Scholarly article on topic 'Significant concentration changes of chemical components of PM1 in the Yangtze River Delta area of China and the implications for the formation mechanism of heavy haze–fog pollution'

Significant concentration changes of chemical components of PM1 in the Yangtze River Delta area of China and the implications for the formation mechanism of heavy haze–fog pollution Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Keywords
{"Haze–fog event" / "PM1 " / "Chemical compositions" / "Mass-size distributions"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Y.W. Zhang, X.Y. Zhang, Y.M. Zhang, X.J. Shen, J.Y. Sun, et al.

Abstract Since the winter season of 2013, a number of persistent haze–fog events have occurred in central-eastern China. Continuous measurements of the chemical and physical properties of PM1 at a regional background station in the Yangtze River Delta area of China from 16 Nov. to 18 Dec., 2013 revealed several haze–fog events, among which a heavy haze–fog event occurred between 6 Dec. and 8 Dec. The mean concentration of PM1 was 212μgm−3 in the heavy haze–fog period, which was about 10 times higher than on clean days and featured a peak mass concentration that reached 298μgm−3. Organics were the largest contributor to the dramatic rise of PM1 on heavy haze–fog days (average mass concentration of 86μgm−3), followed by nitrate (58μgm−3), sulfate (35μgm−3), ammonium (29μgm−3), and chloride (4.0μgm−3). Nitrate exhibited the largest increase (~20 factors), associated with a significant increase in NOx. This was mainly attributable to increased coal combustion emissions, relative to motor vehicle emissions, and was caused by short-distance pollutant transport within surrounding areas. Low-volatility oxidized organic aerosols (OA) (LV-OOA) and biomass-burning OA (BBOA) also increased sharply on heavy haze–fog days, exhibiting an enhanced oxidation capacity of the atmosphere and increased emissions from biomass burning. The strengthening of the oxidation capacity during the heavy pollution episode, along with lower solar radiation, was probably due to increased biomass burning, which were important precursors of O3. The prevailing meteorological conditions, including low wind and high relative humidity, and short distance transported gaseous and particulate matter surrounding of the sampling site, coincided with the increased pollutant concentrations mainly from biomass-burning mentioned above to cause the persistent haze–fog event in the YRD area.

Academic research paper on topic "Significant concentration changes of chemical components of PM1 in the Yangtze River Delta area of China and the implications for the formation mechanism of heavy haze–fog pollution"

Significant concentration changes of chemical components of PMi in the A Yangtze River Delta area of China and the implications for the formation mechanism of heavy haze-fog pollution

Y.W. Zhang a, X.Y. Zhang a'*, Y.M. Zhang a, X.J. Shen a, J.Y. Sun a'b, Q.L. Ma c, X.M. Yu c, J.L. Zhu d, L. Zhang a'e, H.C. Che a,e

a Key Laboratory of Atmospheric Chemistry, Chinese Academy of Meteorological Sciences, Beijing 100081, China

b State Key Laboratory of Cryospheric Sciences, Cold and Arid Region Environmental and Engineering Research Institute, Chinese Academy of Sciences, Lanzhou 730000, China c Lin'an Regional Air Background Station, Lin'an 311307, China d School of Atmospheric Sciences, Nanjing University, Nanjing210093, China e College of Earth Science, University of Chinese Academy of Sciences, Beijing 100049, China

HIGHLIGHTS

• Formation mechanism of a heavy haze-fog event that occurred in the YRD region of china was discussed.

• Nitrate exhibited the largest increase on the heavy haze-fog days.

• Increased biomass burning and atmospheric oxidizing capacity were the major reasons for the significant increase of OA.

• Short-distance transport, low wind speed and high RH were highly responsible for the formation of the pollution event.

ARTICLE INFO ABSTRACT

Since the winter season of 2013, a number of persistent haze-fog events have occurred in central-eastern China. Continuous measurements of the chemical and physical properties of PM1 at a regional background station in the Yangtze River Delta area of China from 16 Nov. to 18 Dec., 2013 revealed several haze-fog events, among which a heavy haze-fog event occurred between 6 Dec. and 8 Dec. The mean concentration of PM1 was 212 ^g m-3 in the heavy haze-fog period, which was about 10 times higher than on clean days and featured a peak mass concentration that reached 298 ^g m- 3. Organics were the largest contributor to the dramatic rise of PMj on heavy haze-fog days (average mass concentration of 86 ^g m-3), followed by nitrate (58 ^g m-3), sulfate (35 ^g m-3), ammonium (29 ^g m-3), and chloride (4.0 ^g m-3). Nitrate exhibited the largest increase (-20 factors), associated with a significant increase in NOx. This was mainly attributable to increased coal combustion emissions, relative to motor vehicle emissions, and was caused by short-distance pollutant transport within surrounding areas. Low-volatility oxidized organic aerosols (OA) (LV-OOA) and biomass-burning OA (BBOA) also increased sharply on heavy haze-fog days, exhibiting an enhanced oxidation capacity of the atmosphere and increased emissions from biomass burning. The strengthening of the oxidation capacity during the heavy pollution episode, along with lower solar radiation, was probably due to increased biomass burning, which were important precursors of O3. The prevailing meteorological conditions, including low wind and high relative humidity, and short distance transported gaseous and partic-ulate matter surrounding of the sampling site, coincided with the increased pollutant concentrations mainly from biomass-burning mentioned above to cause the persistent haze-fog event in the YRD area.

© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Article history:

Received 10 April 2015

Received in revised form 17 June 2015

Accepted 25 June 2015

Available online 22 August 2015

Editor: D. Barcelo

Keywords: Haze-fog event PMi

Chemical compositions Mass-size distributions

1. Introduction

Chemical compositions will undergo singular change when various chemical components of aerosol exist simultaneously in high

* Corresponding author. E-mail address: xiaoye@cams.cma.gov.cn (X.Y. Zhang).

concentration. Previous studies have shown that higher concentrations of NOx and anthropogenic primary organic aerosol (POA) may result in more SOA through promoting the oxidation of biogenic volatile organic compounds (VOCs) (BVOCs) and its transformation to particle phase (Heald et al., 2011; Hoyle et al., 2011). Synergistic reactions between O3 and SO2 on the surface of mineral aerosols at low temperature can also cause the formation rate of sulfate to first increase, and then

http://dx.doi.org/10.1016/j.scitotenv.2015.06.104

0048-9697/© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

decrease, with the descending temperature, which directly affects the total mass of secondary sulfate aerosol formation (Wu et al., 2011). All these factors suggest that aerosol chemical composition changes may become more complicated during a heavy pollution episode.

Moreover, the characteristics of the chemical compositions of aerosols vary according to the prevailing meteorological conditions. For instance, temperature can affect the gas-aerosol equilibrium and the formation of nitrate aerosols significantly. Furthermore, changes to sulfate and nitrate caused by temperature can also influence the formation of ammonium indirectly (Pye et al., 2009). Indeed, heavy and persistent haze-fog events in eastern China have been shown to be caused by both singular static weather conditions and high aerosol loading (Zhang et al., 2013). The associated changes in aerosol chemical components and the formation mechanisms related to haze-fog days, especially heavy haze-fog days, have attracted much attention in the research community (Huang et al., 2014a; Wang et al., 2014e; Zhang et al., 2013).

Recently it has been recognized that the concentration of various chemical components in ambient aerosols over China are the highest in the world, besides those in urban south Asia; and the polluted area covered by high concentrations of particles represents a very large proportion of the total area for China (Zhang et al., 2012a). The Yangtze River Delta (YRD) area is one of four major hazy areas within China. Another three areas are the North China Plain and the Guanzhong Plain, the East China region mainly including the Yangtze River Delta (YRD) area, and the Si Chuan Basin region (Zhang et al., 2012a). Previous studies of aerosol pollution in the YRD region have mainly focused on its urban areas, such as Shanghai (Wang et al., 2014c, 2014d; Yang et al., 2012), Nanjing (Kang et al., 2013) and Hangzhou (Xiao et al., 2011b). For example, by investigating the event that happened in January 2007 in Shanghai using a medium-volume sampler, it was found that more secondary aerosols were generated during polluted days (Fu et al., 2008). In addition, field measurements using aerosol time-offlight mass spectrometry in the winter of 2008 in Shanghai indicated that carbonaceous particles constituted 87% of the total, and that the size distribution tended to be narrower with a higher peak (Yang et al., 2012). However, research on the differences of the various chemical components between clean days, haze-fog days and heavy haze-fog days at background stations in the YRD region is limited.

In the present reported study, continuous field measurements of the mass-size distributions of secondary aerosols of PM-! were carried out from 15 Nov. to 18 Dec. 2013 at Lin'an regional background station in the YRD region. The observed chemical components included inorganic species (i.e. sulfate, nitrate, ammonium and chloride) and organics (i.e. SOA and POA). The main objective of the work was to explore the reasons for the heavy pollution event through the evident variation of the chemical components and their size distributions on heavy haze-fog days, and consequently to better understand the formation mechanisms and main sources of heavy haze-fog pollution in eastern China.

2. Methods

2.1. Sampling site and observation methods

Lin'an regional background station [(30°18'N, 119°44'E); 138.6 m a.s.l] is located in the Henfan township of Lin'an in Zhejiang Province. The site is approximately 50 km away from Hangzhou, 200 km northeast of Shanghai, and 250 km northwest of Nanjing (Fig. 1a), which are the three largest cities in the YRD region. Furthermore, it is surrounded by hills, well covered by vegetation, and the nearby villages represent a negligible pollution source. The site has been demonstrated as ideal for the study of the atmospheric conditions in the YRD region (Zhang et al., 2012a). A quadrupole aerosol mass spectrometer (Q-AMS) was used for the field measurements, with a twin differential mobility particle sizer (TDMPS) and multi-angle absorption photometer (MAAP) in the same inlet system simultaneously. During the experiment, an automatic regenerating absorption aerosol dryer (Tuch et al.,

2009) was used to supply the sampled air for all the observational instruments in the laboratory, to keep the relative humidity (RH) below 30%. A PM10 cut-off head was adopted before the dryer to hold the coarse particles out of the inlet system. The temperature was controlled at about 25 °C in the laboratory.

The key features of the Q-AMS deployed in this study are an aerosol sampling chamber, aerodynamic particle sizing chamber, and particle composition detection chamber. Submicron aerosols are focused into a tight beam of 1 mm by six apertures in the sampling chamber, and then different velocities are achieved by the voltage difference between the sampling chamber and the sizing chamber. The aerodynamic diameter of particles is calculated from the time taken to travel across the sizing chamber. When particles reach the composition detection chamber, the volatile and semi-volatile components are vaporized by a heated surface (600 °C), and then ionized by electron impaction. The signals of the positive ion fragments enlarged by the electronic multiplier are detected by a quadrupole mass spectrometer so that the chemical compositions can be quantified. Detailed information on the Q-AMS can be found in Jayne et al. (2000). The Q-AMS can measure the mass concentrations and size distributions of volatile and semi-volatile components in particles with aerodynamic diameters of between 30 and 1000 nm, a range that includes organics, sulfate, nitrate, ammonium and chloride. The non-volatile components, such as crustal oxides and elemental carbon, cannot be distinguished. Therefore, the particles detected by the AMS should be non-refractory PM-! (NR-PM^. The time resolution of the AMS used in this study was 5 min.

The TDMPS was used here to calibrate the collection efficiency of the AMS (Zhang et al., 2011b). Two differential mobility analyzers and two condensation particle counters (TSI Model 3010 and 3025) made up the TDMPS system, which can characterize particles with a mobility diameter of between 3 and 800 nm. Meanwhile, an aerodynamic particle sizer (TSI Model 3321) was deployed to detect the size distributions of particles with mobility diameters ranging from 500 nm to 10 |am (Shen et al., 2011). Data were collected every 10 min. The mass concentrations of black carbon (BC) were measured using a MAAP (MAAP-5012) at a time resolution of 1 min. The wavelength employed in the MAAP was 670 nm, and the mass absorption efficiency was 6.6 m2 g-1. The data of the reactant gases with a time resolution of 1 min used in this study were all measured using automatic observation instruments, including a CO analyzer (Thermo Model 48C), SO2 analyzer trace level (Thermo Model 43CTL), NO-NO2-NOX analyzer tracer level (Thermo Model 42CTL), and O3 analyzer (Thermo Model 49C). All the meteorological data were collected hourly from automatic meteorological observations at the Lin'an background site.

2.2. Data quality control

A series of calibrations were conducted for the AMS during the experiments to ensure the quality of the data (e.g. electron multiplier calibration, ionization efficiency calibration, and size calibration). In addition, the possibility of the loss of particles in the AMS could not be ignored, and thus collection efficiency (CE) calibration also formed a key part of the quality control process. The CE of particles is impacted by many factors, such as the size of particles, the design of the lens, the RH, the acidity of the particles, and the chemical composition (Middlebrook et al., 2012). The CE was calibrated by comparing the mass concentrations observed by the AMS and TDMPS. To achieve this, the volume concentration measured by the TDMPS first needed to be calculated using the equations VD = nD3 / 6 x ND (where D is the mobility diameter and ND is the number concentration) and Vt = JVD dlogD. As the diameter observed by the AMS is the vacuum aerodynamic diameter, the mobility diameter (D) in these equations should be less than 600 nm. The volume concentrations were then interpolated to make the time series of the AMS and TDMPS consistent. Assuming the aerosols were made up of NH4NO3, (NH4)2SO4, NH4Cl, and organics, whose densities were 1.72, 1.8, 1.7, and 1.3 g cm-3, respectively

Fig. 1. (a) Back trajectories during the whole study period and mass concentrations and proportions of different chemical compositions in air masses from different clusters. (b) Back trajectories during different pollution periods.

(Zhang et al., 2011b), the density of aerosols every 5 min was calculated by the mass concentrations of chemical compositions observed by the AMS. Finally, the mass concentrations of the TDMPS were calculated by the volume concentrations and densities. The linear fitted curve between the mass concentrations of the AMS and TDMPS indicates that the CE was 0.69, with the slope being 0.998 and the correlation 0.98 (Fig. 2).

3. Results and discussion

3.1. Significant increase of nitrate, sulfate, ammonium and organics in PM1 during heavy haze-fog days

As shown in Fig. 3, the patterns of diurnal variation for RH and temperature were opposite, with averages of 56 ± 21% and 8.9 ± 5.1 °C,

0 50 100 150 200 250 300

TDMPSQlgfa|J)

Fig. 2. Collection efficiency calibration by comparing the mass concentrations observed by the AMS and TDMPS.

respectively. The average wind speed was 1.8 ± 1.0 m s- 1.The visibility during the whole period ranged from 0.045 to 31 km, with an average of 7 km. The mass concentration of PMi reached 298 |ag m-3 on 6 December 2013, and the average mass concentration for the whole period was only 63 ± 55 |ag m-3, which was 1.5 times higher than it was in the winter of 2010 in the YRD region (42 ^g m-3) (Huang et al., 2012). This

wintertime concentration in the YRD region was much higher than that reported in Tokyo (Japan; 14.7 |ag m-3; Takegawa et al., 2006), Gwangju (Korea; 12.8 |ag m-3; Park et al., 2012), and Central Europe (22-26 |ag m-3; Lanz et al., 2010), but lower than that reported in Beijing (73-89.3 |jg m-3; Zhang et al., 2014a, 2014b; Zhang et al., 2012b). The mean mass concentration of PM-! in the Pearl River Delta (PRD) region, another area of substantial economic development in China, has been reported to be about 30 |ag m-3 (Huang et al., 2011b; Xiao et al., 2011a). The above data suggest that, besides that in Beijing, the pollution in the YRD region is the most serious in China. In our observations, organics were always the major fraction (average concentration of 29 |ag m-3), followed by nitrate (15 |ag m-3), sulfate (10 |ag m-3), ammonium (7.7 |ag m-3), and chloride (1.1 |ag m-3). Some studies conducted in the cities during this pollution period have detected that the meteorological conditions and a reduced planetary boundary layer (PBL) contributed greatly to this event (Wang et al., 2015; Xu et al., 2015). Moreover, the significant changes in the chemical compositions of fine particles have not been discussed.

Three periods of pollution episodes were defined to compare the changes of the chemical components in PM^ The days from 25 to 30 Nov. were defined as clean days because of the high visibility (average of 14.7 km). The average wind speed and temperature during the clean days were 2.1 m s-1 and 7.5 °C, respectively, and the average RH was 44%. Following the standards of the China Meteorological Administration (CMA, 1979), the days from 21 to 23 Nov. and from 4 to

I laze-Fog Clean Haze-Fog

2013-11-16 2013-11-21 2013-11-26 2013-12-1 2013-12-6 2013-12-11 2013-12-16

Fig. 3. Time series of meteorological data [wind speed (WS), wind direction (WD), relative humidity (RH), temperature (T), visibility (VIS) ] and the mass concentration of non-refractory PM1 (NR-PM1 ) [organics (green line), nitrate (blue line), sulfate (red line), ammonium (yellow line), chloride (pink line) ].

Table 1

Average mass concentration values (units: |jg m-3) of non-refractory PM1 (NR-PM1), including its various chemical compositions, under different pollution conditions.

NR-PM1 Organics Sulfate Ammonium Nitrate Chloride no-/so4-

All 63 29 10 7.7 15 1.1 1.45

Clean 19 9.8 3.2 2.0 3.0 0.3 0.95

Haze-fog 105 48 16 13 26 1.8 1.64

Heavy haze-fog 212 86 35 29 58 4.0 1.64

14 Dec. could be defined as haze-fog days because the visibility was lower than 10 km. The average wind speed and temperature during the haze-fog days were 1.6 m s-1 and 9.4 °C, respectively, and the average RH was 60% (maximum of 92%). Moreover, during 16:00 6 Dec. to 11:00 8 Dec. the visibility was lower than 2 km (average of 0.6 km), and so these days were defined as heavy haze-fog days. The average temperature on these days was 6.9 °C and the average RH reached 81%.

As Table 1 shows, the average mass concentration of PM-i on haze-fog days was 105 |ag m- 3, which was six times higher than on clean days (19 |ag m-3). Meanwhile, the mass concentration on heavy haze-fog days was 212 |ag m-3, which was more than 10 times higher than on clean days, indicating a sharp increase in particles. Moreover, the chemical components all increased significantly during haze-fog days. The mass concentration of organics, which accounted for 47% of PM1, reached 86 |ag m-3 on heavy haze-fog days, which was about nine times higher than on clean days. The mass concentrations of secondary inorganics on heavy haze-fog days also increased obviously. For example, the mass concentrations of sulfate, nitrate and ammonium on heavy haze-fog days were 35,58 and 29 |ag m-3, respectively, which were all more than 10 times higher than those on clean days. By contrast, research on the haze event observed in 2011 has shown that organic matter, sulfate, nitrate and ammonium increased by about 1 -4 times in most of the cities in the YRD (Hua et al., 2015). Among all chemical compositions, nitrate increased most significantly during our measurements,, being about 20 times higher than on clean days. In comparison, the sulfate increased as significantly as nitrate during the heavy haze event that occurred in January 2013 in the YRD (Wang et al., 2014a, 2014b).

3.2. The role of increased emissions from coal combustion, short-distance transport of mixed pollutants, and relatively stable and humid meteorological conditions in the significant increase of inorganic chemical components during heavy haze-fog days

The Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT) was used to explore the air mass transport paths under different conditions. GDAS (Global Data Assimilation System) data were used as the meteorological data for HYSPLIT, which was downloaded from ftp://arlftp.arlhq.noaa.gov/pub/archives/gdas1/. A total of five clusters of 72 h back trajectories during the whole observation period were produced (Fig. 1a), all within the boundary layer (under 1 km). Clusters 1 to 5 accounted for 40%, 29%, 12%, 10% and 9% of the total trajectories, and the corresponding mass concentrations of PM1 were 58, 78,32,118 and 28 |ag m-3, respectively. The concentration of PM1 associated with cluster 4, which passed over urban areas around Hangzhou Bay, such as Shanghai, Jiaxing, Ningbo and Shaoxing, was the highest. Previous research has shown that the air masses transported from these areas around Hangzhou Bay represent one of the most important transport paths for atmospheric pollution in the YRD region (Huang et al., 2011a). Abundant water vapor in the air masses associated with cluster 4 would have promoted aqueous reactions. Nitrate in the cluster-4 air masses contributed 25% of the PM1, which was similar to the proportion of nitrate on haze-fog days. This indicated that cluster 4 was the dominant contributor to the increase of nitrate during the haze-fog days. The high concentrations of PM1 caused by another two short-range air masses (cluster 2 and cluster 1) also brought serious pollution to the YRD region, demonstrating the important role of the

short-distance transport of mixed pollutants in the occurrence of the heavy haze-fog days.

Comparing the air masses under different pollution conditions, there was no long-range transport of air masses observed during heavy haze-fog days, and the characteristics of aerosols at Lin'an were only influenced by local emissions in the YRD region (Fig. 1b). Few air masses from cluster 4 appeared during clean days. The clean days were mainly impacted by long-range-transported air masses from the northwest with cold air. Among the long-range clusters, the air masses from cluster 3 mainly appeared on the clean days, and had little influence on haze-fog days. The air masses from cluster 5 that appeared from 16 Dec. to 18 Dec. brought sufficient water vapor to promote precipitation, since they passed over the ocean.

Organics were always the largest component of PM1 in the air masses from different directions, followed by nitrate, sulfate, ammonium, and chloride, although some differences in proportions were apparent (Fig. 1a). The mass concentrations of chloride were higher in clusters 4 and 5 as a result of their passing across the ocean. The transport distances of clusters 3 and 5 were relatively long. The proportions of organics in the air masses from clusters 3 and 5 were 42% and 43%, respectively, which were lower than those from other clusters. However, the mass concentrations of sulfate and nitrate were relatively higher. In contrast, the transport distances of clusters 1, 2 and 4 were relatively short, and or-ganics accounted for 47%, 48% and 47% of the PM1, respectively.

The changes of nitrate under different pollution conditions (clean days, haze-fog days and heavy haze-fog days) indicated that the mass concentration and the proportion of nitrate increased when the pollution became heavier (Table 1 ). The concentrations of NOx, the main precursor of nitrate, were also continually measured at Lin'an during the observation period. The average mixing ratio of NOx on haze-fog days was 30 ppb, which was approximately three times higher than that on clean days (11 ppb). Peak concentration occurred on heavy haze-fog days, reaching 41 ppb. The increase of NOx nearby led to more nitrate in pollutants, which was transported by short-range air masses (discussed above). According to an analysis of an emissions inventory (Cao et al., 2011 ), the constituents of NOx from the surrounding regions of Lin'an, where the short-range air masses (clusters 1,2 and 4) passed, are about 69%, 62% and 67% contributed by coal combustion, respectively, while only 21%, 34% and 29% are from motor vehicle sources. Therefore, the significant increases of NOx and nitrate transported by the short-range air masses during the heavy haze-fog days can be mainly attributed to increased coal combustion and pollutant transport to Lin'an. In regions such as Anhui Province and the city of Shangqiu in He'nan Province, over which the air masses from cluster 3 passed, only 13% of the NOx was emitted from motor vehicles. Based on another Chinese emissions inventory, for 2010, developed by Tsinghua University (the Multi-resolution Emissions Inventory for China, MEIC) through http://www.meicmodel.org., the ratios of NOx/SO2 from stationary and mobile sources were 0.98 and 28.68, respectively. The ratio of NO-/ SO|- was 0.95 on clean days and 1.64 on heavy haze-fog days, both of which were close to the ratio of NOx/SO2 emitted from coal combustion. NOx and SO2 are the major precursors of the secondary inorganic components NO3- and SO42-. Therefore our results suggest that coal combustion was the main source of the increased NOx and nitrate during the heavy haze-fog days of this study period. In addition, the stability of atmospheric stratification (Zhang et al., 2013) and drop in the height of the boundary layer (Wang et al., 2014a, 2014b; Gao et al.,

2015; Zhang et al., 2015) during heavy haze-fog days lead to pollutants mixing in a smaller space, and thus enhanced mass concentrations. Moreover, the studies on this pollution event have proved that the stable meteorological conditions and reduced PBL contributed to this haze-fog event (Wang et al., 2015; Xu et al., 2015).

The RH measured on haze-fog days, especially heavy haze-fog days, was much higher than that on clean days. The average RH on heavy haze-fog days was 81%, with a maximum of 92% (Fig. 3). Higher RH may enhance NH4NO3 generation via aqueous-phase processes (Li et al., 2014). A previous study showed that the rate of oxidation from SO2 to SO4- on the surface of NaCl particles was much higher when the ambient RH ranged from 75% to 80%, rather than below 75% (Tursic et al., 2004). The mass concentration of sulfate on heavy haze-fog days was observed as 11 times higher than on clean days (Table 1), which agrees with previous findings that the transformation rates of NOx under wet conditions was six times higher than

under dry conditions (Li et al., 2010). Previous research has also shown that the heterogeneous reactions under high RH played an important role in the production of secondary aerosols (Wang et al., 2014c, 2014d; Zheng et al., 2014). Therefore, we propose that high RH also contributes to rapid increases of secondary inorganics and heavy pollution events.

The correlation coefficient on haze-fog days between ammonium and sulfate was 0.93, while that between ammonium and nitrate was 0.97. The combination of NH+ and SO4- or NO- could explain the sharp increase of NH+ on heavy haze-fog days, as well as SO4- and NO-. However, chloride remained at a low proportion, with little variation under different pollution conditions. The fraction of chloride in PMi in the YRD region was only half of that in Beijing (4%) (Zhang et al., 2014a, 2014b), suggesting that coal combustion as a source of chloride contributes less to haze-fog events in winter in the YRD region than in Beijing.

Fig. 4. The (a) mass spectra and (b) time series of the four components [hydrocarbon-like OA (HOA), biomass-burningOA (BBOA), low-volatility oxidized OA (LV-OOA), and semi-volatility oxidized OA (SV-OOA)] and their tracers.

Besides explanations above, the atmospheric oxidizing capacity and other chemical mechanisms promoting the sharp increases of secondary inorganics and secondary organics during haze-fog and heavy haze-fog days will be discussed in the next chapter by analysis of SOA constituents.

3.3. Enhanced atmospheric oxidizing capacity caused by increased biomass burning: another possible cause of the significant increase in OA and other secondary aerosols on heavy haze-fog days

OA was the most important contributor to PM1, with a fraction of 46% during this field campaign. Positive Matrix Factorization (PMF) was applied to analyze the constitution of OA with the data obtained by the Q-AMS. PMF has been widely used in the atmospheric sciences, the detail of which is described elsewhere in the literature (Ulbrich et al., 2009; Zhang et al., 2011a). Four components - hydrocarbon-like OA (HOA), biomass-burning OA (BBOA), low-volatility oxidized OA (LV-OOA), and semi-volatility oxidized OA (SV-OOA) - were identified by their m/z contributions and their correlations with tracers (Fig. 4). The HOA was dominated by peaks of m/z27(C2H+), m/ z41(C3H+), m/z43(C3H+), m/z55(C4H+), m/z57(C4H+), m/z69(C5H+) and m/z71(C5Hri). The BBOA was characterized by masses typical of biomass-burning spectra, such as m/z29(CHO+), m/z60(C2H2O+) and m/z73(C3H5O+). LV-OOA and SV-OOA were all highly oxygenated and identified by the peak of the m/z44(CO+) signal, which is a tracer of deep oxidized organic aerosols. However, the degree of oxidation of LV-OOA is higher than SV-OOA, so it corresponds to a higher peak in m/z44 (Crippa et al., 2013). The temporal correlations of different components with their tracers are presented in Fig. 4b. The time series of LV-OOA showed a similar trend to sulfate, which was considered as a tracer of LV-OOA. The correlation coefficient between them (R = 0.92) was higher than that found in most studies (Huang et al., 2011b; Lanz et al., 2010). The SV-OOA component, which is less-oxygenated than LV-OOA, was correlated well with nitrate (R = 0.77), and a similar trend was also found between HOA and NOx (R = 0.79). BBOA was well correlated with BC (R = 0.80) and CO (R = 0.75).

During this study, LV-OOA was the dominant component (average of 13 |g m-3, which contributed 44% to OA), followed by BBOA (25%), HOA (22%) and SV-OOA (9.3%). The average concentrations of BBOA, HOA and SV-OOA were 7.0, 6.4, and 2.7 |g m-3, respectively. Among the components, POA, which comprised HOA and BBOA, accounted for 47% of the OA. By contrast, LV-OOA and SV-OOA, defined as SOA, shared 53% of OA. This is close to the ratio of SOC to OC at Lin'an in November (0.53) and December (0.47) (Zhang et al., 2012a).

As the results in Table 2 show, LV-OOA and BBOA increased relatively more sharply than the other two components during the haze-fog period. The mass concentration of LV-OOA on heavy haze-fog days (37 |g m-3) was about nine times higher than on clean days (4.2 |g m-3), demonstrating an enhanced oxidation capacity of the atmosphere. The concentration of BBOA increased to 22 |g m-3, which was 11 times that on clean days (2.0 |g m-3). This sharp increase showed that biomass burning was one of the most important contributors to enhanced oxidation during the haze-fog and heavy haze-fog days. Generally, the oxidation capacity of the atmosphere caused by peroxides on clean days should be larger than that on

heavy haze days in China (Zhang et al., 2010). However, abundant precursors of tropospheric O3, such as CO, CH4 and NOx, produced by biomass burning, could generate more O3 in the ambient air (Logan et al., 1981; Mauzerall et al., 1998). In our measurement period, the average peak mixing ratio of O3 during haze-fog days was 57.9 ppb, which was 34% higher than that on clean days (43.3 ppb). The increase of O3, which was caused by biomass burning, possibly would have enhanced the oxidation capacity of the atmosphere during the heavy haze-fog days, thus producing more SOA with low volatility. Indeed, significant increases of LV-OOA on haze-fog days have been observed in some other studies (Sun et al., 2013; Zhang et al., 2014a, 2014b). Therefore, the enhanced atmospheric oxidizing capacity was considered as one of the reasons for the significant increase of LV-OOA and secondary inorganics on the heavy haze-fog days. In fact, there is a mass of residues burned during the post-harvest season in the YRD region (Cheng et al., 2014). The seasonal variation of chemical compositions in aerosols at Lin'an have also been shown to feature a sharp increase in organic compounds and EC after the harvest season in autumn, as well as SOA (Zhang et al., 2012a). The BBOA measured in the winter of 2010 in the YRD region, accounting for 30% of OA, has also been reported as important (Huang et al., 2012). According to the fire point data of NASA during the observation period (https://firms.modaps.eosdis.nasa.gov/ firemap/), there were a lot of fires in Anhui, Zhejiang Province, and the regions around Hangzhou Bay. The analysis of the BBOA from different clusters also showed higher concentrations of BBOA in the air masses from clusters 1, 2 and 4 (Fig. 1a), consistent with the fire point positions. Previous studies have also shown that the pollutants from biomass burning in the YRD region mainly came from local emissions and short-range transport (Cheng et al., 2014).

Zhang et al. (2014b) recently reported a serious underestimation of SOA using a smog chamber approach. The formation rate of SOA from the gaseous precursors represented by toluene will be two to four times higher with an increase in the concentration of seed particles or a change in the seed surface area (Wang et al., 2014c, 2014d). The uptake coefficient of H2O2 can reach 10-5to10-4 when abundant mineral aerosols exist, which is close to the reaction rate of atmospheric radicals (Chen et al., 2008). The mass concentration of PM1 during this study was continually high. A mass of mineral aerosols (22 |g m-3) inJanuary at Lin'an has been observed (Zhang et al., 2012a), among which Fe and Mn were the catalyzers for the production of SOA (Huang et al., 2014b). This may enhance the gas-particle partitioning process and the productivity of SOA, but this idea needs to be further explored.

3.4. Mass-size distributions during clean and heavy haze-fog days

The lognormal distribution was used to fit the mass-size distributions of chemical compositions under different pollution conditions (Fig. 5). The peak sizes and the geometric standard deviations (og) of the fitted distribution are listed in Table 3. The fitting parameter (og) can to some extent indicate the impact of small particles. When og < 2 it is generally the case that most of the aerosols are in accumulation mode and can be mono-mode fitted, while og > 2 may indicate that particles within the fine mode cannot be ignored (Zhang et al., 2014a, 2014b). All these mass-size distributions for the different chemical

Mass concentrations of the four components [hydrocarbon-like OA (HOA), biomass-burning OA (BBOA), low-volatility oxidized OA (LV-OOA), and semi-volatility oxidized OA (SV-OOA) ] in different pollution periods.

Haze-fog

Heavy haze-fog

Mass concentration (|og m 3)

Percent

Mass concentration (|g m 3)

Percent

Mass concentration (|g m 3)

Percent

LV-OOA SV-OOA HOA BBOA

4.2 1.5 2.8 2.0

40% 14% 27% 19%

21 4.3 9.8 12

45% 9.0% 21% 26%

7.2 14

46% 9.1% 17%

Ammonium Lognormal Ammonium -Nitrate ---- Lognormal Nitrate

Sulphate ----Loguormal Sulphate -Orgailies Lognormal Organic«

Fig. 5. Mass-size distributions of organics, nitrate, sulfate and ammonium under different pollution conditions.

components are valuable for a better understanding of the formation mechanism, and can serve as inputs for improving numerical simulations of chemical species of PM^

During all measuring periods, the mass concentrations of particles with sizes between 100 and 1000 nm occupied more than 95% of those particles with sizes between 30 and 1000 nm. Almost all particles were in the accumulation mode; the mode value of sulfate was found to be larger than that of all other chemical components and the peak sizes of organics were the smallest. Of the chemical components, the organics had the largest curve widths. The ag of OA on clean days was 2.01 (Table 2), which implied the impact of small particles. Photochemical reactions of VOCs emitted from surrounding vegetation were one of the possible sources of these fine particles on clean days.

On haze-fog days, the peak sizes of the different chemical components ranged from 601 nm to 687 nm, and the sizes of aerosols increased to 748-839 nm when heavy haze-fog occurred, which was about double that on clean days (378-425 nm). A previous study also showed that the aerosol optical depth at 440 nm rapidly increased when the median effective diameter of fine-mode particles increased from 300 nm to a diameter of 500-600 nm, which would markedly affect the visibility (Bi et al., 2014). The mass-size distributions and the lognormal fitting curves in different pollution conditions (Fig. 5) indicated larger peak sizes and narrower curve widths were associated with the pollution becoming heavier. Previous research has shown that the growth of particles during polluted days plays an important role in the formation of haze event (Guo et al., 2014), and secondary aerosol formation is the main driver of severe haze pollution (Huang et al., 2014a). The formation of secondary aerosols probably contributed a lot to this haze-fog event.

4. Conclusion

The chemical and physical properties of submicron aerosols at Lin'an background site in the YRD region were measured using a Q-AMS from 15 Nov. to 18 Dec. 2013. During the whole study period, the mean concentration ofPM1 was 63 |ag m-3, with organics often being the largest

Table 3

Characteristic values of the mass-size distributions of chemical compositions in different pollution periods.

Clean Haze-fog Heavy haze-fog

Peak % Peak % Peak %

Organics 378 2.01 96.8% 601 1.86 99.5% 748 1.68 98.9%

Sulfate 425 1.95 98.2% 687 1.82 99.2% 839 1.63 98.3%

Nitrate 401 1.91 99.0% 666 1.81 99.6% 814 1.65 99.2%

Ammonium 401 1.91 95.7% 664 1.81 98.8% 802 1.63 97.7%

contributor, accounting for ~46% of PM-i, followed by nitrate (~23%), sulfate (~16%), ammonium (~12%), and chloride (~1.8%).

A heavy haze-fog event occurred during this period and the peak mass concentration of PM1 reached 298 |ag m-3, coinciding with significant increases in organics, accounting for 47% of PM1 and reaching 86 |ag m-3, which was about nine times higher than on clean days. The mass concentrations of secondary inorganics on heavy haze-fog days also increased obviously. The mass concentrations of sulfate, nitrate and ammonium on heavy haze-fog days were 35, 58 and 29 |ag m-3, respectively, which were all more than 10 times higher than those on clean days. Among all the chemical compositions, nitrate increased most significantly, which was about 20 times higher than on clean days.

The sharp increase in nitrate was mainly attributable to the increased coal-combustion emissions during the winter season in the surrounding areas of the site. This was also supported by the fact that the SV-OOA, which is less oxygenated than LV-OOA, was highly correlated with nitrate. This was also true for the high sulfate during the heavy haze-fog event; an enhanced atmospheric oxidizing capacity, probably caused by increased biomass burning, was the major reason for the significant increase of OA, because the LV-OOA was the dominant component (average of 13 |ag m-3), contributing 44% to OA, followed by BBOA (25%), HOA (22%) and SV-OOA (9.3%), during the study period. The enhanced atmospheric oxidizing capacity could of course have increased other secondary aerosols during the heavy haze-fog days.

From the fitted mass-size distributions, the organics were found to have the largest width of the log-normal distribution. The ag of OA on clean days was 2.01 according to the statistics, which implied an impact of fine particles. Photochemical reactions of VOCs emitted from surrounding vegetation were one of the possible sources of these fine particles on clean days. As the pollution became heavier, larger peak sizes and narrower widths of the fitted distribution were found for aged aerosols.

The short-distance transport of mixed precursor gases and aerosols, low wind, and more humid weather conditions were also highly responsible for the formation of the heavy haze-fog event in this study. The air masses transported from these areas, around Hangzhou Bay, represent one of the most polluted transport paths for atmospheric pollution arriving at the Lin'an site. Consequently, the possibility of enhanced aqueous reactions, as well as the gas-to-particle partitioning process through heterogeneous reactions, are also discussed, for further interpreting the increase of PM1 under these conditions.

Acknowledgments

This research was supported by grants from National Key Project of Basic Research (2011CB403401) and the Specific Team Fund of the Jiangsu Collaborative Innovation Center for Climate Change.

References

Bi, J., Huang, J., Hu, Z., Holben, B., Guo, Z., 2014. Investigating the aerosol optical and radiative characteristics of heavy haze episodes in Beijing during January of 2013. J. Geophys. Res. Atmos. 119, 9884-9900.

Cao, G., Zhang, X., Gong, S., An, X., Wang, Y., 2011. Emission inventories of primary particles and pollutant gases for China. Chin. Sci. Bull. 56, 781 -788.

Chen, Z., Wang, H., Zhu, L., Wang, C., Jie, C., Hua, W., 2008. Aqueous-phase ozonolysis of methacrolein and methyl vinyl ketone: a potentially important source of atmospheric aqueous oxidants. Atmos. Chem. Phys. 8, 2255-2265.

Cheng, Z., Wang, S., Fu, X., Watson, J.G., Jiang, J., Fu, Q., et al., 2014. Impact of biomass burning on haze pollution in the Yangtze River delta, China: a case study in summer 2011. Atmos. Chem. Phys. 14,4573-4585.

CMA, 1979. The Ground Meteorological Observation Specification. Meteorological Press, Beijing.

Crippa, M., DeCarlo, P.F., Slowilk J.G., Mohr, C., Heringa, M.F., Chirico, R., et al., 2013. Wintertime aerosol chemical composition and source apportionment of the organic fraction in the metropolitan area of Paris. Atmos. Chem. Phys. 13,961-981.

Fu, Q., Zhuang, G., Wang, J., Xu, C., Huang, K., Li, J., et al., 2008. Mechanism of formation of the heaviest pollution episode ever recorded in the Yangtze River Delta, China. Atmos. Environ. 42, 2023-2036.

Gao, Y., Zhang, M., Liu, Z., Wang, L., Wang, P., Xia, X., et al., 2015. Modeling the feedback between aerosol and meteorological variables in the atmospheric boundary layer during a severe fog-haze event over the North China Plain. Atmos. Chem. Phys. 15, 4279-4295.

Guo, S., Hu, M., Zamora, M.L., Peng, J., Shang, D., Zheng, J., et al., 2014. Elucidating severe urban haze formation in China. Proc. Natl. Acad. Sci. 111, 17373-17378.

Heald, C., Coe, H., Jimenez, J., Weber, R., Bahreini, R., Middlebroolk A., et al., 2011. Exploring the vertical profile of atmospheric organic aerosol: comparing 17 aircraft field campaigns with a global model. Atmos. Chem. Phys. 11,12673-12696.

Hoyle, C.R., Boy, M., Donahue, N.M., Fry, J.L., Glasius, M., Guenther, A., et al., 2011. A review of the anthropogenic influence on biogenic secondary organic aerosol. Atmos. Chem. Phys. 11,321-343.

Hua, Y., Cheng, Z., Wang, S., Jiang, J., Chen, D., Cai, S., et al., 2015. Characteristics and source apportionment of PM 2.5 during a fall heavy haze episode in the Yangtze River Delta of China. Atmos. Environ. (in press).

Huang, C., Chen, C.H., Li, L., Cheng, Z., Wang, H.L., Huang, H.Y., et al., 2011a. Emission inventory of anthropogenic air pollutants and VOC species in the Yangtze River Delta region, China. Atmos. Chem. Phys. 11,4105-4120.

Huang, X.F., He, L.Y., Hu, M., Canagaratna, M.R., Kroll, J.H., Ng, N.L., et al., 2011b. Characterization of submicron aerosols at a rural site in Pearl River Delta of China using an aerodyne high-resolution aerosol mass spectrometer. Atmos. Chem. Phys. 11, 1865-1877.

Huang, X.-F., Xue, L., Tian, X.-D., Shao, W.-W., Sun, T.-L., Gong, Z.-H., et al., 2012. Highly time-resolved carbonaceous aerosol characterization in Yangtze River Delta of China: composition, mixing state and secondary formation. Atmos. Environ. 200-207.

Huang, R.-J., Zhang, Y., Bozzetti, C., Ho, K.-F., Cao, J.-J., Han, Y., et al., 2014a. High secondary aerosol contribution to particulate pollution during haze events in China. Nature 514, 218-222.

Huang, X., Song, Y., Zhao, C., Li, M., Zhu, T., Zhang, Q., et al., 2014b. Pathways of sulfate enhancement by natural and anthropogenic mineral aerosols in China. J. Geophys. Res. Atmos. 119,14165.

Jayne, J.T., Leard, D.C., Zhang, X., Davidovits, P., Smith, KA, Kolb, C.E., et al., 2000. Development of an aerosol mass spectrometer for size and composition analysis of submicron particles. Aerosol Sci. Technol. 33, 49-70.

Kang, H., Zhu, B., Su, J., Wang, H., Zhang, Q., Wang, F., 2013. Analysis of a long-lasting haze episode in Nanjing, China. Atmos. Res. 120-121, 78-87.

Lanz, VA, Prévôt, A.S.H., Alfarra, M.R., Weimer, S., Mohr, C., DeCarlo, P.F., et al., 2010. Characterization of aerosol chemical composition with aerosol mass spectrometry in Central Europe: an overview. Atmos. Chem. Phys. 10,10453-10471.

Li, H., Zhu, T., Zhao, D., Zhang, Z., Chen, Z., 2010. Kinetics and mechanisms of heterogeneous reaction of NO2 on CaCO3 surfaces under dry and wet conditions. Atmos. Chem. Phys. 10,463-474.

Li, Y., Schwandner, F.M., Sewell, H.J., Zivkovich, A., Tigges, M., Raja, S., et al., 2014. Observations of ammonia, nitric acid, and fine particles in a rural gas production region. Atmos. Environ. 83, 80-89.

Logan, JA, Prather, M.J., Wofsy, S.C., McElroy, M.B., 1981. Tropospheric chemistry: a global perspective. J. Geophys. Res. 86, 7210.

Mauzerall, D.L., Logan, J.A., Jacob, D.J., Anderson, B.E., Blake, D.R., Bradshaw, J.D., et al., 1998. Photochemistry in biomass burning plumes and implications for tropospheric ozone over the tropical South Atlantic. J. Geophys. Res. 103,8401.

Middlebrook, A.M., Bahreini, R., Jimenez, J.L., Canagaratna, M.R., 2012. Evaluation of composition-dependent collection efficiencies for the aerodyne aerosol mass spectrometer using field data. Aerosol Sci. Technol. 46, 258-271.

Park, K., Park, J., Lee, S., H-j, Cho, Kang, M., 2012. Real time measurement of chemical composition of submicrometer aerosols at urban Gwangju in Korea by aerosol mass spectrometer. Atmos. Environ. 62, 281 -290.

Pye, H., Liao, H., Wu, S., Mickley, L., Jacob, D., Henze, D., et al., 2009. Effect of changes in climate and emissions on future sulfate-nitrate-ammonium aerosol levels in the United States. J. Geophys. Res. 114, D01205.

Shen, X.J., Sun, J.Y., Zhang, Y.M., Wehner, B., Nowak, A., Tuch, T., et al., 2011. First long-term study of particle number size distributions and new particle formation events of regional aerosol in the North China Plain. Atmos. Chem. Phys. 11,1565-1580.

Sun, Y.L., Wang, Z.F., Fu, P.Q., Yang, T., Jiang, Q., Dong, H.B., et al., 2013. Aerosol composition, sources and processes during wintertime in Beijing, China. Atmos. Chem. Phys. 13, 4577-4592.

Takegawa, N., Miyakawa, T., Kondo, Y., Jimenez, J.L., Zhang, Q., Worsnop, D.R., et al., 2006. Seasonal and diurnal variations of submicron organic aerosol in Tokyo observed using the aerodyne aerosol mass spectrometer. J. Geophys. Res. 111.

Tuch, T., Haudek, A., Müller, T., Nowak A., Wex, H., Wiedensohler, A., 2009. Design and performance of an automatic regenerating adsorption aerosol dryer for continuous operation at monitoring sites. Atmos. Meas. Tech. 2,417-422.

Tursic, J., Berner, A., Podkrajsek B., Grgic, I., 2004. Influence of ammonia on sulfate formation under haze conditions. Atmos. Environ. 38, 2789-2795.

Ulbrich, I., Canagaratna, M., Zhang, Q., Worsnop, D., Jimenez, J., 2009. Interpretation of organic components from Positive Matrix Factorization of aerosol mass spectrometric data Atmos. Chem. Phys. 9,2891-2918.

Wang, H., An, J., Shen, L., Zhu, B., Pan, C., Liu, Z., et al., 2014a. Mechanism for the formation and microphysical characteristics of submicron aerosol during heavy haze pollution episode in the Yangtze River Delta, China. Sci. Total Environ. 490, 501 -508.

Wang, H., Shi, G.Y., Zhang, X.Y., Gong, S.L., Tan, S.C., Chen, B., et al., 2014b. Mesoscale modeling study of the interactions between aerosols and PBL meteorology during a haze episode in China Jing-Jin-Ji and its near surrounding region &ndash; Part 2: aerosols' radiative feedback effects. Atmos. Chem. Phys. Discuss. 14, 28269-28298.

Wang, Y., Li, L., Chen, C., Huang, C., Huang, H., Feng, J., et al., 2014c. Source apportionment of fine particulate matter during autumn haze episodes in Shanghai, China. J. Geophys. Res. Atmos. 119,1903-1914.

Wang, Y., Zhang, Q., Jiang, J., Zhou, W., Wang, B., He, K., et al., 2014d. Enhanced sulfate formation during China's severe winter haze episode in January 2013 missing from current models. J. Geophys. Res. Atmos. 119,10,425-10,440.

Wang, Z., Li, J., Wang, Z., Yang, W., Tang, X., Ge, B., et al., 2014e. Modeling study of regional severe hazes over mid-eastern China in January 2013 and its implications on pollution prevention and control. Sci. China Earth Sci. 57, 3-13.

Wang, M., Cao, C., Li, G., Singh, R.P., 2015. Analysis of a severe prolonged regional haze episode in the Yangtze River Delta, China. Atmos. Environ. 102,112-121.

Wu, L.Y., Tong, S.R., Wang, W.G., Ge, M.F., 2011. Effects of temperature on the heterogeneous oxidation of sulfur dioxide by ozone on calcium carbonate. Atmos. Chem. Phys. 11, 6593-6605.

Xiao, R., Takegawa, N., Zheng, M., Kondo, Y., Miyazaki, Y., Miyakawa, T., et al., 2011a. Characterization and source apportionment of submicron aerosol with aerosol mass spectrometer during the PRIDE-PRD 2006 campaign. Atmos. Chem. Phys. 11,6911-6929.

Xiao, Z.-m., Zhang, Y.-f., Hong, S.-m., X-h, Bi, Jiao, L., Feng, Y.-c., et al., 2011b. Estimation of the main factors influencing haze, based on a long-term monitoring campaign in Hangzhou, China. Aerosol Air Qual. Res. 11, 873-882.

Xu, J., Yan, F., Xie, Y., Wang, F., Wu, J., Fu, Q., 2015. Impact of meteorological conditions on a nine-day particulate matter pollution event observed in December 2013, Shanghai, China. Particuology 20, 69-79.

Yang, F., Chen, H., Du, J, Yang, X., Gao, S., Chen, J., et al., 2012. Evolution of the mixing state of fine aerosols during haze events in Shanghai. Atmos. Res. 104-105,193-201.

Zhang, X., Chen, Z., He, S., Hua, W., Zhao, Y., Li, J., 2010. Peroxyacetic acid in urban and rural atmosphere: concentration, feedback on PAN-NO x cycle and implication on radical chemistry. Atmos. Chem. Phys. 10, 737-748.

Zhang, Q., Jimenez, J.L., Canagaratna, M.R., Ulbrich, I.M., Ng, N.L., Worsnop, D.R., et al., 2011a. Understanding atmospheric organic aerosols via factor analysis of aerosol mass spectrometry: a review. Anal. Bioanal. Chem. 401, 3045-3067.

Zhang, Y.M., Zhang, X.Y., Sun, J.Y., Lin, W.L., Gong, S.L., Shen, X.J., et al., 2011b. Characterization of new particle and secondary aerosol formation during summertime in Beijing, China. Tellus B 63, 382-394.

Zhang, X., Wang, Y., Niu, T., Zhang, X., Gong, S., Zhang, Y., et al., 2012a. Atmospheric aerosol compositions in China: spatial/temporal variability, chemical signature, regional haze distribution and comparisons with global aerosols. Atmos. Chem. Phys. 12, 779-799.

Zhang, Y., Sun, J., Zhang, X., Shen, X., Wang, T., Qin, M., 2012b. Seasonal characterization of components and size distributions for submicron aerosols in Beijing. Sci. China Earth Sci. 56, 890-900.

Zhang, X., Sun, J., Wang, Y., Li, W., Zhang, Q., Wang, W., etal., 2013. Factors contributing to haze and fog in China. Chin. Sci. Bull. Chin. Version 58,1178-1187.

Zhang, J.K., Sun, Y., Liu, Z.R., Ji, D.S., Hu, B., Liu, Q., etal., 2014a. Characterization of submicron aerosols during a month of serious pollution in Beijing, 2013. Atmos. Chem. Phys. 14, 2887-2903.

Zhang, Y.M., Zhang, X.Y., Sun, J.Y., Hu, G.Y., Shen, X.J., Wang, Y.Q., et al., 2014b. Chemical composition and mass size distribution of PM1 at an elevated site in central east China. Atmos. Chem. Phys. 14,12237-12249.

Zhang, L., Wang, T., Lv, M., Zhang, Q., 2015. On the severe haze in Beijing during January 2013: unraveling the effects of meteorological anomalies with WRF-Chem. Atmos. Environ. 104,11-21.

Zheng, B., Zhang, Q., Zhang, Y., He, K.B., Wang, K., Zheng, G.J., et al., 2014. Heterogeneous chemistry: a mechanism missing in current models to explain secondary inorganic aerosol formation during the January 2013 haze episode in North China. Atmos. Chem. Phys. 14,16731-16776.