Health risk assessment of China’s main air pollutants

Background With the rapid development of China’s economy, air pollution has attracted public concern because of its harmful effects on health. Methods The source apportioning of air pollution, the spatial distribution characteristics, and the relationship between atmospheric contamination, and the risk of exposure were explored. The in situ daily concentrations of the principal air pollutants (PM2.5, PM10, SO2, NO2, CO and O3) were obtained from 188 main cities with many continuous air-monitoring stations across China (2014 and 2015). Results The results indicate positive correlations between PM2.5 and SO2 (R 2 = 0.395/0.404, P < 0.0001), CO (R 2 = 0.187/0.365, P < 0.0001), and NO2 (R 2 = 0.447/0.533, P < 0.0001), but weak correlations with O3 (P > 0.05) for both 2014 and 2015. Additionally, a significant relationship between SO2, NO2, and CO was discovered using regression analysis (P < 0.0001), indicating that the origin of air pollutants is likely to be vehicle exhaust, coal consumption, and biomass open-burning. For the spatial pattern of air pollutants, we found that the highest concentration of SO2, NO2, and CO were mainly distributed in north China (Beijing-Tianjin-Hebei regions), Shandong, Shanxi and Henan provinces, part of Xinjiang and central Inner Mongolia (2014 and 2015). Conclusions The highest concentration and risk of PM2.5 was observed in the Beijing–Tianjin–Hebei economic belts, and Shandong, Henan, Shanxi, Hubei and Anhui provinces. Nevertheless, the highest concentration of O3 was irregularly distributed in most areas of China. A high-risk distribution of PM10, SO2 and NO2 was also observed in these regions, with the high risk of PM10 and NO2 observed in the Hebei and Shandong province, and high-risk of PM10 in Urumchi. The high-risk of NO2 distributed in Beijing-Yangtze River Delta region-Pearl River Delta region-central. Although atmospheric contamination slightly improved in 2015 compared to 2014, humanity faces the challenge of reducing the environmental and public health effects of air pollution by altering the present mode of growth to achieve sustainable social and economic development.


Background
Haze is principally formed by an increase in particle size in the atmospheric medium, which affects atmospheric absorption, emission, and scattering of light. PM 2.5 : fine inhalable particles, with diameters that are generally 2.5 micrometers and smaller, and originates from construction sites, unpaved roads, fields, smokestacks or fires, including congregated aerosols (e.g. sulfur dioxide, nitrogen dioxide, carbon monoxide, and so on), black carbon (the incomplete combustion of carbonaceous combustibles) [1], dust, sea salt [2], heavy metals, and polycyclic aromatic hydrocarbon [3]. Haze incidents are a relatively new threat to human health [4], air quality [5], global climate change [6], ecological suitability for human settlement, and regional sustainable development.
Recently, haze has become a principal environmental issue in China. Consequently, the causes of particulate pollution have been discussed widely: e.g., secondary aerosol [5], aerosol optical properties [7], and aerosol chemical components [8]. And the formation and evolution mechanism of haze has been similarly explored [9]: e.g., long-lasting haze occurrences in Nanjing [10], a winter regional haze in the North China Plain [11], and the heavy haze pollution episode over central and eastern China [12]. In addition, we know that understanding the origin of fine particulate matter is essential to finding appropriate strategies to combat haze and the harm it causes. Thus, the source apportioning of fine particulate during the haze events in Shanghai [13], Harbin [14], and Fuzhou [15] was implemented, and the characteristics of atmospheric carbonyls were documented [16] in Beijing.
We have known for some time that haze boosts air pollution, causing significant harm to human health [17]. Previous studies have reported extensively on cardiovascular disease, lung disease, exposure time, mortality, and the mechanisms of biochemistry for haze. Short-term exposure was investigated in metropolitan areas [18], and the effects of dust-haze on mortality were explored [19]. In fact, the relationship between haze and respiratory diseases in Brunei Darussalam were analyzed, and it was found that PM 10 and CO levels have a significant bearing on the incidence of respiratory diseases [20]. In China, Sun et al. [4] explored the relationship between economic development and air pollution, and they found that the variation explained by both total SO 2 emissions and total smoke and dust emissions were 33 and 24% of pertussis (whooping cough), respectively.
However, source apportioning of fine particulate requires considerable investment in time and money. Thus, in this study, we attempt to analyze source apportioning using data mining. Because the risk of exposure to haze across China has been insufficiently discussed, the object of the present study is to address the relationship between atmospheric contamination and human health in China. Specifically, we seek to accomplish the following: (1) to analyze the spatial-temporal distribution of atmospheric contamination over China; (2) to explore the source apportioning of fine particulate in China; and (3) to analyze the relationship between atmospheric contamination and the risk of human exposure in China.

Method of health risk assessment
This study used the risk assessment method of the U.S. Environmental Protection Agency (EPA) [21], which focused on the health risk assessment through inhalation pathway for three kinds of people (adult males, adult females, and children), thus, avoided the effects of population density. The assessment study focuses on the risk of exposure to air pollutants (PM 10 , SO 2, and NO 2 ) in China; R i was the individual health risk for exposure pollution, calculated as Eq. 1 [22]: ADD opt was the average daily dose, calculated as Eq. 2 [23]: where CA was the concentration (mg m −3 ) of air pollutants, the average values of inhalation rate (IR), ED (exposure duration in days) and AT (averaging exposure time in days) were showed in Table 1 [22], and the average weight (BW) was obtained from national physical fitness test communiqués (http://www.gov.cn/test). The Rfd ij (reference dose) values for PM 10 , SO 2, and NO 2 referred to the U.S EPA (https://www3.epa.gov/). Based on the IDW (inverse distance weighted) interpolation method to model the spatial distribution of health risk in China, then, calculated the Rfd ij values and reclassified by the national air quality standard to get the expose risk level of air pollutants.

Tools of analysis
In the present study, the ArcGIS 10.2 (ESRI, Inc., Redlands, CA, USA) was used to draw spatial graphs, and SigmaPlot for Windows 10.0 (Systat Software, Inc., Chicago, IL, USA) was used to conduct correlation and regression analysis. Correlations between different variables were determined using two-tailed Pearson's Correlation at 0.05 levels.

Results
The size of main air pollutants    (Fig. 3d). Compared to other air pollutants, NO 2 had the greatest influence on PM 2.5 . Thus, the contribution rate of SO 2 and NO 2 is as high as 44.7% for PM 2.5 . As illustrated in Fig. 4, the concentration of PM 2.5 had been increasing with the rise of SO 2 , CO, NO 2 , and there are significant relationships between them. High correlation coefficients were noted between SO 2 (R 2 = 0.404,  Fig. 4c) and PM 2.5, which illustrates that the PM 2.5 are significantly associated with SO 2 , and NO 2 , with CO following, but there were no observable relationships between PM 2.5 and O 3 (Fig. 4d).

Spatial patterns of air pollutants in China from 2014 to 2015
According to the spatial distribution of air pollutants in China in 2014 (Fig. 6), we discovered that the concentration of CO ranges from 0.08 mg m −3 to 2.42 mg m −3 , with the maximum distribution around Hebei and Shanxi province, and the minimum distribution in the southeast, northwest, and northeast (Fig. 6a). In contrast, the concentration range of NO 2 was between 2 μg m −3 and 64 μg m −3 , with the maximum distribution occurring in Beijing, Tianjin, Hebei, Shandong, Henan province, and northeastern Xinjiang (Fig. 6b). However, O 3 concentrations ranged from 3 μg m −3 to 198 μg m −3 , with the maximum distribution in eastern China, southern China, and Hubei province; and the minimum distribution regions including Shanxi, Sichuan and Chongqing (Fig. 6c). The concentration of SO 2 ranges from 1 μg m −3 to 113 μg m −3 , with the maximum distribution in northern China and in Shandong province (Fig. 6).
According to Fig. 7a, the concentration of CO ranges from 0.01 mg m −3 to 2.35 mg m −3 in Shanxi, Shandong, Hebei, Henan, Beijing and Tianjin had the maximum values; we also found that the minimum was chiefly located in Heilongjiang, Gansu and Tibet, and the southeast of China (the coastal urban belt). The concentration ranges of NO 2 was between 0.6 μg m −3 and 60 μg m −3 , with maximum values distributed primarily in northeastern China (the Beijing-Tianjin-Hebei-Shanxi-Henan-Shandong region, Fig. 7b). In addition, we found that the concentration of O 3 ranges from 1 μg m −3 to133μg m −3 in Fig. 7c. The higher concentration values were observed in most areas of China, including eastern China, northern and central China (except Hunan province), and the regions of Gansu, Qinghai, Tibet (around Lhasa) and the Pearl River Delta region. The concentration of NO 2 ranged from 0.3 μg m −3 and 80 μg m −3 , with maximum values primarily distributed in Shanxi, Shandong and Hebei provinces (Fig. 7d).
The PM 2.5 concentration ranges from 3 μg m −3 to 103 μg m −3 in 2014 (Fig. 8a), with the maximum values distributed mainly around Hebei province (Beijing-Tianjin and a part of Shandong-Henan-Hubei. The concentration of PM 2.5 ranges from 1 μg m −3 to 106 μg m −3 in 2015  (Fig. 9b). Surprisingly, the high-risk values of NO 2 were mainly distributed in northeast China (Fig. 9c), regions of Hebei-Shandong-Henan-Beijing-Tianjin, a part of Inner Mongolia, the provincial capital cities of Guangzhou, Chengdu, Lanzhou, Xian, Shenyang, Changchun and Harbin, and the central of Jiangsu province. For adult females and children, a similar distribution pattern of the high-risk values for PM 10 /SO 2 /NO 2 was observed. The high-risk values for PM 10 distributed primarily in the central of Inner Mongolia, the south of Hebei province ( Fig. 9d and g). The high-risk values for SO 2 distributed in the central of Shandong province, the border of Shanxi and Hebei province, and a part of Erdos ( Fig. 9e and h). As shown in Fig. 9f and i, the high-risk values of distributed primarily in Beijing-Tianjin-Hebei-Shandong regions, a part of Xinjiang province, and the cities of Chengdu, Shanghai, Wuhan, Wenzhou and Harbin. In 2015, for adult males, the high-risk values of PM 10 were mainly distributed around the border of Hebei-Shandong-Henan province, thus, the cities of Baoding, Hengshui, Xingtai, Handan, Shijiazhuang in Hebei; Liaocheng, Dezhou and Heze in Shandong; and Zhengzhou in Henan (Fig. 10a). As shown in Fig. 10b, the high-risk values of SO 2 occurring in the central of Shanxi (Taiyuan and Linfen) and Shandong province, and a small part of Inner Mongolia (Erdos). The highrisk values of NO 2 mainly distributed in the central and northeastern of China (Fig. 10c), regions of Beijing-Tianjin-Hebei-Shandong-Henan, the central of Jiangsu province, and the cities of Urumchi, Lanzhou, Yanan, Chengdu, Shenyang, Changchun and Harbin, and the main city area of Chongqing. As for adult females and children, the regions of high-risk values for PM 10 /SO 2 /NO 2 were alike. As for PM 10 , the high-risk values distribution primarily in the main city area of Baoding, Hengshui, and Handan ( Fig. 10d and g). The high-risk values for SO 2 occurring in the central of Shandong, the main city area of Taiyuan and Shizuishan ( Fig. 10e and h). As shown in Fig. 10f and i, the high-risk values of NO 2 were chiefly distributed in Beijing-Tianjin-Hebei regions, the central of Shandong and the north of Henan province; and the province capital cities of Urumchi, Lanzhou, Chengdu, Wuhan and Harbin.

Relationships between air pollution and human health
The concentrations of air pollutions were the average values during 2014-2015, the mean rate of total pertussis was calculated from 30 provinces from 2004 to 2014. General linear models analysis illustrated that the mean rate of total pertussis was significantly associated with the average concentrations of PM 2.5 , PM 10 and CO  Fig. 11a, b, and d), and the variation explained by them were 61% (P < 0.06) for the rate of total pertussis. Meanwhile, the rate of total pertussis was related to SO 2 and NO 2 to some extent (Fig. 11c and e). However, there was no significant relationship between O 3 and total pertussis (Fig. 11f ).

Relationships among air pollutants
For this study, the most extensive data for analyzing the concentration and relationship of air pollutants and time-series datasets in China were used, with the aim of understanding pollution and to mitigate the heavy haze on the Chinese mainland. The Min, Median and Max concentration of air pollutants in 2014 and 2015 were presented, after finding that the occurrence frequency narrowed and became more centralized from 2014 to 2015 (Fig. 2). Our study also identified that the main components were NO 2 and SO 2 in PM 2.5 (R 2 = 0.395, R 2 = 0.447) in 2014 (Fig. 3) and (R 2 = 0.404, R 2 = 0.533) in 2015 (Fig. 4) based on regression analysis. The same phenomenon was found in Beijing [9], central and eastern China [12,24,25]. Previous research has observed that the concentration of SO 2 and NO 2 is lowest in the autumn and highest in the winter [26]. The concentration of PM 2.5 is closely correlated with SO 2 and NO 2 in Xi'an [23]. A joint effect of NO 2 and PM was found [27]. Wang not only analyzed the relationships of NO 2 , SO 2, and PM 2.5 but also found that with the increase of sulfate and nitrate, their particle hygroscopicity enhances and drives the formation and evolution of haze pollution [12]. This means that SO 2 and NO 2 play an important role in the enhancement of PM 2.5 [28], and reminds us to explore the relationships between SO 2 , NO 2, and CO. As shown in Fig. 5, the most significant correlation coefficient was NO 2 and SO 2 (R 2 = 0.306), followed by CO and SO 2 (R 2 = 0.289); NO 2 and CO (R 2 = 0.139) were the least pronounced in 2014. Nevertheless, a different significant association order was observed in 2015. Thus, the correlation coefficient of CO and SO 2 (R 2 = 0.386) was greater than NO 2 and CO (R 2 = 0.271) in 2015. This result suggests a strong relationship between CO and SO 2 , and NO 2 and CO; one study reported that the critical factor for formation droplet-mode particles was the availability of the water-vapor contents and precursor gases (SO 2 and NO 2 ) [29]; in other words, under iron and manganese catalysis, the heterogeneous oxidation of SO 2 and NO 2 change into the secondary sulfates (SO 4 2− and NO 3 − ) in the droplet mode [30,31], namely, the complex interaction of SO 4 2− , NO 3 − , congregated aerosols (e.g. sulfur dioxide, nitrogen dioxide, carbon monoxide, and so on), black carbon (the incomplete combustion of carbonaceous combustibles) determined the formation of haze and its particulate size [1,29,32]. NH 3, however, should not be neglected, which may result in the particulate sulfate and nitrate increase [28].

The spatial patterns of air pollutants
Apart from the regression analysis of air pollutants, spatial patterns of yearly average simulation values clearly present different air pollutants concentration distributions in different regions in 2014 and 2015 (Figs. 6 and 7). A few studies found that the Beijing-Tianjin-Hebei-Shandong-Shanxi-Henan regions had the highest concentration of air pollutants [24,25], including the local characteristics of high populations, city traffic, exhaust emissions, and rapid urban expansion [23,33]. In this study, the same spatial patterns of air pollutants were observed (Figs. 6, 7 and 8). Furthermore, we found the concentration of SO 2 , NO 2, and CO in Inner Mongolia cannot be negligible (Fig. 6a, b and d); meanwhile, the Tibet Plateau and coastal areas from Tianjin to Guangxi were affected by O 3 (Fig. 6c) in 2014. In addition, similar spatial patterns of the maximum values were observed in China. However, from 2014 to 2015, the spatial variation of O 3 concentration displayed a rapid increasing trend in China, especially on the Tibet Plateau (Fig. 7c), which may be influenced by its origin and long-distance transport. The PM 2.5 is distributed mainly in the region of Beijing-Tianjin-Shandong-Henan-Hubei in 2014 and 2015. Fortunately, its distribution narrowed and became more centralized in2015, and showed that the extent and area of PM 2.5 were lower than in 2014.
Emissions of PM 2.5 (97%), SO 2 (90%), NO 2 (70%) and CO (32%) were mainly due to the combustion of coal [34]. A number of previous studies have explored the origin and transportation [32,35] of air pollutants [36,37], including the congregated aerosols (e.g. sulfur dioxide, nitrogen dioxide, carbon monoxide, and so on), black carbon (the incomplete combustion of carbonaceous combustibles) [17], dust, sea salt [2], heavy metal, and polycyclic aromatic hydrocarbon [3], which resulted from vehicle exhaust [36], coal consumption, secondary production, stagnant meteorological conditions [13,[38][39][40], biomass open burning [41,42], and petrol stations [43]. Therefore, replacing coal and fossil fuels with cleaner fuels were the fundamental methods of controlling the concentration of air pollutants [39]. Certainly, we need to encourage new technologies and energy sources for automobiles, still a major contributor to air pollution [44]. But uncertainties exist from various sources, particularly the air pollutants in Xinjiang Uygur Autonomous Region, which might cause sand-dust storms. Therefore, ecological conservation projects should be developed and implemented; for instance, building key forest shelterbelts to shield against sandstorms in Xinjiang.
Health risk assessment and human health R i for adult males, adult females, and children, obtained for different pollutants (PM 10 , SO 2, and NO 2 ) in 2014 and 2015. It was observed that the high-risk of PM 10 was mainly distributed in the cities of Baoding, Hengshui, Xingtai, Handan, Shijiazhuang, Liaocheng, Dezhou, Heze and Zhengzhou (Figs. 9a, d, g and 10a, d, g), and Urumchi (Fig. 9a, d and g). The high-risk values of SO 2 were chiefly distributed in the cities of Taiyuan and Linfen, a small part of Erdos, and the central of Shandong province (Figs. 9b, e, h and 10b, e, h). The high-risk values of NO 2 were mainly occurring around in coastal areas from Beijing-Yangtze River Delta region-Pearl River Delta region-central, especially the cities of Urumchi, Lanzhou, A large portion of China's population has been significantly exposed to high-risk areas. Feng et al. [45] evaluated the ILI risk significantly associated with the concentrations of PM in Beijing during the flu season. In Guangzhou, the dust haze clearly increased mortality [19], and the PM 2.5 contains toxic micro-particles that might increase the risk of respiratory disease [46]. Mortality rates increased due to the high PM pollution in Shenzhen, especially for the elderly and male populations [47]. The cardiovascular, nervous system, respiratory and blood vessels of the brain are damaged by exposure to high concentrations of PM 2.5 [48]. Indeed, hemorrhagic stroke was closely associated with PM pollution [49]. Lung and cardiovascular disease are related to PM and NO 2 [50], and NO 2 was identified as the principal pollutant for respiratory diseases [18]. Local residents in Shanghai were exposed to high health risks due  (2015), and the graph a, b, c/d, e, f/g, h, i represent PM 10 , SO 2, and NO 2 for adult male/adult female/children, respectively to NO 2 [26]. Vulnerable people particularly (asthmatics, children, and the elderly), but all people generally, should not be exposed to high concentrations of SO 2 for any length of time [51]. Meanwhile, high concentrations of O 3 will irritate the eyes, nose, and throat, although long-term effects, if any, need further research [50]. One researcher has revealed that individuals exposed to biomass burningimpacted aerosols over the long term increased carcinogenic risk [6]. For these reasons and more, it is a matter of considerable urgency that policies be developed and implemented to mitigate the heavy haze in China.

Limitations of the current study
In this study, although 188 main cities across China used to get the spatial distribution patterns of air pollutions, uncertainties exist for limited data, especially in the sparsely distributed area of Xinjiang, Tibet, and Qinghai. Though significant relationships among main air pollutions were observed, inorganic substance and organic matter also correlated with each other in haze. In addition, PM 2.5 has other sources of crustal materials, such as from Asian dust storms. We analyzed the relationships among gaseous pollutant emissions. Meanwhile, the average parameter values for crowds in Eq. 2 referred from articles rather than actual measurements, led to the above conclusions about health risk in China. In addition, the data of pertussis was collected from the China Statistical Yearbook on Environment, although exposure to pollutions related to the increases in morbidity, accurate and concrete data for long-term effects is urgently needed. Thus, detailed data need collect to define air pollutions and risk assessment of human health in future.

Conclusions
Air pollution is harmful to the environment and to public health. This study focused on the source apportioning and the spatial-temporal characteristics of air pollutants and analyzed the relationship between atmospheric contamination and human exposure risk in China from 2014 to 2015. The main conclusions are as follows: (1)Regression analysis illustrates that there are close positive correlations between PM 2.5 and SO 2 , CO and NO 2 , but weak correlations with O 3 in 2014 and 2015. Additionally, the relationships between SO 2 , NO 2 and CO were significant, suggesting that vehicle exhaust, coal consumption secondary production, stagnant meteorological conditions, and biomass open-burning are the main factors driving the formation and evolution of air pollution. Accompanying the highest concentrations are high-risk areas distributed in provinces of Hebei, Shanxi, and Henan, and along the coast from Beijing-Yangtze River Delta to the Pearl River Delta region. The high-risk of NO 2 occurred in the Beijing-Tianjin-Hebei economic belts.
Measuring air pollutants, tracking contaminant paths and assessing pollutants in different areas with volatile weather conditions are complex challenges and need further research. The objective of this study is to help provide healthy, sustainable development not only for the people of China but for developing and developed nations alike. In particular, this study aims to initiate a constructive forum on the Beijing-Tianjin-Hebei collaborative development.