Nitrate-driven haze pollution during summertime over the North China 1 Plain 2

Compared to the severe winter haze episodes in the North China Plain (NCP), haze pollution during summertime has 14 drawn little public attention. In this study, we present the highly time-resolved chemical composition of submicron particles (PM1) 15 measured in Beijing and Xinxiang in the NCP region during summertime to evaluate the driving factors of aerosol pollution. 16 During the campaign periods (30 June to 27 July, 2015, for Beijing and 8 to 25 June, 2017, for Xinxiang), the average PM1 17 concentrations were 35.0 μg m and 64.2 μg m in Beijing and Xinxiang, respectively. Pollution episodes characterized with 18 largely enhanced nitrate concentrations were observed at both sites. In contrast to the slightly decreased mass fractions of sulfate, 19 semi-volatile oxygenated organic aerosol (SV-OOA), and low-volatile oxygenated organic aerosol (LV-OOA) in PM1, nitrate 20 displayed an almost linearly increased contribution with the aggravation of aerosol pollution in both Beijing and Xinxiang, 21 highlighting the importance of nitrate formation as the driving force of haze evolution in summer. Rapid nitrate production mainly 22 occurred after midnight, with a higher formation rate than that of sulfate, SV-OOA, or LV-OOA. Detailed investigation of nitrate 23 behaviors revealed several factors influencing the rapid nitrate formation in summer: high ammonia emissions in the NCP region, 24 the gas-to-particle equilibrium of ammonium nitrate closely related to variations in temperature and relative humidity, nighttime 25 nitrate production through heterogeneous hydrolysis of dinitrogen pentoxide (N2O5), and regional transport from different air mass 26 origins. Finally, atmospheric particulate nitrate data acquired by mass spectrometric techniques from various field campaigns in 27 Asia, Europe, and North America uncovered a higher concentration and higher fraction of nitrate present in China. Although 28 measurements in Beijing during different years demonstrate a decline in the nitrate concentration in recent years, the nitrate 29 contribution in PM1 still remains high. To effectively alleviate particulate matter pollution in summer, our results call for the urgent 30 need to initiate ammonia emission control measures and further reduce nitrogen oxide emissions over the NCP region. 31


Introduction
Atmospheric aerosol particles are known to significantly impact visibility (Watson, 2002) and human health (Pope et al., 2009;Cohen et al., 2017), as well as affect climate change by directly and indirectly altering the radiative balance of Earth's atmosphere (IPCC, 2007).The effects of aerosols are intrinsically linked to the chemical composition of particles, which are usually dominated by organics and secondary inorganic aerosols (i.e., sulfate, nitrate, and ammonium) (Jimenez et al., 2009).
In recent years, severe haze pollution has repeatedly struck the North China Plain (NCP), and its effects on human health have drawn increasing public attention.Correspondingly, the chemical composition, sources, and evolution processes of particulate Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.matter (PM) have been thoroughly investigated (Huang et al., 2014;Guo et al., 2014;Cheng et al., 2016;Li et al., 2017a), mostly during extreme pollution episodes in winter.Unfavorable meteorological conditions, intense primary emissions from coal combustion and biomass burning, and fast production of sulfate through heterogeneous reactions were found to be the driving factors of heavy PM accumulation in the NCP region (Zheng et al., 2015;Li et al., 2017b;Zou et al., 2017).Although summer is characterized by relatively better air quality compared to the serious haze pollution in winter, fine particle (PM2.5)concentration in the NCP region still remains high during summertime.Through one-year real-time measurements of non-refractory submicron particles (NR-PM1), Sun et al. (2015) showed that the aerosol pollution during summer was comparable to that during other seasons in Beijing, and the hourly maximum concentration of NR-PM1 during the summer reached over 300 μg m -3 .Previous studies focusing on the seasonal variations of aerosol characteristics have noted quite different behaviors of aerosol species in winter and summer (Hu et al., 2017).Therefore, figuring out the specific driving factors of haze evolution in summer would help establish effective air pollution control measures.
Compared to more than 70% reduction of sulfur dioxide (SO2) emissions since 2006 due to the wide application of flue-gas desulfurization devices in power plants and the phase-out of small, high emitting power generation units (Li et al., 2017c), nitrogen oxide (NOx) emissions in China remain high and decreased by less than 20% from 2012 to 2015 (Liu et al., 2016).Therefore, the role of nitrate formation in aerosol pollution is predicted to generally increase as a consequence of high ammonia (NH3) emissions in the NCP region.However, due to the significantly enhanced production of sulfate in extreme winter haze resulting from the high relative humidity (RH) and large SO2 emissions from coal combustion, little attention has been paid to nitrate behaviors.In PM2.5, aerosol nitrate mostly exists in the form of ammonium nitrate, via the neutralization of nitric acid (HNO3) with NH3.HNO3 is overwhelmingly produced through secondary oxidation processes, NO2 oxidized by OH during the day and hydrolysis of N2O5 at night, with the former being the dominant pathway (Alexander et al., 2009).The neutralization of HNO3 is limited by the availability of NH3, as NH3 prefers to react first with sulfuric acid (H2SO4) to form ammonium sulfate with lower volatility (Seinfeld and Pandis, 2006).Because ammonium nitrate is semi-volatile, its formation also depends on the gas-to-particle equilibrium, which is closely related to variations in temperature and RH.A recent review on PM chemical characterization summarized that aerosol nitrate accounts for 16~35% of submicron particles (PM1) in China (Li et al., 2017a).Some studies also pointed out the importance of aerosol nitrate in haze formation in the NCP region (Sun et al., 2012;Yang et al., 2017).However, detailed investigations and the possible mechanisms governing nitrate behaviors during pollution evolution are still very limited.
In this study, we present in-depth analysis of the chemical characteristics of PM1 at urban sites in Beijing and Xinxiang, China during summertime.Based on the varying aerosol composition with the increase of PM1 concentration, the driving factors of haze development were evaluated, and the significance of nitrate contribution was uncovered.In particular, we investigated the chemical behavior of nitrate in detail and revealed the factors favoring rapid nitrate formation during summer in the NCP region.

Sampling site and instrumentation
The data presented in this study were collected in Beijing from 30 June to 27 July, 2015, and in Xinxiang from 8 to 25 June, 2017.
Beijing is the capital city of China, adjacent to Tianjin municipality and Hebei province, both bearing high emissions of air pollutants.The Beijing-Tianjin-Hebei region is regularly listed as one of the most polluted areas in China by the China National Environmental Monitoring Centre.The field measurements in Beijing were performed on the roof of a three-floor building on the campus of Tsinghua University (40.0 °N,116.3 °E).The sampling site is surrounded by school and residential areas, and no major industrial sources are located nearby.Xinxiang is a prefecture-level city in northern Henan province, characterized by considerable were 94 μg m -3 and 84 μg m -3 , respectively.Our sampling in Xinxiang was performed in the mobile laboratory of Nanjing University, deployed in the urban district near an air quality monitoring site (35.3 °N, 113.9 °E).
An Aerodyne Aerosol Chemical Speciation Monitor (ACSM) was deployed for the chemical characterization of NR-PM1, with a time resolution of 15 minutes.Briefly, ambient aerosols were sampled into the ACSM system at a flow rate of 3 L min -1 through a PM2.5 cyclone to remove coarse particles and then a silica gel diffusion dryer to keep particles dry (RH < 30%).After passing through a 100 μm critical orifice mounted at the entrance of an aerodynamic lens, aerosol particles with a vacuum aerodynamic diameter of ~30-1000 nm were directly transmitted into the detection chamber, where non-refractory particles were flash vaporized at the oven temperature (~600 °C) and chemically characterized by 70 eV electron impact quadrupole mass spectrometry.Detailed descriptions of the ACSM technique can be found in Ng et al. (2011).The mass concentration of refractory BC in PM1 was recorded by a multi-angle absorption photometer (MAAP Model 5012, Thermo Electron Corporation) on a 10-min resolution basis (Petzold and Schönlinner, 2004;Petzold et al., 2005).The MAAP was equipped with a PM1 cyclone, and a drying system was incorporated in front of the sampling line.A suite of commercial gas analyzers (Thermo Scientific) were also deployed to monitor variations in the gaseous species (i.e., CO, O3, NO, NOx, and SO2).
For observations in Beijing, the total PM1 mass was simultaneously measured using a PM-714 Monitor (Kimoto Electric Co., Ltd., Japan) based on the β-ray absorption method (Li et al., 2016).Meteorological conditions, including temperature, RH, wind speed, and wind direction, were reported by an automatic meteorological observation instrument (Milos520, VAISALA Inc., Finland).
For measurements in Xinxiang, the online PM2.5 mass concentration was measured using a heated Tapered Elemental Oscillating Microbalance (TEOM series 1400a, Thermo Scientific).The temperature and RH were obtained using a Kestrel 4500 Pocket Weather Tracker.

ACSM data analysis
The mass concentrations of aerosol species, including organics, sulfate, nitrate, ammonium, and chloride, can be determined from the ion signals detected by the quadrupole mass spectrometer (Ng et al., 2011) using the standard ACSM data analysis software (v.1.5.3.0)within Igor Pro (WaveMetrics, Inc., Oregon USA).Default relative ionization efficiency (RIE) values were assumed for organics (1.4), nitrate (1.1), and chloride (1.3).The RIEs of ammonium and sulfate were determined to be 7.16 and 1.08, respectively, through calibration with pure ammonium nitrate and ammonium sulfate.To account for the incomplete detection of aerosol particles (Ng et al., 2011), a constant collection efficiency (CE) of 0.5 was applied to the entire dataset.After all the corrections, the mass concentration of ACSM NR-PM1 plus BC was closely correlated with that of total PM1 obtained by PM-714 in Beijing (r 2 = 0.59; Fig. S1).The slope was slightly higher than 1, which was probably caused by different measuring methods of the different instruments and the uncertainties.For measurements in Xinxiang, the mass concentration of ACSM NR-PM1 plus BC also displayed a good correlation with PM2.5 concentration measured by TEOM, with a slope of 0.83 (r 2 = 0.85; Fig. S1).
Positive matrix factorization (PMF) with the PMF2.exealgorithm (Paatero and Tapper, 1994) was performed on ACSM organics mass spectra to explore various sources of organic aerosol (OA).Only m/z's up to 120 were considered due to the higher uncertainties of larger m/z's and the interference of the naphthalene internal standard at m/z 127-129.In general, signals with m/z > 120 only account for a minor fraction of total signals.Therefore, this kind of treatment has little effect on the OA source apportionment.PMF analysis was performed with an Igor Pro-based PMF Evaluation Tool (PET) (Ulbrich et al., 2009), and the factor, where the former two represented primary sources, and the latter two came from secondary formation processes.In Xinxiang, the identified OA factors included HOA, SV-OOA, and LV-OOA.Procedures for OA source apportionment are detailed in the supplementary materials (Text S1; Tables S1-2; Figs.S2-7).

Air mass trajectory analysis
Back trajectory analysis using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Draxler and Hess, 1998) was conducted to explore the influence of regional transport on aerosol characteristics in Beijing.The meteorological input was adopted from the NOAA Air Resource Laboratory Archived Global Data Assimilation System (GDAS) (ftp://arlftp.arlhq.noaa.gov/pub/archives/).The back trajectories initialized at 100 m above ground level were calculated every hour throughout the campaign and then clustered into several groups according to their similarity in spatial distribution.In this study, a four-cluster solution was adopted, as shown in Fig. S8.

Overview of aerosol characteristics
Summer is usually the least polluted season of the year in the NCP region due to favorable weather conditions and lower emissions from anthropogenic sources (Hu et al., 2017).Figures 1 and 2 show the time series of meteorological parameters, gaseous species concentrations, and aerosol species concentrations in Beijing and Xinxiang.The weather during the two campaigns was relatively hot (average T = 27.1 ± 4.1 °C for Beijing and 26.9 ± 4.0 °C for Xinxiang) and humid (average RH = 55.9 ± 18.5% for Beijing and 63.5 ± 17.2% for Xinxiang), with regular variations between day and night.The average PM1 (= NR-PM1 + BC) concentration Secondary inorganic aerosol, including sulfate, nitrate, and ammonium, dominated the PM1 mass with an average contribution above 50%.The higher nitrate fraction (24% in Beijing and 26% in Xinxiang) is similar to previous observations during summer (Sun et al., 2015;Hu et al., 2016), likely due to photochemical processes being more active than in winter.The mass fraction of OA is lower than that measured during winter in the NCP region (Hu et al., 2016;Li et al., 2017b), in accordance with the large reduction of primary emissions in summer.According to the source apportionment results, OA at both sites is largely composed of secondary factors, in which 44-52% is LV-OOA and 22-23% is SV-OOA (Figs.S4-5).Primary organic aerosol accounts for only 34% and 24% of the total OA in Beijing and Xinxiang, respectively.As there is no need for residential home-heating in summer, which results in lower air pollutant emissions from coal combustion, chloride accounts for a smaller fraction of approximately 1% in total PM1.In addition, the higher temperature during summer drives the partitioning of semi-volatile ammonium chloride into the gas phase, leading to lower concentrations of chloride in the particle phase.
The diurnal variations of aerosol species are similar in the measurements from Beijing and Xinxiang (Fig. S9).Organics demonstrated two pronounced peaks at noon and in the evening.Source characterization of OA suggested that the noon peak was primarily driven by cooking emissions, while the evening peak was a combination of various primary sources, i.e., traffic and cooking.Relatively flat diurnal cycles were observed for sulfate, suggesting that the daytime photochemical production of sulfate may be masked by the elevated boundary layer height after sunrise.Nitrate displayed lower concentrations in the afternoon and higher values at night.To eliminate the effects of different dilution/mixing conditions with the development of boundary layer height, diurnal patterns of the nitrate/sulfate ratio were analyzed to determine the role of chemical processes on nitrate behavior.
The nitrate/sulfate ratio showed the lowest value at approximately 4 pm, indicating that the evaporative loss of particulate NH4NO3 into gaseous NH3 and HNO3 overcame its photochemical production.The nitrate/sulfate ratio peaked at night, revealing the significance of nighttime nitrate formation.During the night, nitrate production is mainly controlled by the heterogeneous hydrolysis of N2O5 (Pathak et al., 2011), which is favored at high RH.A recent study conducted in urban Beijing observed high N2O5 concentrations during pollution episodes and highlighted the vital role of N2O5 chemistry in nitrate formation (Wang et al., 2017).

Enhancement of nitrate formation during pollution episode
To effectively mitigate aerosol pollution through policy-making, the driving factors of the PM increase need to be determined.
Figure 3 illustrates the mass contributions of various species in PM1 as a function of PM1 concentration in Beijing and Xinxiang.
OA dominated PM1 at lower mass loadings (> 40% when PM1 < 20 μg m -3 ), but its contribution significantly decreased with increased PM1 concentration.The source apportionment of OA demonstrated that the large reduction in OA fraction was primarily driven by POA, especially in Beijing.The contribution of SV-OOA and LV-OOA decreased slightly as a result of the photochemical production.The results here are largely different from our winter study in Handan, a seriously polluted city in Northern China, where primary OA emissions from coal combustion and biomass burning facilitated haze formation (Li et al., 2017b).While in Beijing the contribution of sulfate increased slightly at lower PM1 concentrations, the sulfate fraction generally presented a mild decrease with elevated PM1 mass at the two sites.By contrast, nitrate displayed an almost linearly enhanced contribution with increased PM1.This observation is consistent with previous summer measurements in Beijing (Sun et al., 2012) and Nanjing (Zhang et al., 2015), China.Accordingly, the nitrate/sulfate mass ratio steadily increased as PM1 went up.
Notably, the large enhancement of nitrate production mainly occurred after midnight.Figure 4  PM1 mostly increased in the afternoon, nitrate versus PM1 showed steeper slopes from midnight to early morning.The correlation of nitrate with SV-OOA and LV-OOA also indicated that the formation rate of nitrate is considerably higher than that of SV-OOA and LV-OOA after midnight (Fig. S10).Therefore, we further checked the variations in the mass fractions of aerosol species as a function of PM1 concentration for two periods, 0:00 to 11:00 and 12:00 to 23:00.Taking Beijing as an example, both the nitrate contribution in PM1 and the nitrate/sulfate ratio were significantly enhanced for the period of 0:00 to 11:00 (Fig. S11).These results suggest that rapid nitrate formation is mainly associated with nighttime productions, when the heterogeneous hydrolysis of N2O5 dominates the formation pathways along with higher RH and lower temperature.The observed high N2O5 concentrations in urban Beijing further support our hypothesis (Wang et al., 2017).Because the materiality of nitrate formation to haze evolution was observed in both Beijing and Xinxiang, we regard this as the regional generality in summer.Considering the efficient reduction in SO2 emissions in China (Zhang et al., 2012), the results here highlight the necessity of further NOx emission control for effective air pollution reduction in Northern China.

Factors influencing the rapid nitrate formation
Submicron nitrate mainly exists in the form of semi-volatile ammonium nitrate and is produced by the reaction of NH3 with HNO3 in the atmosphere.The formation pathways of HNO3 include the oxidation of NO2 by OH during the day and the hydrolysis of N2O5 at night.Thus, to investigate factors influencing the rapid nitrate formation in summer, the following conditions need to be considered: (1) the abundance of ammonia in the atmosphere, (2) the influence of temperature and RH, and (3) different daytime and nighttime formation mechanisms.Here, we explore nitrate formation processes based on Beijing measurements.
Under real atmospheric conditions, NH3 tends to first react with H2SO4 to form (NH4)2SO4 due to its stability (Seinfeld and Pandis, 2006).Thus, if possible, each mole of sulfate will remove 2 moles of NH3 from the gas phase.NH4NO3 is formed when excess NH3 is available.During the sampling period, the observed molar ratios of ammonium to sulfate were mostly larger than 2 (Fig. 5), corresponding to an excess of NH3.The scatter plot of the molar concentration of excess ammonium versus the molar concentration of nitrate showed that, nitrate was completely neutralized by excess ammonium at most times.When ammonium is in deficit, nitrate may associate with other alkaline species or be part of an acidic aerosol.
Based on the ISORROPIA-Ⅱ thermodynamic model, we performed a comprehensive sensitivity analysis of nitrate formation to the gas-phase NH3 and PM1 sulfate concentrations.Under typical Beijing summer conditions (T = 300.21K,RH = 56%), we assumed that total inorganic nitrate (HNO3 + NO3 -) in the atmosphere was 10 μg m -3 .Total ammonia (gas + particle) and PM1 sulfate concentrations were independently varied and input in the ISORROPIA-Ⅱ model.The predicted equilibrium of the nitrate partitioning ratio (ε(NO3 -) = NO3 -/(HNO3 + NO3 -)) is shown in Fig. 6.At a sulfate concentration from 0.1 to 45 μg m -3 , a 10 μg m - 3 increase of gaseous NH3 generally results in an enhancement of ε(NO3 -) by over 0.1 units, thus increasing the particulate nitrate concentration.Interestingly, for ammonia-rich systems, the existence of more particulate sulfate favors the partitioning of nitrate towards the particle phase.The formation of particulate ammonium nitrate is a reversible process with dissociation constant Kp: Kp equals the product of the partial pressures of gaseous NH3 and HNO3.For an ammonium sulfate-nitrate solution, Kp not only depends on temperature and RH but also on sulfate concentrations, which is usually expressed by the parameter Y (Seinfeld and Pandis, 2006): When the concentration of ammonium sulfate increases compared to that of ammonium nitrate, the parameter Y decreases and the equilibrium product of NH3 and HNO3 decreases.The additional ammonium and sulfate ions make the aqueous system favorable for the formation of ammonium nitrate, by increasing particle liquid water content but not perturbing particle pH significantly.Particle pH is not highly sensitive to sulfate and associated ammonium (Weber et al., 2016;Guo et al., 2017b).Therefore, more ammonium sulfate in the aqueous solution will tend to increase the concentration of ammonium nitrate in the particle phase.
However, compared to the significant influence of gaseous NH3, ε(NO3 -) is weekly sensitive to the sulfate concentration, as shown in Fig. 6.For example, when the ammonia concentration is 10 μg m -3 , a reduction of sulfate from 30 to 20 μg m -3 has little influence on ε(NO3 -).Generally, these results suggest that a decrease in the SO2 emissions may have a positive effect on nitrate reduction, though controlling NH3 emissions appears to be more effective.
The influence of temperature and RH on nitrate formation was also evaluated based on ISORROPIA-Ⅱ simulations by varying temperature and RH separately.As shown in Fig. S12, under typical Beijing summer conditions (T = 30 ℃), ε(NO3 -) remains lower than 0.1, even until RH reaches 80%.When RH > 90%, ε(NO3 -) increases sharply as a function of RH.For T = 0 ℃, representative of Beijing winter conditions, ε(NO3 -) is as high as 0.7, even at low RH. Figure 7 demonstrates the variations in the nitrate/sulfate ratio as a function of temperature and RH in Beijing.The nitrate/sulfate ratio increased with decreasing temperature and increasing RH, which drives the nitrate partitioning towards the particle phase.This is further supported by the variations in the equilibrium constant KAN of Eq. ( 1), which can be calculated as: where T is the ambient temperature in Kelvin, KAN (298) = 3.36 ×10 16 (atm -2 ), a = 75.11,and b = -13.5 (Seinfeld and Pandis, 2006).
Similar to the nitrate/sulfate ratio, the diurnal profile of KAN peaks at night due to the lower temperature and higher RH.
As described in Sect.3.2, the rapid nitrate formation in this study appeared to be mainly associated with its nighttime enhancement.
In addition to the effects of temperature and RH, the nighttime nitrate formation pathways may also play a role.Overnight, particulate nitrate primarily forms via the heterogeneous hydrolysis of N2O5 on the wet surface of aerosol (Ravishankara, 1997).
N2O5 is produced by the reversible reaction between NO2 and the NO3 radical, where NO2 reacts with O3 to form the NO3 radical.
Assuming N2O5 and the NO3 radical are both in steady state considering their short lifetimes (Brown et al., 2006), the nighttime production of N2O5 and HNO3 is proportional to the concentration of NO2 and O3 ([NO2][O3]) (Young et al., 2016;Kim et al., 2017).For the different PM1 concentration bins, we examined the NO2 and O3 data at 0:00 to assess the nighttime HNO3 production rate.It can be seen that [NO2][O3] was obviously enhanced with an increase in the PM1 mass loading (Fig. S13), implying that nitrate formation by the N2O5 pathway favors the driving role of nitrate in haze evolution.
According to the Multi-resolution Emission Inventory for China (MEIC, http://www.meicmodel.org),NOx emissions localized in Beijing are much smaller than emissions in adjacent Hebei, Shandong, and Henan provinces.In Fig. 1, episodes in Beijing, characterized by largely enhanced nitrate concentrations, usually occurred with the change in the wind direction from north and west to south and east, where the highly polluted Hebei, Shandong, and Henan provinces are located.When the relatively clean air masses from north and west returned, aerosol pollution was instantly swept away.Therefore, the importance of regional transport on haze formation in Beijing should also be considered.We examined the association of aerosol concentration and composition with air mass origins determined through cluster analysis of HYSPLIT back trajectories.As illustrated in Fig. 8

Comparison with other regions and policy implications
Figure 9 summarizes the chemical composition of PM1 or NR-PM1 (BC excluded) measured during the summer in Asia, Europe, and North America.Three types of sampling locations were included: urban areas, urban downwind areas, and rural/remote areas.
For further comparison, we classified the datasets into three groups according to the location type and examined their difference in nitrate mass concentrations and mass contributions.Overall, the nitrate concentrations varied from 0.04 μg m -3 to 17.6 μg m -3 in summer, with contributions of 0.9% to 25.2%.Patterns in Fig. 10 demonstrate that the nitrate concentrations in mainland China are usually much higher than those in other areas, consistent with the severe haze pollution in China.In particular, the percentage of nitrate in aerosol particles is generally several times higher in mainland China than in other regions, except for measurements in Riverside, CA, which were conducted near the local highway (Docherty et al., 2011).Compared to rural/remote areas, nitrate shows higher mass concentrations and mass fractions in urban and urban downwind areas, revealing the influence of anthropogenic emissions, i.e., traffic and power plant, on nitrate formation.In Beijing, the capital of China, field measurements among different years show an obvious reduction in the nitrate mass concentration, especially from 2012.This coincides with the decline in satelliteobserved NO2 levels in China after 2011 (Miyazaki et al., 2017) and a 21% decrease in NOx emissions from 2011 to 2015 based on a bottom-up emission inventory (Liu et al., 2017).Detailed analysis by Liu et al. (2017)  obviously enhanced at night along with higher RH, suggesting the increased role of nighttime nitrate production in haze evolution.
Based on cluster analysis via the HYSPLIT model, nitrate formation was also found to depend on regional transport from different air mass origins, in accordance with the spatial distribution of NOx emissions in the NCP region.
Finally, nitrate data acquired from this study were integrated with the literature results, including various field measurements conducted in Asia, Europe, and North America.Nitrate is present in higher mass concentrations and mass fractions in China than in other regions.Due to the large anthropogenic emissions in urban and urban downwind areas, the mass concentrations and mass contributions of nitrate are much higher in these regions than in remote/rural areas.Although the nitrate mass concentrations in Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.industrial manufacturing.In February 2017, the Chinese Ministry of Environmental Protection issued the "Beijing-Tianjin-Hebei and the surrounding areas air pollution prevention and control work program 2017" to combat air pollution in Northern China.The action plan covers the municipalities of Beijing and Tianjin and 26 cities in Hebei, Shanxi, Shandong and Henan provinces, referred to as "2+26" cities. Xinxiang is listed as one of the "2+26" cities.The average PM2.5 concentrations in Xinxiang in 2015 and 2016 Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.results were evaluated following the procedures detailed in Ulbrich et al. (2009) and Zhang et al. (2011).According to the interpretation of the mass spectra, the temporal and diurnal variations of each factor, and the correlation of OA factors with external tracer compounds, a four-factor solution with FPEAK = 0 and a three-factor solution with FPEAK = 0 were chosen as the optimum solutions in Beijing and Xinxiang, respectively.The total OA in Beijing was resolved into a hydrocarbon-like OA (HOA) factor, a cooking OA (COA) factor, a semi-volatile oxygenated OA (SV-OOA) factor, and a less-volatile oxygenated OA (LV-OOA) Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.was35.0 μg m -3 in Beijing and 64.2 μg m -3 in Xinxiang, with the hourly maximum reaching 114.9 μg m -3 and 208.1 μg m -3 , respectively.Several pollution episodes were clearly observed at the two sites, along with largely increased nitrate concentrations.
displays the scatter plots of nitrate versus PM1 as well as sulfate versus PM1 for comparison, both color-coded by the time of day.Though the ratios of sulfate versus Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.
, the aerosol characteristics are quite different for air masses from different regions.Cluster 1 mainly passed through Shanxi and Hebei provinces, and Cluster 2 originated from Hebei, Shandong, and Henan provinces.Consistent with the high air pollutant emissions in these areas, Cluster 1 and Cluster 2 were characterized with high PM1 concentrations and high contributions of secondary aerosols.The nitrate fraction in PM1 was 24% for Cluster 1 and 26% for Cluster 2. In comparison, Cluster 3 and Cluster 4 resulted from longrange transport from the cleaner northern areas and were correspondingly characterized by lower PM1 concentrations.Organics dominated PM1 for Cluster 3 and Cluster 4, with a nitrate contribution of 14% and 16%, respectively.Figure S14 shows the cluster distribution as a function of PM1 concentration.With an increase in the PM1 mass, the contribution of cleaner Cluster 3 and Cluster Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License. 4 significantly decreased.When PM1 concentrations were above 20 μg m -3 , the air masses arriving in Beijing were mostly contributed by Cluster 1 and Cluster 2, which led to rapid nitrate accumulation.
revealed that the NOx decline in China in recent years is mainly driven by the penetration of selective catalytic reduction (SCR) in power plants and strict regulations for vehicle emissions.The large decrease in nitrate concentration in the summer of 2008 was primarily caused by the strict emission control measures implemented during the 2008 Olympic Games(Wang et al., 2010).However, nitrate contributions in China still remain high over the years, especially in urban and urban downwind areas, indicating the importance of nitrate formation in haze episodes.Overall, the higher concentration and, in particular, the higher contribution of nitrate in aerosol particles during summertime call for the urgent need of further NOx reduction measures and NH3 emission control in China.4ConclusionsSummertime field measurements were conducted in both Beijing (30 June to 27 July, 2015) and Xinxiang (8 to 25 June, 2017) in the NCP region, using state-of-the-art online instruments to investigate the factors driving aerosol pollution.The average PM1 concentration was 35.0 μg m -3 in Beijing and 64.2 μg m -3 in Xinxiang, with the hourly maximum reaching 114.9 μg m -3 and 208.1 μg m -3 , respectively.Pollution episodes along with significantly enhanced nitrate concentrations were frequently observed during the campaigns.Secondary inorganic aerosol dominated the PM1 mass, with higher nitrate contributions of 24% in Beijing and 26% in Xinxiang.The diurnal profile of nitrate presented higher concentrations at night and lower values in the afternoon.By eliminating the influences of different dilution/mixing conditions due to boundary layer development, we found that the lower nitrate concentrations in the afternoon were caused by the strong evaporative loss of nitrate at higher temperatures, which overcame the daytime photochemical production of nitrate.With the development of aerosol pollution, OA showed a decreasing contribution to total PM1, despite its obvious domination at lower PM1 mass loadings.The reduction in the OA mass fraction was primarily driven by primary sources, especially in Beijing.Generally, the mass fraction of sulfate also decreased slightly as a function of PM1 concentration.In contrast, nitrate presented an almost linearly enhanced contribution with the elevation of PM1 mass, suggesting Atmos.Chem.Phys.Discuss., https://doi.org/10.5194/acp-2017-1156Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 2 January 2018 c Author(s) 2018.CC BY 4.0 License.theimportant role of nitrate formation in causing high aerosol pollution during summer.Rapid nitrate production mainly occurred after midnight, and the formation rate was higher for nitrate than for sulfate, SV-OOA, or LV-OOA.Comprehensive analysis of nitrate behaviors revealed that abundant ammonia emissions in the NCP region favored nitrate production in summer.According to the ISORROPIA-Ⅱ thermodynamic predictions, ε(NO3 -) is significantly increased when there is more gas-phase ammonia in the atmosphere.Decreased SO2 emissions may have co-beneficial impacts on nitrate reduction.Lower temperature and higher RH shift the equilibrium partitioning of nitrate towards the particle phase, thus increasing the particulate nitrate concentration.Assuming both N2O5 and NO3 radicals are in steady state, [NO2][O3] can be used as an indicator to evaluate the contribution of nighttime N2O5 hydrolysis to nitrate formation.With the anabatic pollution levels, [NO2][O3] Beijing have steadily decreased over the years, its contribution still remains high, emphasizing the significance of further NOx reduction and the initiation of NH3 emission control in China.Most of the previous studies conducted during wintertime reveal that secondary formation of sulfate together with primary emissions from coal combustion and biomass burning are important driving factors of haze evolution in the NCP region.According to this study, in Beijing and Xinxiang, rapid nitrate formation is regarded as the propulsion of aerosol pollution during summertime.Therefore, to better balance economic development and air pollution control, different emission control measures could be established corresponding to the specific driving forces of air pollution in different seasons.Further studies on seasonal variations are needed to test the conclusions presented here and provide more information on haze evolution in spring and fall.477 Figure 1.Time series of meteorological parameters, gaseous species, and submicron aerosol species in Beijing.