1. Introduction
Rapid urbanization tends to lead to excessive levels of suspended particulate matter (PMs) in the urban ambient air, along with frequent smog and haze. Long-term exposure to air pollution not only poses a great threat to humans’ health but also fauna that occur in cities, such as birds [
1,
2]. In addition to observations, an increasing number of state-of-the-art models are used for the description and evaluation of PMs such as the RAMS-CMAQ model and the coupling chemistry-meteorological model [
3,
4,
5].
Urban greenspaces are an integral part of green infrastructure. As a primary function, they provide residents with spaces for outdoor activities, which is closely associated with general health and well-being [
6,
7]. Meanwhile, they also play beneficial roles in improving urban air quality, alleviating urban heat islands and several other health and environmental aspects [
8,
9,
10,
11]. Studies have shown that exposure to greenspaces benefits a person’s respiratory and mental health [
12,
13,
14,
15,
16], while air pollution, social interactions, physical activities and other factors have an indirect role between greenspaces and their surrounding residents’ health and well-being [
17,
18]. However, as the greenspaces in central urban districts tend to border urban arteries and thoroughfares, and the number of motor vehicles in China grows rapidly, the potential beneficial effects of these spaces may be diminished. Against this backdrop, it has been noted that the harm caused by exposure to air pollution is particularly prominent among developed urban districts that are both population centers and emission hotspots [
19].
Towards urban greenspaces, many researchers have carried out studies using varying scales to explore the spatial-temporal distributions of particulate matter, their influencing factors and mechanisms of reduction [
20,
21]. Compared to macro- (like global, regional, national, urban scales) and micro-scale (like microstructures of leaves) studies, there appears to be a lack of small-scale (such as specific urban parks) studies on the influence of different greenspaces’ composition elements on their pedestrian-level PM concentrations and spatial-temporal distributions, which in fact are closely associated to the health and group activities of surrounding residents. Some of these distinguishing factors include landscape features, geographical location and meteorological factors. Due to dynamic meteorological conditions and the heterogeneity of different greenspaces’ ground surface properties, distribution of emission sources, topography and other human activities, PM concentrations vary greatly in time and space between and within greenspaces [
22,
23,
24]. The results are that the reduction effects on PM2.5 concentrations also vary widely at different greenspace locations [
25,
26,
27]. One comparison of several small greenspaces in urban parks, schools and residential areas showed that location and time have an extremely significant influence on the PM2.5 concentrations [
28]. Even within the same park, PM2.5 concentrations in small sites composed of different landscape elements vary significantly while also varying over time [
29]. Other researchers have studied different small-scale sites within a comprehensive park and found that the spatial distribution of PM2.5 concentrations varied minimally [
30]. Overall, existing research shows that PM2.5 concentrations have notable seasonal and daily variations, exhibiting significant differences between macro-regions; meanwhile, the research results of the effect on their spatial-temporal distributions from small-scale greenspaces which are most closely related to people’s daily lives appears non-uniform. This study seeks to better explore this non-uniformity to improve the benefits such greenspaces may provide at reducing PM2.5 concentrations.
The environmental factors affecting PM concentrations at a greenspace site include the percentage of green coverage, sky view factor (SVF), vegetation quantity, plant community structure, vegetation density, canopy density, configuration mode, total greenspace area, canopy volume coverage and plant diversity [
31,
32,
33,
34]. Urban districts with high levels of green coverage can help reduce PM concentrations, and have been shown to be typically negatively correlated with PM concentration levels [
35]. Water environments within greenspaces affect PM2.5 diffusion and deposition as well, which is related to changes in airflow, temperature and humidity caused by the evaporation and cooling of water [
33,
36]. In addition, the impact of vegetation on PM concentration also depends on surrounding traffic density and the relative location of emission sources [
37,
38].
Meteorological parameters, such as rainfall, snowfall, wind speed and direction, temperature and humidity, are another important set of factors affecting PM concentration and exert varying effects on PMs of different sizes [
24,
39,
40,
41,
42]. Within a certain range, PM concentration is negatively correlated with temperature [
43] and positively correlated with relative humidity [
44,
45]. Good natural ventilation and airflow can accelerate the local diffusion of pollutants and dilute their concentration, but high wind speeds can also raise dust from the ground, causing secondary air pollution, while wind direction may affect the location and origin of pollutant [
46,
47]. At the same time, thicker and larger plant canopies can reduce wind speed, typically resulting in an increase in PM concentration [
48]. Finally, evapotranspiration (the sum of evaporation from the land surface plus transpiration from plants) can change a site’s temperature and humidity and further affect the related PM deposition processes.
In summary, several influencing factors of urban greenspaces, including their composition and meteorological conditions, heavily influence their associated PM concentrations. These factors also cause certain differences in the spatial-temporal distributions of PM. Currently, many studies have primarily focused on PM2.5 concentrations by analyzing macro-scale regional variations or, on the micro-scale, by looking at the effects of individual plants. In contrast, few studies have focused on the linkages among meteorological factors, greenspace elements, pollution sources distribution and PM concentrations. Furthermore, there remains great uncertainty among research conclusions regarding the detailed influence of meteorological factors and landscape elements on the concentration of suspended PM.
At present, the global aging problem has drawn great deal of attention in most parts of the world [
18]. In 2020, the total number of elderly people aged 60 and above in China was 264 million, accounting for 18.7% of the total population. Nanjing had 1.77 million people aged 60 and above, accounting for 19.0% of the city’s total population [
49,
50]. China has, accordingly, proposed a national strategy for health aging. Moreover, China’s urbanization development is inevitably leading to an increasingly serious fragmentation of urban greenspace [
51]. Subsequently, fragmented small-scale greenspaces have become an important venue for outdoor activities of residents in high-density urban districts. According to official meteorological data, there are varying degrees of smog and haze among the greenspaces of Nanjing. Long-term exposure to smog and haze poses especially severe health risks to the elderly.
In light of the above discussion and literature review, and while urban land becomes increasingly limited, this paper has chosen to focus on the elderly populations who are the largest user group of greenspaces in China [
52,
53,
54]. With a comprehensive consideration of pollution sources, urban rivers, meteorological conditions and other environmental factors, in addition to consideration of distinctive greenspace characteristics and distance from main urban arteries, this study selected several typical small popular greenspaces dispersed across high-density central urban districts in Nanjing. Sociological research methods such as evidence-based measurements, questionnaires and interviews were adopted, while relevant statistical software was used for the corresponding quantitative analysis. In studying these urban greenspaces, a quantitative and comprehensive analysis was undertaken on the spatiotemporal distribution patterns of the sites’ PM2.5. In addition, further analysis was undertaken on the relationships between environmental factors (temperature and humidity, wind speed and direction), greenspace elements (green coverage, water coverage, airflow openness, SVF) and PM2.5 concentrations. This study also explores the relationships between exposure to urban greenspaces and air pollution on the physical and mental health of the elderly, and thereby hopes to help promote healthy aging and better understanding of how to create the most suitable landscapes for the elderly, while also providing reference for the optimizing of greenspace elements.
In particular, this paper aims to study the following questions:
What are the spatial-temporal distributions characteristics of PM2.5 concentrations in small-scale urban greenspaces in high-density central urban districts?
What are the influencing factors and mechanisms of PM2.5 concentrations in small-scale greenspaces?
How does information on the smog and haze in urban greenspaces affect the decisions of the elderly on going out and visiting such spaces?
What are the relevant implications of this study for urban planning and design?
4. Discussion
4.1. The Greater Impact of Time on PM2.5 Concentrations Compared to Space
This paper found that more than 70% of the selected greenspaces during the observed times had PM2.5 concentrations at moderate to severe levels, and that overall air quality was poor. Among the six sites, the average PM2.5 concentrations of Sites D1 and D2, which are both adjacent to the riverfront, ranked the highest, yet due to the water-loving nature of people, these sites also attracted more visitors who tended to participate in riverside activities. This highlights the conflict between people’s outdoor preferences and behaviors with the potential health risks caused by air quality in greenspaces. From a macro perspective, the differences in directly monitored PM2.5 concentrations across the six sites ranged a limited amount from 0.53% to 2.31%. However, from a micro perspective, this study’s further analysis on the difference values shows that PM2.5 concentrations are significantly correlated with the sites’ spatial location. Although this study’s six sites are dispersed and not located in the same urban park, they all are located near Nanjing’s city center and along the same urban artery. As such, they also are all similarly affected by the influence of traffic pollution. The sites’ diverse landscape features, and the varying behavioral preferences of visitors lead to different visitor densities at the sites, while residents’ activities, such as dancing, jogging and smoking, may cause secondary particular matter. The result is a complex set of dynamic impacts on the PM concentrations among different spaces.
In terms of the temporality trends, the directly measured PM2.5 concentrations of the six sites collectively fluctuated as a whole across the survey’s 23 observed days, and this is assumed to be related to the specific weather conditions and meteorological factors affecting all the sites similarly on any given day. However, according to the temporal distributions of PM2.5 concentrations across different time periods within individual days, concentrations in the greenspaces were found to be greatly affected by road traffic pollution sources, with concentrations in the afternoon being lower than the morning presumably due to the heavier vehicle emissions of the morning rush hour. It is thus recommended that residents of the surrounding areas should avoid going out and visiting the greenspaces during the morning peak hours when PM concentrations are higher, and rather go at 13:30–16:30 when the concentrations are lower. During the morning, Site D2′s 35-m hill appears to provide a shielding effect from the traffic pollution, and thus has lower PM2.5 concentrations compared to the other sites. As an example, this suggests that outdoor morning activities should be carried out in greenspaces that are far from urban arteries, blocked by hills and with relatively large rates of water coverage. In contrast, the PM2.5 concentrations in the afternoon at Sites D1 and D2 are relatively high compared to the other sites, while Site A2 in the center of Hanzhongmen Square had the lowest. Therefore, this suggests that afternoon activities should be carried out in greenspaces in higher degrees of airflow and SVF.
4.2. Smog and Haze Reduction Strategies Based on Optimized Greenspace Elements
This study has found that PM2.5 concentrations of its six research sites were closely correlated with their degree of green coverage, airflow openness and meteorological factors. According to the analysis on meteorological factors in
Section 3.2.2, PM2.5 concentrations had a very significant positive correlation with the sites’ humidity which, in turn, is significantly affected by the sites’ degree of green coverage and the airflow openness. Meanwhile, the analysis of greenspace elements in
Section 3.2.1 showed that PM2.5 concentrations were significantly negatively correlated with airflow openness and positively correlated with green coverage. These findings are consistent with those of Yin (2007) [
66] and Yang (2017) [
67]. However, as described in this study’s introduction, the impacts of green coverage and plant community structure on PM2.5 concentrations remain multifaceted and can even have opposing effects depending on other variables and conditions. The multi-layered composite structures of vegetation including trees, shrubs and grass with a high canopy densities and plant coverage may have higher PM concentrations than more singular-type lawn, shrub and grass environments [
68,
69]. Excessive plant density and canopy density can hinder the dilution and diffusion of PMs and thereby increase their concentrations [
48]. This is because high plant densities and green coverage tend to lead to higher humidity, and relatively poor airflow, impeding the diffusion of PM2.5 and other particles. Moreover, within a certain range of wind speeds, air channels provide the necessary space for air flow, allowing for the significant migration and diffusion of PM and its concentration within a site. However, this study found that there was no significant linear correlation between SVF and PM2.5 concentrations, although SVF was significantly positively correlated with the fully open airflow, thus SVF did appear to indirectly affect PM2.5 concentrations at the sites. This study also found that the sites’ water coverage has little effect on their PM2.5 concentration and humidity, conforming to existing research findings. In China, the main factors of the naturally measured water consumption of urban development land are the evaporation and transpiration of vegetation within greenspaces, known as evapotranspiration [
70]. In summer and autumn, the changes in river evapotranspiration trends are very noticeable while the variations in spring and winter are basically flat [
71]. Considering this and this study’s limited research period during the winter, it is expected that the influence of water coverage on the sites’ humidity is not as great as that of green coverage, and its effect on PM2.5 concentration is also correspondingly weak.
In general, the degree of the sites’ green coverage and airflow openness were the two most important factors affecting greenspace air quality, as well as the two major factors affecting the spatial pattern of the greenspace elements. Based on this, the 23 days of PM2.5 concentration data were ordered from the best to worst according to their air quality, and each site’s average PM2.5 concentration levels for corresponding air quality intervals were calculated. Then, the optimal intervals of green coverage and airflow openness were deduced from the two sites with the lowest average value (
Figure S4,
Table S4).
The mode of air pollution of each greenspace site was then calculated by analyzing the PM2.5 concentration data of the study period. According to the degree of green coverage and airflow openness with the best performance in the above air quality intervals, the following improvement measures are proposed for the current greenspace elements of the six sites:
Adjust airflow. For example, in Site A2, the degree of fully open airflow can be reduced to between 71–77% by adding small landscape buildings, while semi-open airflow can be decreased to between 45–60% by planting tall trees.
Adjust green coverage. For example, in Site B, the degree of green coverage can be increased to between 37% and 47% by planting more vegetation, while airflow can be maintained at its current status.
Jointly adjust airflow openness and green coverage. For example, in Site C, the fully open airflow can be lowered to 71–77% by adding small landscape buildings, and the semi-open airflow can be reduced to 45–60% by planting tall trees. Meanwhile, the degree of green coverage can be increased to 37–47%.
Similar suggestions can be made for the other sites based on this analysis. Furthermore, other measures can be taken at all sites to improve the thermal and humidity environments by adjusting the types of underlying surfaces that form the site and improving the spatial pattern of vegetation, so as to indirectly help reduce the PM concentrations and improve air quality.
4.3. Elderly Visitor’s Weak Sensitivity to Smog and Haze in Urban Greenspaces and Corresponding Potential Risks to Their Physical and Mental Heath
This study has found that elderly visitors of the research sites are not sensitive to the smog and haze in such greenspaces. Due to the decline in physical function among elderly populations, exposure to PM2.5 can cause greater and more severe harm to their physical and mental health (
Text S1). In addition, towards their decisions regarding outdoor activities, this study found the elderly are relatively sensitive to changes in climate compared to smog and haze, while the comfort index appears to play an intermediary role in the process of smog and haze affecting crowd activities.
Using questionnaires and interviews, this study explored the overall perception of the elderly population towards smog and haze and their decision process towards going out to visit a greenspace. Questions included whether they had a sense of the current day’s air pollution levels and what they felt about the actual PM2.5 concentrations levels, and whether there is certain PM2.5 concentration level that they considered a special concern. Among 200 interviewees, six stated that they felt sensitive or physically unwell due to smog and haze from a recent day, and six others expressed quantitative knowledge of smog and haze. This paper focuses on the answers of these 12 respondents to explain a typical type of elderly visitor of the greenspaces. As shown in
Table S5, among these 12 elderly persons, the ratio of male to female was 1:5, suggesting that women may be more sensitive to air quality and may tend to pay more attention to health problems related to smog and haze than men. For most of the set of elderly interviewees with quantitative knowledge of smog and haze, they stated their maximum acceptable value for PM2.5 concentrations when going out was 150 μg/m³, which is considered moderately polluted. A few stated they would go to the park only when the air was good quality or lightly polluted at 100 μg/m³ or less, and they may relax their standards if it is a sunny day. The other six interviewees were able to describe the exact dates on which they felt sensitive or uncomfortable due to smog and haze. The measured concentrations on these dates were all verified as being above 250 μg/m³, but they still insisted in going out to visit the greenspaces. This example shows how in their healthy pursuit of exposure to greenspaces, the elderly can also inadvertently expose themselves to unhealthy air pollution.
5. Conclusions
This study found that the air quality of small greenspaces dispersed within high-density central urban districts of Nanjing is non-ideal and poses a threat to human health. In terms of spatial and temporal distribution, overall, there is no significant difference in PM2.5 concentrations between different greenspaces. However, when it comes to the difference values, the distance from a shared urban artery has a notable influence on air quality between the different sites. In addition, the PM2.5 concentrations of different greenspaces show significant variations in their temporal distributions. Due to the traffic pollution caused mainly by vehicle emissions during the morning rush hour period, the PM2.5 concentrations in the morning are higher than in the afternoon. In terms of physical factors, greenspace elements such as green coverage and airflow openness show remarkably positive and negative correlations with PM2.5 concentrations respectively. SVF presents a significantly positive correlation with the degree of fully open airflow, which indirectly affects the PM2.5 concentration at the sites. Meteorological factors such as temperature and humidity are very significantly positively correlated with PM2.5 concentrations, while comfort level has a significantly negative correlation. Furthermore, humidity is also significantly influenced by green coverage and airflow openness. Based on the above findings, the optimal green coverage and airflow openness for different air quality intervals were calculated and were used to formulate optimized smog and haze reduction strategies. According to the mode of air quality in the six greenspaces during the study period, corresponding improvement measures were then proposed based on the aforementioned strategies.
The elderly are the main users of greenspaces yet this study’s questionnaire and interview findings showed they tend to be insensitive to smog and haze when deciding whether to visit a greenspace for physical and social activities. Instead, they are more likely to be relatively sensitive to changes in the local climate. The comfort level exerts a notable mediating effect in the process of PM2.5 concentration affecting greenspace visitor activities. Currently, the well-being of the elderly has drawn global attention and this study provides relevant insights regarding the spatial and temporal distribution of PM2.5 concentrations in China’s greenspaces where the surrounding elderly communities tend to carry out many of their daily activities. With these insights, this paper intends to spark discussion for improving greenspace quality and overall livable conditions of high-density central urban districts in the hope of providing a theoretical support and reference for elderly-oriented greenspace construction and indicators in the future.