well-being. Perceived Green Justice: Assessed using the core question, “Do you think urban park green spaces in Changsha are fairly distributed?” to measure perceptions of fairness in green resource allocation. The use of a single item to measure green justice in this study has been empirically justified by prior research (72, 73). Green Space Exposure Scale (74, 75): Defined as the ease of access to public green space, measured through two items: “Walking time from home to the nearest park” and “Frequency of visits to the nearest park per month.” Note that walking time is reverse-coded—a higher score indicates longer distance and thus lower exposure. The construct composed of two items is also permitted in measurement practice and is commonly found in research in related fields. All items were rated using a 5-point Likert scale (1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree; Table 1).

Table 2 presents the demographic profile of the 524 respondents. The age distribution is primarily concentrated in the 41–54 age group (29.0%), followed by 18–23 (18.9%), 31–40 (16.8%), and 55–64 years (16.4%). Respondents under 18 or over 65 years of age accounted for a relatively small proportion. In terms of gender, females constituted a slightly higher percentage (58.4%) than males (41.6%). Regarding marital status, the majority were married (60.7%), followed by unmarried individuals (34.0%) and divorced respondents (5.3%). Household size was predominantly composed of three-member (32.2%) and four-member (29.0%) families, with larger households (five or more members) making up 23.5%. In terms of housing type, group-owners (purchased housing with family or roommates) accounted for 35.9%, followed by solo homeowners (25.6%), and dormitory residents (22.5%). Regarding length of residence, over half of respondents (56.9%) had lived in their current residence for more than 5 years, while 18.1% reported living there for less than 1 year, indicating relatively stable residential patterns.

Table 3 outlines the socioeconomic characteristics of respondents. The sample shows diverse occupational backgrounds, with students comprising the largest group (26.0%), followed by farmers, workers, and employees (21.8%). Education levels were relatively high, with 33.8% holding a bachelor’s degree and 9.4% holding a master’s degree or above. Respondents with a high school or vocational diploma made up 25.4%. Regarding annual household income, the majority fell below CNY 100,000 (Chinese Yuan), with 29.6% earning less than CNY 50,000 and 26.3% earning between CNY 50,000 and 100,000. High-income households (above CNY 300,000) represented less than 10%. Monthly expenditures were mostly concentrated in the CNY 2,000–10,000 range, with 34.9% spending CNY 2,000–5,000 and 29.2% spending CNY 5,000–10,000. High-expenditure households (over CNY 15,000 per month) accounted for 9.9%. In terms of household financial status, 36.6% reported a “balanced income and expenditure,” while 14.7% experienced “occasional surplus,” and 10.9% reported “consistent surplus.” Conversely, 17.4% indicated it was “difficult to balance income and expenditure,” and 20.4% reported it was “occasionally difficult.” Most respondents owned one residential property (46.8%), while 30.3% owned two properties; only 11.1% were non-homeowners. Over half of the respondents reported owning one car, with 22.7% owning two cars; non-car owners accounted for 21.0%.

Figure 3 illustrates respondents’ park visitation behaviors and the spatial attributes of green space use. Figure 3a shows that community parks were the most frequently visited (47.1%), followed by comprehensive parks (30.9%) and linear parks (15.7%), with specialized parks accounting for only 6.3%. Figure 3b presents the size of the nearest parks: small parks (less than 1 hectare) were most common (50.0%), followed by medium-sized parks (1–10 hectares, 36.5%), while large and extra-large parks were less common (11.8 and 1.7%, respectively). Figure 3c displays park visit motivations. Over half of respondents (50.6%) visited parks primarily to enjoy nature and breathe fresh air. Others cited exercise (23.3%) and recreation (14.5%) as main reasons, while smaller percentages visited for family time (9.5%) and socializing with friends (2.1%).

In addition, respondents were divided into three groups based on their daily activity space range, defined by walking distance from home to the nearest park: Short-distance group (500 m–1 km): 158 individuals (30.2%). Medium-distance group (1–3 km): 215 individuals (41.0%). Long-distance group (≥3 km): 151 individuals (28.8%). This classification was used for subsequent Multi-Group SEM analyses.

This study, grounded in environmental justice theory and the activity space perspective, systematically investigates the formation mechanism of urban park green justice, with particular attention to differences in perceptions among groups with varying activity ranges. The findings largely support the research hypotheses: the environmental quality of urban park green spaces not only significantly affects residents’ green space exposure and perceived social benefits but also enhances their well-being through the perception of green justice. This suggests that high-quality green environments serve as a prerequisite for attracting public use and fostering positive social experiences, while perceived green justice plays a potential mediating role between environmental conditions and mental health outcomes.

Furthermore, the study finds that both annual household income and household income–expenditure balance have significant positive effects on green space exposure. This indicates that groups with stronger economic conditions are more likely to increase their opportunities for green space access. This result is consistent with the findings of Yang et al., who, based on spatial justice, social stratification theory, and Rawls’ theory of justice, argue that structural disparities exist in access to urban green spaces and the associated health benefits across socioeconomic groups. In this structure, high-SES groups enjoy significantly better opportunities for accessing green spaces and thus gain more health-related advantages (75). This inequality highlights the necessity for urban green space planning and allocation to place greater emphasis on the equity demands and practical accessibility of socially disadvantaged groups, thereby ensuring the fair effectiveness of environmental resources in promoting mental well-being.

Further multi-group analysis reveals that activity space plays a significant structural role in the mechanism of urban green justice. First, across all three distance-based groups, the path from “green space exposure → perceived green justice” remains significantly positive, indicating that higher levels of green space exposure are more likely to elicit stronger perceptions of fairness in the distribution of urban green spaces. In other words, greater exposure to green space can stimulate individuals’ subjective evaluation of justice.

Second, the paths from “green space quality to green space exposure” and “green space quality to perceived social benefits” are also significant in all groups. This suggests that regardless of differences in spatial location, high-quality green spaces consistently promote both the frequency of green space use and residents’ recognition of its social benefits. These three paths form a stable structural mechanism in the relationship between green justice and well-being, demonstrating strong cross-group consistency.

However, beyond these three core paths, the remaining path relationships do not reach statistical significance across the groups. This may reflect that under different activity space conditions, the relationships between certain psychological perceptions and well-being are not yet robust, possibly constrained by daily usage habits, levels of participation in social activities, and variations in park functions. For instance, within the short-distance group, community parks have become an integral part of daily life, and their social functions may have been “internalized,” thus no longer significantly influencing perceptions of justice. This suggests that traditional static models may underestimate the mechanisms of “functional saturation” and “diminishing marginal effects” in everyday green space use.

In the long-distance group, green space usage is more “occasional” and “purpose-driven,” typically involving infrequent “green tourism” focused on landscape experiences and recreational functions. This group pays less attention to fairness, resulting in a non-significant path from perceived green justice to well-being, reflecting a disconnect between a “consumption-oriented” usage model and a “belonging-oriented” path to Well-being.

In addition, within the mid-distance group, a significant positive path from “household annual income to green space exposure” is observed, which is not present in the other two groups. This indicates that the green space usage behavior of the mid-distance group is more strongly influenced by economic conditions. This may reflect that accessing green spaces at this range often requires more reliance on transportation or other activity resources, and economically better-off individuals are more capable of overcoming spatial barriers, thereby increasing their level of green space exposure.

At the practical level, the research findings provide strategic guidance for the differentiated design of urban green spaces. First, the descriptive analysis indicates that small community parks and “enjoying nature” are the main usage scenarios at present. Therefore, priority should be given to enhancing the ecological and landscape quality of community parks to improve residents’ sense of access to daily leisure and natural experiences. Second, the measurement model further demonstrates that landscape elements such as plant color and spatial layering are particularly critical to residents’ experiences. Future green space design should focus on optimizing these visual and spatial features, for example, by incorporating plant combinations with seasonal variation, enhancing visual coherence and spatial layering, and creating multifunctional, walkable spatial structures to increase visual appeal and spatial engagement.

Targeted strategies should be adopted for different activity range groups. For the short-distance group, which mainly focuses on spatial layering and aesthetic ambiance, it is recommended to enhance spatial diversity and layering in community parks, such as introducing undulating terrain, small water features, shaded walkways, and quiet resting areas, to create spaces suitable for exploration and contemplation that fulfill the needs for frequent, short-term, deep contact with nature. The mid-distance group pays more attention to aesthetic ambiance and species diversity, while also considering accessibility and economic affordability. For this group, regional parks should be enriched with diverse plant communities, ecological education facilities, and trail networks, while improving public transportation and pedestrian connections to enhance accessibility and reduce travel barriers. The long-distance group is particularly concerned with plant color and spatial layering and tends to use parks primarily for leisure and visual enjoyment. Therefore, large urban or suburban parks should be designed with large-scale, colorful, and open landscapes and equipped with facilities suitable for families and group activities to meet the needs of recreational users or short-distance tourists.

It is worth noting that this study categorizes activity range groups based on self-reported travel distance and usage behaviors of respondents, focusing on the concept of “functional distance in everyday use” rather than fixed divisions based solely on geometric distance. Thus, the same park may indeed present different spatial relationships for different groups. The core purpose of this research is to identify differences in individuals’ actual usage and perceptions in real-life contexts. In terms of design and policy, this concept of “relative distance” emphasizes the need to pay closer attention to residents’ actual travel patterns, activity trajectories, and psychological perceptions. Urban green space planning should comprehensively consider walkability, transportation accessibility, and functional zoning, promoting a transformation from single-use “neighborhood green spaces” to “multifunctional, multi-layered green space networks” that can meet the diverse needs of different population groups.

At the theoretical level, this study expands the perspective of environmental justice theory and green equity research. It is the first to systematically integrate the complex pathways between green space quality, green space exposure, perceived justice, perceived social benefits, and well-being, addressing the limitations of previous studies that focused on single pathways. Furthermore, by introducing groupings based on spatial distance, this study reveals group-based differences in residents’ perceptions and usage of green space, offering important theoretical innovations.

Despite providing systematic evidence, this study has several limitations. First, the data were primarily collected in Changsha, which introduces a certain degree of regional limitation. Future research could be extended to different cities or cross-national comparisons to test the broader applicability of the model. Second, the study is based on cross-sectional survey data. Although it reveals the relational pathways among urban green space quality, green space exposure, perceived justice, and subjective well-being, it does not allow for direct causal inference and cannot fully capture the dynamic evolution and spatiotemporal complexity of green equity. Future studies could incorporate longitudinal tracking, experimental designs, or natural experiments to further verify the causal direction and temporal features of these pathways. Third, although random sampling was employed at the selected sites, we acknowledge that some vulnerable groups may still be underrepresented. Future research is therefore recommended to adopt strategies such as stratified sampling to enhance the inclusiveness and representativeness of the sample. In addition, as green space exposure was measured using only two items, its convergent validity proved insufficient. Future studies may improve its reliability and validity by increasing the number of items, refining the scale structure, and expanding the sample size. Finally, this study relies on self-reported data and questionnaire surveys, mainly focusing on subjective perceptions and behavioral characteristics, while lacking detailed measurements of objective behavioral trajectories and dynamic spatial exposure. Future research could introduce emerging technologies such as mobile signaling data or spatiotemporal activity tracking, integrating dynamic activity spaces with multi-source data to reconstruct spatial justice models. This would allow for a more precise exploration of the complex mechanisms underlying green equity.