Aerosol and Air Quality Research, 18: 2086–2107, 2018 Copyright © Taiwan Association for Aerosol Research ISSN: 1680-8584 print / 2071-1409 online doi: 10.4209/aaqr.2018.06.0240

Characterization of Air Quality Index for both Handan and Cities, China

Rong Zhao1, Kangping Cui1*, Weiwei Wang1, How-Ran Chao2**

1 School of Resources and Environmental Engineering, University of Technology, Hefei 246011, China 2 Emerging Compounds Research Center, Department of Environmental Science and Engineering, National Pingtung University of Science and Technology, Pingtung 91201, Taiwan

ABSTRACT

In this study, the atmospheric PM10, SO2, NO2, CO, and O3concentration from 2015-2017, in Handan and Kaifeng were investigated. Besides, the seasonal variations of AQI values and their corresponding primary pollutants of six AQI grades were also discussed. In Handan, the daily AQI values ranged from 22 to 500 in 2015, from 19 to 500 in 2016, and from 27 to 500 in 2017, for which the corresponding average values were 143, 132 and 151, respectively. In Kaifeng, the daily AQI ranged from 25 to 496 in 2015, from 20 to 420 in 2016, and from 21 to 434 in 2017, for which the corresponding average values were 129, 124 and 131, respectively. During the three-year study, in Handan, the average proportion of grades I, II, III, IV, V and VI were 4.0%, 35.0%, 25.6%, 31.7%, 2.5% and 1.2% in spring; were 4.0%, 35.1%, 22.6%, 35.8%, 2.5% and 0% in summer; were 14.2%, 31.0%, 15.3%, 35.8%, 3.7% and 0% in fall, and were 3.3%, 26.1%, 20.0%, 37.8%, 8.1% and 4.7% in winter. In Kaifeng, the average proportion of grades I, II, III, IV, V and VI were 2.6%, 34.0%, 24.4%, 36.1%, 2.6% and 0.3% in spring; were 18.2%, 41.9%, 21.8%, 17.0%, 1.1% and and 0% in summer; were 8.6%, 34.3%, 17.4%, 34.8%, 4.9% and 0% in fall, and were 2.1%, 15.1%, 7.8%, 44.5%, 26.2% and 4.3% in winter. Overall, the air quality in the two cities were the worst in winter. The AQI values were between 101 and 150, where in Handan, PM2.5, PM10, and O3 were the primary air pollutants over the three years. In Kaifeng, the primary air pollutants were PM2.5, PM10, NO2, and O3 in 2015; were PM2.5, PM10, and O3 in 2016, and were PM2.5, PM10, and O3 in 2017. When AQIs ranged between 151 and 200, in Handan, the primary air pollutants were PM2.5, PM10, and O3 in both 2015 and 2016, and were PM2.5, PM10, NO2, and O3 in 2017. In Kaifeng, PM2.5, PM10, and O3 were the primary air pollutants from 2015-2017. When AQIs were between 201 and 300, in Handan, the primary air pollutants were PM2.5 and PM10 in 2015 and were PM2.5 and O3 in 2017. In Kaifeng, the primary air pollutants comprised PM10 in 2015, PM2.5 and PM10 in 2016, and PM2.5 and O3 at the same time. When the AQI values were between 301 and 500, which did not occur in Handan in 2015, the primary air pollutants comprised PM10 in 2016, PM2.5 and PM10 in 2017, and in Kaifeng from 2015-2017, PM2.5, and PM10 were the primary air pollutants. This study investigated the AQI values and corresponding primary pollutants in each season in more detail, the control strategies for these air pollutants will be more precisely.

Keywords: AQI; PM10; PM2.5; SO2; NOx; CO; O3.

INTRODUCTION of China, the concentrations of PM2.5, PM10, NO2, CO, SO2, and O3 in January 2013 exceeded the limit by 68.9%, A high extent of economic development, urbanization 46.9%, 12.3%, 14.9% and 1.2%, respectively (http://www. and industrialization in the past three decades, China has zhb.gov.cn/). Therefore, a heavy loading of air pollutant faced serious air pollution problems (Lin et al., 2010; Hu emission (Chen et al., 2013; Brauer et al., 2016) has caused et al., 2015a). According to the monthly report on urban air public concern world-wide (Wang et al., 2015a, b). quality issued by the Ministry of Environmental Protection Recently, the term “air pollution and human health” is often being discussed. The Air Quality Index (AQI) is based on the ambient air quality standards and the impact of various pollutants on human health, ecology, and the * Corresponding author. environment. Several air pollutant concentrations that are Tel.: +86-186-5595-3355 routinely monitored are reduced to a single, conceptual E-mail address: [email protected] index value. It classifies the degree of air pollution and air ** Corresponding author. quality, and it is suitable for representing the short-term air Tel.: +886-877-03202 quality status and trends of the city. The AQI is divided E-mail address: [email protected] into six categories of health concern.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2087

According to the World Health Organization (WHO), (http://www.mep.gov.cn/), with the highest concentration –3 the AQI value is determined by the standard concentrations of PM2.5 (over 700 µg m per hour average) and persistent of six pollutants, including sulfur dioxide, nitrogen dioxide, haze ( et al., 2014a, b). From et al.’s (2017) study on carbon monoxide, ozone, and PM2.5 and PM10. The WHO Kaifeng PM2.5 and dioxins, it can be seen that the pollution of estimates that approximately 7 million people die each ambient air also presents a very severe hazard. Therefore, this year from exposure to fine particles that can permeate the study mainly discusses the pollution from PM10, NO2, CO, lungs and the cardiovascular system in polluted air, resulting SO2, and O3 in Kaifeng and Handan, and an AQI analysis in diseases such as stroke, heart disease, lung cancer, and is conducted. The purpose of this study is to better understand chronic obstructive pulmonary disease, including pneumonia the characteristics of air pollution in Handan and Kaifeng, and respiratory tract infection. Coarse particles (PM10) and providing better suggestions for air pollution control. fine particles (PM2.5) have been relating to hospital admissions for respiratory (Brunekreef and Forsberg, 2005) METHODS and cardiovascular disease (Bell, 2012). Atmospheric carbon monoxide not only destroys the neurological function of Data for this study was obtained from the data bank for the heart but also can affect the central nervous system and three-years from January 2015 to December 2017 in both even cause suffocation leading to death (Yang et al., 2012). Handan and Kaifeng cities. The PM mass concentration The main sources of SO2 is the combustion of coal, a (including daily PM2.5 and PM10) and gaseous pollutants process highly common in China (Kurokawa et al., 2013). (including daily SO2, NO2, CO, and 8h -averaged O3) were Long-term exposure to sulfur dioxide may cause respiratory obtained from the China Air Quality Online Monitoring symptoms and illness and may also aggravate asthma. and Analysis Platform (http://www.aqistudy.cn/). Ozone (O3) is an important greenhouse gas contributing to global warming (Johnson et al., 1992; Berntsen et al., 1997; Air Quality Index (AQI) Sitch et al., 2007) and has a direct adverse impact on As shown in Equation (1), the sub-AQIs of the six ecosystems and human health (Solomon et al., 2000; Silva standard pollutants are first calculated using the observed et al., 2013; Lelieveld et al., 2015). O3 can be produced by concentrations (Shen et al., 2017; She et al., 2017). The photochemical reactions in the presence of NOx and overall AQI represents the maximum of sub-AQI for all volatile organic compounds (VOCs), which work as peroxy contaminants. When the AQI is above 50, the highest radical sources, especially in polluted areas. contributor to AQI is defined as the primary contaminant of CO is emitted along with NOx and VOCs during the the day, as in Eq. (2). (Shen et al., 2017; She et al., 2017): process of combustion. The lifetime of CO is longer than one month, so its concentration generally remains high IIhigh low even after being transported to remote locations. Therefore, IAQIp  CPlowlow C I (1) CChigh low monitoring of O3 and CO concentrations provides useful information leading to an understanding of the long-range transport of polluted air to remote areas (Kato et al., 2016). AQI = max(I1, I2, …, In) (2) In recent years, China has taken many measures to control the concentration of atmospheric pollutants. September 3, IAQIp: the air quality sub index for air pollutant P; th 2015 was the 70 anniversary of China’s Second World CP: the concentration of pollutant P; War victory (V-Day). In commemoration of the anniversary Clow: the concentration breakpoint that is ≤ CP; celebration, China imposed strict controls on emissions Chigh: the concentration breakpoint that is ≥ CP; from and its surrounding areas from August 20 to Ilow: the index breakpoint corresponding to Clow; September 3, 2015. In order to improve Beijing's air Ihigh: the index breakpoint corresponding to Chigh. quality during the annual parade, the number of cars in the The six standard air pollutants have a dramatic impact streets has been reduced by half, and 10,000 companies on health. The daily AQI value is calculated from the 24- have been asked to restrict production or to shut down hour average concentration of SO2, NO2, PM2.5, PM10, CO construction activities (Li et al., 2016). Li et al. (2016) and the daily maximum 8-hour concentration of O3. From discussed the observed data of SO2, NOx, and PM and the US Environmental Protection Agency (US EPA) AQI, found that SO2, NOx, and PM1 concentrations decreased by according to the scope of the AQI value, air quality can be 61–77%, 48–57%, and 64% during V-Day, respectively. divided into six categories: Grade I: 0–50 (good, green); Compared with the same period of the previous year, the Grade II: 51–100 (moderate, yellow); Grade III: 101–150 NO2 concentration decreased by an average of 44.6%. (unhealthy for sensitive group ; orange); Grade IV: 151–200 The Plain is considered to be the hardest-hit (unhealthy; red); Grade 5: 201–300 (very unhealthy; purple); area (Zhang et al., 2015) due to the combined results of Grade VI: 300–500 (hazardous; maroon) (Hu et al., 2015b; high air pollutant emissions caused by heating and frequent Lanzafame et al., 2015; She et al., 2017; Zhao et al., 2018). stable meteorological conditions in the area (Wang et al., 2015b; Elser et al., 2016). It is located on the edge of the RESULTS AND DISCUSSION national atmosphere, adjacent to , , , and Provinces (Wei et al., 2010). Handan was one of PM10 Concentration the top ten most polluted cities in China from 2013 to 2017 The rapid economic development and increase in the

2088 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 number of vehicles play a crucial role in the increase of ambient air of Handan and Kaifeng between 2015 and PM10 pollution (Han et al., 2005; Tang et al., 2005; Huang 2017 are shown in Figs. 1(a), 1(b), and 1(c), respectively. et al., 2010). Because PM10 consists of multiple chemical For Handan, in 2015, the monthly average PM10 components over a broad-sized spectrum (Hajek and Olej, concentration was in the range of 115.3–298.6 µg m–3, and –3 2015), the risks of PM10 related to human health cannot be in 2016, it was between 83.0 and 331.3 µg m ; while in –3 ignored. The monthly average PM10 concentration sin the 2017, it was in the range of 109.2–285 µg m . Overall, the

Fig. 1(a). Monthly average atmospheric PM10 concentration in Handan and Kaifeng in 2015.

Fig. 1(b). Monthly average atmospheric PM10 concentration in Handan and Kaifeng in 2016.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2089

Fig. 1(c). Monthly average atmospheric PM10 concentration in Handan and Kaifeng in 2017. average annual value of three years was 167.2 µg m–3, fall, and winter were 154.0, 135.5, 136.5, and 242.9 µg m–3, 150.6 µg m–3, and 157.4 µg m–3, respectively, with a three- respectively; and those in 2016 were 148.5, 90.2, 151.8, year average value of 158.4 µg m–3, which was 7.9 times and 211.7 µg m–3, respectively; while those in 2017 were of magnitude higher than the WHO air quality regulated 151.7, 116.8, 138.8, and 222.2 µg m–3, respectively. On the –3 standard (20 µg m ). Over the three years examined, the three-year basis, we can see that the concentration of PM10 maximum daily average concentration occurred in December basically was in the following order: winter > spring > fall –3 (886.0 µg m ) of 2016, and the minimum also occurred in > summer. The maximum PM10 concentration occurred in July (19.2 µg m–3) of the same year. This is mainly because the winter (242.9 µg m–3) of 2015, which was approximately in December, all northern regions of China entered a 1.8 times of magnitude higher than the minimum occurring –3 period requiring heating, and the concentration of PM10 in the summer (135.5 µg m ) of 2015. Similarly, in 2016, increased sharply due to the large amount of pollutants the maximum PM10 concentration occurred in the winter discharged during this period as well as adverse weather (211.7 µg m–3), which was approximately 2.3 orders of conditions. magnitude higher than the minimum occurring in the –3 As for Kaifeng, in 2015, the monthly average PM10 summer (90.2 µg m ). In 2017, the maximum PM10 concentration was in the range of 78.0 and 204.0 µg m–3, concentration occurred in the winter (222.2 µg m–3), which with those in 2016 ranging between 65.2 and 197.1 µg m–3, was approximately 1.9 orders of magnitude higher than the and those in 2017 ranging from 66.2 to 184.7 µg m–3. minimum occurring in the summer (116.8 µg m–3). Overall, the average annual values for the three years were For Kaifeng in 2015, the average PM10 concentrations in 129.0 µg m–3, 124.0 µg m–3, and 116.6 µg m–3, respectively, spring, summer, fall and winter were 131.1, 93.2, 114.4, with a three-year average value of 123.2 µg m–3, which and 177.3 µg m–3, respectively, and those in 2016 were was 6.2 times of magnitude higher than the WHO air 137.1, 71.9, 111.9, and 175.0 µg m–3, respectively; while quality regulated standard (20 µg m–3). The average annual those in 2017 were 127.0, 78.5, 103.5, and 157.5 µg m–3, concentration declined slightly over the three years, but the respectively. On the three-year basis, similar to the Handan, pollution is still very serious. The average daily maximum Kaifeng’s PM10 concentration was also in the following occurred in January (607.0 µg m–3) 2016, and the minimum order: winter > spring > fall > summer. The maximum –3 –3 occurred in October (16.0 µg m ) 2017. This is also related PM10 concentration occurred in the winter (177.3 µg m ) to the large amount of coal production because Kaifeng of 2015, which was approximately 1.9 times of magnitude entered into the period during which heating is required, and higher than the minimum occurring in the summer –3 adverse weather conditions have also aggravated the (93.2 µg m ) of 2015. Next, in 2016, the maximum PM10 –3 pollution of PM10. concentration occurred in the winter (175.0 µg m ), which With regard to seasonal variations, in Handan during was approximately 2.4 times of magnitude higher than the –3 2015, the average PM10 concentrations in spring, summer, minimum occurring in the summer (71.9 µg m ). Finally,

2090 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 in 2017, the maximum PM10 concentration occurred in the concentration of SO2 in Kaifeng also showed an annual winter (157.5 µg m–3), which was approximately 2.0 times decreasing trend. The concentration in 2017 decreased by of magnitude higher than the minimum occurring in the 34.0% compared to 2015. Over the three-year period, for the –3 summer (78.5 µg m ). 24-hour average maximum concentration of SO2 occurred The transportation, power plants, industry and domestic in January (21.4 ppb) 2015, while the minimum occurred heating sources, agriculture, and biomass open burning in October (0.2 ppb) in the same year. It can be seen that were identified as the major local PM10 sources (Wang et the concentration of SO2 in Kaifeng met the national al., 2008). A comparison of the PM10 pollution situation for secondary standard (52.5 ppb). Kaifeng and Handan indicates that the pollution in Handan The annual reduction in the SO2 concentration may be was more serious than that in Kaifeng. This is related to the related to the in Power 195 factory installation of flue gas fact that Handan is an important heavy industry city in desulfurization (FGD) system, the use of clean combustion Hebei Province. In addition, the maximum value of PM10 technology in industrial boiler, applied in the transportation occurs in winter, and the minimum value occurs in new vehicle standards, and use clean fuel in the residential summer. This is related to heating coal in winter in northern sector (Wang et al., 2014). Total sulfur dioxide emissions China. The burning of large amounts of coal and unfavorable from China's coal-fired power plants are estimated to have weather conditions increase the concentration of PM10. reduced by 54% between 2006 and 2010 due to the Previous research has proven that most pollution incidents installation of the FGD system (Liu et al., 2015). In addition, occur frequently in the heating season and continue for a Handan City took measures such as coal-to-gas conversion, long time (Zheng et al., 2015; et al., 2017). coal-to-coal conversion, and non-dissolving coalification Fine particulate matter (PM2.5) is considered as an in the main city area. important air pollutant with significant adverse effects on As to seasonal variations, for Handan, in 2015, the human health, including the effects on the heart, nervous average SO2 concentrations were 12.9, 10.6, 11.4 and system and vascular system (Chan et al., 2006; Zeller et al., 28.7 ppb in spring, summer, fall, and winter, respectively, 2006; Leiva G et al., 2013). Previous research has shown and those in 2016 were 13.4, 6.9, 12.6, and 25.9 ppb, that the PM2.5-to-PM10 ratio reveals that changes in the PM2.5 respectively; while those in 2017 were 13.4, 6.0, 7.0, and concentration are similar to changes in the PM2.5/PM10 25.5 ppb, respectively. ratio (Fang et al., 2016). In addition, it is benefit to clarify the As for Kaifeng, in 2015, the average SO2 concentrations characteristics of regional air quality of both PM2.5 and PM10, were 10.6, 5.0, 9.2, and 18.1 ppb in spring, summer, fall, by using the PM2.5/PM10 ratios, in general, ranging from and winter, respectively, and those in 2016 were 10.0, 3.3, 0.5 to 0.9 (Wang et al., 2012). Therefore, the concentration of 8.6, and 17.1 ppb, respectively; while those in 2017 were PM2.5 can be controlled by reducing the PM10 concentration. 7.9, 2.0, 6.5, and 12.1 ppb, respectively. It was obvious that the concentrations of SO2 both in Handan and Kaifeng SO2 Concentration are in the following grade order: winter > spring > fall > SO2 is absorbed by the aerosol moisture and oxidized by summer. High concentrations of SO2 occurred in winter, and 2– H2O2 or O3 to SO4 (Seinfeld et al., 1998). In high the lowest concentrations occurred in summer. Similar conditions (fog or wet haze) in many other cities, liquid conclusions have also been reached in a study by Zhao et chemicals in the pollution process play an important role in al. (2017) on SO2 in northern China cities. Sulfur dioxide sulphate pollution. (Sun et al., 2006; Du et al., 2011; Sun pollution caused by human activities, such as burning of et al., 2013). fossil fuels, metal smelting and biomass burning (Kettle The monthly average SO2 concentration in Handan and and Andreae, 2000; Halmer et al., 2002; Vijay et al., 2004; Kaifeng from 2015–2017 are illustrated in Figs. 2(a), 2(b), Dentener et al., 2006; Lee et al., 2008), the burning of coal and 2(c). is a common source in China (Kurokawa et al., 2013; Kato The monthly SO2 concentrations in Handan ranged et al., 2016). Therefore, increased coal burning in winter between 8.6 and 34.8 ppb, with an average of 15.9 ppb in promotes an increase in SO2. In warm seasons, higher 2015; ranged between 5.6 and 31.6 ppb, with an average of temperature favor the diffusion of pollutants, and stronger 14.6 ppb in 2016; and ranged between 3.7 and 31.3 ppb, rainfall can help remove pollutants. with an average of 12.9 ppb in 2017. From 2015 to 2017, the average annual value of SO2 declined slightly with a CO Concentration decrease of 18.9% in 2017 compared to 2015. Over the three- If carbon monoxide in the atmosphere is prone to year period, the maximum 24-hour average concentration of oxidative stress and inflammation, human can develop SO2 occurred in July (24.7 ppb) 2015, while the minimum symptoms such as headaches, dizziness, and even nausea occurred in October (0.25 ppb) 2017.It can be seen that the because they inhale carbon monoxide. Atmospheric carbon concentration of SO2 in Handan met the national secondary monoxide not only destroys the neurological functions of standard (52.50 ppb). the heart but also affects the central nervous system, and As for Kaifeng, the monthly concentration of SO2 ranged even causes suffocation that may lead to death (Yang et al., from 3.6–21.4 ppb, with an average of 10.7 ppb in 2015. In 2012). However, few studies have investigated the CO 2016, the concentration ranged from 2.4–18.8 ppb, with an concentration in northern cities. The monthly mean average of 9.7 ppb; in 2017 the concentration was between concentrations of CO in Handan and Kaifeng are shown in 1.0 and 12.4 ppb, with an average of 7.1 ppb. The Figs. 3(a), 3(b), and 3(c) from 2015 to 2017, respectively.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2091

2015 Handan Kaifeng 80.0 70.0 60.0 50.0 (ppb)

40.0 2

SO 30.0 20.0 10.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 2(a). Monthly average atmospheric SO2 concentration in Handan and Kaifeng in 2015.

2016 Handan Kaifeng 80.0 70.0 60.0 50.0

(ppb) 40.0 2

SO 30.0 20.0 10.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 2(b). Monthly average atmospheric SO2 concentration in Handan and Kaifeng in 2016.

2017 Handan Kaifeng

80.0 70.0 60.0 50.0 (ppb)

2 40.0 SO 30.0 20.0 10.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 2(c). Monthly average atmospheric SO2 concentration in Handan and Kaifeng in 2017.

2092 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

2015 Handan Kaifeng 8.0 7.0 6.0 5.0 4.0 (ppm)

CO 3.0 2.0 1.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 3(a). Monthly average atmospheric CO concentration in Handan and Kaifeng in 2015.

2016 Handan Kaifeng 8.0 7.0 6.0

(ppm) 5.0

CO 4.0 3.0 2.0 1.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 3(b). Monthly average atmospheric CO concentration in Handan and Kaifeng in 2016.

2017 Handan Kaifeng 8.0 7.0 6.0 5.0 4.0 (ppm)

CO 3.0 2.0 1.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 3(c). Monthly average atmospheric CO concentration in Handan and Kaifeng in 2017.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2093

In Handan, the monthly mean CO concentration ranged concentrations are always accompanied by higher air from 0.74 to 2.72 ppm in 2015, was between 0.94 and 2.68 temperatures, which is in accordance with the report of ppm in 2016, and ranged between 0.83 and 2.89 ppm in Kato et al. (2016). 2017, while the corresponding annual average values were 1.28, 1.44, and 1.29 ppm, respectively. The annual mean O3 Concentration concentrations of CO in 2016 increased about 20.5% As a result of urban expansion and industrial compared with those observed in 2015, and decreased development, tropospheric ozone levels have been rising approximately 10.4% from 2016 to 2017. Over the three- (Vingarzan, 2004). Since 1980, as China and low-latitude year period, the 24-hour average maximum concentration countries become more industrialized, emissions have of CO occurred in December (6.96 ppm) 2016, while the begun to shift to equatorwrd. (Zhang et al., 2016). minimum occurred in April (0.24 ppm) 2015. It can be O3 precursors such as VOCs and NOx have dramatically seen that the maximum value of CO exceeded the national increased in the past decade (Akimoto, 2003; Vingarzan, secondary standard (3.20 ppm). 2004; Richter et al., 2005; Monks et al., 2009) and will For Kaifeng, the monthly mean CO concentrations varied continue to increase (Ou et al., 2010; Wang et al., 2013; from 0.95 to 1.83 ppm in 2015, were 0.76 to 1.79 ppm in Yang et al., 2015; Wu et al., 2017).Transportation is one 2016, and were 0.67 to 1.90 ppm in 2017, while the source of major anthropogenic emissions of NOx (Zhang et corresponding annual average values were 1.26, 1.25 and al., 2009; Liu et al., 2015, 2016; Li et al., 2017a; Saikawa 0.97 ppm, respectively. The mean CO concentrations have et al., 2017). In a recent evaluation of Chinese O3 data, Wang decreased continuously. During the observed three-year et al. (2017) found that Beijing, , , and period, the 24-hour average maximum concentration of CO had O3 that exceeded the WHO standard of 100 occurred in December (4.56 ppm) 2015, while the minimum µg m–3 on more than 30% of the days in 2013–2015. In this occurred in July (0.4 ppm) 2017. It can be seen that the study, Figs. 4(a), 4(b), and 4(c) showed the variations in the maximum value of CO exceeded the national secondary monthly average O3 concentrations in Handan and Kaifeng standard (3.20 ppm) by approximately 42.5%. A comparison in 2015, 2016, and 2017, respectively. of the concentration of CO in Handan City and Kaifeng In Handan, the monthly average concentrations of O3 City shows that the concentration in Handan City was ranged between 14.32 and 54.55 ppb, with an average of slightly higher than that of Kaifeng City. 37.45 ppb in 2015; they ranged between 14.13 and As for seasonal variation sin CO concentrations, in 68.37 ppb, with an average of 42.15 ppb in 2016, and were Handan, in 2015, the concentrations of CO were 0.91, 0.88, between 17.02 and 88.41 ppb, with an average of 48.67 ppb 1.22, and 2.11 ppm in spring, summer, fall, and winter, in 2017. O3 concentrations rose significantly from 2015 to respectively, and those in 2016 were 1.05, 1.05, 1.38, and 2017, by approximately 29.9%. As can be seen from the 2.30 ppm, respectively; while those in 2017 were 1.02, figure, the concentration of O3 is in an inverted U-shaped 0.94, 1.01, and 2.21 ppm, respectively. As for Kaifeng, in distribution. On the basis of the three-year period, the 2015, the concentrations of CO were 1.19, 1.02, 1.16, and maximums all occurred in June (the concentrations were 1.69 ppm in spring, summer, fall, and winter, respectively, 54.59, 68.35, and 88.44 ppb, respectively). In addition to and those in 2016 were 1.22, 0.92, 1.16, and 1.71 ppm, the minimum value of 2017, which occurred in January respectively; while those in 2017 were 0.86, 0.70, 0.87, (17.03 ppb), the other minimum occurred in December (the and 1.45 ppm, respectively. The above results indicated concentrations were 14.36 and 14.13 ppb). As a whole, the that the CO concentration varied significantly from season three-year average O3 concentrations ranged between 14.16 to season, where the cold season always achieved the and 88.46 ppb, and with an average of 42.74 ppb. The results maximum values, and the warm season displayed the show that the O3 concentration in Handan met the WHO air minimum one, while the values were very similar in spring quality regulated standard of 46.67 ppb for O3. The increase and summer, and were both among the middle grades. In in O3 concentration may be related to the rapid growth in general, the three-year mean CO concentration in Handan the number of vehicles, which discharge a significant amount in winter (2.21 ppm) was 2.3 times of magnitude higher of VOCs. Because the increase in NO2 concentration was than that in summer (0.96 ppm), and the values in Kaifeng caused by the growth in the number of vehicles in Handan in summer (0.88 ppm) were 45.7% times of magnitude from 83.56 million in 2015 to 132.83 million in 2017 that lower than those in winter (1.62 ppm). Previous studies led to an increase in VOCs, resulting in the grade evaluation have shown that there is a strong correlation between obtained for O3. PM2.5 and CO (r = 0.745), which implies that the CO In Kaifeng, the monthly mean concentrations of O3 ranged emission is accompanied by the emission of PM2.5. An between 14.20 and 56.04 ppb, with an average of 35.91 ppb increase in concentration is probably because of the burning in 2015; ranged between 18.26 and 64.15 ppb and with an of fossil fuels and biomass (Fang et al., 2016). Zhao et al. average of 44.98 ppb in 2016; and were between 23.03 and (2017) studied the seasonal variations in PM2.5 in both 80.45 ppb, with an average of 49.80 ppb in 2017. O3 Handan and Kaifeng cities from 2015-2017. It was found concentrations rose significantly from 2015 to 2017, by that CO and PM2.5 were well-correlated. Therefore, the approximately 38.7%. Similar to Handan City, the monthly high concentration of CO in winter was still related to the average concentration of O3 in Kaifeng City also exhibited increase in coal burning. In addition, the low concentration a U-shaped distribution. Over the three-year period, the in summer was related to air temperature since a lower CO maximum average concentration occurred in June (80.41 ppb)

2094 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

2015 Handan Kaifeng 100.00

80.00

60.00 (ppb)

3

O 40.00

20.00

0.00 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 4(a). Monthly average atmospheric O3 concentration in Handan and Kaifeng in 2015.

2016 Handan Kaifeng 120.00

100.00

80.00 (ppb)

3 60.00 O

40.00

20.00

0.00 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 4(b). Monthly average atmospheric O3 concentration in Handan and Kaifeng in 2016.

2017 Handan Kaifeng 140 120 100 80 (ppb)

3 60 O 40 20 0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 4(c). Monthly average atmospheric O3 concentration in Handan and Kaifeng in 2017.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2095 of 2017, for which the value was 5.7 times higher than the progress of photochemical reactions. The spring peak likely minimum occurring in December (14.26 ppb) of 2015. On reflects the dominance of global background O3, which a whole, the three-year average O3 concentration ranged typically peaks in spring at remote sites in the Northern between 14.22 and 80.42 ppb, with an average of 43.59 ppb. Hemisphere (Zhang and Jaffe, 2017). In addition, previous The results show that the O3 concentration in Kaifeng met studies (Kato et al., 2016; Li et al., 2017b; Gong et al., the WHO air quality regulated standard of 46.6 ppb for O3. 2018) have indicated that the atmospheric relative humidity It can be seen that the three-year average for Kaifeng City is negatively correlated with the O3 concentration, so low was similar to that of Handan City. relative humidity is conducive to the formation of O3 As to seasonal variability, in Handan, in 2015, the seasonal during the dry fall period. average concentrations of O3 of spring, summer, fall, and winter were 42.73, 51.75, 35.01, and 20.13 ppb, respectively; NO2 Concentration while in 2016, those values were 53.33, 58.02, 34.74, and Transportation is one of the major anthropogenic sources 22.38 ppb, respectively, and in 2017, were 58.36, 72.35, of NOx emissions (Zhang et al., 2009; Liu et al., 2015, 40.02, and 23.72 ppb, respectively. As for Kaifeng, the 2016; Li et al., 2017a; Saikawa et al., 2017). Adame et al. seasonal average O3 concentrations in 2015 were 53.02, (2014) found that a total of 211544 tons/year of NOx was 41.06, 28.05, and 21.72 ppb in spring, summer, fall, and emitted, with traffic and industrial activities as the main winter, respectively; in 2016, they were 57.32, 59.23, 37.36, sources, at 36% and 34%, respectively in Andalusia in and 25.70 ppb in spring, summer, fall, and winter, 2007. From 2015 to 2017, the monthly mean concentration respectively; in 2017, they were 61.35, 67.34, 42.32, and of NO2 in Handan and Kaifeng are shown in Figs. 5(a), 28.36 ppb in spring, summer, fall, and winter, respectively. 5(b), and 5(c), respectively. From the data, it can be clearly seen that the three-year O3 In Handan, in 2015, the monthly average NO2 concentration due to seasonal changes in Handan City and concentration ranged between 15.6 and 38.0 ppb, and the Kaifeng City had obvious regularity. Overall, the three- annual average value was 23.0 ppb; while in 2016, those year mean O3 concentration were 51.47, 60.74, 36.77, and values ranged between 16.0 and 42.2 ppb, and the annual 22.09 ppb in spring, summer, fall, and winter, respectively average value was 26.4 ppb. In 2017, the values ranged in Handan and were 57.23, 55.96, 35.97, and 25.23 ppb in from 14.1 to 39.0 ppb, and the annual average value was spring, summer, fall and winter, respectively, in Kaifeng. 25.0 ppb. Over the three-year period, the maximum average This indicated that the grades for Handan was in the of NO2 occurred in December (42.2 ppb) of 2016, where following order: summer> spring > fall > winter; while in the value increased by 199% from the minimum value that Kaifeng they were: spring> summer> fall > winter. The occurred in July (14.1 ppb) of 2017. results for the O3 concentrations in 16 cities in China from In Kaifeng in 2015, the monthly average concentration 2014 to 2016 conducted by Gong et al. (2018) showed that of NO2 ranged between 10.0 and 31.0 ppb, with an average O3 grades in the four seasons of Handan were similar to those of 19.6 ppb; while in 2016, those values ranged between of Chengdu and , for which the concentrations 10.0 and 31.0 ppb, with an average of 19.5 ppb, and in were 53.85 ppb (spring), 62.44 ppb (summer), 30.63 ppb 2017, they ranged between 10.0 and 29.2 ppb, with an (fall), 22.77 ppb (winter), and were 48.02 ppb (spring), average of 18.8 ppb. It can be seen that the maximum NO2 57.81 ppb (summer), 30.86 ppb (fall), and 17.08 ppb concentrations during the three-year period were all around (winter), respectively. In contrast, the grades for O3 in 31.0 ppb and the minimum values were all 10.0 ppb. The Kaifeng were similar to those for and Lhasa, for annual average NO2 concentration decreased slowly over which the concentrations were 55.54 ppb (spring), 53.98 ppb the three-year period. The 2016 value decreased by 0.5% (summer), 46.33 ppb (fall), and 29.50 ppb (winter), and as compared to that in 2015, and the 2017 value decreased 58.56 ppb (spring), 52.68 ppb (summer), 43.24 ppb (fall), by 3.6% as compared to 2016. and 39.11 ppb (winter), respectively. It can be seen that the With regard to the seasonal variations, in Handan, the low values of O3 occur in the winter and that the highest seasonal NO2 concentrations during spring, summer, fall, values of O3 occur in the warmer periods, especially summer. and winter were 19.7, 18.0, 26.0 and 28.3 ppb in 2015. In addition, the study by Gao et al. (2017) also confirmed They were 25.0, 18.3, 27.3 and 35.0 ppb in 2016, and 23.7, that 3/4 out of 16 cities had the same situation. This may be and 35.0 ppb, 25.0 and 34.3 ppb in 2017, respectively. We related to the haze phenomenon that frequently occurred in can see that typically, the seasonal NO2 exhibited U shape winter. Cold weather in winter increased the consumption patterns, with lower values from May to August with higher of coal, thereby increasing the concentration of particulate values in November and December. The concentration of matter in the atmosphere. It is known from previous studies NO2 was in the following order: winter > fall > spring > that higher mass concentrations of particles reduce radiation summer. During the three-year period, the maximum and the decrease in radiation intensity, which will alter occurred in the winter (35.0 ppb) of 2016, for which the photochemical reactions and oxidant concentrations. The value was 2.1 times higher than the minimum that occurred concentration of O3 was reduced by 91.3%, dropping from in summer (17.0 ppb) in 2017. –3 –3 42.4 µg m with a clean grade to 3.7 µg m with a For Kaifeng, the seasonal NO2 concentrations during severely polluted grade from November 16, 2015 to March spring, summer, fall, and winter were 16.3, 12.3, 21.3 and 14, 2016 (Zhang et al., 2017). On the contrary, accelerated 28.4 ppb in 2015. They were 18.6, 11.0, 20.7 and 27.7 ppb radiation and high temperatures in summer accelerate the in 2016 and 17.0, 10.3, 21.3 and 26.3 ppb in 2017,

2096 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

2015 Handan Kaifeng 90.0 80.0 70.0 (ppb)

60.0 50.0 40.0 30.0 Concentration

2 20.0

NO 10.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 5(a). Monthly average atmospheric NO2 concentration in Handan and Kaifeng in 2015.

2016 Handan Kaifeng 90.0 80.0 70.0 60.0 50.0 40.0 30.0 Concentration(ppb)

2 20.0

NO 10.0 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 5(b). Monthly average atmospheric NO2 concentration in Handan and Kaifeng in 2016.

2017 Handan Kaifeng 90.0 80.0 70.0 60.0 50.0 40.0 30.0

Concentration(ppb) 20.0 2 10.0 NO 0.0 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Month

Fig. 5(c). Monthly average atmospheric NO2 concentration in Handan and Kaifeng in 2017.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2097 respectively. Similar to Handan, the concentraion of NO2 vehicles was been calculated to be –0.86 (95% confidence in Kaifeng also exhibited U shape patterns. The maximum interval, two-tailed) during the 2000–2015, the correlation occurred in the winter (28.4 ppb) of 2015, for which the between vehicle NOx emissions and the concentration of value was 2.8 times higher than the minimum that occurred NO2 was 0.86 (95% confidence interval, two-tailed), reduce in the summer (10.3 ppb) of 2017. vehicle NOx emissions have significant influence on Beijing A comparison of the concentrations of NO2 in Handan NO2 concentration decline (Cheng et al., 2017). In terms of and Kaifeng shows that the NO2 pollution in Handan was seasonal variations, during the three-year period, the average slightly worse than that in Kaifeng. This was mainly related NO2 concentrations in fall and winter were relatively higher to the pollution from vehicles. Previous research shows that than those in spring and summer, and the higher monthly the increase of NO2 concentration in the atmosphere caused average NO2 concentration was significantly related to by the large amount of nitrogen oxides from the increase of PM2.5 pollution events (Yang et al., 2015). vehicles in China will directly affect the ecological environment and adversely affect human health (Boersma AQI Analysis et al., 2009; Sachin et al., 2009). It can be seen from the For this reason of health effects, the Chinese government data provided by the statistics bureaus of Handan and set the Technical Regulation on the Ambient Air Quality Kaifeng (http://tj.hd.gov.cn; http://www.ha.stats.gov.cn) Index (AQI) in 2012, which included PM2.5 concentration that the number of vehicles in Handan increased by 59%, for the first time (Du et al., 2017). The AQI index is used from 83.49 million in 2015 to 132.76 million in 2017. In to characterize the degree of air contamination and its Kaifeng, the number of vehicles in Handan increased by associated health risks. In the period from 2015 to 2017, 41.1%, from 38.4 million in 2015 to 54.2 million in 2017. the number of days for the six AQI categories in the The number of vehicles in Handan City was much higher different seasons of the Handan and Kaifeng are shown in than the number in Kaifeng. The correlation coefficient Figs. 6(a)–6(f), and the cumulative number of days of between the Beijing NO2 concentration and the number of primary pollutants are shown in Tables 1(a)–1(b).

A B

C D Fig. 6(a). The number fractions of the six AQI categories for Handan in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2015.

2098 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

Fig. 6(b). The number fractions of the six AQI categories for Kaifeng in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2015.

In Handan, the daily AQI ranged from 22 to 500, with were 2.2%, 33.3%, 23.7%, 40.8%, 0%, and 0% in 2015, an annual value of 143 in 2015; these values ranged from respectively; were 7.6%, 32.6%, 29.3%, 27.2%, 2.2%, and 19 to 500, with an annual average value of 132 during 1.1% in 2016, respectively, and were 2.2%, 39.1%, 23.9%, 2016 and ranged from 27 to 500, with an annual average 27.2%, 5.4%, and 2.2% in 2017, respectively. From 2015 value of 151 during 2017. In Kaifeng, the daily AQI ranged to 2017, except for a slightly higher percentage of grade IV from 25 to 496, with an annual value of 129 in 2015; it (40.8%) in 2015, the other grades were similar. The state ranged from 20 to 420, with an annual value of 124 during of ambient air quality was mainly concentrated in grades 2016, and it ranged from 21 to 434, with the annual value II–IV, which showed that there was mild pollution in of 131 in 2017. Handan City had a maximum of 500 AQI Handan in the spring. From Table 1(a), it was obvious that during the three years under consideration, where the values Handan was polluted by particulate matter (PM2.5 and were higher than the maximum (496, 420, and 434) in PM10) in the spring. This was related to the wind-driven Kaifeng City. This indicated that the air pollution situation sand in the spring. in Handan was more serious than that in Kaifeng City. In In summer, the distribution of the six AQI classes in addition, the values of the AQI in the two cities fluctuated Handan, as shown in Fig. 6(a)–(B), Fig. 6(c)–(B), and significantly, which was related to the local pollution Fig. 6(e)–(B) with grades I, II, III, IV, V and VI were 3.3%, emission characteristics and meteorological conditions. 41.1%, 15.6%, 40%, 0%, and 0% in 2015, respectively; In this study, we analyzed the seasonal pollution of six were 8.7%, 35.9%, 27.2%, 28.2%, 0%, and 0% in 2016, categories in Handan and Kaifeng. respectively, and were 0, 28.3%, 25%, 39.1%, 7.6%, and Fig. 6(a)–(A), Fig. 6(c)–(A), and Fig. 6(e)–(A) present 0% in 2017, respectively. From 2015–2016, most of the the distribution of six AQI classes in Handan in the spring, days that belonged to grades II–IV, for which the value for which the proportion of grades I, II, III, IV, V and VI was as high as 90% or more. In 2107, the number of days

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2099

Fig. 6(c). The number fractions of the six AQI categories for Handan in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2016. belonged to Grade IV increased significantly. It can be I = Grade V > Grade VI and others appeared Grade II > seen that the air in the summer was relatively good, which Grade IV > Grade I > Grade III > Grade V > Grade VI. In was related to summer conditions that favor the diffusion general, the number of days of Grade II and Grade IV were of pollutants. Table 1(a) shows that PM2.5 remains the most the highest, with the number of days with these two Grades frequently occurring primary pollutant, followed by PM10 accounting for more than 50% of the whole year, indicating and O3. The frequency of occurrence of O3 as the primary that the air pollution grade in Handan City was categorized pollutant in the summer was significantly higher than in as moderately polluted in fall. Table 1(a) shows that in other seasons. Severe O3 pollution in summer may be 2015, with the exception that PM10 and PM2.5 were the attributed to higher VOC emissions, which provide a high primary pollutants most days, the number of days that NO2 concentration of precursor for the formation of O3, during was the primary pollutant was significantly increased. which high air temperature and intense solar radiation can However, in 2016 and 2017, the number of days with NO2 promote the photochemical production of O3. This result is as the primary pollutant became zero again, due to the consistent with previous research results (Atkinson and dominant source of pollution being particulate matter. This Arey, 2003; Zhang and Ying et al., 2011; Li et al., 2012; indicated that the pollution from NO2 in fall did not always He et al., 2017; Shen et al., 2017). occur. This may be related to the automobile exhaust gas In the fall of 2015, the distribution of the six AQI classes, and gas phase conditions at the time. as shown in Fig. 6(a)–(C), Fig. 6(c)–(C), and Fig. 6(e)–(C) As for the distribution of six AQI categories of winter for grades I, II, III, IV, V and VI were 19.6%, 32.6%, throughout the observed three-year period, in Handan, 17.4%, 27.2%, 3.2%, and 0%, respectively; in 2016, these Fig. 6(a)–(D), Fig. 6(c)–(D), and Fig. 6(e)–(D) show that fractions were 18.7%, 33.0%, 16.5%, 24.2%, 6.6%, and 1.0%, during 2015, grades I, II, III, IV, V, and VI were 0%, 6.7%, respectively, and in 2017, these fractions were 4.4%, 23.1%, 6.7%, 58.9%, 16.7%, and 11.0%, respectively; while in 12.1%, 56.0%, 4.4%, and 0%, respectively. Obviously, the 2016, the proportions were 7.7%, 32.6%, 29.3%, 27.2%, number of days with each grade in 2017 was in the 2.2%, and 1.0%, respectively, and in 2017, the proportions following order: Grade IV > Grade II > Grade III > Grade were 2.2%, 39.1%, 23.9%, 27.2%, 5.4%, and 2.2%,

2100 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

A B

C D Fig. 6(d). The number fractions of the six AQI categories for Kaifeng in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2016. respectively. In 2015, Grade IV accounted for more than showed an increase from 2015 to 2016, but the proportion half of the spring, followed by grades V and VI. There from 2016 to 2017 decreased. This showed that the were no good days in the winter in Handan City. The environmental quality grade in the spring of 2016 was worse Grade IV index indicated an unhealthy environment. This than that in the other two years. From Table 1(b), different mean that everyone may begin to experience negative from Handan City, apart from the frequent occurrence of health effects, and members of sensitive groups may face PM2.5 and PM10 as primary pollutants, the proportion of serious sick. Moreover, Table 1(a) shows that PM2.5 was gaseous pollutant O3 also accounted for some proportion. the most frequent primary air pollutant. This was mainly The above analysis of O3 shows that O3 pollution often related to the massive amount of coal burning and occurs in spring and summer. During this period, particulate unfavorable meteorological conditions in winter (Sun et matter is lighter, and the radiation is stronger, which is al., 2014). favorable for the occurrence of photochemical reactions. It is well known that PM2.5 is an inhaled lung particle. Fig. 6(b)–(B), Fig. 6(d)–(B), and Fig. 6(f)–(B) showed Due to its large specific surface area, PM2.5 is prone to the the proportion of grades I, II, III, IV, V, and VI in summer, accumulation of a large amount of toxic and harmful which were 19.4%, 38.7%, 17.2%, 24.7%, 0%, and 0% in substances, stays in the atmosphere for a long time, and is 2015, respectively. They were 20%, 45.6%, 24.4%, 10%, far from the transportation distance. It has a greater impact 0%, and 0% in 2016, respectively and were 15.2%, 41.3%, on human health and the atmospheric environment than 23.9%, 16.3%, 3.3%, and 0% in 2017, respectively. other pollutants. In Kaifeng, Fig. 6(b)–(A), Fig. 6(d)–(A), Between 2015 and 2017, the number of days of Grade V and Fig. 6(f)–(A) show that in spring, the proportion of grades and Grade VI were basically zero, and the number of days I, II, III, IV, V and VI were 1.1%, 33.3%, 31.2%, 34.4%, with Grade II indexes was the highest. In addition, 0%, and 0% in 2015, respectively; they were 5.6%, 39.3%, compared to other seasons, there were more summer day 20.2%, 30.3%, 4.6%, and 0% in 2016, respectively, and ranked as Grade I. From Table 1(b), it can be seen that, were 1.1%, 29.3%, 21.7%, 43.5%, 3.3%, and 1.1% in 2017, unlike Handan, PM10 was the most important pollutant in respectively. From 2015 to 2017, each grade basically terms of number of days with this pollutant, and in 2017

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2101

Fig. 6(e). The number fractions of the six AQI categories for Handan in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2017.

O3 replaced PM10 as the most frequent primary pollutant. particulate pollution. The pollution of NO2 and O3 only Strong radiation and high temperatures in summer accelerate occurred occasionally for a few days. photochemical reactions and promote the formation of O3. Finally, in winter, Fig. 6(b)–(D), Fig. 6(d)–(D), and Therefore, it is urgent to take effective actions to prevent Fig. 6(f)–(D) shows that during 2015, the proportion of increases in O3 in the summer. grades I, II, III, IV, V, and VI was 0%, 18.9%, 7.8%, Fig. 6(b)–(C), Fig. 6(d)–(C), and Fig. 6(f)–(C) illustrated 44.4%, 23.3%, and 5.6%, respectively; while in 2016, in 2015, the proportion of grades I, II, III, IV, V, and VI in those proportions were 3.0%, 12.1%, 10.1%, 36.4%, 33.3% the fall were 11.2%, 34.8%, 12.4%, 38.2%, 3.4%, and 0%, and 5.1%, respectively, and in 2017, they were 3.3%, respectively; while in 2016, these fractions were 12.5%, 14.3%, 5.5%, 52.8%, 22.0%, and 2.1%, respectively. From 34.1%, 17.0%, 28.4%, 8.0%, and 0%, respectively, and in 2015–2017, the proportions of different AQI classes in 2017, they were 13.3%, 28.9%, 16.7%, 37.8%, 3.3%. and Kaifeng show that AQI >150 days exceeded 70% each 0%, respectively. Comparing the proportions of different year mainly in terms of the VI and V grades. The pollution AQI categories from 2015 to 2017, Grade I days grew of the ambient air has a severe impact on humans. From by18.8%. There was a slight reduction in Grade II, which Table 1(b), similar to Hnadan, it is shown that PM2.5 is the decreased by 17.0% from 2015 to 2017. Grade III showed most frequent primary air pollutant. an increasing trend initially and then decreased, and Grade IV exhibited an opposite trend. Grade V maintained a very CONCLUSION low occupancy rate for all three years. The number of days in Grade VI were all zero. From Table 1(a), it can be seen The results of this study on the atmospheric deposition that the city of Kaifeng was mainly dominated by in Handan and Kaifeng can be summarized as follows:

2102 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

Fig. 6(f). The number fractions of the six AQI categories for Kaifeng in (A) Spring, (B) Summer, (C) Fall, and (D) Winter in 2017.

1. Over the three-year period, the PM10 concentrations in ppb, respectively. Comparing the concentration of NO2 –3 Handan were in range of 83.0 and 331.3 µg m ; in in Handan and Kaifeng, we can know that NO2 pollution Kaifeng, the concentrations were from 65.2 to in Handan was slightly worse than Kaifeng. In addition, –3 204.0 µg m . The average concentrations of Handan and for annual seasonal variation, the concentration of NO2 Kaifeng were 158.4 and 123.2 µg m–3, respectively. The exhibited U shape patterns both in Handan and Kaifeng. PM10 level in Handan was higher than that in Kaifeng. 4. During the observed three-year period, in Handan and As for seasonal variations of PM10 basically were in the Kaifeng the average CO concentration ranged 0.74– following order: winter > spring > fall > summer. In 2.89 ppm and 0.67–1.90 ppm, respectively; the summer of Handan (114.2 µg m–3) was 49.4% in corresponding average values were 1.34 and 1.16 ppm, magnitude lower than that in winter (225.6 µg m–3); as respectively. The high concentration of CO in winter was for Kaifeng in summer (81.2 µg m–3) was 52.2% in related to the increase of coal burning. Besides, the low magnitude lower than that in winter (169.9 µg m–3). concentration in summer was related to air temperature, a 2. In Handan, the average concentrations of SO2 for the lower CO concentration was always accompanied with a three years were between 3.7 and 34.8 ppb; and those higher air temperature. values were in range of 1.0–21.4 ppb in Kaifeng, with 5. The three-year average O3 concentration ranged from the corresponding average values were 14.5 and 9.2 ppb, 14.13 to 88.41 ppb, and with an average of 42.76 ppb in respectively. The result indicated that the pollution of Handan, and ranged between 14.20 and 80.45 ppb, and Handan by SO2 was more serious than that of Kaifeng City. with an average of 43.56 ppb in Kaifeng. The monthly 3. During the three-year period, the average concentrations average concentration of O3 both in Kaifeng and Handan of NO2 were between 14.1 and 42.2 ppb, in Handan. In showed inverted U-shaped. Besides, the maximum value Kaifeng the concentration were from 10.0 to 31.0 ppb. of O3 concentration always occurred in spring and The corresponding average values were 24.8 and 19.3 summer, and the minimum value was in winter.

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2103

Table 1(a). Cumulative number of days of primary pollutants for Handan from 2015–2017. AQI Spring Summer Fall Winter Year Class PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 2015 51–100 0 31 0 0 8 33 0 17 0 29 18 1 0 6 0 0 101–150 2 18 0 2 4 7 0 3 1 12 0 3 1 5 0 0 151–200 33 4 0 1 34 0 0 1 25 0 0 0 53 0 0 0 201–300 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > 300 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2016 51–100 5 18 0 7 0 18 0 14 0 23 0 6 0 8 0 0 101–150 11 9 0 7 13 0 0 12 0 11 0 4 4 2 0 0 151–200 17 4 0 4 23 0 0 12 22 0 0 0 35 0 0 0 201–300 2 0 0 0 0 0 0 0 0 6 0 0 7 1 0 0 > 300 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2017 51–100 0 34 1 1 11 6 0 6 0 19 0 2 0 20 0 0 101–150 0 16 0 6 9 1 0 13 3 4 0 4 1 3 0 0 151–200 18 0 0 7 15 0 21 8 1 46 0 5 29 0 0 0 201–300 1 0 0 4 0 0 0 7 4 0 0 0 25 0 0 0 > 300 0 2 0 0 0 0 0 0 0 0 0 0 7 2 0 0

Table 1(b). Cumulative number of days of primary pollutants for Kaifeng from 2015–2017. Year AQI Spring Summer Fall Winter Class PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 PM2.5 PM10 NOx O3 2015 51–100 5 23 0 3 0 32 0 3 0 30 2 0 0 17 0 0 101–150 4 15 0 11 11 8 0 0 9 2 0 0 4 3 1 0 151–200 31 1 0 0 17 1 0 5 34 0 0 0 40 0 0 0 201–300 0 0 0 0 0 0 0 0 0 0 0 0 0 22 0 0 > 300 0 0 0 0 0 0 0 0 0 0 0 0 4 1 0 0 2016 51–100 0 21 0 14 0 11 0 30 0 20 0 10 0 10 0 0 101–150 2 10 0 6 9 0 0 13 8 4 0 3 1 1 0 0 151–200 24 3 0 0 5 0 0 4 25 0 0 0 33 3 0 0 201–300 3 1 0 0 0 0 0 0 0 7 0 0 33 1 0 0 > 300 0 1 0 0 0 0 0 0 0 0 0 0 5 0 0 0 2017 51–100 7 11 0 8 0 8 0 30 6 13 0 6 0 13 0 0 101–150 3 7 0 10 2 0 0 20 7 2 0 7 0 5 0 0 151–200 27 4 0 9 3 12 0 21 30 0 0 4 48 0 0 0 201–300 1 0 0 2 0 0 0 3 3 0 0 0 20 0 0 0 > 300 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0

6. In the spring of Handan, the state of ambient air quality was as the highest primary pollutant. was mainly concentrated on Grade II–IV, which showed 9. In fall, in general, the number of days in Grade II and that there was mild pollution in Handan in spring. PM2.5 Grade IV were the highest, with these two Grades and PM10 were the most frequent primary pollutants; accounting for more than 50% of the whole year both in different from Handan City, apart from the frequent two cities. The primary pollutants were PM2.5 and PM10. occurrence of PM2.5 and PM10 as primary pollutants, the The pollution of NO2 and O3 only occured occasionally proportion of gaseous pollutants O3 also accounts for the for a few days. proportion. 10. In the winter in Handan, the fraction of excellent air 7. In the spring in Handan, the state of ambient air quality quality was zero, but the heavy pollution was improved. was mainly grades II–IV, which showed that there was Similar to Handan, there were very few days with mild pollution in Handan in spring. PM2.5 and PM10 excellent conditions in Kaifeng City. PM2.5 and PM10 were the most frequent primary pollutants; different were the typical primary air pollutants. NO2 appeared from Handan City, apart from the frequent occurrence of only once in the Kaifeng in 2015. PM2.5 and PM10 as the primary pollutants, the proportion of gaseous O3 also accounted for a proportion of the REFERENCES pollution. 8. In summer, the AQI was mainly grades I–II in both Adame, J.A., Hernández-Ceballos, M.A., Sorribas, M., cities, which indicated that the quality of the ambient air Lozano, A. and Morena, B.A.D.L. (2014). Weekend- was good or moderate. Different from other seasons, O3 weekday effect assessment for O3, NOx, CO and PM10 in

2104 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

Andalusia, Spain (2003-2008). Aerosol Air Qual. Res. Canonaco, F., Li, G.H., Bozzetti, C., Daellenbach, K.R., 14: 1862–1874. Huang, Y., Zhang, R.J., Li, Z.Q., Cao, J.J., Baltensperger, Akimoto, H. (2003). Global air quality and pollution. U., ElHaddad, I. and Prévôtl, A.S.H. (2016). New Science 302: 1716–1719. insights into PM2.5 chemical composition and sources in Atkinson, R. and Arey, J. (2003). Atmospheric degradation two major cities in china during extreme haze events of volatile organic compounds. Chem. Rev. 103: 4605– using aerosol mass spectrometry. Atmos. Chem. Phys. 4638. 16: 3207–3225. Bell, M.L. (2012). Assessment of the health impacts of Fang, C.S., Zhang, Z., Jin, M.Y., Zou, P.C. and Wang, J. particulate matter characteristics. Research Report 161. (2016). Pollution characteristics of PM2.5 aerosol during Health Effects Institute, Boston, MA, USA. haze periods in , China. Aerosol Air Qual. Berntsen, T.K., Isaksen, I.S.A., Myhre, G., Fuglestvedt, Res. 17: 888–895. J.S., Stordal, F., Larsen, T.A., Freckleton, R.S. and Gong, X., Hong, S. and Jaffe, D.A. (2018). Ozone in Shine, K.P. (1997). Effects of anthropogenic emissions China: Spatial distribution and leading meteorological on tropospheric ozone and its radiative forcing. J. factors controlling O3 in 16 Chinese cities. Aerosol Air Geophys. Res. 102: 28101–28126. Qual. Res., in Press. Boersma, K.F., Jacob, D.J., Trainie, M., Rudich, Y., Hajek, P. and Olej, V. (2015). Predicting common air DeSmedt, I., Dirksen, R. and Eskes, H.J. (2009). quality index-the case of Czech microregions. Aerosol Validation of urban NO2 concentrations and their diurnal Air Qual. Res. 15: 544–555. and seasonal variations observed from the SCIAMACHY Halmer, M.M., Schmincke, H.U. and Graf, H.F. (2002). and OMI sensors using in situ surface measurements in The annual volcanic gas input into the atmosphere, in Israeli cities. Atmos. Chem. Phys. 9: 3867–3879. particular into the stratosphere: A global data set for the Brauer, M., Freedman, G., Frostad, J., Van Donkelaar, A., past 100 years. J. Volcanol. Geotherm. Res. 115: 511– Martin, R.V., Dentener, F. and Balakrishnan, K. (2016). 528. Ambient air pollution exposure estimation for the global Han, L.H., Zhuang, G.S., Sun, Y.L. and Wang, Z.F. (2005). burden of disease 2013. Environ. Sci. Technol. 50: 79– Local and non-local sources of airborne particulate 88. pollution at Beijing-The ratio of Mg/Al as an element Brunekreef, B. and Forsberg, B. (2005). Epidemiological tracer for estimating the contributions of mineral evidence of effects of coarse airborne particles on aerosols from outside Beijing. Sci. China, Ser. B Chem. health. Eur. Respir. J. 26: 309–318. 48: 253–264. Chan, C.C., Chuang, K.J., Chien, L.C., Chen, W.J. and He, J.J., Gong, S.L., Yu, Y., Yu, L.J., Wu, L., Mao, H.J., Chang, W.T. (2006). Urban air pollution and emergency Song, C.B., Zhao, S.P., Liu, H.L., Li, X.Y. and Li, R.P. admissions for cerebrovascular diseases in , (2017). Air pollution characteristics and their relation to Taiwan. Eur. Heart J. 27: 1238–1244. meteorological conditions during 2014-2015 in primary Chen, P.F., Zhang, J.Q., Bi, X.H., Feng, Y.C. and Wu J.H. Chinese cities. Environ. Pollut. 223: 484–496. (2013). Pollution characteristics and sources of heavy http://tj.hd.gov.cn; http://www.ha.stats.gov.cn metals in PM10 and PM2.5 in city. Acta Sci. Nat. http://www.aqistudy.cn Univ. Nankaiensis 46: 1–7. http://www.mep.gov.cn/hjzl/zghjzkgb/lnzghjzkgb Cheng, N.L., Li, Y.T., Sun, F., Chen, C., Wang, B., Li, Q., https://www.airnow.gov/ Wei, P. and Cheng, B.F. (2017). Ground-grade NO2 in Hu, G., Zhong, N. and Ran, P. (2015a). Air pollution and urban Beijing: Trends, distribution, and effects of emission COPD in China. J. Thorac. Dis. 7: 59–66. reduction measures. Aerosol Air Qual. Res. 18: 343– Hu, J.L., Ying, Q., Wang, Y.G. and Zhang, H.L. (2015b). 356. Characterizing multi-pollutant air : Dentener, F., Kinne, S., Bond, T., Boucher, O., Cofala, J., Comparison of three air quality indices. J. Environ. Int. Generoso, S., Ginoux, P., Gong, S., Hoelzemann, J.J., 15: 17–25. Ito, A., Marelli, L., Penner, J.E., Putaud, J.P., Textor, C., Huang, Q., Cheng, S.Y., Li, Y.P., Li, J.B., Chen, D.S., Wang, Schulz, M., Van der Werf, G.R. and Wilson, J. (2006). H.Y. and Li, S. (2010). An integrated MM5-CAMx Emissions of primary aerosol and precursor gases in the modeling approach for assessing PM10 contribution from years 2000 and 1750 prescribed data-sets for AeroCom. different sources in Beijing, China. J. Environ. Inf. 15: Atmos. Chem. Phys. 6: 4321–4344. 47–61. Du, H., Kong, L., Cheng, T., Chen, J., Du, J., Li, L., Xia, Johnson, C., Henshaw, J. and McInnes, G. (1992). Impact X., Leng, C. and Huang, G. (2011). Insights into of aircraft and surface emissions of nitrogen oxides on summertime haze pollution events over Shanghai based tropospheric ozone and global warming. Nature 355: on online water-soluble ionic composition of aerosols. 69–71. Atmos. Environ.45: 5131–5137. Kato, S., Shiobara, Y., Uchiyama, K., Miura, K., Okochi, H., Du, P.R., Du, R., Lu, Z.D., Ren, W.S. and Fu, P.Q. (2017). Kobayashi, H. and Hatakeyama, S. (2016). Atmospheric Variation of bacterial and fungal community structures CO, O3, and SO2 measurements at the summit of Mt. in PM2.5 collected during the 2014 APEC summit periods. Fuji during the Summer of 2013. Aerosol Air Qual. Res. Aerosol Air Qual. Res. 18: 444–455. 16: 2368–2377. Elser, M., Huang, R.J., Wolf, R., Slowik, J.G., Wang, Q.Y., Kettle, A. and Andreae, M. (2000). Flux of dimethylsulfide

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2105

from the oceans: A comparison of updated data sets and Ma, Q.X., Wu, Y.F., Zhang, D.Z., Wang, X.J., Xia, Y.J., flux models. J. Geophys. Res. 105: 26793–26808. Liu, X.Y., Tian, P., Han, Z.W., Xia, X.G., Wang, Y. and Kurokawa, J., Ohara, T., Morikawa, T., Hanayama, S., Zhang, R.J. (2017). Roles of regional transport and Janssens-Maenhout, G., Fukui, T., Kawashima, K. and heterogeneous reactions in the PM2.5 increase during Akimoto, H. (2013). Emissions of air pollutants and winter haze episodes in Beijing. Sci. Total Environ. 599– greenhouse gases over Asian regions during 2000–2008: 600: 246–253. Regional emission inventory in Asia (REAS) version 2. Monks, P.S., Granier, C., Fuzzi, S., Stohl, A., Williams, Atmos. Chem. Phys. 13: 11019–11058. M.L., Akimoto, H., Amann, M., Baklanov, A., Lanzafame, R., Monforte, P., Patane, G. and Strano, S. Baltensperger, U., Bey, I., Blake, N., Blake, R.S., Carslaw, (2015). Trend analysis of Air Quality Index in Catania K., Cooper, O.R., Dentener, F., Fowler, D., Fragkou, E., from 2010 to 2014. Energy Procedia 82: 708–715. Frost, G.J., Generoso, S., Ginoux, P., Grewe, V., Guenther, Lee, C., Richter, A., Weber, M. and Burrows, J.P. (2008). A., Hansson, H.C., Henne, S., Hjorth, J., Hofzumahaus, SO2 Retrieval from SCIAMACHY using the Weighting A., Huntrieser, H., Isaksen, I.S.A., Jenkin, M.E., Kaiser, Function DOAS (WFDOAS) technique: Comparison J., Kanakidou, M., Klimont, Z., Kulmala, M., Laj, P., with Standard DOAS retrieval. Atmos. Chem. Phys. 8: Lawrence, M.G., Lee, J.D., Liousse, C., Maione, M., 6137–6145. McFiggans, G., Metzger, A., Mieville, A., Moussiopoulos, Leiva, G.M.A., Santibaez, D.A., Ibarra, E.S. and Matus, N., Orlando, J.J., O'Dowd, C.D., Palmer, P.I., Parrish, C.P. (2013). A five-year study of particulate matter D.D., Petzold, A., Platt, U., Pöschl, U., Prévôt, A.S.H., (PM2.5) and cerebrovascular diseases. Environ. Pollu. Reeves, C.E., Reimann, S., Rudich, Y., Sellegri, K., 181: 1–6. Steinbrecher, R., Simpson, D., ten Brink, H., Theloke, Lelieveld, J., Evans, J.S., Fnais, M., Giannadaki, D. and J., van der Werf, G.R., Vautard, R., Vestreng, V., Pozzer, A. (2015). The contribution of outdoor air Vlachokostas, C. and von Glasow, R. (2009). Atmospheric pollution sources to premature mortality on a global composition change-global and regional air quality. scale. Nature 525: 367–371. Atmos. Environ. 43: 5268–5350. Li, H., Zhang, Q., Duan, F., Bo, Z. and Kebin, H. (2016). Ou, X., Zhang, X. and Chang, S. (2010). Scenario analysis The “Parade Blue”: Effects of short-term emission control on alternative fuel/vehicle for China’s future road on aerosol chemistry. Faraday Discuss. 189: 317–335. transport: Life-cycle energy demand and GHG emissions. Li, J.Y., Zhang, H.L. and Ying, Q. (2012). Comparison of Energy Policy 38: 3943–3956. the SAPRC07 and SAPRC99 photochemical mechanisms Richter, A., Burrows, J.P., Nuss, H., Granier, C. and during a high ozone episode in Texas: Differences in Niemeier, U. (2005). Increase in tropospheric nitrogen concentrations, OH budget and relative response factors. dioxide over China observed from space. Nature 437: Atmos. Environ. 54: 25–35. 129–132. Li, K., Chen, L., Han, K., Lv, B., Bao, K., Wu, X., Guo, X. Sachin, D.G., Vander, A. and Beig, G. (2009). Satellite and Cen, K. (2017a). Smog chamber study on aging of derived trends in NO2 over the major global hotspot combustion soot in isoprene/SO2/NOx system: Changes regions during the past decade and their inter-comparison. of mass, size, effective density, morphology and mixing Environ. Pollut. 157: 1873–1878. state. Atoms. Res. 184: 139–148. Saikawa, E., Kim, H., Zhong, M., Avramov, A., Zhao, Y., Li, M., Zhang, Q., Kurokawa, J.I., Woo, J.H., He, K., Lu, Janssens-Maenhout, G., Kurokawa, J.I., Klimont, Z., Z., Ohara, T., Song, Y., Streets, D.G., Carmichael, G.R., Wagner, F., Naik, V., Horowitz, L.W. and Zhang, Q. (2017). Cheng, Y., Hong, C., Huo, H., Jiang, X., Kang, S., Liu, Comparison of emissions inventories of anthropogenic F., Su, H. and Zheng, B. (2017b). Mix: A mosaic Asian air pollutants and greenhouse gases in China. Atmos. anthropogenic emission inventory under the international Chem. Phys. 17: 6393–6421. collaboration framework of the MICS-Asia and HTAP. Seinfeld, J.H., Pandis, S.N. and Noone, K. (1998). Atmos. Chem. Phys. 17: 935–963. Atmospheric chemistry and physics: From air pollution Lin, J., Nielsen, C.P., Zhao, Y., Lei, Y., Liu, Y. and to climate change. Phys. Today 51: 88–90. McElroy, M.B. (2010). Recent changes in particulate air She, Q.N., Peng, X., Xu, Q., Long, L.B., Wei, N. Liu, M. pollution over China observed from space and the and Jia, W.X. (2017). Air quality and its response to ground: Effectiveness of emission control. Environ. Sci. satellite-derived urban form in the Yangtze River Delta, Technol. 44: 7771–7776. China. Environ. Res. Sci. 75: 297–306. Liu, F., Zhang, Q., Tong, D., Zheng, B., Li, M., Huo, H. Shen, F.Z., Ge, X.L., Hu, J.L., Nie, D.Y., Tian, L. and and He, K.B. (2015). High-resolution inventory of Chen, M.D. (2017). Air pollution characteristics and technologies, activities, and emissions of coal-fired health risks in Henan Province, China. Environ. Res. power plants in China from 1990 to 2010. Atmos. Chem. Sci. 156: 625–634. Phys. 15: 13299–13317. Silva, R.A., West, J.J., Zhang, Y., Anenberg, S.C., Lamarque, Liu, F., Zhang, Q., Ronald, J.v.d.A., Zheng, B., Tong, D., J.F., Shindell, D.T., Collins, W.J., Dalsoren, S., Faluvegi, , L., Zheng, Y. and He, K. (2016). Recent reduction G., Folberth, G., Horowitz, L.W., Nagashima, T., Naik, in NOx emissions over China: Synthesis of satellite V., Rumbold, S., Skeie, R., Sudo, K., Takemura, T., observations and emission inventories. Environ. Res. Bergmann, D., Cameron-Smith, P., Cionni, I., Doherty, Lett. 11: 114002. R.M., Eyring, V., Josse, B., MacKenzie, I.A., Plummer,

2106 Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018

D., Righi, M., Stevenson, D.S., Strode, S., Szopa, S. and Prod. 112: 1627–1634. Zeng, G. (2013). Global premature mortality due to Wang, Z.S., Wu, T., Shi, G.L., Fu, X., Tian, Y.Z., Feng, anthropogenic outdoor air pollution and the contribution Y.C. and Zhang, W.J. (2012). Potential source analysis of past climate change. Environ. Res. Lett. 8: 034005. for PM10 and PM2.5 in autumn in a northern city in Sitch, S., Cox, P.M., Collins, W.J. and Huntingford, C. China. Aerosol Air Qual. Res. 12: 39–48. (2007). Indirect radiative forcing of climate change Wei, W.X. and Wei, W.X. (2010). Spatial-temporal through ozone effects on the land-carbon sink. Nature characteristics of haze in Hebei province. Meteorol. 448: 791–794. Mon. 36: 77–82. Solomon, P., Cowling, E., Hidy, G. and Furiness, C. Wei, Z., Wang, L.T., Chen, M.Z. and Zheng, Y. (2014a). (2000). Comparison of scientic findings from major The 2013 severe haze over the Southern Hebei, China: ozone field studies in north America and Europe. Atmos. PM2.5 composition and source apportionment. Atmos. Environ. 34: 1885–1920. Pollut. Res. 5: 759–768. Sun, Y., Wang, Z., Fu, P., Jiang, Q., Yang, T., Li, J. and Ge, Wei, Z., Yang, J., Wang, L.T., Wei, W., Zhang, F.F. and Su, X. (2013). The impact of relative humidity on aerosol J. (2014b). Characteristics of the severe haze episode in composition and evolution processes during wintertime Handan City in January. Acta Sci. Circumst. 34: 1118– in Beijing, China. Atmos. Environ. 77: 927–934. 1124. Sun, Y., Jiang, Q., Wang, Z., Fu, P., Li, J., Yang, T. and Wu, Y., Zhang, S., Hao, J., Liu, H., Wu, X., Hu, J., Walsh, Yin, Y. (2014). Investigation of the sources and M.P., Wallington, T.J., Zhang, K.M. and Stevanovic, S. evolution processes of severe haze pollution in Beijing (2017). On-road vehicle emissions and their control in in January 2013. J. Geophys. Res. 119: 4380–4398. China: A review and outlook. Sci. Total Environ. 574: Sun, Y.L., Zhuang, G.S., Tang, A.H., Wang, Y. and An, 332–349. Z.S. (2006). Chemical characteristics of PM2.5 and PM10 Yang, C.X., Yang, H.B., Guo, S., Wang, Z.S., Xu, X.H., in haze-fog episodes in Beijing. Environ. Sci. Technol. Duan, X.L. and Kan, H.D. (2012). Alternative ozone 40: 31448–3155. metrics and daily mortality in : The China Air Tang, A., Zhuang, G., Wang, Y., Yuan, H. and Sun, Y. Pollution and Health Effects Study (CAPES). Sci. Total (2005). The chemistry of precipitation and its relation to Environ. 426: 83–89. aerosol in Beijing. Atmos. Environ. 39: 3397–3406. Yang, X.F., Liu, H., Man, H.Y. and He, K.B. (2015). Vijay, S., Molina, L.T. and Molina, M.J. (2004). Characterization of road freight transportation and its Estimating Air pollution emissions from fossil fuel use impact on the national emission inventory in China. in the electricity sector in Mexico. Integrated Program Atmos. Chem. Phys. 15: 2105–2118. on Urban, Regional and Global Air Pollution, Prepared Zeller, M., Giroud, M., Royer, C., Benatru, I., Besancenot, for North American Commission for Environmental J.P., Rochette, L. and Cottin, Y. (2006). Air pollution and Cooperation. cardiovascular and cerebrovascular disease epidemiologic Vingarzan, R. (2004). A review of surface ozone data. Presse Med. 35: 1517–1522. background levels and trends. Atmos. Environ. 38: Zhang, H.L. and Ying, Q. (2011). Contributions of local 3431–3442. and regional sources of NOx to ozone concentrations in Wang, H.L., Zhuang, Y.H., Wang, Y., Sun, Y.L., Yuan, H., Southeast Texas. Atmos. Environ. 45: 2877–2887. Zhuang, G.S. and Hao, Z.P. (2008). Long-term Zhang, L. and Jaffe, D.A. (2017). Trends and sources of monitoring and source apportionment of PM2.5/PM10 in ozone and sub-micron aerosols at the Mt. Bachelor Beijing, China. J. Environ. Sci. China 20: 1323–1327. Observatory (MBO) during 2004–2015. Atmos. Environ. Wang, S., Xing, J., Zhao, B., Jang, C. and Hao, J. (2014). 165: 143–154. Effectiveness of national air pollution control policies Zhang, Q., Quan, J.N., Tie, X.X., Li, X., Liu, Q., Gao, Y. on the air quality in metropolitan areas of China. J. and Zhao, D.L. (2015). Effects of meteorology and Environ. Sci. 26: 13–22. secondary particle formation on visibility during heavy Wang, S., Paul, M.J. and Dredze, M. (2015a). Social media haze events in Beijing, China. Sci. Total Environ. 502: as a sensor of air quality and public response in China. 578–584. J. Med. Internet Res. 17: e22. Zhang, Q., Streets, D.G., Carmichael, G.R., He, K.B., Huo, Wang, W.N., Cheng, T.H., Gu, X.F., Chen, H., Guo, H., H., Kannari, A., Klimont, Z., Park, I.S., Reddy, S., Fu, J.S., Wang, Y., Bao, F.W., Shi, S.Y., Xu, B.R., Zuo, X., Meng, Chen, D., Duan, L., Lei, Y., Wang, L.T. and Yao, Z.L. C. and Zhang, X.C. (2017). Assessing spatial and temporal (2009). Asian emissions in 2006 for the NASA INTEX- patterns of observed ground-level ozone in China. Sci. B mission. Atmos. Chem. Phys. 9: 5131–5153. Rep. 7: 3651. Zhang, Y., Cooper, O.R., Gaudel, A., Thompson, A.M., Wang, Y., Shen, L., Wu, S., Mickley, L., He, J. and Hao, J. Nédélec, P., Ogino, S.Y. and West, J.J. (2016). (2013). Sensitivity of surface ozone over China to 2000- Tropospheric ozone change from 1980 to 2010 dominated 2050 global changes of climate and emissions. Atmos. by equatorward redistribution of emissions. Nature Environ. 75: 374–382. Geosci.9: 875–879. Wang, Y.T., Sun, M.X., Yang, X.C. and Yuan, X.L. Zhao, H.J.,Che, H.Z., Wang, Y.Q., Dong, Y.S., Ma, Y.J., Li, (2015b). Public awareness and willingness to pay for X.X., Hong, Y., Yang, H.B., Liu, Y.C., Wang, Y.F., Gui, tackling smog pollution in China: A case study. J. Clean. K., Sun, T.Z., Zheng, Y. and Zhang, X.Y. (2017). Aerosol

Zhao et al., Aerosol and Air Quality Research, 18: 2086–2107, 2018 2107

vertical distribution and typical air pollution episodes Pöschl, U., Cheng, Y.F. and He, K.B. (2015). Exploring over northeastern China during 2016 analyzed by the severe winter haze in Beijing: The impact of ground-based lidar. Aerosol Air Qual. Res. 18: 918–937. synoptic weather, regional transport and heterogeneous Zhao, R., Cui, K.P., Wang, W.W., Wang, L.C. and Yan, P. reactions. Atmos. Chem. Phys. 15: 2969–2983. (2018). Atmospheric PM2.5 and total PCDD/Fs-WHO2005- TEQ Level: A case of Handan and Kaifeng Cities, China. Aerosol Air Qual. Res. 18: 994–1007. Received for review, May 16, 2018 Zheng, G.J., Duan, F.K., Su, H., Ma, Y.L., Cheng, Y., Revised, June 28, 2018 Zheng, B., Zhang, Q., Huang, T., Kimoto, T., Chang, D., Accepted, July 7, 2018