No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
The influence of neighborhood-level urban morphology on PM2.5 variation based on random forest regression
Abstract
To improve the atmospheric environment by optimizing urban morphology, this study develops a random forest (RF) model to investigate the influence of urban morphology on PM2.5 variations via the relative importance of urban morphology and the nonlinear response relationship between urban morphology and PM2.5. Two indices—reduction range (C↓) and rate (C˅) of PM2.5 concentrations—are defined to evaluate the temporal variations of PM2.5. Results show that RF models are more accurate and perform better than multiple linear regression models, with R² ranging from 0.861 to 0.936. Five out of nine urban morphological indicators have the most significant contribution to PM2.5 reduction. For each indicator, the nonlinear response relationship shows similar trends in general, despite of the difference at the higher pollution level. Building evenness index and water body area ratio have a similar response such that C↓ and C˅ sharply increase and tend to be stable when they reach at 0.05 and 8 %, respectively. With the increase in vegetated area ratio, the change of C↓ presents an inverted V-shape trend with the turning point of about 20 %; however, the change of C˅ greatly differs from the pollution level. A higher density of the low-rising buildings with one to three floors will lead to a small reduction rate but a greater reduction range of PM2.5. Floor area ratio values generally show a negative and nonlinear influence on C↓ and C˅. This study provides useful implications for planners and managers for PM2.5 reduction through neighborhood morphology optimization.
... Our previous studies examined the influences of neighborhood green space and urban morphology on PM2.5 separately, providing evidence for the effects of urban green space coverage and morphological patterns and gray space forms on PM2.5 [18,32,35]. As a series of studies, hourly PM2.5 concentrations from 2016 to 2017 were collected from monitoring stations to ensure the consistency of PM2.5 data. ...
... Because of the different locations of 37 neighborhoods, the uncontrolled factors (such as weather conditions, PM2.5 background concentration of five cities) were kept at similar levels to remove the external impact factors as much as possible and to greatly minimize evident differences, thereby performing an intercomparison. Consistent with our previous studies [32,35], three relative indicators, the range (Cin and Cde), duration (Δtin and Δtde), and rate (Cin' and Cde'), of the increase and decrease in PM2.5 concentration were calculated. To investigate pollution-level differences in the effect of the neighborhood-level built environment on PM2.5, four pollution levels, including slight (PM2.5 level ranging from 75 μg/m 3 to 114 μg/m 3 ), moderate (PM2.5 level ranging from 115 μg/m 3 to 150 μg/m 3 ), heavy (PM2.5 level greater than 150 μg/m 3 ), and overall pollution, were analyzed based on Chinese ambient air quality standards. ...
... Our previous studies examined the influences of neighborhood green space and urban morphology on PM 2.5 separately, providing evidence for the effects of urban green space coverage and morphological patterns and gray space forms on PM 2.5 [18,32,35]. As a series of studies, hourly PM 2.5 concentrations from 2016 to 2017 were collected from monitoring stations to ensure the consistency of PM 2.5 data. ...
Air pollution, especially PM2.5 pollution, still seriously endangers the health of urban residents in China. The built environment is an important factor affecting PM2.5; however, the key factors remain unclear. Based on 37 neighborhoods located in five Chinese megacities, three relative indicators (the range, duration, and rate of change in PM2.5 concentration) at four pollution levels were calculated as dependent variables to exclude the background levels of PM2.5 in different cities. Nineteen built environment factors extracted from green space and gray space and three meteorological factors were used as independent variables. Principal component analysis was adopted to reveal the relationship between built environment factors, meteorological factors, and PM2.5. Accordingly, 24 models were built using 32 training neighborhood samples. The results showed that the adj_R2 of most models was between 0.6 and 0.8, and the highest adj_R2 was 0.813. Four principal factors were the most important factors that significantly affected the growth and reduction of PM2.5, reflecting the differences in green and gray spaces, building height and its differences, relative humidity, openness, and other characteristics of the neighborhood. Furthermore, the relative error was used to test the error of the predicted values of five verification neighborhood samples, finding that these models had a high fitting degree and can better predict the growth and reduction of PM2.5 based on these built environment factors.
... RF models can handle complex nonlinear relationships and intercorrelations between input variables. Further, existing studies showed that RF regression models improved the accuracy of model prediction compared to traditional methods and can be used for classification and regression [61]. Importantly, RF models can measure variables' importance through permutation [62]. ...
... Importantly, RF models can measure variables' importance through permutation [62]. RF regression models have been employed to examine the relative contribution of various built environmental indicators on land surface temperature [63], PM 2.5 variations [61], and carbon emissions [64]. Therefore, an RF model can be adopted to investigate the significance of different UGS indicators and compare the contribution of UGS spatial pattern and quantity. ...
Since 2019, COVID-19 has triggered a renewed investigation of the urban environment and disease outbreak. While the results have been inconsistent, it has been observed that the quantity of urban green spaces (UGS) is correlated with the risk of COVID-19. However, the spatial pattern has largely been ignored, especially on the community scale. In high-density communities where it is difficult to increase UGS quantity, UGS spatial pattern could be a crucial predictive variable. Thus, this study investigated the relative contribution of quantity and spatial patterns of UGS on COVID-19 risk at the community scale using a random forest (RF) regression model based on (n = 44) communities in Wuhan. Findings suggested that 8 UGS indicators can explain 35% of the risk of COVID-19, and the four spatial pattern metrics that contributed most were core, edge, loop, and branch whereas UGS quantity contributed least. The potential mechanisms between UGS and COVID-19 are discussed, including the influence of UGS on residents’ social distance and environmental factors in the community. This study offers a new perspective on optimizing UGS for public health and sustainable city design to combat pandemics and inspire future research on the specific relationship between UGS spatial patterns and pandemics and therefore help establish mechanisms of UGS and pandemics.
... Random forest models are insensitive to multicollinearity, and their results are relatively robust to missing and unbalanced data. In addition, it performs well in analyzing data with small sample sizes (Chen et al., (2021a)). ...
Developing urban underground space for transportation and commercial purposes has become a vital strategy for sustainable urban growth, especially in dense urban contexts. The vitality of underground space has received wide attention as a key indicator of its efficiency and performance. The spatial vitality of underground space is affected by many factors, including function, spatial configuration, transport environment, and pedestrian-oriented design features. This study takes the underground space of Shanghai Hongqiao Business District (Phase I), one of the most successful integrated railway station-city development projects, as a case to explore the spatial and temporal distribution characteristics of spatial vitality and its influencing factors. Multilevel linear models were used to analyze the relationship between pedestrian flow and its influencing factors. Random forest models were used to investigate the relative importance of influencing factors on spatial vitality. This study verified that the distance to the transportation hub and spatial configurations calculated by space syntax are the two most essential factors for vitality. The number of retail stores and walkway width impact the vitality on weekends. The results shed light on our understanding of the urban vitality in the underground space and provide some tentative suggestions for the planning and design of underground space in high-density cities.
... Generally, air pollution is significantly positively correlated with floor area ratio and building density. High building density and complex block interfaces can cause higher friction coefficients on the urban subsurface, which can further reduce air circulation velocity and lead to air pollutants accumulation [65,67]. ...
Air pollution is a significant global environmental issue. Nevertheless, the importance of rational urban planning in mitigating it is frequently disregarded. Conducting air quality optimization simulations that consider UFIs (urban form indicators) is an effective approach for air quality-friendly urban planning. Despite its potential, this technique is still in its infancy. This study aims to identify research gaps and uncertainties by summarizing the current research status of key steps in air quality optimization simulation: selecting UFIs, analyzing impact mechanisms of UFIs on air pollution, building air quality models, and proposing optimization strategies through multiple-scenario predictions. Our findings indicate a lack of consistency in selecting UFIs, and an indicator system considers their impact on air quality and urban planning standards proposed by us might be a viable option. Increases in the proportion of construction land, industrial land, and floor area ratio significantly contribute to air pollution, whereas factors such as forest land, public green space, and sky openness have a noteworthy alleviating effect. Typically, statistical modeling is preferred at the city scale while CFD simulation techniques are used at smaller scales. However, future air quality improving through UFIs optimization still lacks a multi-scale nested tool. Thus, further research is recommended to explore combined impacts of various UFIs on different air pollutants under typical scenarios, and develop intelligent, big data-driven air quality models and tools. This review might contribute to transforming air quality optimization that consider UFIs from fragmented academic research to practical application, thereby assisting in global air pollution mitigation efforts.
... Generally, air pollution is significantly positively correlated with floor area ratio and building density. High building density and complex block interfaces can cause higher friction coefficients on the urban subsurface, which can further reduce air circulation velocity and lead to air pollutants accumulation [65,67]. ...
While significant reductions in certain air pollutant concentrations did not induce obvious mitigations of health risks, a shift from air quality management to health risk prevention and control might be necessary to protect public health. This study thus constructed an Air Quality Health Index (AQHI) for respiratory (Res-AQHI), cardiovascular (Car-AQHI), and allergic (Aller-AQHI) risk groups using mixed exposure under multi-air pollutants and portrayed their distribution and variation at multiple spatiotemporal scales using spatial analysis in GIS with the medical big data and air pollution remote sensing data by taking Hunan Province in China as a case. Results showed that the AQHIs constructed for specific health-risk groups could better express their risks than common AQHI and AQI. Moreover, based on the spatiotemporal association of health and environmental information, the allergic risk group in Hunan provided the highest health risk mainly affected by O3. The following cardiovascular and respiratory risk groups can be significantly attributed to NO2. Moreover, the spatiotemporal heterogeneity of AQHIs within regions was also evident. On the annual scale, the population in the air health risk hotspots for respiratory and cardiovascular risk decreased, while allergic risks increased. Meanwhile, on seasonal scale, the hotspots for respiratory and cardiovascular risks expanded significantly in winter while completely disappearing for allergic risk. These findings suggest that disease specific AQHIs effectively disclose the health effects of multi-air pollutants and their subsequently varied spatiotemporal distribution patterns.
... Morphological spatial pattern analysis (MSPA) has been used to identify cores with important ecological functions in rural ecosystem space, and eliminate the fine and fragmented patches among them [48][49][50][51][52]. In this study, Conefor 2.6 was used to measure the patch importance index of cores in order to select the parts with better ecological functions and better connectivity levels. ...
The inadequate consideration of livable rural spatial morphology in rural planning has impeded the further advancement of the rural social system, resulting in a challenge for rural residents to establish an appealing living experience that distinguishes itself from urban areas. This situation calls for an urgent exploration of livable spatial morphology based on human-centered principles, as well as an investigation of planning spatial morphology optimization mechanisms that consider ecological backgrounds and human settlement needs. In response to this issue, this study employs the theory of flow space and constructs a framework for the optimization of rural spatial methodology. By integrating ecological and sociological analysis methods, the study identifies the “flow” structure of spatial association in rural ecosystems through ecological network analysis, and identifies the “flow” structure of behavioral association in rural human systems through social network analysis. Based on these findings, the complex network morphologies are evaluated and screened. To test the effectiveness of this framework, the study examines the spatial morphology of four planning options through case empirical analysis in Zepan Village, Hebei Province, China. The research results demonstrate that the framework can help achieve the goal of optimizing rural spatial morphology, improve existing planning practices that prioritize single plans and disregard the selection of multiple plans, and serve as an effective tool to aid planners in tackling complex planning problems by balancing scientific principles and empirical values.
... Research on the relationship between air pollutants and the urban built environment focuses on two main aspects, namely, physical space and ecological space. Studies on physical space focus on the influence of land use and urban morphology on the distribution of pollutant concentrations (Chen et al., 2021a;Liu et al., 2021;Shi et al., 2018). Many studies have been conducted to provide evidence of the impacts of urban land use and urban form on air quality. ...
Concentrations of airborne particulate matter (PM) are influenced by land cover types. Water bodies in cities can influence the spatial distribution of air pollution by altering the microclimate. However, the influence of water bodies on PM2.5 concentrations is complex and requires further exploration, especially at the microscale. In this study, a geographically weighted regression (GWR) model was constructed to explore the spatially heterogeneous effect of a river on the PM2.5 concentrations in nearby riverside neighborhoods using mobile monitoring in Wuhan. The results showed that the GWR model was applicable at the neighborhood scale and had a good explanatory performance for PM2.5 concentrations, with a higher R2 (0.71). The river had multiple effects on the PM2.5 concentrations in the riverfront neighborhoods and strongly affected the air quality in neighborhoods within an 800 m distance due to wind infiltration, with the strongest effect at a 500–700 m distance. Moreover, considering spatial nonstationarity, the effect of the large river on street-level air quality largely depended on its effect on wind, and good ventilation conditions could amplify that effect. Commercial, road intersections, second-level roads and parks were identified as sensitive environmental factors affecting the river’s influence on PM2.5 concentrations. In addition, urban parks had a greater mitigating effect on PM2.5 pollution than did water bodies in this study. These results help to clarify the impact of rivers on air quality and provide a theoretical basis for urban design to mitigate PM2.5 pollution.
Fine particulate matter (PM 2.5 ) is a major source of air pollution and exerts serious impacts on human health. The 3D urban landscape patterns can significantly affect the diffusion and emissions of PM 2.5 . However, studies on the relationships between 3D urban landscape patterns and PM 2.5 pollution across different seasons remain understudied. With the ground-level air pollutants estimated by the remote sensing and fine-scale building information, this study applied the multiscale geographically weighted regression model to explore such relationships. Wuhan, the largest metropolis in Central China, was selected as the study area for the application of our methodology. The results showed that the direction, degree, and scale of the effect of 3D urban landscape patterns on PM 2.5 pollution varied across seasons. For building height, the standard deviation of building height had a significant positive correlation with PM 2.5 all year round. For building density, the building count density showed a significant positive correlation with PM 2.5 in general, with the bandwidth in winter and autumn smaller than in spring and summer. The building plan area fraction exerted both positive and negative influences on PM 2.5 , dependent on season and location. The bandwidth of it gradually increased from spring to winter, with the effect changing from local to regional scale. For building volume, the floor area ratio showed a significant negative correlation with PM 2.5 in winter and autumn, and a localized effect was found, especially in winter. The findings of this study provide practical implications for urban planning and policy making to mitigate PM 2.5 pollution in the rapidly urbanizing regions.
Greenery may be effective in mitigating particulate matter (PM) pollution. However, most previous studies have explored the effects of greenery on air quality from a two-dimensional rather than a three-dimensional (3D) perspective. In this study, a geographically weighted regression (GWR) model was constructed to explore the effect of street greenery on street-level PM 2.5 concentrations using street view imagery from a 3D perspective. Mobile monitoring of PM 2.5 concentrations was conducted in Wuhan. The results demonstrated that GWR is suitable for exploring street-level air quality. From a 3D perspective, street greenery strongly affects the air quality of the surrounding area within 300 m. Moreover, considering spatial nonstationarity, the negative impact of street greenery on air quality depends on the pollution and ventilation conditions. High pollution and canyon effects amplify the negative impact of street greenery on pollution dispersion. Thus, the negative effect of street greenery on air quality is sensitive to the density of second-level roads and of commercial areas, the number of intersections and the distance to rivers. Dense planting trees in these areas must be reconsidered based on the possibility of ventilation loss. These results provide insights into street-level PM 2.5 pollution and can contribute to the evidence-based design of urban green infrastructure.
The implementation of short-term traffic restriction policies (TRPs) during major events positively influences the traffic emission reduction. However, few studies explore the impact of diesel vehicle emissions on air quality during short-term TRP. Also, the intertwined influences of short-term TRP and Spring Festival remains unclear. Based on Beijing 2022 Olympic Games, this study analyzed the spatiotemporal changes in urban air quality and diesel vehicle emission during short-term TRP. The results showed that the TRPs and Spring Festival contributed 50 % to the improvement of air quality and reduction of diesel vehicle emissions, respectively. The "interruption-recovery" pattern of short-term TRPs is characterized by a longer duration and a slower decline/recovery rate. Additionally, the individual contribution rate of diesel vehicle emissions to urban air pollutants was 15-20 % higher than that of meteorological factors during short-term TRPs.
Introduction
There is strong evidence in Western cities that neighborhood-level characteristics of the built environment are linked with a higher risk of overweight or obesity among residents. Due to the rapid urbanization during the last several decades, many types of residential community have formed and coexisted in the major cities of China. These communities provide residents with different built-environment features and lifestyles. However, it remains unclear whether the community type affects the risk of overweight and obesity among residents.
Methods
The present study investigated the associations of built-environment characteristics and the bodyweight status (normal vs. overweight or obese) of 4,440 residents from three main types of community (i.e., commercial, work-unit, and traditional communities) in Shijingshan district, Beijing. Multilevel logistic regression and the random forest approach were adopted to investigate both the significance and relative importance of neighborhood-level factors of the built environment.
Results
The results of multilevel logistic regression suggest that the community type has a significant association with the obesity risk. In addition, the land-use mix, the number of water features, the number of supermarkets and groceries, street intersections, and the normalized difference vegetation index are negatively related to the odds of obesity. The number of transit stops is positively associated with the odds of obesity. Random forest analysis reveals significant disparities in the relative importance of population structure and the built environment factors among the three types of community. Furthermore, we find a notable difference between the results of the multilevel logistic model and random forest model. Hence, both the significance and relative importance of neighborhood-level factors of the built environment should be considered.
Conclusions
The findings indicate that the community type has a significant association with the risk of overweight. Tailored policies and urban renewal interventions should be developed for different community types.
The distribution transformer voltage may be overloaded, which may lead to the aging of distribution transformer components, shorten the service life of distribution transformer components and even affect the daily life of community residents and the operation of enterprises. A large amount of real data are collected, and the factors that affect the heavy overload of distribution transformer are comprehensively considered from multiple angles, so as to establish a model for future prediction and early maintenance to reduce losses. First, the collected data is analyzed by attributes and preprocessed to improve the quality of the data. Then, the time attributes are generalized according to seasons, months, holidays and weekends. The test results show that the data prediction value is more accurate when generalized according to seasons. For the prediction model, the gradient lifting decision tree algorithm is selected to establish the model, and then the parameters are further optimized, and finally the model is evaluated. Lastly, the prediction accuracy of the model reaches a high level, and it can be determined that the prediction is close to the objective fact. The model can be used to predict the heavy overload of distribution transformer voltage, so as to reduce the loss caused by abnormal conditions of relevant equipment for the enterprises.
The intraurban distribution of PM2.5 concentration is influenced by various spatial, socioeconomic, and meteorological parameters. This study investigated the influence of 37 parameters on monthly average PM2.5 concentration at the subdistrict level with Pearson correlation analysis and land-use regression (LUR) using data from a subdistrict-level air pollution monitoring network in Shenzhen, China. Performance of LUR models is evaluated with leave-one-out-cross-validation (LOOCV) and holdout cross-validation (holdout CV). Pearson correlation analysis revealed that Normalized Difference Built-up Index, artificial land fraction, land surface temperature, and point-of-interest (POI) numbers of factories and industrial parks are significantly positively correlated with monthly average PM2.5 concentrations, while Normalized Difference Vegetation Index and Green View Factor show significant negative correlations. For the sparse national stations, robust LUR modelling may rely on a priori assumptions in direction of influence during the predictor selection process. The month-by-month spatial regression shows that RF models for both national stations and all stations show significantly inflated mean values of R² compared with cross-validation results. For MLR models, inflation of both R² and R²CV was detected when using only national stations and may indicate the restricted ability to predict spatial distribution of PM2.5 levels. Inflated within-sample R² also exist in the spatiotemporal LUR models developed with only national stations, although not as significant as spatial LUR models. Our results suggest that a denser subdistrict level air pollutant monitoring network may improve the accuracy and robustness in intraurban spatial/spatiotemporal prediction of PM2.5 concentrations.
Characterizing urban morphology is critical for urban climate examination and modeling. However, high-resolution urban morphological datasets are lacking, especially in Chinese cities that undergoing fast-urbanization. We proposed a two-step rasterization method to develop an urban morphological dataset at a resolution of 100 m for 36 major cities in China. The morphological dataset includes building height, width, fraction, and street width as well as sky view factor and frontal area index. We then characterized the spatial patterns of these morphological parameters. Generally, the derived morphological parameters match the raw vector data well in terms of both magnitude and spatial distribution at the block, district, and city scales. The morphological parameters show large spatial variations within and across cities. On the city scale, the city center showed a larger building height and frontal area index, but smaller street width and sky view factor, compared to the city edge. Across cities, the morphological parameters generally show latitudinal variations, with higher building height and frontal area index, but smaller street width and sky view factor in the south. This new morphological dataset provides fundamental data to examine urban climate mechanism, classify urban land use, and drive urban climate model.
Air pollution has a serious fallout on human health, and the influences of the different urban morphological characteristics on air pollutants cannot be ignored. In this study, the relationship between urban morphology and air quality (wind speed, CO, and PM2.5) in residential neighborhoods at the meso-microscale was investigated. The changes in the microclimate and pollutant diffusion distribution in the neighborhood under diverse weather conditions were simulated by Computational Fluid Dynamics (CFD). This study identified five key urban morphological parameters (Building Density, Average Building Height, Standard Deviation of Building Height, Mean Building Volume, and Degree of Enclosure) which significantly impacted the diffusion and distribution of pollutants in the neighborhood. The findings of this study suggested that three specific strategies (e.g. volume of a single building should be reduced, DE should be increased) and one comprehensive strategy (the width and height of the single building should be reduced while the number of single buildings should be increased) could be illustrated as an optimized approach of urban planning to relief the air pollution. The result of the combined effects could provide a reference for mitigating air pollution in sustainable urban environments.
Many studies link long-term fine particle (PM2.5) exposure to mortality, even at levels below current US air quality standards (12 μg/m3). These findings have been disputed, with claims that the use of traditional statistical approaches does not guarantee evidence of causality. Leveraging 16 years of data—68.5 million Medicare enrollees and 570 million observations—we provide strong evidence of the causal link between long-term PM2.5 exposure and mortality under a set of assumptions necessary for causal inference. Using five distinct statistical approaches, we found that a decrease of 10 μg/m3 PM2.5 leads to a statistically significant 6%–7% decrease in mortality risk. Based on these models, lowering the air quality standard to 10 μg/m3 would save 143,257 lives (95% confidence interval 115,581–170,645) in one decade. Our study provides the most comprehensive evidence to date of the link between long-term PM2.5 exposure and mortality, even at levels below current standards.
Estimating air pollution exposure has long been a challenge for environmental health researchers. Technological advances and novel machine learning methods have allowed us to increase the geographic range and accuracy of exposure models, making them a valuable tool in conducting health studies and identifying hotspots of pollution. Here, we have created a prediction model for daily PM 2.5 levels in the Greater London area from 1st January 2005 to 31st December 2013 using an ensemble machine learning approach incorporating satellite aerosol optical depth (AOD), land use, and meteorological data. The predictions were made on a 1 km × 1 km scale over 3960 grid cells. The ensemble included predictions from three different machine learners: a random forest (RF), a gradient boosting machine (GBM), and a k-nearest neighbor (KNN) approach. Our ensemble model performed very well, with a tenfold cross-validated R 2 of 0.828. Of the three machine learners, the random forest outperformed the GBM and KNN. Our model was particularly adept at predicting day-today changes in PM 2.5 levels with an out-of-sample temporal R 2 of 0.882. However, its ability to predict spatial variability was weaker, with a R 2 of 0.396. We believe this to be due to the smaller spatial variation in pollutant levels in this area.
Air pollution in the form of particulate matter (PM) is becoming one of the greatest current threats to human health on a global scale. This paper firstly presents a Bayesian space–time hierarch piecewise regression model (BSTHPRM) which can self-adaptively detect the transitions of local trends, accounting for spatial correlations. The spatiotemporal trends of the approximately anthropogenic PM2.5 removed natural dust (PM2.5_No Dust) concentrations and the corresponding population's PM2.5_No Dust exposure (PPM2.5E) in the global continent from 1998 to 2016 were investigated by the presented BSTHPRM. The total areas of the high and higher PM2.5_No Dust-polluted regions, whose spatial relative magnitude of PM2.5_NoDust pollution to the global continental overall level was between 1.89 and 14.68, accounted for about 13.4% of the global land area, and the corresponding exposed populations accounted for 56.0% of the global total population. The spatial heterogeneity of the global PM2.5_NoDust pollution increased generally from 1998 to 2016. The areas of hot, warm, and cold spots with increasing trends of PM2.5_NoDust concentration initially contracted and then later expanded. The local trends of the global continental PM2.5_NoDust concentrations and PPM2.5E can be parted into three changing stages, early, medium, and later stages, using the BSTHPRM. The area proportions of the regions experiencing a decreasing trend of PM2.5_NoDust concentrations and PPM2.5E were greater in the medium stage than in the early and later stages. The local trends of PM2.5_NoDust concentration and PPM2.5E in the two higher PM2.5_NoDust polluted areas, northern India and eastern and southern China, increased in the early stage and then decreased in the medium stage. In the later stage (recent years), northern India displayed a stronger increasing trend; nevertheless, the follow-up decreasing trend still occurred in eastern and southern China. In the first two stages, more than half of the areas in Europe experienced a decreasing trend of PM2.5_NoDust concentration and PPM2.5E; later, more than half of areas in Europe exhibited increasing trends in the later stage. North America and South America experienced a similar local trend of PPM2.5E to Europe. The PPM2.5E trend in Africa generally increased during the study period. Keywords: Spatiotemporal trends, Global PM2.5 pollution, Population's PM2.5 exposure, Bayesian statistics
Haze, especially PM 2.5 , poses a serious threat to public health in China. PM 2.5 primarily originates from urban activities, and built environment may affect its formation and dispersion. Previous studies were based on limited data from ground-monitoring stations, and high resolution pollution maps are unavailable for statistical analyses. In this study, a 1 km*1 km wall-to-wall map of PM 2.5 concentration is developed with remote sensing data in Wuhan, China, and spatial statistics are used to figure out the influence of the built environment on PM 2.5 concentrations. In terms of land cover, high-rise high-density building areas have the largest impact on PM 2.5 concentrations, and the effect of forestland on the concentrations is not obvious in winter. In terms of land use, industrial lands are unrelated to air pollution in the downtown, while transportation has become a main source of PM 2.5 pollution. In terms of urban form, floor area ratio and building density are positively associated with PM 2.5 concentrations, and different types of road densities have different effects on air pollution. Finally, the implications of the study for urban planning and development are given. It is necessary to develop a polycentric urban structure to balance high population density and reduce traffic emissions in downtown areas. Road and bus networks should be optimized simultaneously to reduce traffic emissions and “small blocks and narrow roads” may be considered as an alternative for urban development. The spatial morphology of streets and buildings should be considered during urban design and urban renewal. In general, the study contributes to the application of remote sensing in urban planning and development, and remotely sensed PM 2.5 concentration data could provide further findings than the air pollution data obtained from ground monitoring and “bottom-up” models in past studies.
This case study was conducted to quantify the effects of urban greenspace patterns on particle matter (PM) concentration in Zhengzhou, China by using redundancy and variation partitioning analysis. Nine air-quality monitoring stations (AQMS) were selected as the central points. Six distances of 1 km, 2 km, 3 km, 4 km, 5 km, and 6 km were selected as the side lengths of the squares with each AQMS serving as the central point, respectively. We found: (1) the fine size of PM (PM2.5) and coarse size of PM (PM10) among four seasons showed significant differences; during winter, the concentration of PM2.5 and PM10 were both highest, and PM2.5 and PM10 concentration in summer were lowest. (2) To effectively reduce the PM2.5 pollution, the percentage of greenspace, the differences in areas among greenspace patches, and the edge complexity of greenspace patches should be increased at distances of 2 km and 3 km. To effectively reduce PM10, the percentage of greenspace at a distance of 4 km, the edge density at distances of 2 km and 4 km, and the average area of greenspace patches at a distance of 1 km should be increased. (3) Greenspace pattern significantly affected PM2.5 at a distance of 3 km, and PM10 at a distance of 4 km. From shorter distance to longer distance, the proportion of variance explained by greenspace showed a decline–increase–decline–increase trend for PM2.5, and a decline–increase–decline trend for PM10. At shorter distances, the composition of greenspace was more effective in reducing the PM pollution, and the configuration of greenspace played a more important role at longer distances. The results should lead to specific guidelines for more cost-effective and environmentally sound greenspace planning.
Understanding how urban form is related to air pollution is important to urban planning and sustainability, but the urban form-air pollution relationship is currently muddled by inconsistent findings. In this study, we investigated how the compositional and configurational attributes of urban form were related to different air pollution measures (PM2.5, API, and exceedance) in 83 Chinese cities, with explicit consideration of city size and seasonality. Ten landscape metrics were selected to quantify urban form attributes, and Spearman's correlation was used to quantify the urban form-air pollution relationship. Our results show that the urban form and air pollution relationship was dominated by city size and moderated by seasonality. Specifically, urban air pollution levels increased consistently and substantially from small to medium, large, and megacities. The urban form-air pollution relationship depended greatly on seasonality and monsoons. That is, the relationship was more pronounced in spring and summer than fall and winter, as well as in cities affected by monsoons. Urban air pollution was correlated more strongly with landscape composition metrics than landscape configuration metrics which seemed to affect only PM2.5concentrations. Our study suggests that, to understand how air pollution levels are related to urban form, city size and seasonality must be explicitly considered (or controlled). Also, in order to mitigate urban air pollution problems, regional urban planning is needed to curb the spatial extent of built-up areas, reduce the degree of urban fragmentation, and increase urban compactness and contiguity, especially for large and megacities.
Deposition of particulate matter (PM) from air is influenced by land use types and their attributes. Urban lakes wetlands can promote this deposition through their effects on the microclimate. To document these effects we investigated air PM10 and PM2.5 concentrations in transects around by monitoring PM and land use of 16 urban lake wetlands in the spring in Wuhan. The results showed that (1) lake wetland land-use regression models for air PM10,2.5 were successfully developed with an adjusted R²(PM10) (0.365–0.982) and adjusted R²(PM2.5) (0.483–0.969). Based on the lake wetland land use regression (LW LUR) model, the traffic variable was determined to be the strongest predictor of PM10,2.5, followed by distance to the city centre and relative humidity for PM10 and open green space and distance to the city centre for PM2.5. (2) The landscape descriptors of the lake wetland (less than 0.2 km² lake area) significantly affected PM10,2.5 in the central areas of the city, and air PM10,2.5 concentration was negatively correlated with wetland area (WA), landscape shape index (LSI), and proportion of non-built up area in the 500-m buffer zone surrounding the lake wetland (PB). (3) The lake wetlands showed decreasing air PM10,2.5 concentration, and the maximum air PM10,2.5 concentration gap between the wetlands and their surroundings occurred at 12:00–14:00 am and was from 10 to 50 μg/m³ for PM10 and from 5 to 15 μg/m³ for PM2.5. Effective understanding of these PM10,2.5 effects can contribute to better planning of urban development when considering the rising concerns of global air pollution and continued rapid urbanization.
In the last years many accurate decision support systems have been constructed as black boxes, that is as systems that hide their internal logic to the user. This lack of explanation constitutes both a practical and an ethical issue. The literature reports many approaches aimed at overcoming this crucial weakness sometimes at the cost of scarifying accuracy for interpretability. The applications in which black box decision systems can be used are various, and each approach is typically developed to provide a solution for a specific problem and, as a consequence, delineating explicitly or implicitly its own definition of interpretability and explanation. The aim of this paper is to provide a classification of the main problems addressed in the literature with respect to the notion of explanation and the type of black box system. Given a problem definition, a black box type, and a desired explanation this survey should help the researcher to find the proposals more useful for his own work. The proposed classification of approaches to open black box models should also be useful for putting the many research open questions in perspective.
The city of London, UK, has seen in recent years an increase in the number of high-rise/multi-storey buildings ("skyscrapers") with roof heights reaching 150 m and more, with the Shard being a prime example with a height of ∼310 m. This changing cityscape together with recent plans of local authorities of introducing Combined Heat and Power Plant (CHP) led to a detailed study in which CFD and wind tunnel studies were carried out to assess the effect of such high-rise buildings on the dispersion of air pollution in their vicinity. A new, open-source simulator, FLUIDITY, which incorporates the Large Eddy Simulation (LES) method, was implemented; the simulated results were subsequently validated against experimental measurements from the EnFlo wind tunnel. The novelty of the LES methodology within FLUIDITY is based on the combination of an adaptive, unstructured, mesh with an eddy-viscosity tensor (for the sub-grid scales) that is anisotropic. The simulated normalised mean concentrations results were compared to the corresponding wind tunnel measurements, showing for most detector locations good correlations, with differences ranging from 3% to 37%. The validation procedure was followed by the simulation of two further hypothetical scenarios, in which the heights of buildings surrounding the source building were increased. The results showed clearly how the high-rise buildings affected the surrounding air flows and dispersion patterns, with the generation of "dead-zones" and high-concentration "hotspots" in areas where these did not previously exist. The work clearly showed that complex CFD modelling can provide useful information to urban planners when changes to cityscapes are considered, so that design options can be tested against environmental quality criteria.
Data samples (temperature and thermal drifts) obtained in motorized spindle thermal experiments are usually small and contain much information. Some of the information cannot be fully comprehended by most regression modeling methods when modeling with small training samples; hence, the modeled thermal error predictors can seriously lack robustness, especially for the thermal tilt angle (vital for the boring accuracy of precision boring machines) predictors. To solve this problem, the LS-MLR (least-squares multivariable linear regression), the GA-SVR (genetic algorithm-support vector machine for regression), and the RFR (random forest regression) regression modeling methods are applied to construct the thermal error predictors (models) with the training sample, and the predictors are then evaluated with the testing sample. Comparisons of the three modeling methods are carried out afterwards; it is pointed out that predicting ability of the bias-corrected RFR predictor for thermal elongation is close to the GA-SVR predictor, and as for the thermal pitch and the thermal yaw, predicting ability of the RFR predictors are superior to the LS-MLR and the GA-SVR predictors. Finally, the evaluation results also indicate that the proposed RFR thermal error modeling method is promising to be used in motorized spindle thermal error predicting in machining processes even when the modeling data are poor.
In recent years, China has experienced severe and persistent air pollution associated with rapid urbanization and climate change. Three years' time series (January 2014 to December 2016) concentrations data of air pollutants including particulate matter (PM 2.5 and PM 10) and gaseous pollutants (SO 2 , NO 2 , CO, and O 3) from over 1300 national air quality monitoring sites were studied to understand the severity of China's air pollution. (67%, 70%) and 88% (97%, 98%) of the population in China lived in areas that meet the level of annual PM 2.5 , PM 10 , NO 2 , and SO 2 standard metrics from Chinese Ambient Air Quality Standards-Grade II. The annual PWA concentrations of PM 2.5 , PM 10 , O 3 _8 h, NO 2 , SO 2 , CO in the Northern China are about 40.4%, 58.9%, 5.9%, 24.6%, 96.7%, and 38.1% higher than those in Southern China, respectively. Though the air quality has been improving recent years, PM 2.5 pollution in wintertime is worsening, especially in the Northern China. The complex air pollution caused by PM and O 3 (the third frequent major pollutant) is an emerging problem that threatens the public health, especially in Chinese mega-city clusters. NOx controls were more beneficial than SO 2 controls for improvement of annual PM air quality in the northern China, central, and southwest regions. Future epidemiologic studies are urgently required to estimate the health impacts associated with multi-pollutants exposure , and revise more scientific air quality index standards.
Fine particulate matter (PM2.5) pollution has become one of the greatest urban issues in China. Studies have shown that PM2.5 pollution is strongly related to the land use pattern at the micro-scale and optimizing the land use pattern has been suggested as an approach to mitigate PM2.5 pollution. However, there are only a few researches analyzing the effect of land use on PM2.5 pollution. This paper employed land use regression (LUR) models and statistical analysis to explore the effect of land use on PM2.5 pollution in urban areas. Nanchang city, China, was taken as the study area. The LUR models were used to simulate the spatial variations of PM2.5 concentrations. Analysis of variance and multiple comparisons were employed to study the PM2.5 concentration variances among five different types of urban functional zones. Multiple linear regression was applied to explore the PM2.5 concentration variances among the same type of urban functional zone. The results indicate that the dominant factor affecting PM2.5 pollution in the Nanchang urban area was the traffic conditions. Significant variances of PM2.5 concentrations among different urban functional zones throughout the year suggest that land use types generated a significant impact on PM2.5 concentrations and the impact did not change as the seasons changed. Land use intensity indexes including the building volume rate, building density, and green coverage rate presented an insignificant or counter-intuitive impact on PM2.5 concentrations when studied at the spatial scale of urban functional zones. Our study demonstrates that land use can greatly affect the PM2.5 levels. Additionally, the urban functional zone was an appropriate spatial scale to investigate the impact of land use type on PM2.5 pollution in urban areas.
As fine particle (FP) pollution is harmful to humans, previous studies have focused on the mechanisms of FP removal by forests. The current study aims to compare the FP removal capacities of urban forests and wetlands on the leaf, canopy, and landscape scales. Water washing and scanning electron microscopy are used to calculate particle accumulation on leaves, and models are used to estimate vegetation collection, sedimentation, and dry deposition. Results showed that, on the leaf scale, forest species are able to accumulate more FP on their leaf surface than aquatic species in wetlands. On the canopy scale, horizontal vegetation collection is the major process involved in FP removal, and the contribution of vertical sedimentation/emission can be ignored. Coniferous tree species also showed stronger FP collection ability than broadleaf species. In the landscape scale, deposition on the forest occurs to a greater extent than that on wetlands, and dry deposition is the major process of FP removal on rain-free days. In conclusion, when planning an urban green system, planting an urban forest should be the first option for FP mitigation.
With the increase in urbanization and energy consumption, PM2.5 has become a major pollutant. This paper investigates the impact of road patterns on PM2.5 pollution in Beijing, focusing on two questions: Do road patterns significantly affect PM2.5 concentrations? How do road patterns affect PM2.5 concentrations? A land-use regression model (LUR model) is used to quantify the associations between PM2.5 concentrations, and road patterns, land-use patterns, and population density. Then, in the condition of excluding other factors closely correlated to PM2.5 concentrations, based on the results of the regression model, further research is conducted to explore the relationship between PM2.5 concentrations and the types, densities, and layouts of road networks, through the controlling variables method. The results are as follows: (1) the regression coefficient of road patterns is significantly higher than the water area, population density, and transport facilities, indicating that road patterns have an obvious influence on PM2.5 concentrations; (2) under the same traffic carrying capacity, the layout of "a tight network of streets and small blocks" is superior to that of "a sparse network of streets and big blocks"; (3) the grade proportion of urban roads impacts the road patterns' rationality, and a high percentage of branch roads and secondary roads could decrease PM2.5 concentrations. These findings could provide a reference for the improvement of the traffic structure and air quality of Beijing.
This paper presents the relations established between urban morphology and variations in observed air quality within a city centre. A dataset from four monitoring stations in a medium-sized city was collected and clustered by the amount of traffic-generated air pollutants. For each cluster, the relationship between the configuration of the open spaces and average PM10 concentration was established. Results show the impact of urban geometry on the outdoor pollutant concentration, concluding that increasing Sky View Factor and Ratio of Open Spaces lead to a decrease in the PM10 concentration.
Fine particulate matter (PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
) pollution can cause serious public health problems worldwide. A novel geographically and temporally weighted neural network constrained by global training (GC-GTWNN) is proposed in this article for the remote sensing estimation of surface PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
. The global neural network (NN) is trained to learn the overall effect of the influencing variables on surface PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
, and the local geographically and temporally weighted NN (GTWNN) addresses the spatiotemporal heterogeneity of the relationship between PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
and the influencing variables. Specifically, a global NN is trained with all samples collected from the entire study domain and period. Then, initialized with the global NN, the GTWNN models are built for each location and time and fine-tuned via spatiotemporally localized samples. Meanwhile, the geographically weighted loss function is designed for GTWNN. The proposed GC-GTWNN modeling is tested with a case study across China, which integrates satellite aerosol optical depth, surface PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
measurements, and auxiliary variables. Cross-validation results indicate that a remarkable improvement is observed from the global NN to GC-GTWNN modeling (
$R^{2}$
value increasing from 0.49 to 0.80), and GC-GTWNN modeling also notably outperforms the conventionally popular PM
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub>
estimation models.
Accurate understanding of the relationship between urban land morphology and the concentration of PM2.5 is essential for achieving high-quality development of urban agglomerations. Based on a mechanism framework of "Internal-External driving force", 19 Chinese urban agglomerations at different development levels were analysed using the geographically weighted regression model to evaluate the impacts of urban land morphology on PM2.5 concentrations in years 2000–2017. The results show: (1) The PM2.5 average concentrations of all 19 urban agglomerations continue to increase from 30 μg/m³ in 2000 to 52 μg/m³ in 2007 but decreased to 34 μg/m³ in 2017. The changes in PM2.5 concentrations vary for urban agglomerations at different development levels. Spatial differences in PM2.5 concentrations are significant, forming a pattern that decreases from the centre to the periphery regions; (2) The urban land morphology of the entire urban agglomeration areas has undergone significant changes. The fractal dimension index (from 4.150 to 2.731) and the compactness (from 0.647 to 0.635) showed a downward trend, while the shape indices (from 1.421 to 1.606) demonstrated an increasing trend. National-level urban agglomerations are more compact and more complex in shape, while more fragmented are regional and local urban agglomerations; (3) Different parameters of urban land morphology have varying effects on PM2.5 concentration varies and at different development levels of urban agglomerations. The combination of urban land morphology, socio-economic factors, and natural elements has a complex effect on PM2.5 concentrations. It can contribute to understanding the linkage between urban land morphology and PM2.5, providing references for future studies.
Urban areas contribute substantially to human exposure to ambient air pollution. Numerous statistical prediction models have been used to estimate ambient concentrations of fine particulate matter (PM2.5) and other pollutants in urban environments, with some incorporating machine learning (ML) algorithms to improve predictive power. However, many ML approaches for predicting ambient pollutant concentrations to date have used principal component analysis (PCA) with traditional regression algorithms to explore linear correlations between variables and to reduce the dimensionality of the data. Moreover, while most urban air quality prediction models have traditionally incorporated explanatory variables such as meteorological, land use, transportation/mobility, and/or co-pollutant factors, recent research has shown that local emissions from building infrastructure may also be useful factors to consider in estimating urban pollutant concentrations. Here we propose an enhanced ML approach for predicting urban ambient PM2.5 concentrations that hybridizes cascade and PCA methods to reduce the dimensionality of the data-space and explore nonlinear effects between variables. We test the approach using different durations of time series air quality datasets of hourly PM2.5 concentrations from three air quality monitoring sites in different urban neighborhoods in Chicago, IL to explore the influence of dynamic human-related factors, including mobility (i.e., traffic) and building occupancy patterns, on model performance. We test 9 state-of-the-art ML algorithms to find the most effective algorithm for modeling intraurban PM2.5 variations and we explore the relative importance of all sets of factors on intraurban air quality model performance. Results demonstrate that Gaussian-kernel support vector regression (SVR) was the most effective ML algorithm tested, improving accuracy by 118% compared to a traditional multiple linear regression (MLR) approach. Incorporating the enhanced approach with SVR algorithm increased model performance up to 18.4% for yearlong and 98.7% for month-long hourly datasets, respectively. Incorporating assumptions for human occupancy patterns in dominant building typologies resulted in improvements in model performance by between 4% and 37%. Combined, these innovations can be used to improve the performance and accuracy of urban air quality prediction models compared to conventional approaches.
Air pollution is one of the hot issues that attracted widespread attention from urban and society management. Air quality prediction is to issue an alarm when severe pollution occurs, or pollution concentration exceeds a specific limit, contributing to the measure-taking of relevant departments, guiding urban socioeconomic activities to promote sustainable urban development. However, existing methods have failed to make full use of the temporal features from spatiotemporal correlations of air quality monitoring stations, and achieved poor performances in long-term predictions (up to or above 24h-predictions). In this study, we proposed a deep learning framework to predict air quality in the following 24 h: a neural network with a temporal sliding long short-term memory extended model (TS-LSTME). The model integrated the optimal time lag to realize sliding prediction through multi-layer bidirectional long short-term memory (LSTM), involving the hourly historical PM 2.5 concentration, meteorological data, and temporal data. We applied the proposed model to predict the next 24 h average PM 2.5 concentration in Jing-Jin-Ji region, with the most severe air pollution in China. The proposed model had better stability and performances with high correlation coefficient R 2 (0.87), compared to the multiple linear regression (MLR), the support vector regression (SVR), the traditional LSTM, and the long short-term memory extended (LSTME) models. Moreover, the proposed model can achieve PM 2.5 concentration predictions with high accuracy in long-term series (48 h and 72 h). We also tested the model to predict O 3 concentration. The proposed model could be applied for other air pollutants. The proposed methods can significantly improve air quality prediction information services for the public and provide support for early warning and management of regional pollutants.
This study explores the relationship between urban form at the metropolitan and neighborhood scale and fine particulate matter (PM2.5) concentrations by establishing multi-level regression models. This study has different assumptions about urban form depending on the scale used. While at the metropolitan scale, the urban form is related to the change in travel behavior, at the neighborhood scale, it is related to proximate emission sources such as roads, emissions facilities, employment centers, etc. The study shows that at the metropolitan scale, higher urban fragments, population density, and road density are associated with higher PM2.5 concentrations; higher job-resident balance and accessibility to schools are associated with lower PM2.5. At the neighborhood scale, a higher density of the nearby emission sources and higher accessibility to destinations are associated with higher PM2.5 concentrations. Urban fragments and land use mix have consistent impacts on air quality compared to preexisting studies. While population density and road density have two conflicting assumptions regarding PM2.5, in this study, the net effect of population density and road density on air quality is negative. Accessibility to destinations has different associations with PM2.5 depending on the scale of urban form measurement. At the metropolitan scale, high accessibility to destinations lessens PM2.5 concentrations by reducing vehicle distance. On the other hand, at the neighborhood scale, high accessibility to destinations makes it close to areas, which concentrate air emissions. We suggest that urban planners and decision-makers establish different strategies depending on urban form types when there are urban development plans.
The integration of satellite-derived aerosol optical depth (AOD) and station-measured PM2.5 (particulate matter with an aerodynamic diameter of less than 2.5 μm) provides a promising approach for the monitoring of PM2.5. Previous models have generally only considered either the spatiotemporal heterogeneities of the AOD-PM2.5 relationship or the nonlinear relationship between AOD and PM2.5. In this paper, to simultaneously allow for the nonlinearity and spatiotemporal heterogeneities of the AOD-PM2.5 relationship, the geographically and temporally weighted neural network (GTWNN) model is proposed for the satellite-based estimation of ground-level PM2.5. The GTWNN model represents the nonlinear AOD-PM2.5 relationship via a generalized regression neural network, and is separately established for individual locations and times, to address the spatiotemporal heterogeneities of the AOD-PM2.5 relationship. Meanwhile, a spatiotemporal weighting scheme is incorporated in the GTWNN model to capture the local relations of samples for the training of the AOD-PM2.5 relationship. By the use of the Moderate Resolution Imaging Spectroradiometer (MODIS) AOD product, meteorological data, and MODIS normalized difference vegetation index (NDVI) data as input, the GTWNN model was verified using ground station PM2.5 measurements from China in 2015. The GTWNN model achieved sample-based cross-validation (CV) and site-based CV R² values of 0.80 and 0.79, respectively, and it outperformed the geographically and temporally weighted regression model (CV R²: 0.75 and 0.73) and the daily geographically weighted regression model (CV R²: 0.72 and 0.72). The proposed model implements the combination of geographical law and a neural network, and will be of great use for remote sensing retrieval of environmental variables.
The complex mixtures of local emission sources and regional transportations of air pollutants make accurate PM2.5 prediction a very challenging yet crucial task, especially under high pollution conditions. A symbolic representation of spatio-temporal PM2.5 features is the key to effective air pollution regulatory plans that notify the public to take necessary precautions against air pollution. The self-organizing map (SOM) can cluster high-dimensional datasets to form a meaningful topological map. This study implements the SOM to effectively extract and clearly distinguish the spatio-temporal features of long-term regional PM2.5 concentrations in a visible two-dimensional topological map. The spatial distribution of the configured topological map spans the long-term datasets of 25 monitoring stations in northern Taiwan using the Kriging method, and the temporal behavior of PM2.5 concentrations at various time scales (i.e., yearly, seasonal, and hourly) are explored in detail. Finally, we establish a machine learning model to predict PM2.5 concentrations for high pollution events. The analytical results indicate that: (1) high population density and heavy traffic load correspond to high PM2.5 concentrations; (2) the change of seasons brings obvious effects on PM2.5 concentration variation; and (3) the key input variables of the prediction model identified by the Gamma Test can improve model's reliability and accuracy for multi-step-ahead PM2.5 prediction. The results demonstrated that machine learning techniques can skillfully summarize and visibly present the clusted spatio-temporal PM2.5 features as well as improve air quality prediction accuracy.
Accurate and high temporal resolution predictions of fine particulate matter (PM2.5) and nitrogen oxides (NOx) concentrations are crucial for pollution control, air pollutant exposure and epidemiological studies. This study aimed to develop machine learning algorithms (MLAs)-based models for predicting hourly street-level PM2.5 and NOx concentrations at three roadside stations in Hong Kong. We comprehensively evaluated and compared the performance of six common MLAs including Random Forest (RF), Boosted Regression Trees (BRT), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Generalized Additive Model (GAM), and Cubist and hence applied the most suitable MLAs to apportion the contributions from emission and non-emission factors to hourly street-level PM2.5 and NOx concentrations. The results show that RF outperforms other MLAs with ten-fold cross validation (CV) R² values higher than 0.81 and 0.62 for PM2.5 and NOx predictions, respectively. BRT, XGBoost and Cubist presented comparable predictive performance, with CV R² of 0.79–0.83 (for PM2.5 predictions) and 0.59–0.71 (for NOx predictions). SVM and GAM had worse predictions than other MLAs. The external validation R² values for RF and BRT models were more than 0.62 and 0.51 for PM2.5 and NOx concentration predictions, respectively. Non-emission factors contributed 84% and 65% to the predictions of street-level PM2.5 and NOx concentrations, respectively. Non-local pollution and temperature were the major non-emission factors, whereas private cars were the major emission contributor. This study highlights the capability of MLAs to produce high temporal resolution air pollution predictions, which can supplement traditional methods (e.g., land use regression) in generating accurate and high-temporal-resolution estimations of air pollution concentration.
Previous studies on air pollution exposure have mostly focused on the urban, regional, national and global scales, but the outcomes cannot well support risk assessments in urban communities. To determine how the urban form and meteorology influence the air pollution distribution at a neighborhood scale (2 km*2 km), we performed a fine-scale investigation of a typical air pollutant (i.e., PM 2.5) by mobile measurements in two communities in the inner and outer cities of Shanghai, China. The PM 2.5 mass concentrations and potential impacting factors, such as PM 2.5 background levels, road networks, traffic volumes, meteorological parameters, building heights and land use types, were collected at a 10 m spatial resolution, and their relationships were analyzed using both a generalized additive model (GAM) and land use regression (LUR). The modeling results showed that the GAM outperformed LUR in both study areas, with a higher adjusted R 2 and a lower RMSE. The PM 2.5 was found to have drastic variations at the neighborhood scale, which was primarily driven by spatial patterns of the PM 2.5 background levels and traffic volumes. The GAM-based PM 2.5 concentration surface clearly disclosed the heterogeneous variation of the PM 2.5 mass concentrations in each community, and demonstrated the prominent influence from the nearby highway, especially in the central urban area. This study provides the first empirical evidence of the exposure heterogeneity of the air pollution on a neighborhood scale, which can assist planners and policymakers in evaluating planning strategies with a full consideration of reducing local air pollution.
PM2.5 poses a serious threat to public health, however its spatial concentrations are not well characterized due to the sparseness of regulatory air quality monitoring (AQM) stations. This motivates novel low-cost methods to estimate ground-level PM2.5 at a fine spatial resolution so that PM2.5 exposure in epidemiological research can be better quantified. Satellite-retrieved aerosol products are widely used to estimate the spatial distribution of ground-level PM2.5. However, these aerosol products can be subject to large uncertainties due to many approximations and assumptions made in multiple stages of their retrieval algorithms. Therefore, estimating ground-level PM2.5 directly from satellites (e.g. satellite images) by skipping the intermediate step of aerosol retrieval can potentially yield lower errors because it avoids retrieval error propagating into PM2.5 estimation and is desirable compared to current ground-level PM2.5 retrieval methods. Additionally, the spatial resolutions of estimated PM2.5 are usually constrained by those of the aerosol products and are currently largely at a comparatively coarse 1 km or greater resolution. Such coarse spatial resolutions are unable to support scientific studies that thrive on highly spatially-resolved PM2.5. These limitations have motivated us to devise a computer vision algorithm for estimating ground-level PM2.5 at a high spatiotemporal resolution by directly processing the global-coverage, daily, near real-time updated, 3 m/pixel resolution, three-band micro-satellite imagery of spatial coverages significantly smaller than 1 × 1 km (e.g., 200 × 200 m) available from Planet Labs. In this study, we employ a deep convolutional neural network (CNN) to process the imagery by extracting image features that characterize the day-to-day dynamic changes in the built environment and more importantly the image colors related to aerosol loading, and a random forest (RF) regressor to estimate PM2.5 based on the extracted image features along with meteorological conditions. We conducted the experiment on 35 AQM stations in Beijing over a period of ∼3 years from 2017 to 2019. We trained our CNN-RF model on 10,400 available daily images of the AQM stations labeled with the corresponding ground-truth PM2.5 and evaluated the model performance on 2622 holdout images. Our model estimates ground-level PM2.5 accurately at a 200 m spatial resolution with a mean absolute error (MAE) as low as 10.1 μg m⁻³ (equivalent to 23.7% error) and Pearson and Spearman r scores up to 0.91 and 0.90, respectively. Our trained CNN from Beijing is then applied to Shanghai, a similar urban area. By quickly retraining only RF but not CNN on the new Shanghai imagery dataset, our model estimates Shanghai 10 AQM stations' PM2.5 accurately with a MAE and both Pearson and Spearman r scores of 7.7 μg m⁻³ (18.6% error) and 0.85, respectively. The finest 200 m spatial resolution of ground-level PM2.5 estimates from our model in this study is higher than the vast majority of existing state-of-the-art satellite-based PM2.5 retrieval methods. And our 200 m model's estimation performance is also at the high end of these state-of-the-art methods. Our results highlight the potential of augmenting existing spatial predictors of PM2.5 with high-resolution satellite imagery to enhance the spatial resolution of PM2.5 estimates for a wide range of applications, including pollutant emission hotspot determination, PM2.5 exposure assessment, and fusion of satellite remote sensing and low-cost air quality sensor network information.
A thorough understanding of the spatiotemporal variations of PM2.5 concentrations is crucial for mitigating PM2.5 pollution with effective measures; however, few studies have investigated this issue based on the neighborhood scale. This study investigated the spatiotemporal variations of PM2.5 concentrations in 40 sites from five megacities by using hourly PM2.5 concentrations derived from air quality monitoring stations. Two relative indicators—the range and rate of change in PM2.5 concentration—were calculated to exclude the background levels of PM2.5 in different cities. Therefore, the differences between both the regions and between the sites within the same city were taken into consideration. Results showed that the within-city differences in PM2.5 concentration gradually increased from 2015 to 2017, and the differences were greater at a lower pollution level. Neighborhood-level PM2.5 concentration fluctuated mainly between 80% and 120% of each city's overall PM2.5 level. Combined with urban structure, the distribution of PM2.5 concentration during the four seasons presented three spatial patterns: the PM2.5 concentration was higher in the transition area, the PM2.5 concentration was higher in the core area than other areas, and the PM2.5 concentration had a relatively even distribution. Furthermore, the regional differences in the variations of PM2.5 concentrations depended on the PM2.5 concentrations differences among the cities; and the higher the PM2.5 pollution level, the greater was the observed regional difference. In addition, the range and rate of change in PM2.5 concentration had no significant differences among most sites at different pollution levels.
PM2.5 (particles <2.5 μm in aerodynamic diameter) has become the primary pollutant in the air of most cities in China, and it is an important index reflecting the degree of air pollution. In this study, the response of the PM2.5 concentration in the air to multiple factors reflecting the meteorological, underlying surface and socioeconomic conditions in the Yangtze River Delta region from 2001 to 2010 was investigated by Spearman correlation analysis, multivariate analysis of variance (MANOVA) and lasso regression. In consideration of the characteristics of natural conditions and intensity of human activities in the Yangtze River Delta region, we designed six spatial scales to explore finely the effects of each factor on PM2.5 concentration. The results may provide decision support for the cross-regional air pollution risk identification. The main conclusions are as follows: (1) In different buffer zones, the dominant factors affecting the PM2.5 concentration are different. The buffer zones of 30, 40 and 50 km are the most effective areas for socioeconomic factors to affect the PM2.5 concentration. (2) The physical properties of underlying surfaces have significant effects on the PM2.5 concentration. Forestland can reduce PM2.5 concentrations in air to a certain extent, while land for construction has the opposite effect. (3) The influence of natural factors on the PM2.5 concentration in air is greater than that of socioeconomic factors in the Yangtze River Delta region, but the influence of socioeconomic factors on the PM2.5 concentration in buffer zones of 30, 40 and 50 km can not be ignored. The WS (the wind speed), PF (the proportion of forestland), PLC (the proportion of land for construction), P (the precipitation), NLI (the night light index), and PD (the population density) are the six main factors affecting the PM2.5 concentration in air.
Urban forests play a vital role in terms of environmental quality related to particulate matter (PM) and studies on this area are increasing. This review discusses how urban forests influence atmospheric PM at three scales, including the single tree scale, stand scale and regional scale. Additionally, the PM analysis was divided into vertical and lateral directions at the stand scale. As individuals, trees capture particles mainly by the leaves, the extent of which is determined by the characteristics in foliar structure such as hair, trichomes, wax, stoma, shape and others. At the stand level, the effects of urban forest vegetation showed differences in the two directions (vertical and lateral) and the predicted values of deposition velocity in the vertical direction differed with various models, which resulted from the input of numerous influencing factors (e.g., meteorological factors, plant species, building design and others). At the regional scale, the removal capacity of atmospheric particles by urban forests was dependent on the pollution level, vegetation coverage area, leaf area index and underlying surface type. The PM removal capacity of different vegetation types was usually in the following order: coniferous forest > evergreen forest > deciduous forest. The analyses at different scales indicated that the roles of urban forests on PM could not be simply deemed positive or negative, because they need to be considered by combining various factors (e.g., plant species, meteorological condition and city planning and design) at different scales. Therefore, multi-scale analyses on the effects of atmospheric particles could help to better understand the roles of urban forests as a complex system and provide the foundation for future studies.
Currently, urban air pollution becoming the major issue of urban physical environment in high density urban area. Many wind-calm zones or vortex zones can be formed where air pollution is retained or concentrated, which affects the environment and human health. To investigate this problem, the authors analyzed changes of wind velocity, direction, and air pollutant flow caused by changes in building height, volume, form, and density using a simulated three-dimensional (3D) conceptual model. The authors conducted an empirical study based on a representative high-density urban area. The results reveal that the actual 3D-simulated environment is complex. The wind environment changes continuously, and the retention or flow of the air pollutants changes drastically as well. The corner wind zones surrounding high-rise buildings may even generate new dust pollution due to the overly high wind speed. In this process, building height volume, layout, and orientation all significantly influence the flow and distribution of air pollution. Based on theoretical and empirical study, this paper discussed spatial planning strategies on the macroscopic city level, mesoscopic block level, and microscopic building level intended to promote the rapid dispersion of air pollutants by controlling the wind environment through optimizing the urban form.
Aerosol optical depth (AOD) from polar orbit satellites and meteorological factors have been widely used to estimate concentrations of surface particulate matter with an aerodynamic diameter <2.5 μm (PM2.5). However, estimations with high temporal resolution remain lacking because of the limitations of satellite observations. Here, we used AOD data with a temporal resolution of 1 h provided by a geostationary satellite called Himawari 8 to overcome this problem. We developed a stacking model, which contained three submodels of machine learning, namely, AdaBoost, XGBoost and random forest, stacked through a multiple linear regression model. Then, we estimated the hourly concentrations of PM2.5 in Central and Eastern China. The accuracy evaluation showed that the proposed stacking model performed better than the single models when applied to the test set, with an average coefficient of determination (R2) of 0.85 and a root-mean-square error (RMSE) of 17.3 μg/m3. Model precision reached its peak at 14:00 (local time), with an R2 (RMSE) of 0.92 (12.9 μg/m3). In addition, the spatial and temporal distributions of PM2.5 in Central and Eastern China were plotted in this study. The North China Plain was determined to be the most polluted area in China, with an annual mean PM2.5 concentration of 58 μg/m3 during daytime. Moreover, the pollution level of PM2.5 was the highest in winter, with an average concentration of 73 μg/m3.
Rapid urbanization in China has led to increasingly serious air pollution, mainly particulate matter (PM) and NOx, which has caused significant negative impacts on human health. In typical water network cities such as Wuhan, the purification of air pollution through the ecological effect of water bodies is an effective way to implement improvements. In this study, we conducted correlation analysis and established linear regression models to explore the relationship between water surface area and PM2.5, PM10, and NO2 concentrations by using pollutant monitoring data 2017, combined with water body information in remote sensing images. We found that the stability of concentration of particulate matters was low in areas where free water exists, which indicates that water has a role in absorption and minimization of PM. We also found that the concentrations of the three pollutants were strongly and negatively correlated with the free surface water coverage, with NO2 > PM2.5 > PM10, and the concentration of PM2.5, PM10 and NO2 was decreased by 2.45%, 2.96% and 9.65% respectively when water surface area increased by 10%. In addition, increasing water surface coverage in a specific size region can most effectively reduce pollutant concentrations.
The rapid urbanization of China related to fine particulate (PM2.5) emissions has attracted global attention, and exploring how urban forms affect PM2.5 concentration has become a hot topic for sustainable development of cites. Here, we combined multisource data and econometric models to quantify the relationships between urban forms and PM2.5 concentration in 250 Chinese cities, with explicit consideration of urban area size, population size, economic structure, and geographical location. The results showed that urban expansion had significant positive impacts on PM2.5 concentration in medium-sized (50–150 km²) and very large-sized cities (>250 km²). Urban form compactness tended to be beneficial for the reduction in PM2.5 concentration in large cities (5 million < population < 10 million). The urban form-PM2.5 concentration relationships were also dominated by economic structure change. That is, urban form irregularity/compactness played an important role in PM2.5 concentration within second cities (the proportion of second industry > 50%) and other cities compared to third cities (the proportion of third industry > 50%). Moreover, a compact and connected urban form was found to be more beneficial for reducing PM2.5 emissions in the northern region than in other regions. This study illustrates that the urban form-PM2.5 concentration relationships are sensitive to the differences in urban types and suggests that flexible urban planning strategies can actually help to reduce PM2.5 concentration in Chinese cities.
Atmospheric particulate matter (PM)pollution is becoming a growing global problem with the rapid process of urbanization. Urban green space (UGS)can effectively alleviate PM; however, few studies have investigated the effects of the UGS morphological pattern on PM, especially from a spatial strategy perspective. This study probed the contribution and strength of UGS on variation of PM 2.5 concentration based on morphological spatial pattern analysis (MSPA). Three relative indicators (range, duration, and rate)were used to represent PM 2.5 changes, and seven MSPA classes (core, islet, perforation, edge, loop, bridge, and branch)were performed to measure UGS morphological patterns. Stepwise regression analysis was used to build the PM 2.5 estimation models and partial correlation analysis was used to further analyze how well different MSPA classes influence PM 2.5 . Results showed that MSPA classes and meteorological factors combined can explain more of PM 2.5 increase variance at a high PM 2.5 level, and 40.7–81.4% for PM 2.5 reduction variance, and meteorological factors contributed more to PM 2.5 increase and reduction. Higher proportions of the core and bridge were conducive to restrict the growth and promote the reduction of PM 2.5 concentration, however, a higher proportion of perforation, islet, and edge showed opposite results. The effects of loop and branch were complex. In addition, higher air temperature and lower relative humidity were effective in reducing PM 2.5 . Wind speed, also a significant factor, had an unstable influence. The study results may provide important insights and effective spatial strategies for urban managers to mitigate PM 2.5 .
A “call to action” has been issued for scholars in landscape and urban planning, natural science, and public health to conduct interdisciplinary research on the human health effects of spending time in or near greenspaces. This is timely in light of contemporary interest in municipal tree planting and urban greening, defined as organized or semi-organized efforts to introduce, conserve, or maintain outdoor vegetation in urban areas. In response to injunctions from scholars and urban greening trends, this article provides an interdisciplinary review on urban trees, air quality, and asthma. We assess the scientific literature by reviewing refereed review papers and empirical studies on the biophysical processes through which urban trees affect air quality, as well as associated models that extend estimates to asthma outcomes. We then review empirical evidence of observed links between urban trees and asthma, followed by a discussion on implications for urban landscape planning and design. This review finds no scientific consensus that urban trees reduce asthma by improving air quality. In some circumstances, urban trees can degrade air quality and increase asthma. Causal pathways between urban trees, air quality, and asthma are very complex, and there are substantial differences in how natural science and epidemiology approach this issue. This may lead to ambiguity in scholarship, municipal decision-making, and landscape planning. Future
research on this topic, as well as on urban ecosystem services and urban greening, should embrace epistemological and etiological pluralism and be conducted through interdisciplinary teamwork.
Airborne particulate matter (PM) has been a major threat to air quality and public health in major cities in China for more than a decade. Green space has been deemed to be effective in mitigating PM pollution; however, few studies have examined its effectiveness at the neighborhood scale. In this study, the authors probe the contributions from different landscape components in the green space (i.e., tree, grass), as well as the spatial scale of planning on fine PM (PM 2.5 ) concentrations in urban neighborhoods. PM 2.5 data including 37 samples from five megacities were collected from the National Environmental Monitoring Centre in China. Results showed that, neighborhood green space greatly contributed to the spatial variation in PM 2.5 . The total green space coverage, tree coverage, and grass coverage were all negatively correlated with PM 2.5 concentration (p < 0.05). The higher green space coverage the site had, the lower the daily mean, daily minimum, and daily maximum of PM 2.5 concentration were there. Tree coverage, in particular, was effective in reducing the PM 2.5 concentrations, and, more importantly, its effectiveness was more significant with the higher ambient PM 2.5 level. According to the examination on the effect of spatial scale, the capability for a neighborhood green space to attenuate PM 2.5 pollution would be vanished when its size smaller than 200 m, and would be maximized when its size within 400–500 m. These results will contribute to the evidence-based design and management of green space to mitigating urban PM pollution.
Particulate matter with aerodynamic diameter less than 2.5 μm (PM2.5) is a complex mixture of chemical constituents emitted from various emission sources or through secondary reactions/processes; however, PM2.5 is regulated mostly based on its total mass concentration. Studies to identify the impacts on climate change, visibility degradation and public health of different PM2.5 constituents are hindered by limited ground measurements of PM2.5 constituents. In this study, national models were developed based on random forest algorithm, one of machine learning methods that is of high predictive capacity and able to provide interpretable results, to predict concentrations of PM2.5 sulfate, nitrate, organic carbon (OC) and elemental carbon (EC) across the conterminous United States from 2005 to 2015 at the daily level. The random forest models achieved high out-of-bag (OOB) R² values at the daily level, and the mean OOB R² values were 0.86, 0.82, 0.71 and 0.75 for sulfate, nitrate, OC and EC, respectively, over 2005–2015. The long-term temporal trends of PM2.5 sulfate, nitrate, OC and EC predictions agreed well with their corresponding ground measurements. The annual mean of predicted PM2.5 sulfate and EC levels across the conterminous United States decreased substantially from 2005 to 2015; while concentrations of predicted PM2.5 nitrate and OC decreased and fluctuated during the study period. The annual prediction maps captured the characterized spatial patterns of the PM2.5 constituents. The distributions of annual mean concentrations of sulfate and nitrate were generally regional in the extent that sulfate decreased from east to west smoothly with enhancement in California and nitrate had higher concentration in Midwest, Metro New York area, and California. OC and EC had regional high concentrations in the Southeast and Northwest as well as localized high levels around urban centers. The spatial patterns of PM2.5 constituents were consistent with the distributions of their emission sources and secondary processes and transportation. Hence, the national models developed in this study could provide supplementary evaluations of spatio-temporal distributions of PM2.5 constituents with full time-space coverages in the conterminous United States, which could be beneficial to assess the impacts of PM2.5 constituents on radiation budgets and visibility degradation, and support exposure assessment for regional to national health studies at county or city levels to understand the acute and chronic toxicity and health impacts of PM2.5 constituents, and consequently provide scientific evidence for making targeted and effective regulations of PM2.5 pollution.
Land cover-landscape pattern affects the atmospheric environment directly or indirectly, and the understanding of the atmospheric environment response to land cover - landscape pattern is of great significance to the maintenance and improvement of ecological environment. In this paper, such 9 landscape metrics as PLAND, PD, LPI, ED, MPS, AWMSI, CONTAG, SHDI and SHEI were selected by remote sensing inversion of PM2.5 data and land use data in long time series. The correlation analysis and multiple stepwise regression analysis were also used to analyze the effect of land use and landscape pattern on PM2.5 in Yangtze River Delta. The results showed that: (1) PM2.5 concentration was increasing in Yangtze River Delta from 1998 to 2015; (2) PM2.5 concentration was negatively correlated with the forest land and grassland, while positively correlated with the urban construction land; (3) At the level of landscape type, the urban construction land, water body and farm land PLAND, LPI, ED, MPS, AWMSI were positively correlated, the urban construction land and water body PD were positively correlated with PM2.5, the farm land PD was negatively correlated with PM2.5, and the forest land and grassland PLAND, PD, LPI, ED, MPS and AWMSI were negatively correlated with PM2.5. (4) In the integral landscape of land use, AWMSI was negatively correlated with PM2.5 concentration. It is of great significance to control PM2.5 pollution from the perspective of land use planning and contributed to an estimation methods of PM2.5 concentrations using land use type and land use landscape metrics in the absence of missing PM2.5 monitoring data.
In high-density cities, optimization of their compact urban forms is important for the enhancement of pollution dispersion, improvement of the air quality, and healthy urban living. This study aims to identify critical building morphological design factors and provide a scientific basis for urban planning optimization. Through a long-term mobile monitoring campaign, a four-month (spanning across summer and winter seasons) spatiotemporal street-level PM2.5 dataset was acquired. On top of that, the small-scale spatial variability of PM2.5 in the high-density downtown area of Hong Kong was mapped. Seventeen building morphological factors were also calculated for the monitoring area using geographical information system (GIS). Multivariate statistical analysis was then conducted to correlate the PM2.5 data and morphological data. The results indicate that the building morphology of the high-density environment of Hong Kong explains up to 37% of the spatial variability in the mobile monitored PM2.5. The building morphological factors with the highest correlation to PM2.5 concentration are building volume density, building coverage ratio, podium layer frontal area index and building height variability. The quantitative correlation between PM2.5 and morphological factors can be adopted to develop scientifically robust and straightforward optimization strategies for planners. This will allow considerations of pollution dispersion to be incorporated in planning practices at an early stage.
Background:
Daily mean concentration cannot fully address the hourly variations of air pollution within one day. As such, we proposed a new indicator, daily exceedance concentration hours (DECH), to explore the acute cardiovascular effects of ambient PM2.5 (particles with aerodynamic diameters less than 2.5μm). The DECH in PM2.5 was defined as daily total concentration-hours >25μg/m3.
Methods:
A generalized additive model with a quasi-Poisson link was applied to estimate the associations between day-to-day variation in PM2.5 DECH and day-to-day variation in cardiovascular mortality in six subtropical cities in Guangdong Province, China.
Results:
The analysis revealed significant associations between PM2.5 DECHs and cardiovascular mortality. A 500μg/m3∗h increase in PM2.5 DECHs at lag03 was associated with an increase of 4.55% (95% confidence interval (CI): 3.59%, 5.52%) in cardiovascular mortality, 4.45% (95% CI: 2.81%, 6.12%) in ischemic cardiovascular mortality, 5.02% (95% CI: 3.41%, 6.65%) in cerebrovascular mortality, and 3.00% (95% CI: 1.13%, 4.90%) in acute myocardial infarction mortality. We further observed a greater mortality burden using PM2.5 DECHs than daily mean PM2.5 (6478 (95% CI: 5071, 7917) VS 5136 (95% CI: 3990, 6305)).
Conclusion:
This study reveals that PM2.5 DECH is one important exposure indicator of ambient PM2.5 to measure its cardiovascular mortality effects in Pearl River Delta region; and that using daily mean concentration could under-estimate the mortality burden compared with this new indicator.
The spatial-temporal characteristics of aerosol loading over the Yangtze River Basin, China during 2001–2015 were investigated using moderate resolution imaging spectroradiometer (MODIS), multi-angle imaging spectroradiometer (MISR), and ground-level particulate matter (PM) data. Aerosol optical depth (AOD) >0.8 occurs in the Yangtze River Delta, central China, and the Sichuan Basin, while AOD <0.3 occurs over higher-elevation areas in the western Sichuan Plateau and the source regions of the Yangtze River. The western Sichuan Plateau is characterized by fine-mode natural aerosols, while the source region of the Yangtze River is more influenced by dust aerosols. The Sichuan Basin, central China, and the Yangtze River Delta are dominated by anthropogenic aerosols. The spatial distribution of ground-level PM2.5 and its ratio to PM10 confirm the high aerosol loading and dominance of small particles over the middle and lower Yangtze Basin. Over the Yangtze River Delta, central China, and the Sichuan Basin, AOD varies seasonally from low in the cold months to high in the warm months, being suppressed by rainfall and wind during the Asian summer monsoon. Precipitation increases aerosol loading in the western Sichuan Plateau but reduces dust particles in the source region of the Yangtze River. There is no significant AOD trend over most areas of the Yangtze River Basin during 2001–2015, while strong decreasing trends are found over most of the middle and lower Yangtze Basin during 2011–2015. These decreasing trends may relate to changes in annual precipitation, wind speed, and air-pollution control policies. The NO2 and SO2 emissions decreased by 16.51 and 23.40% over major provinces and cities of the Yangtze River Basin from 2011 to 2015. An increase in rainfall over the middle and lower Yangtze Basin and a better pollutant diffusion condition in the Sichuan Basin also favour the decreasing AOD during this period.
As the major engine of economic growth in China, the Pearl River Delta (PRD) region is one of the most urbanized regions in the world. Rapid development has brought great wealth to its citizens; however, at the same time, increasing emissions of ambient pollutants from vehicles and industrial combustions have caused considerable air pollution and negative health effects for the region's residents. In this study, the concentration response function method was applied together with satellite-retrieved particulate matter (PM10 and PM2.5) concentration data to estimate the health burden caused by this pollutant from 2004 to 2013. The value of statistical life was used to calculate the economic loss due to the negative health effects of particulate matter pollution. Our results show that in the whole PRD region, the estimated number of deaths from the four diseases attributable to PM2.5 was the highest in 2012, at 45,000 (19,000–61,000); the number of all-cause hospital admissions due to PM10 was the highest in 2013, reaching up to 91,000 (0–270,000) (excluding Hong Kong). Among the 10 cities, the capital city Guangzhou suffered the most from ambient particulate matter pollution and had the highest mortality and morbidity over the 10 years. The cost of mortality in this region was the highest in 2012, at 46,000 million USD, or around 6.1% of local total gross domestic product (GDP). The positive spatial relationship between the degree of urbanization and the particulate matter concentration proves that the urbanization process does worsen air quality and hence increases the health risks of local urban citizens. It is recommended that local governments further enhance their control policies to better guarantee the health and wealth benefits of local residents.