Urban green space plays a significant role in protecting the environment and purifying the air. In a city, green infrastructures are interconnected in a large system, which can play a positive role in the ecological protection of the city; meanwhile, in the main urban areas of large cities with complex land use types, the distribution of green infrastructure is disorderly, and how to effectively determine the actual effect of green space on PM2.5 at an urban small scale is still a challenge. Smog in southern Chinese cities primarily occurs in spring and winter, so we chose Nanjing, one of the largest megacities in southern China, as a case study, and chose spring and winter as the research time series. We took 11 air monitoring points as the core and delimited a radius of 500 meters as the research area, tracked and monitored the changes in PM2.5 values at 11 stations for 2 months, and then conducted a correlation analysis between this change and the NDVI (Normalized Difference Vegetation Index) values of the 11 regions. The analysis shows that in the main urban area, green space can have an obvious influence on the ambient air state within 500 m. In addition, when the NDVI value was high, the PM2.5 concentration in the corresponding area was low, and the correlation reached 0.89 on average (1 is a perfect correlation); the mean regression equation was y = 56.3–0.088x. In addition, the correlation was stronger when the PM2.5 concentration was higher, indicating that green space in the main urban area can play an obvious ecological efficiency role in the extremely polluted atmospheric environment. These research results reveal for the first time the impact of green space on air particles in small-scale urban space, which is conducive to better optimization of green space layout to better give full play to the ecological function of green space in response to public health events.
Research Article
Correlation between NDVI and PM2.5 Concentrations in a Small-Scale Urban Area
https://doi.org/10.21203/rs.3.rs-3146587/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Landscape architecture
PM2.5 concentration
green space
correlation analysis
one-way ANOVA
PM2.5, a common particulate matter in the air, is known for its harm to public health. The PM2.5 concentration can, to some extent, represent the level of air quality in a city [1–3]. COVID-19, which began to spread in early 2020, has triggered a global focus on public health and a rethinking of the human environment [4]. Research by scholars at Harvard University has shown that when PM2.5 levels in the air increase by 1 microgram per cubic meter, the death rate of COVID-19 patients increases by 15%[5]. In addition to the risk of virus transmission, PM2.5 is also an important environmental factor contributing to early death and many chronic diseases such as cardiovascular disease, respiratory disease, and cancer [6–8]. Apart from PM2.5, air pollution, extreme weather and global warming are all becoming increasingly important in today's society. These adverse climate issues affect people's physical and mental health, activities, and even human behavior, such as crime on a global scale, so many countries are looking to natural factors to find solutions. As an important ecological carrier in the city, urban green space is important in public activities, cultural display, and landscaping. With the deepening of ecological civilization construction in China, urban green space is expected to play a more important role in promoting health and the construction of the public health system [9]. China is still in the stage of industrialization and urbanization, inevitably bring higher carbon emissions and heat island effect. However, heat island effect will in turn promote energy consumption and carbon emission. This is a vicious cycle, which is not only a problem of cities, but also a problem that needs to be solved by the joint efforts of countries and regions [10–11]. Given that PM2.5 is always closely related to carbon emissions, China is working with other regions to control air pollution with the goal of "carbon neutrality and carbon peak". This requires multidisciplinary coordination and joint exploration on the air pollution formation rules[12].
At present, many studies have shown that green space plays an important role in urban heat and climate comfort[13–14], which directly affect people's feelings and health[15–16]. But in terms of the reduction of air particles, most studies are limited to the micro level[17]: the reduction of particulate air pollutants by different plant species or plant communities [18]. Vegetation can greatly reduce particulate pollutants in the air when it acts as a green barrier for cities [19]. Small plant communities were reported to significantly decrease the PM in the surrounding environment, and this was most obvious in December [20]. These functions originate from the vegetation itself—their species, height, leaf size and so on all affect the PM2.5 concentration [21], and the shape of plant leaves, the waxy layer on the leaf surface, and the roughness of the leaf surface can affect the adsorption of PM[22]. At the macro level, scholars have studied the spatial and temporal effects of suburban vegetation on the regulation and purification of PM2.5 particles and have thus guided the design optimization of recreational parks for residents [23]. Researchers have explored the effect of the size and shape of different green patches on the reduction of air particulate matter [24]: through monitoring the forest belt with different widths of 20–60 m, it was found that the purification efficiency of PM2.5 in the forest belt gradually increased with increasing width, and the purification efficiency was the highest at 12:30–5:00 PM [25]. Studies have also taken urban blocks as the object and found that within a buffer zone of 500 m and 1000 m with the monitoring point as the center, the concentration of atmospheric particulate matter had a significant negative correlation with the green space area and a significant positive correlation with the building area[26]. Moreover, the relationship between land use LPI and MPS has been discussed at the landscape scale [27–28], but the contribution of green space to PM2.5 reduction in complex small-scale urban areas was not specifically analyzed.
As an important megacities in eastern China, many scholars have studied the PM2.5 concentration distribution in Nanjing. Research shows that PM2.5 concentration distribution in the urban area of Nanjing is closely related to land use types, and the spatial distribution of PM2.5 is primarily influenced by the density of food and beverage outlets[29]. Scholars have also studied the influence of Nanjing's urban form characteristics on the spatial change of PM2.5 through the collection of a large amount of air data, and urban trees was found to be the most important urban morphological characteristic in explaining the PM2.5 air quality, followed by the height-normalized roughness length as the second important parameter[30]. But these studies have not taken more accurately study on the actual relationship between green space and PM2.5.
The purpose of this study is to explore the precise relationship between PM2.5 and green space at a small scale in Nanjing through data analysis. The types of land use in large cities are often intricate, especially in the central urban areas of modern cities, where the spatial forms are even more complex and diverse [31–32]. A variety of land use types are accompanied by green spaces, that is, some are scattered and irregularly distributed; some are along roads; some are in residential areas; some are blocky; and some are along river belts [33]. The distribution complexity determines that the study of their impact on air quality needs to be considered from a more systematic scale. We selected the NDVI to express the green space area. The NDVI is obtained from the vegetation spectrum of a satellite map and can reflect the quality and area of plants in urban green spaces at different times [34]. The NDVI has a positive and negative bidirectional spectrum that can accurately distinguish green and non-green urban space [35]. In this study, several air monitoring stations were established in the main urban areas of the city, and a correlation analysis was conducted between the PM2.5 value obtained by long-term monitoring and the average NDVI value of the surrounding environment of the monitoring station in the corresponding period. This method can more truly reflect the actual impact of urban green space on PM2.5 and reveal the corresponding relationship, to promote the optimization of urban green space layout and provide support for the improvement of the public respiratory health system.
2.1 Study area
The study area is the main urban area of Nanjing, China, which is the capital of Jiangsu Province and one of China’s megacities. In 2022, the GDP of Nanjing was 1,635,532 billion yuan. There are 9,423,400 permanent residents in Nanjing, among which 8,188,900 live in urban areas. According to the latest figures from the Office for National Statistics, the population density of Nanjing main urban areas ranks among the top in China(http://www.stats.gov.cn/tjsj/pcsj). In terms of green space construction, Nanjing has been awarded the highest grade of "National Ecological Garden City", meaning that Nanjing has a relatively systematic and perfect green space system, which is representative of the quality of the urban green space. In the context of unprecedented concern for urban public health and respiratory health, every possible measure needs to be taken to reduce the risk of airborne transmission of a virus in the center of a large, densely populated city, and green space should be given more attention as a potentially effective factor. Therefore, the choice of the main urban area of Nanjing (Fig. 1) as the research area indicates the realistic demand and the urgency of this research. Compared with urban areas, urban downtown is more in line with the definition of "small-scale space", that is, there are higher population density and municipal facilities in a smaller space, and people's activities are more diverse, which will also bring higher PM2.5 emissions. Therefore, it is more meaningful to study the impact of greening on PM2.5 in urban downtown.
2.2 Analysis methods
The main research method adopted was to conduct a correlation between the parameterized data of NDVI images within 500 m of 11 monitoring points and the changes in PM2.5 values of these 11 monitoring points through variance analysis and linear regression analysis, to reflect the relationship between the state of the urban green space and the spatial distribution of PM2.5 in the main urban area (Fig. 2). This limit of 500 meters is based on previous research, i.e., that the air environment within a comfortable walking distance of residents (500 m) plays an important role in people's daily quality of life [36]. After the locations of the 11 monitoring sites were determined, a buffer zone with a radius of 500 m was constructed with each monitoring site as the center, forming 11 research areas with an area of about 78 hectares. The daily PM2.5 data of the 11 monitoring stations were collected in April and December and compared with the NDVI parameter values of the corresponding 11 research areas in the corresponding months to explore the relationship between the green space status and the PM2.5 values among the 11 samples, to reflect the influence of the green space status (NDVI) on PM2.5 in the real environment of an urban main city. In addition, we conducted a longitudinal comparison of the relationship between the green space and air quality in spring and winter [37] and obtained the difference in the contribution of the green space to air quality in different seasons.
2.2.1. Analysis of the significant difference
Due to the large data sample size, significant difference analysis can preliminarily determine the correlation between the green space state and air state. SPSS 20 was used to conduct independent sample t tests and one-way analysis of variance (ANOVA). The least significant difference method (LSD) was used to conduct a single-factor comparison of daily PM2.5 data at different monitoring points, to determine whether there was a significant difference in the PM2.5 data of different periods and different monitoring points.
2.2.2. Correlation analysis: Unitary Linear Regression
Further calculations were performed to reveal the precise relationship between the green space status and the PM2.5 concentration. As two random variables, the green space state and air state can be described by a correlation coefficient. We used the Pearson correlation coefficient to determine whether there was a linear correlation between the two elements and used the function COR in R software to calculate the Pearson correlation coefficient as follows:
Then, a linear regression was established between NDVI and PM2.5 data in the 11 study areas. Regression analysis is an analytical method to model the relationship between one or more independent variables and dependent variables using the least square function called a linear regression equation. In this study, the independent variable X axis represents the NDVI, and the dependent variable Y axis represents the air PM2.5 index. The correlation between the two can be approximated by a straight line.
2.3. Data collection and processing
2.3.1 PM2.5 data in the main urban areas of Nanjing
Open source data and an instrument field survey were used to obtain accurate PM2.5 data in the main urban areas. At present, there are 13 National Air Quality Monitoring Stations in Nanjing. After the monitoring stations were screened, nine monitoring points were obtained after excluding four monitoring points in Liuhe and Gaochun, which were far from the central urban areas. All nine points are located in the main urban areas and have different green backgrounds. The data of each monitoring point are the daily average values of China's air Quality online Monitoring and Analysis platform (www.aqistudy.cn) and the Air’s Voice website (http://www.air-level.com).
To make the results more obvious, two self-test points were added based on the 9 state-controlled points for comparative analysis, namely, Yuzui Wetland Park and the Xiamafang area of Purple Mountain National Forest Park, which both have a good green space environment, to prove whether different green substrates have an impact on the air in the main urban area. A hand-held laser air detector was used for data acquisition of self-test points, with a precision of 0.1 µg/m3. PM2.5 data of each group of time series were collected in the months when there was no wind (wind speed less than 2 m/s) and no precipitation.
In terms of the data acquisition time, according to the "Nanjing Environmental Status Bulletin" issued by Nanjing Municipal Bureau of Ecology and Environment, in 2022, the number of days of ambient air quality in Nanjing built-up areas that reached the second-level standard was 304, and the proportion that reached the standard was 83.1%. In addition, the number of days that did not reach the second-level standard was 62, and the annual mean value of PM2.5 was 31 µg/m3. The proportion of days with light or moderate pollution was 16.9%, and the days with moderate or higher pollution were mainly in winter and spring (Fig. 3). According to these seasonal dynamics and based on the seasonal characteristics of plants in Nanjing, December and April were selected for the study. After long-term monitoring, the PM2.5 data of 11 monitoring points in December and April were successfully collected (Fig. 4 ).
2.3.2 Accurate NDVI data acquisition of the main urban areas
The NDVI data required for the study should meet two criteria[38]: firstly, the research scale is limited to the level of the main urban area, which requires high resolution and clarity of NDVI images; secondly, the research is limited to the time series of April and December, which requires that the images can accurately reflect the seasonal differences. The NDVI images used in this study were obtained by programming the "Pie-Engine" platform(https://engine.piesat.cn/engine/home). "Pie-engine" is a geospatial data analysis and calculation platform independently developed by China Aerospace Hongtu Information Technology Co., LTD. Pie-engine Studio combines all kinds of open satellite remote sensing images, geographical factor data, and other spatial data, thereby enabling users to study the algorithm model at any scale and adopt interactive programming verification to quickly explore the surface features and find changes and trends.
Based on the atmospheric-corrected reflectance data of the lower atmosphere obtained by ESA's Sentinel-2 Level-2A, we wrote the code of NDVI according to the formula: NDVI = (BAND8-BAND4)/(BAND8 + BAND4), and we obtained the long-term series NDVI images of Nanjing for April and December[39].
After the image acquisition, the NDVI images from the 11 study areas were parameterized by Matlab, which is good at parameterization of images to achieve various editing functions. In the process of parameterization of the NDVI images, image graying and binarization functions in Matlab are mainly used. NDVI image goes through three steps in MATLAB. First, importing NDVI images from the "Pie-engine" into MATLAB, then standardized RGB images are obtained and a three-dimensional pixel matrix is formed. Second, the standardized RGB images of the 11 study areas were transformed into gray-scale images, at this time, each pixel in the image was represented by the intensity between 0 (black) and 255 (white), and the values between 0 and 255 were used to represent different gray levels. At last, the gray level images in the 11 study areas were converted into binary images, that is, the image was divided into the background and target according to the gray characteristics of the image (Fig. 5). When we use the IM2BW function to transform the grayscale image into a binary image, we need to set a threshold value, which was used as the average grayscale value of the image in this study. In the selection method of the threshold, we adopted the histogram bimodal method proposed by Prewitt in 1996 [40].
2.4 Data analysis and processing
2.4.1. Analysis of significant differences
Due to the large amount of data, significant difference analysis can identify anomalies in the data [41–43]. We first calculated the monthly mean value and peak value of PM2.5 at each monitoring point and then conducted a one-way ANOVA on the mean value and peak value changes of PM2.5 at the 11 monitoring points in April and December [44].
|
VAR00001 |
N |
alpha = subset of 0.05 |
||
|---|---|---|---|---|
|
1 |
2 |
3 |
||
|
10 |
3 |
34 |
||
|
6 |
3 |
35 |
||
|
3 |
3 |
36 |
36 |
|
|
9 |
3 |
36 |
36 |
|
|
11 |
3 |
36 |
36 |
|
|
5 |
3 |
37 |
37 |
37 |
|
7 |
3 |
38 |
38 |
38 |
|
2 |
3 |
39.3 |
39.3 |
39.3 |
|
1 |
3 |
42 |
42 |
|
|
4 |
3 |
43 |
||
|
8 |
3 |
43 |
||
|
Significance |
0.174 |
0.075 |
0.059 |
|
|
Sum of squares |
DF |
Mean Square |
F |
Significance |
|
|---|---|---|---|---|---|
|
Between groups |
316.850 |
10 |
31.685 |
5.418 |
0.000 |
|
Within groups |
128.670 |
22 |
5.848 |
||
|
Sum |
445.510 |
32 |
|
Points |
N |
alpha = subset of 0.05 |
|||
|---|---|---|---|---|---|
|
1 |
2 |
3 |
4 |
||
|
10 |
3 |
44 |
|||
|
6 |
3 |
51 |
|||
|
2 |
3 |
54 |
54 |
||
|
9 |
3 |
55 |
|||
|
5 |
3 |
56 |
|||
|
7 |
3 |
57 |
|||
|
1 |
3 |
63 |
|||
|
4 |
3 |
63 |
|||
|
11 |
3 |
63.3 |
|||
|
8 |
3 |
64.3 |
|||
|
3 |
3 |
66 |
|||
|
Significance |
1.00 |
0.07 |
0.18 |
0.25 |
|
|
Sum of squares |
DF |
Mean Square |
F |
Significance |
|
|---|---|---|---|---|---|
|
Between groups |
1358.73 |
10 |
135.87 |
45.29 |
0.000 |
|
Within groups |
66 |
22 |
3 |
||
|
Sum |
1424.73 |
32 |
2.4.2. Correlation analysis
The R function of SPSS software was used to calculate the multi-directional correlation. First, the correlation analysis between the NDVI value and the daily average value and peak value of all monitoring days was carried out (Table 5), and the overall correlation was obtained. Second, the correlation between the NDVI value and the average value and peak value in April and December was analyzed in pairs (Table 6), and the correlation between NDVI and PM2.5in the two months was obtained (Fig. 6–7, R is the correlation coefficient, P is the test accuracy, less than 0.01 indicates that the correlation is extremely reliable); at the same time, the correlation difference between the two seasons was obtained.
|
NDVI value |
PM2.5 average value |
PM2.5 peak value |
|
|---|---|---|---|
|
NDVI value |
1 |
−0.890** |
−0.928** |
|
PM2.5 average value |
−0.890** |
1 |
0.869** |
|
PM2.5 peak value |
−0.928** |
0.869** |
1 |
| Note: ** indicates a significant correlation at the 0.01 level (bilateral). | |||
Table 6. Correlation between the NDVI value, PM2.5 average value, and peak value in April and December
|
NDVI value |
PM2.5 average value |
PM2.5 peak value |
|
|
NDVI value |
1 |
−0.900** |
−0.910** (Dec) |
|
PM2.5 average value |
−0.760** (Apr) |
1 |
0.937** (Dec) |
|
PM2.5 peak value |
−0.903** (Apr) |
0.637** |
1 |
Note: ** indicates a significant correlation at the 0.01 level (bilateral).
3.1. One-way analysis of variance (ANOVA)
The results of one-way ANOVA are shown in Fig. 8, where the blue value is the monthly average PM2.5 concentration of the monitoring site, and the orange value is the monthly peak PM2.5 concentration of the monitoring site. The significance of the monthly average PM2.5 concentration in different regions is indicated from A to D. It was found that although these monitoring points were all in the main urban area and were very close to each other, there was a significant difference in the PM2.5 concentration between different monitoring points. Statistical results in spring show that the PM2.5 concentration of monitoring sites in downtown areas such as Maigao Bridge, Shanxi Road, and Zhonghuamen exceeded 40 µg/m³, which was significantly higher than that in other areas, while the values of monitoring sites such as Yuzui Wetland and Zijinshan Ximafang were 34 µg/m³ and 35 µg/m³, which were the lowest among all monitoring sites. In addition to these monitoring sites, other 6 sites like Ruijin Road, Olympic Sports Center and Caochangmen, the difference in NDVI between them was small, and the PM2.5 concentration we obtained was also in the middle position among all the 11 monitoring sites. A comparison of the data in winter showed the same pattern, which preliminarily indicates that the air condition in the main urban area changes with the change in the surrounding green environment. According to the longitudinal analysis of variance in Fig. 9, the average monthly PM2.5 concentration in April was 39.6 µg/m³, while the average monthly PM2.5 concentration in December was as high as 58 µg/m³, which was about 1.5 times higher than the former. An independent sample t test showed that the mean concentration of PM2.5 was significantly different between the two months (P < 0.01). Besides atmospheric circulation, the main reason for the significant difference between seasons may be that plants grow better in April, while some trees shed their leaves in December, which can further verify that green space can effectively improve the air condition of small-scale urban areas.
3.2. Linear regression analysis
Correlation analysis showed that the NDVI value was significantly negatively correlated with the average and peak value of PM2.5 in the corresponding region, with overall correlation coefficients of − 0.89 and − 0.928, respectively. There was a significant positive correlation between the average concentration of PM2.5 and the peak value of PM2.5, with a correlation coefficient of 0.869. In addition, unitary linear regression analysis showed that the NDVI value, average PM2.5 value, and peak value showed a simple linear regression. Taking the NDVI value as the independent variable(x), and taking the average PM2.5 value(Y1) and peak PM2.5 value as the dependent variable, the regression equations were Y1 = 56.3 − 0.088x and Y2 = 83.2 − 0.09x.
To be specific, in April 2022, according to the study of the 11 monitoring points in the main urban area of Nanjing, the correlation coefficient between the PM2.5 average value, peak value, and NDVI were − 0.76 and − 0.903, while in December, these two values were − 0.91 and − 0.937, respectively. In both April and December, NDVI showed a significant negative correlation with the PM2.5 concentration and a very significant negative correlation with the PM2.5 peak value.
4.1. Spatial dimension
In the downtown area of Nanjing, urban green space has a significant impact on the air state of the surrounding 500 m. While PM2.5 levels are significantly higher in downtown areas than in the suburbs, we found that the higher the NDVI, the lower the PM2.5 concentration in the surrounding area. The correlation reached 0.89 on average (1 is a perfect correlation), indicating that in the densely populated downtown area, it is feasible to improve respiratory health and reduce the transmission risk of various viruses by increasing the green area and improving the quality of the green space. Based on this conclusion, urban planning departments can improve the urban environment by increasing the density of plants in densely populated areas [45–46]. For urban greening departments, the formula Y = 56.3 − 0.088x found in the study can be used to preliminarily determine the greening shortage in the main urban areas through the calculation of regional PM2.5 monitoring data, to carry out targeted planning. In addition, public health departments can also predict the natural transmission risk of viruses in the downtown area according to the change in the PM2.5 values in different areas [47] and make corresponding adjustments to the green space to reduce this risk [48].
4.2. Time dimension
Unlike most current studies that only focus on the influence of green space on PM2.5 concentrations at different scales [49], our study focuses on the small-scale scope of the urban center and analyzes it from both time and space dimensions, which can more realistically reflect the real influence of urban greening on urban air quality.
With respect to the time dimension, April and December were the two months with the highest PM2.5 concentration in Nanjing in 2022, and the average concentration in December was higher than that in April, which may be due to many reasons, including the influence of atmospheric circulation. However, the average correlation between the green space and PM2.5 concentration in December was − 0.91, while that in April was only − 0.76. This shows that the influence of the urban green space on PM2.5 in winter was stronger than that in spring. To further discuss whether green space can play a greater role in a more severe air environment, the study compared the NDVI and the PM2.5 peak values in April and December and found that the correlation between the NDVI and PM2.5 peak values in each month was higher than the average values, and the correlation in December was higher than that in April, which further indicated the conjecture of this study: In seasons or places with high PM2.5 concentration, the influence of greening status on PM2.5 becomes more obvious, that is to say, the higher the PM2.5 concentration, the greater the ecological efficiency of the urban green space[50]. This indicates that even in the winter, plant physiological function is limited but also more obvious.
4.3. Possible reasons dicussion
Firstly, different urban areas will affect the flow of PM2.5. After drawing the previous research conclusions, we conducted a preliminary survey on the land use in the 11 study areas, and found that among the 11 study areas, the population density and commercial density in the study area with high correlation between NDVI and PM2.5 were higher than that in the study area with a low correlation, which meant higher carbon emissions. In addition, the use rate of large trees in these areas is significantly higher than that in areas with low correlation, and among all tree types, evergreen trees such as camphor and magnolia magnolia account for a large proportion. These large trees may have a greater impact on the reduction of PM2.5 than the grassland and shrubs in the suburbs.
The second possible reason is, the physiological function of plants. The middle of December was the time when PM2.5 was generally high in Nanjing, according to the observation of trees in the study area at this time, it was found that camphor trees and ginkgo biloba, which are commonly used in Nanjing, were in the last deciduous period, and they still had strong physiological functions[51]. It can be inferred that the leaves of large trees may play more obvious ecological functions when the air pollution is heavy. In addition, since plant types, varieties and cluster planting forms may also have a significant impact on local environmental PM2.5[52], this guess can be further studied through the control variable method of the key factors affecting the change of PM2.5 in the main urban area.
4.4. Limitations
The study has two limitations. First, to make the comparison more obvious, we chose the two months with the highest average PM2.5 concentration in a year. If continuous monitoring is carried out for more than one year, the correlation between the two months may be further confirmed. Second, the average NDVI value was selected to represent the status of the green space in the 11 study areas, without including the connectivity of each green space, the scale of individual green spaces, and other indicators; this may impact the concentration and diffusion of PM2.5 [53], which can be further verified in subsequent studies.
This study revealed the relationship between the green space and air quality in a small-scale area of Nanjing, indicating that the greening of the main urban area could play a positive role in reducing the PM2.5 concentration within 500 meters. The results indicates that the urban green space can play a more obvious role under severe air pollution, and further concluding the mathematical relationship between the two factors. This can bring some inspiration to the planning of green spaces in the main urban area[54]. In densely populated urban areas, public respiratory health can be improved by increasing the number and quality of plants. In southern China, where winter air pollution is a frequent problem, the planting of evergreen trees can also improve the air quality. In the context of frequent public health events, PM2.5 is being studied more as a viral vector, existing studies in the United States have shown that greenbelt exposure has a significant impact on virus transmission in a small-scale area and may reduce the mortality rate in areas with a high population density [55]. Although no causal relationship can be drawn from the reduction of PM2.5, this paper and many domestic and foreign studies have shown that the quality of the urban green space could have important public health benefits in responding to the future pandemics. In this paper, the correlation between NDVI and PM2.5 in small-scale areas was studied. Subsequently, indicators such as the size, shape, and connectivity of green spaces can be introduced to further verify the impact of green space on air particles [56], and detailed data on NDVI and PM2.5 in more Chinese cities could be introduced to further analyze the correlation, so as to get a more precise relationship between the two factors.
Author Contributions: Research design, P.T.; methodology, P.T.; data acquisition, X.Y. and X.S.; validation, P.T.; data analysis, P.T.; investigation, P.T. and H.Y.; resources, H.Y.; data curation, P.T.; writing— original draft preparation, P.T.; writing—review and editing, P.T.; visualization, X.Y.; supervision, X.S.; project administration, H.Y.; funding acquisition, H.Y. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by “Science and Technology Think Tank Young Talent Program of China Association for Science and Technology”（2022-162）, “Natural Science Foundation of Jiangsu Universities”(grant number 21KJB220004) , Social Science Application Research Coordinated Innovation Base project of Jiangsu(22XTA-38), “Jiangsu Grassroots Social Governance research Collaborative innovation base” , “13th Five-Year Plan project” of Jiangsu Open University.
Data Availability Statement: The data presented in this study are available on request from the author. The data are not publicly available due to [privacy].
Conflicts of Interest: The authors declare no conflict of interest.
- Marangon D, Traversi D, D'Agostino AM, Gea M, Schilir T (2022) The north-western Italy air quality monitoring network: improving experience of pm2.5 assessment with mutagenicity assay. Environ Res 1:110699
- Zhang Y, Liu X, Zhang L, Tang A, Collett JL (2022) Evolution of secondary inorganic aerosols amidst improving PM2.5 air quality in the North China Plain. Environ Pollut D22:117027
- Chu HJ, Ali MZ, He YC (2020) Spatial calibration and PM2.5 mapping of low-cost air quality sensors. Sci Rep 10(1)
- Wright L, Burton A, Mckinlay A, Steptoe A, Fancourt D (2022) Public opinion about the UK government during COVID-19 and implications for public health: A topic modeling analysis of open-ended survey response data. PLoS ONE 17(4):e0264134
- Field RD, Moelis N, Salzman J, Bax A, Ausiello D, Woodward SM (2022) Inhaled Water and Salt Suppress Respiratory Droplet Generation and COVID-19 Incidence and Death on US Coastlines. Mol Front J 05(01n02):17–29
- Wang Y, Li C, Zhang X, Kang X, Du L (2022) Exposure to PM2.5 aggravates Parkinson's disease via inhibition of autophagy and mitophagy pathway. Toxicology 152770
- Mehrotra R, Group IS (2020) B.O.D. Health and economic impact of air pollution in the states of India: the Global Burden of Disease Study 2019. Lancet Planet Health
- Li G, Zhang H, Hu M, He J, Yang W, Zhao H, Zhu Z, Zhu J, Huang F Associations of combined exposures to ambient temperature, air pollution, and green space with hypertension in rural areas of Anhui Province, China: A cross-sectional study 2022
- Wolch JR, Byrne JA, Newell JP (2014) Urban green space, public health, and environmental justice: The challenge of making cities 'just green enough' Landscape & Urban Planning 125
- Wei G, Bi M, Liu X, Zhang ZK, He BJ (2023) Investigating the impact of multi-dimensional urbanization and FDI on carbon emissions in the belt and road initiative region: Direct and spillover effects. J Clean Prod 384,135608
- Yang J, Wang YC, Xue B, Li YF, Xiao XM, Xia JH (2022) (Cecilia).; He,B.J. Contribution of urban ventilation to the thermal environment and urban energy demand: Different climate background perspectives. Sci Total Environ 795
- Wei G, Zhang ZK, Ouyang X, ,Shen Y, Jiang SN, Liu BL, He BJ (2022) Delineating the spatial-temporal variation of air pollution with urbanization in the Belt and Road Initiative area. Environ Impact Assess Rev 91
- He BJ, Wang JS, Zhu J, Qi JD (2022) Beating the urban heat: Situation, background, impacts and the way forward in China. Renew Sustain Energy Rev 161:112350
- He BJ, Ding L (2020) Deo Prasad. Wind-sensitive urban planning and design: Precinct ventilation performance and its potential for local warming mitigation in an open midrise gridiron precinct. J Building Eng 29:101145
- Song CR, Yang J, Wu F, Xiao XM, Xia JH, Li XM (2022) Response characteristics and influencing factors of carbon emissions and land surface temperature in Guangdong Province,China. Urban Clim 46,101330
- Zhou CS, Zhang DH, Cao YW, Wang YZ, Zhang GJ (2022) Spatio–temporal evolution and factors of climate comfort for urban human settlements in the Guangdong–Hong Kong–Macau Greater Bay Area. Front Environ Sci 10:1001064
- Baraldi R, Neri L, Costa F, Facini O, Rapparini F, Carriero G (2018) Ecophysiological and micromorphological characterization of green roof vegetation for urban mitigation. Urban Forestry & Urban Greening S1618866717304740
- Chen M, Dai F, Yang B, Zhu S (2019) Effects of neighborhood green space on PM2.5 mitigation: Evidence from five megacities in China. Build Environ
- Brantley HL, Hagler G, Deshmukh PJ, Baldauf RW (2014) Field assessment of the effects of roadside vegetation on near-road black carbon and particulate matter. Sci Total Environ 468–469(jan.15):120–129
- Tomson N, Michael RN, Agranovski IE (2022) Removal of particular air pollutants by Australian vegetation potentially used for green barriers. Atmospheric Pollution Research 12(32):101070
- Cza B, Ap C, Yca B, Hga B, Yca B, Yza B (2019) Effect of spatial heterogeneity of plant communities on air PM10 and PM2.5 in an urban forest park in Wuhan, China. Urban Forestry & Urban Greening 46(2):126487
- Prusty B, Mishra PC, Azeez PA (2005) Dust accumulation and leaf pigment content in vegetation near the national highway at Sambalpur. Orissa India Ecotoxicology & Environmental Safety 60(2):228–235
- Wang Y, Xuan K, Xue L, Xie S (2016) Temporal and spatial effect of atmospheric PM2.5 and PM10 regulated by forest vegetation in the suburban park of Chongqing. Ecol Environ Sci 10
- Tiwari A, Kumar P (2020) Integrated dispersion-deposition modelling for air pollutant reduction via green infrastructure at an urban scale. Sci Total Environ 723:138078
- Yang LI, Wang SY, Xiao H, Zhang LX (2019) Study on Testing Method of Purification Efficiency of Fresh Air Purifier PM2.5. Standard Sci
- Dai F, Chen M, Zhu S, Chen H, Hu H (2020) Effects of Different Green Coverage in Block Scale on PM 10 and PM 2.5 Removal——A Case Study of the Main City of Wuhan. Chin Landsc Archit 36(3):109–114
- Su W, Zhang SJ, Lai XY, Gu XR, Lai SN, Huang GX, Zhang ZJ, Liu YQ (2017) Spatiotemporal dynamics of atmospheric PM2.5 and PM10 and its influencing factors in Nanchang, China. Chin J Appl Ecol 28(1):257–265
- Wang M, Yang Y, Guo T (2022) Measurement of Urban–Rural Integration Level in Suburbs and Exurbs of Big Cities Based on Land-Use Change in Inland China: Chengdu. Land 10
- Zhao JD, Yang L, Ren HY (2022) Study on spatial and temporal distribution characteristics of PM2.5 in the main urban area of Nanjing and influencing factors. Earth and Environmental Science 01
- Tom VK, Pauli P;Shahzad,G.;Jiang,Lin.;Wang,B.;Zhou,D.;Qin,W.;Nie,W.;, Veli-Matti K, Tuukka P, Sun JN, Markku (2022) K.;Aijun,D. The effect of urban morphological characteristics on the spatial variation of PM2.5 air quality in downtown Nanjing. Environmental Science:Atmospheres 07
- Zeng LY, Hang J, Wang XM, Shao M (2022) Influence of urban spatial and socioeconomic parameters on PM 2.5 at subdistrict level: A land use regression study in Shenzhen, China. J Environ Sci 114:485–502
- Zhang DH, Zhou CS, He BJ (2022) Spatial and temporal heterogeneity of urban land area and PM2.5 concentration in China. Front Urban Climate 45:101268
- Nazmfar H (2018) Spatial Analysis of Parks and Urban Green spaces by using Copras Models and GIS (Case study: 22 regions of Tehran).
- Moreno R, Ojeda N, Azocar J, Venegas C, Inostroza L (2020) Application of NDVI for identify potentiality of the urban forest for the design of a green corridors system in intermediary cities of Latin America: Case study, Temuco, Chile. Urban Forestry & Urban Greening 55(1396):126821
- Wu H, Ye LP, Shi WZ (2014) Assessing the effects of land use spatial structure on urban heat islands using HJ-1B remote sensing imagery in Wuhan, China. Int J Appl Earth Observations Geoinf 32(1):67–78
- Lei Y, Duan Y, He D, Zhang X, Chen L, Li Y (2018) Effects of Urban Greenspace Patterns on Particulate Matter Pollution in Metropolitan Zhengzhou in Henan. China Atmos 9(5):199
- Mallia DV, Lin JC, Urbanski S et al (2015) Impacts of upwind wildfire emissions on CO, CO2, and PM2.5 concentrations in Salt Lake City, Utah. J Geophys Res Atmos 120(1):147–166
- Jiang YL, Chen BW, Huang YF, Cui J, Guo Y (2022) Crop planting area extraction based on Google Earth Engine and NDVI time series difference index. J Geo-information Sci 23(5):938–947
- Shahabi H, Ahmad BB, Mokhtari MH, Zadeh MA (2012) Detection of urban irregular development and green space destruction using normalized difference vegetation index (NDVI), principal component analysis (PCA) and post classification methods: A case study of Saqqez city. Int J Phys Sci 7(17):2587–2595
- Dhagdi ST, Deshmukh PR (2012) New Technique for Keyframe Extraction Using Block Based Histogram. Int J Adv Res Comput Sci 3(3):794–796
- González-Rodríguez G, Colubi A, MÁ, Gil (2012) Fuzzy data treated as functional data: A one-way ANOVA test approach. Comput Stat Data Anal 56(4):943–955
- E O (2013) Methodology and Application of Oneway ANOVA. Sci Educ Publishing
- Slusky DJG (2017) Significant Placebo Results in Difference-in-Differences Analysis: The Case of the ACA's Parental Mandate. East Econ J 43(4):580–603
- Tian LG, Meng QH, Wang LP, Dong JH, Wu H (2015) The value of peak to peak, mean, variance and mean square with illumination variation in Sansevieria diagram. PLoS ONE
- Chow JC, Watson JG, Mauderly JL, Costa DL, Dockery DW (2006) 2006 critical review discussion - Health effects of fine particulate air pollution: Lines that connect. J Air Waste Manag Assoc 56(6):1368–1380
- Zhang DH, Zhou CS, Zhou Y (2022) Bahram Zikirya. Spatiotemporal relationship characteristic of climate comfort of urban human settlement environment and population density in China. Front Ecol Evol 10:953725
- Fann N, Lamson AD, Anenberg SC, Wesson K, Risley D, Hubbell BJ (2012) Estimating the National Public Health Burden Associated with Exposure to Ambient PM2.5 and Ozone. Risk Anal 23(1):81–95
- Natasha G, Vernon M (2006) Assessment of Public Health Risks Associated with Atmospheric Exposure to PM2.5 in Washington, DC, USA. Int J Environ Res Public Health 3(1):86–97
- Chen M, Dai F, Yang B, Zhu S (2019) Effects of neighborhood green space on PM2.5 mitigation: Evidence from five megacities in China. Build Environ 156:33–45
- Kerimray A, Azbanbayev E, Kenessov B, Plotitsyn P, Alimbayeva D, Karaca F (2020) Spatiotemporal variations and contributing factors of air pollutants in almaty, Kazakhstan. Aerosol Air Qual Res 20:1340–1352
- Tanino KK, Kalcsits L, Silim S, Kendall E, Gray GR (2010) Temperature-driven plasticity in growth cessation and dormancy development in deciduous woody plants: a working hypothesis suggesting how molecular and cellular function is affected by temperature during dormancy induction. Plant Mol Biol 73(s1–2):49–65
- Heshmat N, Khaniabadi YO, Goudarzi G, Jourvavd M, Nikmehr K (2016) Investigation on the dust dispersion (PM10 and PM2.5) by doroud cement plant and study of its individual exposure rates. J Ilam Univ Med Sci 24(1):64–75
- Wanghe K, Guo X, Luan X, Li K (2019) Assessment of urban green space based on bio-energy landscape connectivity: a case study on Tongzhou district in Beijing, China. Sustainability 11, 4943
- Yang J, Luo X, Jin C, Xia JC (2020) Spatiotemporal Patterns of Vegetation Phenology along the Urban–Rural Gradient in Coastal Dalian, China. Urban Forestry & Urban Greening 126784
- Heo S, Desai MU, Lowe SR, Bell ML (2022) Impact of changed use of greenspace during COVID-19 pandemic on depression and anxiety. Int J Environ Res Public Health 18(11):5842
- Lin Y, An W, Gan M, Shahtahmassebi AR, Ye Z, Huang L (2022) Spatial grain effects of urban green space cover maps on assessing habitat fragmentation and connectivity. Land 10