Part I: PM2.5 and Polychlorinated Dibenzo-p-dioxins and Dibenzofurans (PCDD/Fs) in the Ambient Air of Southern China

The atmospheric PM2.5, PM2.5/PM10, PCDD/Fs-WHO2005-TEQ, and PCDD/F (polychlorinated dibenzo-p-dioxins and dibenzofuran) phase distributions of 23 cities in southern China, during 2014–2016, were investigated in this study. In general, the cities with higher latitudes had higher PM2.5 concentrations than those with lower latitudes. During 2014– 2016, the lowest three-year average concentrations of PM2.5 occurred at Sanya and Haikou and were 16.4 and 21.7 μg m, respectively; while the highest concentrations of PM2.5 was occurred at Wuhan and Luzhou and were 68.8 and 63.1 μg m, respectively. During 2015–2016, the PM2.5 concentrations of most of cities decreased, but those of five cities (Chengdu, Luzhou, Nanchang, Qujing and Quanzhou) increased, indicting that the air quality of these five cities was still not well controlled. The average RM values of the 23 cities were 5.20, 4.49 and 4.13 in 2014, 2015 and 2016, respectively, which revealed that the PM2.5 concentrations in the cities of southern China slowly decreased, although they were still far above the WHO air quality regulated standard (10 μg m). In general, a city with a higher PM2.5 concentration was also had a higher PM2.5/PM10 ratio. Among the 23 cities, the six highest three-year averages of total-PCDD/Fs-WHO2005-TEQ concentrations were 0.0665, 0.0633, 0.0625, 0.0600, 0.0528 and 0.0526 pg-WHO2005-TEQ m in Chengdu, Wuhan, Nanjing, Hefei, Luzhou and Hangzhou, respectively. During 2014, the six cities (Hefei, Nanjing, Wuhan, Guiyang, Shanghai and Chengdu) with the lowest temperatures in winter (an average of 5.4°C), their average particle phase fractions of total-PCDD/FsWHO2005-TEQ that were approximately 76%, 53%, 71% and 93% in the spring, summer, fall and winter, respectively; while, the six cities (Haikou, Fuzhou, Guangzhou, Nanning, Nanchang and Changsha) with the highest temperatures in summer (an average of 16.5°C), had average particle phase fractions of total-PCDD/Fs-WHO2005-TEQ that were approximately 61%, 42%, 57% and 81% in the spring, summer, fall and winter, respectively. The results of this study provide information showing the trends of both atmospheric PM2.5 and PCDD/Fs in the cities of southern China. In addition, this study provided the overview relating to the PM2.5 and PCDD/Fs in ambient air of southern China, which was not reported in previous studies. The results of this study were of great importance to present the trends of air quality in China. It is also useful for the establishment of control strategies in the future.


INTRODUCTION
Due to its harmful properties, the issue of atmospheric PM 2.5 has received much more attention in recent years.PM 2.5 is also known as fine particles, which refers to the particulate matter (PM) with aerodynamic diameters of less than 2.5 µm.The characteristics of PM are connected to this size (including physical characteristics such as columnar optical depth, size distribution, single scattering albedo, refractive index, and), while the chemical composition has a direct influence the related toxicity (NASA Facts, 2005;Kahn et al., 2009;Rosenfeld et al., 2014;Cheng et al., 2010).A length of 2.5 µm is equal to 3.6% of the diameter of a human hair, and this means that PM can be easily inhaled deeply in human lungs and directly penetrate the pulmonary alveolar cells, thus mobbing into the blood circulatory system (Yang et al., 2017).PM 2.5 can absorb toxic constituents, and have many adverse health effects due to its higher surface/mass ratio compared to other PM (WHO, 1999;EPA, 2015).A large amount of research shows that individuals exposed to high levels of PM 2.5 have a significantly higher risk of cardiovascular and pulmonary disease than others.In addition, PM 2.5 can interfere with atmospheric visibility by scattering and/or absorbing solar light, which impacts the amount and spectral distribution of the incoming solar radiation, and so the Earth's radiation budget (Kalaiarasan et al., 2017;Qi et al., 2017;Haywood and Boucher, 2000;Ramanathan et al., 2001;Kanniah and Yaso, 2010).
Some studies have shown the relationship between PM 2.5 /PM 10 and carbon content, such as organic carbon (OC) and elemental carbon (EC), and other chemical indexes, like water-soluble ions, and elements including Be, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Cd, Sb, Hg and Pb.A study carried out in southern Taiwan showed that As, Se, Cd, Sb, Hg were the most enriched in PM 2.5 (Kuo et al., 2007).Another study at a city in a semi-arid location in northeast China found that the mass level of PM 2.5 corresponded to dust storm events, and was much higher in dust storm periods than otherwise.The total mass concentration of nine kinds of water-soluble ions (SO 4 2-, NO 3 -, Cl -, F -, Na + , NH 4 + , K + , Mg 2+ and Ca 2+ ) accounted about 17% of the PM 2.5 mass in the region (Shen et al., 2011).In the city of Tianjin, the concentrations of OC and EC in PM 2.5 and PM 10 were all relatively higher in both winter and fall and lower in summer and spring, the similar conclusions have been obtained in a study of the indoor and outdoor carbonaceous pollution in Shaanxi, China (Gu et al., 2010;Zhu et al., 2010).A study of contamination of polycyclic aromatic hydrocarbons bound to PM 10 /PM 2.5 in Xiamen, China, showed that between December 20 and 29, 2004, the concentrations of total 16 polycyclic aromatic hydrocarbons were 5.20-28.1 ng m -3 in PM 10 and 3.04-11.3ng m -3 in PM 2.5 .The concentrations of OC and EC were 9.83-15.6µg m -3 and 1.41-4.73µg m -3 , respectively, as associated with PM 10 .This suggests that the highest concentration of Σ16PAHs always occurred at areas with factories or traffic, and thus industry, vehicles and coal combustion are the major contributors of PAHs in the ambient air in Xiamen (Wang et al., 2007).
Atmospheric aerosols that include PM 2.5 matters refer to small liquid and solid particles suspended in the air (Wilson et al., 2002), and these can be emitted from multiple sources and have complex compositions.Sources such as industrial activities, energy production, construction, urban waste treatment and vehicle exhausts constitute anthropogenic sources, while dust storms, volcanic and oceans activities are natural sources.Due to the characteristics of ambient aerosols, they can significantly affect air quality, the global environment and human health.
Toxic substances such as dioxins, polycyclic aromatic hydrocarbons and heavy metals often attached to PM 2.5 .Polychlorinated dibenzo-p-dioxin and dibenzofurans (PCDD/Fs) are unintentional by-products of combustion processes and many industrial activities, such as waste incineration, metal production, power and heating facilities and chemical manufacture processes.After being emitted from combustion sources, they are distributed in both gas and particle phases in the atmosphere.Climate change and increasing climate variability have the potential to influence the emission, distribution and degradation of PCDD/Fs (Bogdal et al., 2010;Chi et al., 2014Chi et al., , 2015)).Factors that have been proposed by the United Nations Environment Programme as the key ones influencing the environmental fate of transportation of PCDD/Fs include the re-evaporation of secondary sources, wind, precipitation, ocean currents, the melting of ice caps and mountains glaciers at the poles, the frequency of extreme weather events, the degradation and transportation of POPs, environmental partitioning and biotic transportation (UNEP/AMAP, 2011; Chi et al., 2016).
Many studies have been carried on PM 2.5 , PCDD/Fs, and the relationships between them.The foci of previous works can be classified into several categories: PM 2.5 emission sources, regional distribution, health impacts, analysis and simulation of concentration and chemical compositions.With the industrial production rate developing very rapidly in China, air pollution has become one of the main problems the country faces.In recent years, haze episodes have occurred frequently in northern Chinese cities, such as Beijing, Tianjin, and Hebei (Lang et al., 2012).Some researchers focused on the PM 2.5 pollution circumstance in recent years of Chinese special sites.A study measured the PM 2.5 concentration in ambient air of three sites in Shijiazhuang during 18-22 January 2016, aimed to understand the chemical characteristics and potential source regions of PM 2.5 in Shijiazhuang, China.The results showed that the potential sources of PM 2.5 in Shijiazhuang were from the Beijing-Tianjin region and Shandong Province (Chen et al., 2017).Other researchers investigated the long-term trend (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) of PM 2.5 and found that it generally decreased from 2000 to 2015, with the changes seen in the most polluted season (winter) being related to emission control effects and meteorological conditions (Lang et al., 2017).As one of the largest cities in China, Beijing suffers from a great number of air pollution problems, and haze incidents have been occurring more frequently in recent years (Zheng et al., 2014).A study showed that the sum of the concentrations of the seventeen 2,3,7,8-PCDD/Fs that were examined, and those of the concentrations of the thirteen 2,3,7,8-PBDD/Fs that were analyzed, were 1499-2799 fg m -3 (95.4-175.4fg I-TEQ m -3 ) and 1171-2424 fg m -3 (42.2-109.3fg-I-TEQ m -3 ), respectively.Moreover, the PXDD/Fs were mainly (90%) in the particulate phase in the ambient air samples collected in a suburban part of Beijing.The fraction of the total PXDD/Fs concentration that was in the particulate phase increased as the particle size decreased.More than 80% of the total PXDD/F concentrations were in the form of particles of d ae < 2.5 µm (Zhang et al., 2015).
In light of these earlier work, the objectives of this study were to investigate the (1) PM 2.5 concentrations, (2) R M ratio, (3) PM 2.5 /PM 10 ratio, (4) total-PCDD/F-WHO 2005 -TEQ concentrations, and (5) PM 2.5 -bound total PCDD/Fs-WHO 2005 -TEQ content in the ambient air of northern China.The results of this study are expected to provide useful information to help in finding more efficient strategies to improve the air quality in China.

METHODS
This study examined a total of 22 cities belonging to 14 provinces and two municipalities in northern China.These cities were Harbin and Qiqihar in Heilongjiang province, Changchun in Jilin, Dalian and Shenyang in Liaoning, Shijiazhuang and Baoding in Hebei, Hohhot in Inner Mongolia, Jinan and Linyi in Shandong, Taiyuan in Shanxi, Yinchuan in Ningxia, Lanzhou in Gansu, and Lhasa in Tibet, along with the municipalities of Beijing and Tianjin.Both the PM 2.5 and PM 10 concentrations in each city were obtained from the website of Air Quality Monitoring Stations (The Real Atmosphere Network).
Since PCDD/Fs exist in both particle and gas phases, the partitioning fraction will affect the atmospheric deposition processes.The partitioning between the two phases is highly dependent on their vapor pressure and the ambient temperature.The gas-particle partitioning is described by the following equation (Yamasaki et al., 1982;Pankow, 1987;Pankow andBidleman, 1991, 1992) where K p : temperature-dependent partitioning constant (m 3 µg -1 ); TSP: concentration of total suspended PM (µg m -3 ), as calculated by the factor TSP/PM 10 = 1.24 (Sheu et al., 1996) where RI: gas chromatographic retention indexes (GC-RI), referring to Hale et al. (1985) and Donnelly et al. (1987); T: ambient temperature (K) (Hung et al., 2002).
The cities chosen in this study are either provincial capitals or other densely populated cities.We evaluated each concentration of the 17 PCDD/Fs congeners by using its proportion in the total PCDD/F mass concentration, which is deduced from the atmospheric concentration of PCDD/Fs during four seasons at commercial areas in Taiwan (Wang et al., 2017).The total PCDD/F concentration is the average of the results of using two formulas that demonstrate the relationship between PM 10 values and total PCDD/F mass concentrations.One was presented by Huang et al. (2011) shown below as Eq. ( 4) and the other was presented by Lee et al. (2016) shown as Eq. ( 5).The correlation coefficients of the two equations are as high as 0.9438 and 0.9855, respectively.y = 0.0472 + 0.0138x (4) where y: total PCDD/Fs concentration (pg m -3 ); x: PM 10 concentration (µg m -3 ).
To obtain the toxicity equivalent (TEQ), the concentrations of PCDD/F congeners are multiplied by their respective TEF values.There are two TEF schemes, the International Toxicity Equivalent (I-TEF) and World Health Organization TEF (WHO 2005 -TEF) (Cheruiyot et al., 2016).In this study, the WHO 2005 -TEF was used to calculate TEQ concentration, as this is the stricter scheme.The total TEQ of PCDD/Fs is the sum of all the individual TEQ.

PM 2.5 Concentration
The PM 2.5 concentration reflects the air pollutant status of a region.In addition, the PM 2.5 concentration influences both air visibility and human health.The corresponding PM 2.5 concentrations during the period between 2014-2016 in 22 cities of northern China are shown in Tables 1(A)-1(D).Among the 22 cities, the lowest annual average PM 2.5 concentrations during 2014-2016 were all found to be at Lhasa in Tibet and Qiqihar in Heilongjian, which had concentrations of 23.5 and 37.5 µg m -3 , 24.6 and 37.2 µg m -3 and 27.4 and 35.3 µg m -3 , in 2014, 2015, and 2016, respectively.While during 2014 and 2015, the highest annual average PM 2.5 concentrations both occurred in Baoding City in Hebei, with the levels of 120 and 106 µg m -3 , respectively; however in 2016, the highest annual average PM 2.5 concentration was 98.5 µg m -3 at Shijiazhuang City in Hebei.During 2014-2016, the lowest three-year average concentrations of PM 2.5 occurred at Lhasa and Qiqihar, and were 25.2 and 36.6 µg m -3 , respectively.While the highest concentrations of three-year average PM 2.5 occurred at Shijiazhuang and Baoding City, as 102 and 106 µg m -3 , respectively.
The six cities with the highest three-year average PM 2.5 concentrations (HAC) among the 22 cities were Baoding, Shijiazhuang, Zhengzhou, Jinan, Beijing and Linyi, with a range between 79.1 and 106 µg m -3 , and an average of 89.7 µg m -3 .The six cities with the lowest three-year average PM 2.5 concentrations (LAC) among the 22 cities were Lhasa, Qiqihar, Hohhot, Dalian, Yinchuan and Sinning, with a range between 25.2 and 52.5 µg m -3 , and an average of 41.6 µg m -3 .These results show that the six cities with the highest three-year average PM 2.5 concentrations (HAC) levels were approximately 2.2 times higher than that the LAC.The six cities with the highest PM 2.5 concentrations were Wuhan, Hefei, Chengdu, Luzhou, Changsha and Nanjing, with PM 2.5 concentrations between 47.8 and 80.5 µg m -3 , and an average of 64.4 µg m -3 .The six cities with the lowest PM 2.5 concentrations were Sanya, Haikou, Quanzhou, Fuzhou, Kunming and Shenzhen, with levels between 13.6 and 32.9 µg m -3 , and an average of 26.0 µg m -3 (Tang et al., 2017).The average PM 2.5 concentrations of the six northern cities with highest and lowest PM 2.5 concentration are 1.39 and 1.60 times higher than those seen in the south, respectively.The high PM 2.5 concentrations in Shijiazhuang and Baoding City may be due to them both being highly industrialized, with a lot of coal and other fossil fuel combustion for heating.In Hebei province, there are large amount of steel, building materials, petrochemical and power production facilities, which all increase PM 2.5 emissions.In addition, the north wind in winter and spring caused by terrain conditions is not conducive to the diffusion of regional pollutants.Moreover, the south wind in summer and autumn blocks the southwest air stream of Hebei Province, which has a great influence on the air quality of this area.A related study pointed that the low temperature of a region will work against vertical ventilation and prevent the dispersion of contaminants (Tang et al., 2017).The temperature in the north is lower than in the south, especially in winter, and this produces a greater need to burn coal for heating, which thus could make the air quality worse in the north.Among the 22 cities in northern China examined in this study, the lowest PM 2.5 concentrations were all found to be at Lhasa, with the , and 2016, respectively.As mentioned above, the highest PM 2.5 concentrations were found to be at Baoding, with levels of 120 and 106 µg m -3 in 2014 and 2015, respectively, while during 2016 the highest levels were at Shijiazhuang, with a PM 2.5 concentration of 98.5 µg m -3 .However, in southern China, the highest PM 2.5 concentrations were at Wuhan, with levels of 80.5 and 68.8 µg m -3 , respectively; in contrast, in 2016 the highest level was 64.0 µg m -3 in Luzhou (Tang et al., 2017).These results show that in 2014, 2015, and 2016, the average PM 2.5 concentration of these six cities in the north were 1.50, 1.54 and 1.54 times higher than those of south, respectively.It can also be seen that the lowest PM 2.5 concentrations in southern China were much lower than those in northern China.
As for the three-year average PM 2.5 concentration, the highest three-year concentration in northern China was found to be in Baoding, at 106 µg m -3 , with the lowest at 25.2 µg m -3 , in Lhasa.The highest three-year average PM 2.5 concentration in southern China was found to be in Wuhan, at concentration of 68.8 µg m -3 , which is approximately 65% lower than that of Baoding.It can be seen that the PM 2.5 concentrations in northern China are much higher than those in southern China.One of the reasons for this is that the northern part of China is colder, and thus coal is burned for heat for a longer period of time, with the lower temperature also preventing pollutant diffusion, thus making the PM 2.5 concentration higher for the whole year.
Comparing the annual average PM 2.5 concentrations in the 22 northern cities during 2014-2016, it can be found that the concentration was highest in 2014, followed by 2015, and lowest in 2016.As for any reductions, Shijiazhuang was the city with the highest decrease in the PM 2.5 concentration between 2014 and 2015, falling from 118 to 88.1 µg m -3 , falling at a rate of 25.3%.While the PM 2.5 concentration in Shijiazhuang increasing to 98.5 µg m -3 in 2016, at a rate of 11.8%.The annual average PM 2.5 concentrations of the 22 cities in this study in 2014, 2015 and 2016 were 71.5, 64.8, and 61.4 µg m -3 , with the relative standard deviations (RSD) being 33.0%, 32.3% and 29.9%, respectively.The results also showed that the annual average concentration of PM 2.5 in 2015 decreased by 7.0 µg m -3 (9.8%) compared to 2014, while for 2016, it decreased by 3.3 µg m -3 (5.1%) compared to 2015.This means during these three years the PM 2.5 levels in the atmosphere have fallen, although still high.The air quality has thus improved, but there is still much more work to do here.
With regard to 2015 and 2016, the PM 2.5 concentrations of most cities decreased in the latter year.For example, that in Baoding decreased from 106 to 92.1 µg m -3 , while in Zhengzhou it fell from 96.1 to 78.5 µg m -3 , in Jinan from 89.8 to 75.4 µg m -3 , in Beijing from 80.2t o 72.8 µg m -3 , in Linyi from 77.3to 67.0 µg m -3 , and in Nanyang from 73.6 to 60.9 µg m -3 .Moreover, some cities saw an increase from 2015 to 2016.For example, the level in Shijiazhuang increased from 88.1 to 98.5 µg m -3 , that in Taiyuan from 60.3 to 66.3 µg m -3 , in Yinchuan from 48.3 to 54.8 µg m -3 , in Lanzhou from 50.1 to 53.7 µg m -3 , in Lhasa from 24.6 to 27.4 µg m -3 , in Sinning from 47.8 to 48.7 µg m -3 , in Urumqi from 64.6 to 73.1 µg m -3 , in Weinan from 58.8 to 76.0 µg m -3 , and in Xian from 57.8 to 71.4 µg m -3 .It can thus be noted that the Beijing-Tianjin-Hebei region still had a serious air pollution problem in 2016, although the PM 2.5 level in 2016 was still lower than 2015 and 2014.
With regard to the four seasons, defined as spring (March, April, May), summer (June, July, August), fall (September, October, November) and winter (January, February and December), the PM 2.5 concentration was always highest in winter and in summer.For example, during 2016, the average PM 2.5 concentrations in Shijiazhuang were 65.6, 45.5, 123 and 160 µg m -3 in spring, summer, fall and winter, respectively.In Baoding, the average PM 2.5 concentrations were 70.2, 54.3, 103 and 141 µg m -3 , respectively.In Beijing, they were 71.7, 58.4,79.6 and 81.4 µg m -3 , respectively; while in Xian they were 53.8, 39.5, 49.5 and 101 µg m -3 .These results may be because the temperatures were lower in these areas in winter and higher in summer, and the transport of atmospheric contaminants through the vertical current was hindered by the cold air during the former.
It should be noted that among the 22 cities in northern China examined in this study, most saw remarkable declines in their PM 2.5 concentrations in October 2015 compared with the same period in 2014.This may be due to the various memorial activities that took place with regard to the Anti-Japanese War and World War II, which took place from August 20 to September 3, 2015.During these activities, China's environmental protection administration took powerful action to stop the operation of factories and so improved the air quality of the Beijing-Tianjin-Hebei region, thus enabling to reach the best quality observed during the time examined in this work.

The R M of Atmospheric PM 2.5
The World Health Organization (WHO) standard was chosen, when the R M of different cities was examined, and this is 10 µg m -3 .The R M value is the ratio of PM 2.5 concentration to limitation, which can intuitively suggest the air pollution conditions.Comparisons of the R M in 22 cities for the years 2014, 2015 and 2016, are shown in Figs.1(A)-1(C), respectively.These reveal that in both 2014 and 2015 the highest R M values occurred in Baoding, at 12.0 and 10.6, respectively.However, for 2016, the level at Baoding decreased to 9.21, and Shijiazhuang rose to first place with a level of 9.85.Over the three years, Lhasa always had the lowest R M values among the 22 cities, at 2. 35, 2.46 and 2.74 in 2014, 2015 and 2016, respectively.It can be seen from the figures that the top six R M among the 22 cities for each year ranged between 7.91 and 10.6, and averaged 8.97.This suggests that the average PM 2.5 concentration in this period was as high as 8.97 times the WHO air quality standard.Moreover, the six lowest R M values in three years loaded were in the range between 2.52 and 5.25, and averaged 4.21.When comparing the figures seen in 2014-2016, the average R M values were 7.15, 6.48 and 6.14 in 2014, 2015 and 2016, respectively.The fall in the R M value during 2015 and 2016 suggests that the PM 2.5 concentrations in these cities in northern China slowly decreased, but were still high, and most cities would have to fall by at least 80% to meet the WHO standard, at which levels they are considered harmlessness to human health.
The R M values for southern China for the years 2014, 2015 and 2016 were 5.2, 4.3 and 4.1, respectively.Compared with northern China, the average R M values of southern China were much lower.However, the ambient air quality for the whole of China has a long way to go to meet the WHO standard, although it has slowly improved.Moreover, the air pollution situation is better in the south than in the north, and in 2016 than in 2015.However, in order to meet the WHO standard for PM 2.5 , both the north and south need to make substantial changes.

PM 2.5 /PM 10 Ratio
The PM 2.5 /PM 10 ratio reflects the proportion of PM in the ambient air, and thus the impact of the air on human health.Due to its small particle diameter, PM 2.5 has a long standing time and propagation distance, and thus more significant impacts on the environment and human health.During 2014, the six highest annual averages of the PM 2.5 /PM 10 concentration ratios were 0.72 in Beijing, 0.64 in Tianjin, 0.61 in Harbin, 0.61 in Dalian, 0.60 in Zhengzhou and 0.58 in Shenyang; in 2015, the six highest annual averages of the PM 2.5 /PM 10 concentration ratios were 0.81 in Beijing, 0.63 in Harbin, 0.62 in Shenyang, 0.60 in Changchun, 0.59 in Baoding and 0.58 in Tianjin; and in 2016, the six Fig. 1(A).The R M of atmospheric PM 2.5 in various cities (2014).highest annual averages of the PM 2.5 /PM 10 concentration ratios were 0.76 in Beijing, 0.65 in Harbin, 0.64 in Tianjin, 0.62 in Baoding, 0.58 in Changchun and 0.58 in Dalian.
In 2014, the six lowest annual averages of the PM 2.5 /PM 10 concentration ratios were 0.38 in Hohhot, 0.41 in Urumqi, 0.42 in Lhasa, 0.45 in Yinchuan, 0.49 in Xian and 0.50 in Lanzhou; in 2015, the six lowest annual averages of the PM 2.5 /PM 10 concentration ratios were 0.39 in Hohhot, 0.43 in Lhasa, 0.43 in Lanzhou, 0.43 in Yinchuan, 0.45 in Xian and 0.46 in Urumqi; and in 2016, these were 0.35 in Lhasa, 0.41 in Lanzhou, 0.43 in Hohhot, 0.44 in Sinning, 0.48 in Yinchuan and 0.50 in Xian.The three-year averages of the PM 2.5 /PM 10 concentration ratios in northern China were in the range of 0.40 and 0.76, and averaged 0.54.
It was found that, in general, a city with a high PM 2.5 concentration always had a high PM 2.5 /PM 10 ratio, too.The top six cities in northern China with the highest three-year average PM 2.5 concentrations were Baoding, Shijiazhuang, Zhengzhou, Jinan, Beijing and Linyi, which an average PM 2.5 concentration of 89.7 µg m -3 , with an average PM 2.5 /PM 10 ratio of 0.59.However, during 2014-2016, among the 22 cities examined in this work, the six with the lowest three-year average PM 2.5 concentrations were Lhasa, Qiqihar, Hohhot, Dalian, Yinchuan and Sinning City, which had an average PM 2.5 concentration of 42.1 µg m -3 , and with an average PM 2.5 /PM 10 ratio of 0.48.A higher the PM 2.5 /PM 10 ratio means the PM in the ambient air are more harmful to people.
For comparison, the highest three-year average total-PCDD/Fs-WHO 2005 -TEQ concentration in southern China occurred in Chengdu, which was 0.067 pg-WHO 2005 -TEQ m -3 , but in the north the highest was for Shijiazhuang, with a level nearly 1.60 times that of Chengdu, at 0.107 pg-WHO 2005 -TEQ m -3 .The lowest three-year average total-PCDD/Fs-WHO 2005 -TEQ concentration in the south was 0.020 pg-WHO 2005 -TEQ m -3 in Sanya, while in the north it was 1.8 times higher in Qiqihar, at 0.036 pg-WHO 2005 -TEQ m -3 .By comparing the highest and lowest values, it can be seen that the total-PCDD/Fs-WHO 2005 -TEQ concentrations in northern China were much higher than those in the south.Tang et al. (2017) reported that one of the reasons why the southern cities with higher latitudes (like Wuhan and Hefei) have a relatively high total-PCDD/Fs-WHO 2005 -TEQ concentrations compared to other cities in the south is because they are located downstream of the air pollution current flowing from the northern cities, and may be affected by the long-range transport of pollutants from these (Tang et al., 2017).Wuhan and Hefei are in Hubei and Anhui provinces, respectively, and located along the edges of Henan and Shaanxi provinces.Moreover, the three-year average of total-PCDD/Fs-WHO 2005 -TEQ concentration in Zhengzhou (in Henan province) was 0.095 pg-WHO 2005 -TEQ m -3 , while in Xian (in Shaanxi province) it was 0.077 pg-WHO 2005 -TEQ m -3 , both figures being at the higher end among the 22 northern cities.This proves that the air quality of cities at higher latitudes of southern China was affected by the northern cities, which had higher levels of atmospheric pollutants.
Both Baoding (0.121 pg-WHO 2005 -TEQ m -3 ) and Shijiazhuang (0.119 pg-WHO 2005 -TEQ m -3 ) had higher total PCDD/Fs-WHO 2005 -TEQ concentrations than industrial areas (0.096 pg I-TEQ Nm -3 ) in Taiwan (Wang et al., 2010).This demonstrates that the total PCDD/Fs-WHO 2005 -TEQ concentration is correlated with the strength of industrial activities in specific areas.It is thus obvious that the total-PCDD/F-WHO 2005 -TEQ concentrations in the industrial areas were much higher than those in urban or rural areas.
As shown by Eduljee and Dyke (1996), industrial processes in the UK contributed 535-955 g total-PCDD/Fs I-TEQ yr -1 , which was approximately 90% of the total PCDD/Fs-I-TEQ emission inventory.The same study indicated that local industrial activity was always a key factor impacting the air quality level of the total-PCDD/Fs-I-TEQ concentration.
The highest monthly average total-PCDD/Fs-WHO 2005 -TEQ concentrations occurred during January, 2014, at Shijiazhuang, and during January, 2015, at Baoding (averaging 0.205 and 0.170 pg-WHO 2005 -TEQ m -3 , respectively), while for 2016, it occurred during October in Shijiazhuang, with the total-PCDD/Fs-WHO 2005 -TEQ concentration of 0.261 pg-WHO 2005 -TEQ m -3 .However, the lowest monthly average concentration for 2014 occurred in July, (0.017 pg WHO 2005 -TEQ m -3 ) in Lhasa, while in 2015 and 2016 it occurred during June in Qiqihar and during August at Dalian (0.018 and 0.019 pg WHO 2005 -TEQ m -3 , respectively).With regard to the seasonal variation, during  -3 , with an average of 61.4 µg m -3 ), followed by 2015 (ranging between 24.6 and 106 µg m -3 , with an average of 64.8 µg m -3 ), and the highest in 2014 (ranging between 23.5 and 120 µg m -3 , with an average of 71.5 µg m -3 ).Overall, from 2014 to 2016, the atmospheric PM 2.5 decreased in northern China.3. When comparing the PM 2.5 concentration in 2015 with those in 2016, it was found that the levels in most of the cities decreased, but there were still some cities in which they increased.For example, the values in Shijiazhuang rose from 88.1 to 98.5 µg m -3 , in Taiyuan from 60.3 to 66.3 µg m -3 , in Yinchuan from 48.3 to 54.8 µg m -3 , in Lanzhou from 50.1 to 53.7 µg m -3 , in Lhasa from 24.6 to 27.4 µg m -3 , in Sinning from 47.8 to 48.7 µg m -3 , in Urumqi from 64.6 to 73.1 µg m -3 , in Weinan from 58.8 to 76 µg m -3 , and in Xian from 57.8 to 71.4 µg m -3 .This means that in 2015 and 2016 these cities all still had poor air pollution control.4. With regard to monthly averages, during December the WHO 2005 -TEQ concentrations were 0.036, 0.037, 0.045, 0.055, 0.056 and 0.060 pg-WHO 2005 -TEQ m -3 in Qiqihar, Lhasa, Dalian, Harbin, Changchun and Hohhot, respectively.8.As for the PM 2.5 -bound total PCDD/Fs-WHO 2005 -TEQ content, during 2014, the PM 2.5 -bound total PCDD/Fs-WHO 2005 -TEQ content of 12 cities (the six with the highest PM 2.5 concentrations, and six with the lowest PM 2.5 concentrations) ranged between 1.00 and 0.444 ng-WHO 2005 -TEQ g -1 and averaged 0.672 ng-WHO 2005 -TEQ g -1 .The six cities with the lowest PM 2.5 concentrations were Hohhot, Lhasa, Yinchuan, Changchun, Sinning and Dalian, at 1.00, 0.855, 0.762, 0.661, 0.688 and 0.570 ng-WHO 2005 -TEQ g -1 , respectively.The six cities with the highest PM 2.5 concentrations in 2014 were Shijiazhuang, Urumqi, Zhengzhou, Tianjin, Beijing, and Harbin, at 0.587, 0.859, 0.444, 0.501, 0.463 and 0.568 ng-WHO 2005 -TEQ g -1 , respectively.9.This study presents the results of a systematic analysis of PM 2.5 and PCDD/Fs levels in northern China, and thus provides a theoretical basis for proposing better air pollution control strategies and improving the atmospheric environment in China.