characterization of submicron aerosols and effect on visibility during a severe haze fog episode in yangtze river delta china

Zhanga a Key Laboratory of Atmospheric Chemistry of CMA, Institute of Atmospheric Composition, Chinese Academy of Meteorological Sciences, Beijing, China b State Key Laboratory of Cryosp

Characterization of submicron aerosols and effect on visibility during a

severe haze-fog episode in Yangtze River Delta, China

X.J Shena, J.Y Suna,b,*, X.Y Zhanga, Y.M Zhanga, L Zhanga, H.C Chea, Q.L Mac,

X.M Yuc, Y Yuec, Y.W Zhanga

a Key Laboratory of Atmospheric Chemistry of CMA, Institute of Atmospheric Composition, Chinese Academy of Meteorological Sciences, Beijing, China

b State Key Laboratory of Cryospheric Sciences, Cold and Arid Region Environmental and Engineering Research Institute, Chinese Academy of Sciences,

Lanzhou, China

c Lin'an Atmospheric Background Station, Meteorological Bureau of Zhejiang Province, China

h i g h l i g h t s

Particle number size distribution characteristic during haze-fog episode

Submicron aerosol influence on light extinction coefficient is evaluated

Secondary aerosol formation controlling haze-fog episode

The effect of air mass origin on haze-fog formation episode

a r t i c l e i n f o

Article history:

Received 26 March 2015

Received in revised form

1 September 2015

Accepted 2 September 2015

Available online 7 September 2015

Keywords:

Particle number size distribution

Light extinction of submicron aerosol

Secondary aerosol formation

Severe haze-fog

Air mass origin

a b s t r a c t

Particle size, composition and optical properties were measured at a regional atmosphere background station in the Yangtze River Delta (YRD) to understand the formation and evolution of haze-fog episodes

in Jan 2013 The peak of particle number size distribution was in the size range of 80e100 nm during the measurements PM1mass concentration contributed 84% to the total particle mass (PM10) Based on visibility and ambient relative humidity, three types of weather conditions (i.e., clear, haze and fog) were classified in this study The extinction coefficients of PM1and PM10under dry conditions were simulated

by the Mie model Under dry conditions, PM1was found to contribute approximately 91% to the light extinction coefficient of PM10 However, the PM1 with the assumption of dry state was found to contribute approximately 85% to the ambient extinction coefficient of PM10during clear conditions, 58% during haze conditions and approximately 41% during fog conditions The variation of the dry PM1

contribution was related to the water uptake of particles under different relative humidity conditions

A severe haze-fog event on Jan 14e17 was discussed in more detail as a case study Two episodes were chosen to show that nitrate and organics dominated the aerosol component during the severe haze-fog episode and were related to secondary aerosol formation and air mass origin Nitrate played a more dominant role than sulfate in heavy haze formation in the YRD region, which was different from the North China Plain region

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

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

1 Introduction

The Yangtze River Delta (YRD), including Shanghai and its

neighboring 8 cities in Jiangsu Province and 7 cities in Zhejiang

Province, is one of the most densely populated and economically developed regions in China Complex and regional air pollution, such as high concentrations of ozone, anthropogenic gaseous pol-lutants (SO2, NOx, NH3and VOCs) and particulate matter (Wang

et al., 2001), has been a severe issue in this region Haze-fog days

atmospheric environment research (Zhang et al., 2012) Haze-fog formation is closely related to meteorological conditions and high aerosol mass loading (Wang et al., 2014a), which has a significant impact on the visibility, public health and even the global climate

* Corresponding author Key Laboratory of Atmospheric Chemistry of CMA,

Institute of Atmospheric Composition, Chinese Academy of Meteorological

Sci-ences, Beijing, China.

E-mail address: jysun@cams.cma.gov.cn (J.Y Sun).

Contents lists available atScienceDirect Atmospheric Environment

j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / lo c a t e / a t m o s e n v

http://dx.doi.org/10.1016/j.atmosenv.2015.09.011

1352-2310/© 2015 The Authors Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

Atmospheric Environment 120 (2015) 307e316

(Chan and Yao, 2008; Che et al., 2014).

Several severe haze-fog periods were observed in the YRD

re-gion with major aerosol components, including sulfate, nitrate,

coefficient in this region is high due to the aerosol mass

concen-tration and is also affected by the aerosol chemical component,

particle number size distribution (PNSD), and water vapor in the

atmosphere (Pan et al., 2010) The secondary organic aerosol (SOA)

played an important role in the haze formation (Zhang et al., 2012)

The transformation of SO2and NO2, which mostly originated from

fossil fuel combustion and vehicle emissions (Geng et al., 2009),

contributed much to the high concentration of secondary nitrate

and sulfate in the YRD (Fu et al., 2008) In Jan 2013, a large area in

China, including the North China Plain (NCP), Central East China

(CEC), and part of Southern China, experienced extremely severe

and persistent haze pollution One study in urban cities, including

Beijing, Shanghai, Guangzhou and Xi'an, reported that the haze was

driven by high aerosol mass concentration, which was contributed

2014a) It was also found that this severe large area haze episode

was accompanied by low visibility, high particle mass loading and

aerosol optical depth (Wang et al., 2014b), and also by modification

of fog processing on the particle size (Huang et al., 2014b) Thefield

campaign proved that the contribution of secondary species to

the haze-fog episode compared with non-haze-fog days in

Shanghai, and nitrate mass even exceeded sulfate mass during the

episode (Jansen et al., 2014)

In this work, an intensive measurement of submicron aerosol

physical and chemical characteristics was performed in Jan 2013 in

YRD region We focused on the study of PNSD, with auxiliary data

including meteorological factors, visibility, reactive gas and aerosol

chemical components, as well as the optical model, to illustrate the

formation and evolution of the heavy haze-fog episode We

addressed PNSD characteristics under different conditions and

tried to estimate the contribution of dry submicron aerosols to the

extinction coefficient under dry and ambient conditions

Further-more, the PNSD evolution processes, as well as the variation of

corresponding chemical composition and air mass will be

illus-trated in detail following a haze-fog episode case study

2 Experimental methods

2.1 Site description

The measurements were made in Jan 2013 at the Lin'an regional

atmospheric background station (30170N, 119450E, 138.6 m asl.),

which is nearly 50 km west of Hangzhou, the capital of Zhejiang

province (Fig 1) The site is approximately 200 km southwest of

Shanghai Approximately 10 km to the south of the Lin'an station is

the Lin'an Township, with a population of approximately 50,000 As

one of the regional Global Atmosphere Watch stations in China, the

Lin'an station lies on the top of a hill surrounded by patches of pine

and bamboo forest There are very limited local pollution sources

nearby; thus, the station can represent the background atmosphere

of the economically developed YRD region

2.2 Instrumentation

Measurements were conducted inside a laboratory with

regu-lated temperature Sampling air was collected through a PM10inlet

placed on the roof of the room, with aflow rate of 16.7 l/min The

et al., 2009) The dried air then went through the splitter to the

TDMPS (Twin Differential Mobility Particle Sizer, TROPOS), APS

(Aerodynamic Particle Sizer, TSI 3321), MAAP (Multi-Angle Ab-sorption Photometry, Thermo 5012), AMS (Aerosol Mass Spec-trometer, Aerodyne) and a hygroscopicity measurement system based on a wet and a dry nephelometer (TSI, Model 3563) The TDMPS were used to measure PNSD with electrical mobility diameter in the range of 3e800 nm The APS measured the particle number size distributions with aerodynamic diameters from 0.5 to

We followed the recommended standard inversion routine to derive PNSDs from the measured electrical mobility distribution (Wiedensohler et al., 2012) The PNSDs derived by APS system were converted from aerodynamic to mobility diameters using a particle

components And the APS data with mobility diameter larger than

800 nm were selected to combine with the TDMPS data

integrating nephelometer at the wavelengths of 450, 550 and

700 nm (Anderson et al., 1996), with a time resolution of 1 min The MAAP (Multi-Angle Absorption Photometer, Thermo 5012) deter-mined absorption coefficients directly and converted them to mass concentrations of black carbon (BC) with an assumed mass ab-sorption efficiency of 6.6 m2g1(Petzold and Sch€onlinner, 2004)

compo-nents, including sulfate, nitrate, ammonium, organic and chloride, was conducted by an Aerodyne Quadropole Aerosol Mass Spec-trometer (Q-AMS), which provided high time-resolution (5 min) information to characterize the size-resolved composition of PM1 Details of the AMS have been described in a previous publication (e.g., Canagaratna et al., 2007) More details about AMS set up,

(2010)

43CTL) The time resolution is 5 min The maintenance and cali-bration of the instruments, as well as the correction of the data have been described byLin et al (2008)

Meteorological data, including atmospheric temperature (T), relative humidity (RH), precipitation, wind speed and wind direc-tion, were monitored by an automatic weather station (type DZZ4, Jiangsu Radio Scientific Institute CO., LTD, China) Visibility was measured by a Vaisala FD12 visibility meter with a time resolution

of 15 s During the measurements, we noticed that the old RH

Fig 1 The location of Lin'an station (star) as well as the surrounding major cities (circles).

X.J Shen et al / Atmospheric Environment 120 (2015) 307e316 308

sensor had a low bias To correct this bias, a new RH sensor has been

operated in parallel with the old one since July 2014 Details of

material

2.3 Modalfitting

Atmospheric aerosol size distributions are often described as

the sum of three log-normal distributions (Birmili et al., 2001),

(2005):

dN

d log Dp¼Xn

i¼1

Ni ffiffiffiffiffiffi

2p

p

logsi exp

0

@ 

 log Dp logDp;i2 2ðlog siÞ2

1

where Niis the number concentration, Dp; iis the geometric mean

diameter (GMD), andsiis the standard deviation of the ith

used two modes (Aitken mode and accumulation mode) in the

fitting process, as the nucleation mode particle number

concen-tration during the measurement period was usually much lower

than the other two modes

3 Results and discussion

3.1 Meteorological condition and particle number/mass

concentration

larger than 10 km (Fig 2a) The worst visibility during the

mea-surement period, lower than 1 km, occurred when it snowed and

rained and when RH was larger than 95% The mean value and standard deviation of RH during the measurement period was

occurred in the CEC, NCP and YRD regions During this long-lasting period, the surface pressurefield showed a uniform pattern, which indicated the wind speed was low and unfavorable for the hori-zontal and vertical exchange of water vapor and pollutants (Zhang

et al., 2014) Based on the surface measurements, wind direction was predominately northeast and southwest Furthermore, wind

pre-cipitation are excluded in the following discussion to segregate the

mentioned

75mg m3, which was the criterion value of the second grade of air

polluted condition Based on this criterion, there were 21 polluted days out of 31 days in Jan 2013 The mass concentration of PM1,

chemical components and an assumption of spherical shape The time series of PM1, PM2.5and PM10mass concentration with 10 min resolution are given inFig 2d The time series show that PM1, PM2.5

under severe polluted conditions (e.g., Jan 15) On clear days, the

PM2.5mass concentration was even less than 35mg m3, which was the criterion (daily mean value) of thefirst grade of the air quality

in China The average and standard deviation of PM1, PM2.5and

104± 46mg m3, respectively, during the measurement The mass ratios of PM1/PM10, PM1/PM2.5and PM2.5/PM10were 0.84, 0.92 and 0.91, respectively The mass ratio of PM1/PM10could be as low as 0.60 when coarse mode particle number concentration increased and also could reach to 0.95 due to the obvious scavenging process

Fig 2 The time series of visibility and precipitation (a), evolution of PNSD with the D p (circles) of the dominant mode (b); particle number concentrations in different modes (c) and

of large particles under fog conditions Generally speaking, mass of

total particles was dominated by submicron particles (PM1) during

the entire period

Fig 2b shows that PNSD had a large variation from day to day

Particle number concentrations were concentrated below the size

from 1000 to 13,000 cm3, with a mean value of 4700 cm3 The

accumulation mode number concentration was in the range of

1300e12,000 cm3, with a mean value of 4780 cm3 The

nucle-ation mode number concentrnucle-ation was usually as low as several

hundred per cubic centimeter, except when a NPF event occurred

Dal Maso et al (2005) This method required a sharp increase of

nucleation mode particle number concentration, subsequent

growth to larger sizes and duration of a few hours Only one NPF

event could be identified during the entire study period, occurring

on Jan 20 The hourly maximum of particle number concentration

of nucleation mode increased to a value of approximately 104cm3,

much higher than the mean value during the study period,

490 cm3

3.2 The classification of clear, haze, fog conditions

The relationship of visibility, RH and PM2.5mass concentration

in this study is shown inFig 3 In this work, three typical weather

conditions, clear, haze and fog, were considered The criteria

(WMO, 2005) for classifying these conditions were: (1) for the clear

condition, visibility 5 km; (2) for the haze condition, visibility

5 km and RH 95%; (3) for the fog condition, visibility <1 km and

vis-ibility data with an hour resolution Theoretically, particle will be

activated only at the saturation higher than 100% However, the RH

sensor doesn't work well at such high RH Furthermore, it has been

revealed that the particles of size 1e10mm can be activated into cloud droplets with a relative humidity of less than 100% based on

(Kulmala et al., 1997) Therefore, a threshold value of 95% was chosen as the best choice to identify the fog condition in this study For the haze condition, PM2.5varied from less than 50mg m3to more than 200mg m3; levels at which aerosols could play major roles in the reduction of visibility With an elevated RH, the hy-groscopic growth of aerosols had a larger impact on visibility due to water uptake (Zhang et al., 2015) As observed fromFig 3, the PM2.5

mass concentration in fog conditions was relatively low, which suggests that some particles could be activated and grow out the range of PM10 Therefore, the degradation of visibility was not sensitive to the PM2.5mass concentration

Classification results for clear, haze, and fog conditions are

concentration and occurrence frequency during the corresponding conditions Statistical results were derived based on hourly

that the haze condition occurred most frequently in Jan 2013, which also showed the highest mean mass concentration of

115mg m3compared with the other conditions Due to the diurnal variation of RH (normally higher at night and lower during the day), haze could transform to fog, and fog to haze However, this trans-formation process was not easily and clearly distinguished; thus, this condition is referred to as a haze-fog episode in this study 3.3 Characteristics of particle size distribution on different conditions

The mean PNSD and particle volume size distribution (PVSD) in clear, haze and fog conditions are given inFig 4 The modefitting

clear conditions was dominant in the Aitken mode For haze and fog days, PNSD was characterized by the accumulation mode, with

mode shifted from 78 nm under the clear condition to the larger size of 137 nm under fog condition Particle growth probably relates

to atmospheric aging processes, including oxidation (Ivleva et al.,

(Riemer et al., 2004) processes For PVSD (Fig 4b), particle volume

condi-tions and was highest under haze condicondi-tions, indicating highest particle mass loading under this condition

3.4 The effect of PM1on visibility The extinction coefficient of particles could be estimated using the PNSD measurement, particle refractive indices and the Mie

2011, 2012), and then converted to visibility using a modified Koschmieder relation (Schichtel et al., 2001):

where VIScalis the calculated visibility and sextis the extinction coefficient in an ambient state, the sum of scattering and absorp-tion by particles and gases However, in this work, extincabsorp-tion by gas was not taken into consideration The standard Koschmieder con-stant of 3.92 (Seinfeld and Pandis, 1998) is replaced by 1.9 because the real visual targets are not black, are too small in angular size and are located only at quantized distances away from the observer (Carrico et al., 2003; Cheng et al., 2008)

In the Mie calculations, an assumption of internal mixing of all chemical components in the particles was applied The extinction

Fig 3 The relationship of visibility vs RH colored by PM 2.5 mass concentration The

visibility is in log scale.

Table 1

The classification of clear, haze and fog days based on visibility and RH, as well as the

corresponding mean PM 2.5 mass concentration and occurrence frequency.

Type VIS (km) RH (%) PM 2.5 (mg m3) Frequency (%)

X.J Shen et al / Atmospheric Environment 120 (2015) 307e316 310

coefficientsext at the wavelength of 550 nm was calculated by

details given in thesupplementary material The linearfit result of

slope and R2(square of correlation coefficient) was 1.10 and 0.99,

respectively, indicating they were in agreement, with the

calcu-lated result being 10% higher (Fig 5a) A sensitivity study was

conducted by using the varied volume fraction of chemical

component as the input of the Mie model, with the result showing

that the uncertainty induced by the assumption of constant volume

fraction input in simulating the optical properties could be 5%

higher than an assumption of varied volume fraction The details of

well

Furthermore, if we simulated the optical properties in the

ambient atmosphere, the hygroscopic characteristics of aerosols

had to be considered At Lin'an station, the enhancement factor,

wavelength of 550 nm from a wet and a dry nephelometer in

values of f(RH) at RH values of 85%, 80%, 70%, 60% and 50%, and

gave parameterization results for f(RH) based on the equation

fðRHÞ ¼ 1 þ a·RHb (Kotchenruther and Hobbs, 1998) The

param-eters a and b were given for different polluted cases and were 1.24

and 5.46 for a locally polluted case and 1.20 and 3.90 for a

northerly polluted case (Zhang et al., 2015) In this work, the

mean value of the two conditions was applied and thus f(RH)

corresponding to an ambient RH condition could be derived

However, when RH was higher than 85%, f(RH) might be

underestimated

The extinction coefficients of PM1(sext, 1mm, dry, cal) and of PM10

(sext, 10mm, dry, cal) under dry conditions were calculated using the

RH Furthermore, the extinction coefficient of PM10under ambient conditions (sext, 10mm, amb, cal) could be derived fromsext, 10mm, dry, cal

by multiplying the enhancement factor f(RH) at ambient RH Based

on thesext, 10mm, amb, caland the modified Koschmieder relation, the calculated visibility (VIScal) was derived and is shown inFig 5b VIScal showed a quite similar pattern to the measured visibility (VISmea) except that when the measured visibility was lower than

1 km and the corresponding RH was higher than 95%, the calculated visibility was always higher than the measured visibility Under this condition, the VIScalwas approximately a factor of 2e20 higher than the VISmea, which indicated that the visibility was overestimated Once activated, particles may grow freely and cause a significant enhancement in extinction coefficient The ambient extinction

co-efficient in fog therefore cannot be simulated with f(RH) because f(RH) is based on hygroscopic growth Furthermore, the contribu-tion by snow or rain droplets was not estimated, which was the larger contributor to reducing visibility during precipitation Another explanation is that the measurement of PNSD was underestimated, especially for the accumulation mode particles During fog conditions, there could be many particles being

impactor

The mean ratio ofsext, 1mm, dry/sext, 10mm, amband the standard deviation of contribution of PM1(PM1,contrib) was 85± 11% under clear conditions, 58± 18% under haze conditions and 41 ± 1% under fog conditions, as shown in Fig 5c This ratio could reflect the

extinction of aerosols at ambient RH conditions, which mixed the optical contribution of dry PM1and the influence of RH And the

under haze and fog conditions could be inferred based on the dif-ference of sext, 1mm, dry/sext, 10mm, ambbetween clear condition It could be roughly estimated that the contribution of the hygroscopic growth of PM1particles to the light extinction was 27% under haze condition and 44% under fog condition, respectively Because the contribution of PM1extinction to PM10extinction under dry con-dition had been evaluated (~90%), the ratio ofsext, 1mm, dry/sext, 10mm, ambcould be regarded as an indicator of the influence of RH The result was comparable with the study conducted in the NCP region that high aerosol volume concentration was responsible for low visibility at RH < 90% and that for RH > 90% low visibility was usually caused by the increase of RH (Chen et al., 2012) In addition,

we used the available nephelometer measurement data (Jan

scattering albedo at the wavelength of 550 nm, which was 0.84± 0.08 during these days

Fig 4 The mean PNSD and PVSD under different conditions.

Table 2

Characteristic parameters to represent PNSD belong to different conditions The

symbols mean: geometric mean diameters D p;n (nm), particle number concentration

N (cm3), and geometric standard deviationsn The subscript of n indicated the

mode fitting parameters for number size distribution, respectively The number of 1

and 2 in the subscript indicated there two modes fitted base on the PNSD data.

Type Number size distribution

3.5 Case study of a severe haze-fog period

In Jan 2013, extensive and long-lasting haze-fog phenomena

occurred due to poor meteorological conditions and accumulating

pollutants Two major heavy haze-fog episodes were observed,

occurring on Jan 7e17, and Jan 23e28, respectively FromFig 2, it

can be observed that there were cycles of a few days when the mass

concentration of aerosols built up gradually until the pollution was

removed by strong wind or precipitation In the following

discus-sion, we chose Jan 14e17 as a representative case to discuss the

haze-fog formation and evolution in more detail

sur-face and volume concentration, as well as the meteorological

conditions during this severe haze-fog period is shown in Fig 6

There were two episodes, marked with P1 and P2, when haze

occurred during daytime and transformed into fog during the night

During these episodes, the visibility decreased significantly and the

surface and volume concentration built up under haze conditions

The time period was chosen as 16:00 on Jan 14 to 06:00 on Jan 15

for the P1 episode and 13:00 on Jan 15 to 09:00 on Jan 16 for the P2

episode, respectively During the P1 episode, Dpshowed a

contin-uous increase from about 100 nm to 150 nm, with a growth rate of

6.8 nm h1from 18:00 to 23:00 on Jan 14 During the P2 period, Dp

also showed a growth trend, with a growth rate of 6.1 nm h1from

18:00 on Jan 15 to 02:00 on Jan 16 The wind vector showed

dominant wind directions of northeast and southwest, with

occa-sional calms

The particle surface and volume concentration were derived

based on the PNSD measurements, with the assumption of

spher-ical particles The surface and volume concentration could be

enhanced significantly by a factor of ~2 and 2e3 during these two

episodes The increased surface concentration not only enhanced

the light extinction ability of particles; it also provided a larger

surface area for heterogeneous reactions The high volume con-centration indicates increased aerosol mass loading

72 h backward trajectories were calculated using the HYSPLIT 4 model (Draxler and Rolph, 2003) on Jan 14e17 at 08:00, 14:00 and 20:00 LT, with a terminal height of 500 m above the ground level It showed that most of the back trajectories on Jan 14e16 traveled at the low height and affected by the mixing layer Model results (Fig 7) on Jan 14 confirmed that the air mass originated from the Shanghai metropolitan area (by a prevailing northeasterly air mass), traveled a relatively short distance and crossed the boundary area between Jiangsu and Zhejiang, which contained high mass loading of anthropogenic emissions On Jan 15, an air mass coming from the East Sea also passed over the same area, but turned to the south and circled around Hangzhou Bay before arriving at the station The air mass on Jan 15 could favor more local pollutants being entrained and taken to the station The areas along the banks

of the Yangtze River and Hangzhou Bay have many industrial zones for manufacturing, steel production and chemical production, which were the main contributors of air pollutants (Huang et al.,

2011) On Jan 16 and 17, the air mass advected from the north of the station On Jan 17, an air mass originated from Mongolia, traveling a longer distance, which resulted in improvement of visibility in the afternoon As discussed above, the particles in this study period built up under stagnant meteorological conditions over the course of several days followed by quick removal associ-ated with the air mass from a clean region

Gaseous pollutants (SO2and NOx) and chemical components of non-refractory PM1(NH4þ, NO3, SO4 and Organics) are shown

inFig 8 However, chemical component data were not available after 18:00 on Jan 16 To segregate possible physical effects, such as evolution of boundary layer and transport, the gaseous pollutants and aerosol chemical components were normalized to the mass concentration of CO (Su et al., 2008) If the normalized values were

Fig 5 The time series of calculated extinction coefficientssext , 1/10 m m, dry/amb, cal and measured extinction coefficientssext , 10 m m, dry, mea (a), the calculated visibility (VIS cal ) and measured visibility (VIS mea ) with precipitation during this period (b), the ambient RH and the contribution of dry PM 1 to the ambient PM 10 extinction (c) during the one month measurement.

X.J Shen et al / Atmospheric Environment 120 (2015) 307e316 312

dramatically changed, this would indicate that the air masses could

be different Elevated concentrations of SO2and NOxwere observed

in the P1 and P2 episodes, indicating that emissions of SO2and NOx

were much more abundant in haze-fog episodes The P1 and P2

episodes had higher nighttime concentrations, presumably due to

continuous emissions at night, slower photochemical reaction

processes, and weaker vertical mixing Excluding the vertical

mixing effect, peaks of SO2and NOxwere observed approximately

18:00e20:00 on Jan 14 and 15, respectively

In the P1 episode, the maximum mass concentration of each

chemical composition, NH4 þ, NO

3 , SO

48, 20 and 39 mg m3, respectively, appearing at approximately

22:00 on Jan 14 At the start time of the P1 episode (16:00 on Jan

14), organics were the dominant species, with their contribution to

the sum of the non-refractory components being 40% During the

evolution of the heavy P1 episode, the contribution by the organics

enhanced, with the fraction changing from 18% to 37% While the

contribution of SO4 decreased, NH4þchanged only slightly The

results suggested that the nitrate and organics contributed most

significant during the P1 haze-fog episode They also revealed that

the mass concentration of major secondary aerosol compounds

increased to a different extent in the P1 episode The normalized

mass concentration of all of the species increased from 18:00 to

24:00 on Jan 14 This variation indicated that secondary aerosol

formation, including inorganic and organic aerosol, contributed to

the high mass loading and the particle growth

In the P2 episode, the PM1mass concentration reached an even

higher value, approximately 200mg m3 It can be observed that

during this episode the contribution of organics was even higher At

the beginning of the P2 episode, approximately 14:00, the decrease

in the mass concentration ratio of [organics]/[CO] and [SO4 ]/[CO] occurred, indicating an air mass change, which is consistent with the back trajectory analysis results It is probable that the air mass circling around Hangzhou and the nearby cities could enhance the air pollution The analysis above showed that different haze periods could be formed due to different types of aerosols and their for-mation processes For the P2 episode, the contribution by organics aerosols was more dominant

The result in the YRD region was different from that in the NCP region, where secondary sulfate played a more important role than nitrate under polluted conditions (Wiedensohler et al., 2009) and haze episodes (Wang et al., 2014b) The difference might be caused

by the various emissions sources in the YRD and the NCP Especially

in the cold season, coal burning for heating would be a major SO2 emitter in the NCP region In the YRD region, nitrate from industrial and traffic emissions would probably be a larger contributor to the fine particles

4 Conclusion

A consecutive series of haze-fog episodes occurred in a large area in the Yangtze River Delta in Jan 2013 During this period, PM1,

could contribute 84% to the mass concentration of PM10during the period Three types of conditions were classified based on visibility and RH: clear, haze and fog days The particle number size distri-bution on haze and fog days shifted to a larger size than on clear days The largest number and volume concentrations occurred

Fig 6 The wind vector (a), PNSD (b), RH and visibility (c), the GMD of the dominant mode of PNSD (d) and surface, volume concentration derived by PNSD for PM 1 (e) during the period Jan 14e17.

during haze conditions The extinction coefficient of PM1/PM10

under dry conditions was derived based on the Mie model, which

extinction under dry conditions Thus, the contribution of PM with

the assumption of dry state to the ambient extinction coefficient due to water uptake was evaluated as 85% during clear conditions, 58% during haze conditions and 41% during fog conditions

A typical heavy haze-fog event, occurring on Jan 14 to 17, was

Fig 7 The back trajectory (a) arriving at Lin'an station at different times from Jan 14 to 17 with the terminal height of 500 m agl and the running altitude (b).

Fig 8 The volume concentration of SO 2 and NO x (a), mass concentration of chemical components (c), as well as the normalized concentration of gases (b) and chemical com-ponents (d) by volume concentration of CO during the period of Jan 14 to Jan 16.

X.J Shen et al / Atmospheric Environment 120 (2015) 307e316 314

selected to illustrate the aerosol characteristics during formation

and evolution of haze-fog There were two separate episodes,

marked with P1 and P2, showing significant decrease of visibility

and increase of particle mass loading During these two episodes,

the geometric mean diameter of particle number size distribution

increased, with a growth rate of approximately 6.1 and 6.8 nm h1,

respectively The study also revealed that the growth process was

mainly due to the nitrate and organics components, driven by the

processes of secondary aerosol formation and air mass origin

Ni-trate was found to play a more important role in haze formation in

the YRD region, which is different from the NCP region where

sulfate was a larger contributor The difference suggests that the

dominant emission source should be different between the YRD

and the NCP regions Thus, a better understanding of the

relation-ships between secondary aerosol formation and haze-fog pollution

is crucial to the control of regional air quality for different region in

China

Acknowledgments

This work was supported by the National Natural Science

Foundation of China (41475118, 41405132), National Basic Research

Program of China (2011CB403401), National Science& Technology

Pillar Program (2014BAC16B01), key project of CAMS (2013Z007,

2013Y004) and International Science and Technology Cooperation

Project of China (2009DFA22800) It was also supported by the

CMA Innovation Team for Haze-fog Observation and Forecasts The

authors would also like to thank the Lin'an atmospheric

back-ground station staff for operating and maintaining the instruments

Appendix A Supplementary data

Supplementary data related to this article can be found athttp://

dx.doi.org/10.1016/j.atmosenv.2015.09.011

References

Anderson, T.L., Covert, D.S., Marshall, S.F., Laucks, M.L., Charlson, R.J., Waggoner, A.P.,

Ogren, J.A., Caldow, R., Holm, R.L., Quant, G., Sem, J., Wiedensohler, A.,

Ahlquist, N.A., Bates, T.S., 1996 Performance characteristics of a

High-sensitivity, three-wavelength total scatter/backscatter nephelometer J Atmos.

Ocean Technol 13, 967e986

Birmili, W., Wiedensohler, A., Heintzenberg, J., Lehmann, K., 2001 Atmospheric

particle number size distribution in central Europe: statistical relations to air

masses and meteorology J Geophys Res 106 (D23), 32005e32018

Bohren, C.F., Huffman, D.R., 1998 Absorption and Scattering of Light by Small

Par-ticles John Wiley, Hoboken, N J

Canagaratna, M., Jayne, J., Jimenez, J.L., Allan, J.A., Alfarra, R., Zhang, Q., Onasch, T.,

Drewnick, F., Coe, H., Middlebrook, A., Delia, A., Williams, L., Trimborn, A.,

Northway, M., Kolb, C., Davidovits, P., Worsnop, D., 2007 Chemical and

micro-physical characterization of aerosols via Aerosol mass Spectrometry Mass

Spectrom Rev 26, 185e222

Carrico, C.M., Bergin, M.H., Xu, J., Baumann, K., Maring, H., 2003 Urban aerosol

radiative properties: measurements during the 1999 Atlanta Supersite

Experi-ment J Geophys Res 108 (D7), 8422

Chan, C.K., Yao, X.H., 2008 Air pollution in mega cities in China Atmos Environ 42,

1e42

Che, H., Xia, X., Zhu, J., Li, Z., Dubovik, O., Holben, B., Goloub, P., Chen, H., Estelles, V.,

Cuevas-Agullo, E., Blarel, L., Wang, H., Zhao, H., Zhang, X., Wang, Y., Sun, J.,

Tao, R., Zhang, X., Shi, G., 2014 Column aerosol optical properties and aerosol

radiative forcing during a serious haze-fog month over North China Plain in

2013 based on ground-based sunphotometer measurements Atmos Chem.

Phys 14, 2125e2138 http://dx.doi.org/10.5194/acp-14-2125-2014

Chen, J., Zhao, C.S., Ma, N., Liu, P.F., G€obel, T., Hallbauer, E., Deng, Z.Z., Ran, L.,

Xu, W.Y., Liang, Z., Liu, H.J., Yan, P., Zhou, X.J., Wiedensohler, A., 2012.

A parameterization of low visibilities for hazy days in the North China Plain.

Atmos Chem Phys 12, 4935e4950

http://dx.doi.org/10.5194/acp-12-4935-2012

Cheng, Y.F., Eichler, H., Wiedensohler, A., Heintzenberg, J., Zhang, Y.H., Hu, M.,

Herrmann, H., Zeng, L.M., Liu, S., Gnauk, T., Brüggemann, E., He, L.Y., 2006.

Mixing state of elemental carbon and non-light absorbing aerosol components

derived from in situ particle optical properties at Xinken in Pearl River Delta of

Cheng, Y.F., Wiedensohler, A., Eichler, H., Heintzenberg, J., Tesche, M., Ansmann, A., Wendisch, M., Su, H., Althausen, D., Herrmann, H., Gnauk, T., Brüggemann, E.,

Hu, M., Zhang, Y.H., 2008 Relative humidity dependence of aerosol optical properties and direct radiative forcing in the surface boundary layer at Xinken

in Pearl River Delta of China: an observation based numerical study Atmos Environ 42, 6373e6397

Dal Maso, M., Kulmala, M., Riipinen, I., Wagner, R., Hussein, T., Aalto, P.P., Lehtinen, K.E.J., 2005 Formation and growth of fresh atmospheric aerosols: eight years of aerosol size distribution data from SMEAR II, Hyyti€al€a Finlal Boreal Environ Res 10, 323e336

Draxler, R.R., Rolph, G.D., 2003 HYSPLIT (HYbrid Single-particle Lagrangian Inte-grated Trajectory) NOAA Air Resources Laboratory, Silver Spring, MD Model access via NOAA ARL READY Website http://www.arl.noaa.gov/ready/hysplit4 html

Fu, Q.Y., Zhuang, G.S., Wang, J., Xu, C., Huang, K., Li, J., Hou, B., Lu, T., Streets, D.G.,

2008 Mechanism of formation of the Heaviest pollution episode ever recorded

in the yangtze river Delta, China Atmos Environ 42, 2023e2036

Geng, F.H., Zhang, Q., Tie, X.X., Huang, M.Y., Ma, X.C., Deng, Z.Z., Yu, Q., Quan, J.N., Zhao, C.S., 2009 Aircraft measurements of O 3 , NOx, CO, VOCs, and SO 2 in the Yangtze River Delta region Atmos Environ 43, 584e593

Huang, C., Chen, C.H., Li, L., Cheng, Z., Wang, H.L., Huang, H.Y., Streets, D.G., Wang, Y.J., Zhang, G.F., Chen, Y.R., 2011 Emission inventory of anthropogenic air pollutants and VOC species in the Yangtze River Delta region, China Atmos Chem Phys 11, 4105e4120

Huang, R.-J., Zhang, Y., Bozzetti, C., Ho, K.-F., Cao, J., Han, Y., D€allenbach, K.R., Slowik, J.G., Platt, S.M., Canonaco, F., Zotter, P., Wolf, R., Pieber, S.M., Bruns, E.A., Crippa, M., Ciarelli, G., Piazzalunga, A., Schwikowski, M., Abbaszade, G., Schnelle-Kreis, J., Zimmermann, R., An, Z., Szidat, S., Baltensperger, U., Haddad, I.E., Prev^ot, A.S.H., 2014a High secondary aerosol contribution to particulate pollution during haze events in China Nature 514, 218e222

Huang, K., Zhuang, G., Wang, Q., Fu, J.S., Lin, Y., Liu, T., Han, L., Deng, C., 2014b Extreme haze pollution in Beijing during January 2013: chemical characteris-tics, formation mechanism and role of fog processing Atmos Chem Phys Discuss 14, 7517e7556

Ivleva, N.P., Messerer, A., Yang, X., Niessner, R., P€oschl, U., 2007 Raman micro spectroscopic analysis of changes in the chemical structure and reactivity of soot in a diesel exhaust after treatment model system Environ Sci Technol 41, 3702e3707

Jansen, R.C., Shi, Y., Chen, J.M., Hu, W.J., Xu, C., Hong, S.M., Li, J., Zhang, M., 2014 Using hourly measurements to explore the role of secondary inorganic aerosol

in PM2.5 during haze and fog in Hangzhou, China Adv Atmos Sci 31, 1427e1434

Kotchenruther, R.A., Hobbs, P.V., 1998 Humidification factors of aerosols from biomass burning in Brazil J Geophys Res 103 (D24), 32,081e32, 090

Kulmala, M., Laaksonen, A., Charlson, R.J., Korhonen, P., 1997 Clouds without su-persaturation Nature 388, 336e337

Lin, W., Xu, X., Zhang, X., Tang, J., 2008 Contributions of pollutants from North China Plain to surface ozone at the Shangdianzi GAW station Atmos Chem Phys 8, 5889e5898

Ma, N., Zhao, C.S., Nowak, A., Müller, T., Pfeifer, S., Cheng, Y.F., Deng, Z.Z., Liu, P.F.,

Xu, W.Y., Ran, L., Yan, P., G€obel, T., Hallbauer, E., Mildenberger, K., Henning, S.,

Yu, J., Chen, L.L., Zhou, X.J., Stratmann, F., Wiedensohler, A., 2011 Aerosol optical properties in the North China Plain during HaChi campaign: an in-situ optical closure study Atmos Chem Phys 11, 5959e5973 http://dx.doi.org/10.5194/ acp-11-5959-2011

Ma, N., Zhao, C.S., Müller, T., Cheng, Y.F., Liu, P.F., Deng, Z.Z., Xu, W.Y., Ran, L., Nekat, B., van Pinxteren, D., Gnauk, T., Müller, K., Herrmann, H., Yan, P., Zhou, X.J., Wiedensohler, A., 2012 A new method to determine the mixing state

of light absorbing carbonaceous using the measured aerosol optical properties and number size distributions Atmos Chem Phys 12, 2381e2397

Pan, L., Che, H.Z., Geng, F.H., Xia, X., Wang, Y.Q., Zhu, C.Z., Chen, M., Gao, W., Guo, J.P.,

2010 Aerosol optical properties based on ground measurements over the Chinese Yangtze Delta Region Atmos Environ 44, 2587e2596

Petzold, A., Sch€onlinner, M., 2004 Multi-angle absorption photometry-a new method for the measurement of aerosol light absorption and atmospheric black carbon J Aerosol Sci 35, 421e441

Riemer, N., Vogel, H., Vogel, B., 2004 Soot aging time scales in polluted regions during day and night Atmos Chem Phys 4, 1885e1893

Saathoff, H., Naumann, K.H., Schnaiter, M., Sch€ock, W., M€ohler, O., Schurath, U., Weingartner, E., Gysel, M., Baltensperger, U., 2003 Coating of soot and (NH 4 ) 2 SO 4 particles by ozonolysis products of a-pinene J Aerosol Sci 34, 1297e1321

Schichtel, B.A., Husar, R.B., Falke, S.R., Wilson, W.E., 2001 Haze trends over the United States 1980e1995 Atmos Environ 35, 5205e5210

Seinfeld, J., Pandis, S., 1998 Atmospheric Chemistry and Physics From Air Pollution

to Climate Change, first ed John Wiley & Son, Inc, Hoboken, New Jersey

Su, H., Cheng, Y.F., Cheng, P., Zhang, Y.H., Dong, S.F., Zeng, L.M., Wang, X.S., Slanina, J., Shao, M., Wiedensohler, A., 2008 Observation of nighttime nitrous acid (HONO) formation at a non-urban site during PRIDE-PRD 2004 in China Atmos Environ.

42, 6219e6232

Sun, J., Zhang, Q., Canagaratna, M.R., Zhang, Y., Ng, N.L., Sun, Y., Jayne, J.T., Zhang, X., Zhang, X., Worsnop, D.R., 2010 Highly time- and size-resolved characterization

of submicron aerosol particles in Beijing using an aerodyne aerosol mass spectrometer Atmos Environ 44, 131e140

Tuch, T.M., 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 Technol 2, 417e422

Wang, Y.S., Yao, L., Wang, L.L., Liu, Z.R., Ji, D.S., Tang, G.Q., Zhang, J.K., Sun, Y., Hu, B.,

Xin, J.Y., 2014a Mechanism for the formation of the January 2013 heavy haze

pollution episode over central and eastern China Sci China Earth Sci 57, 14e25.

http://dx.doi.org/10.1007/s11430-013-4773-4

Wang, H.L., An, J.L., Shen, L.J., Zhu, B., Pan, C., Liu, Z.R., Liu, X.H., Duan, Q., Liu, X.,

Wang, Y.S., 2014b Mechanism for the formation and microphysical

character-istics of submicron aerosol during heavy haze pollution episode in the Yangtze

River Delta, China Sci Total Environ 490, 501e508

Wang, T., Cheung, V.T.F., Li, Y.S., Anson, M., 2001 Ozone and related gaseous

pol-lutants in the boundary layer of eastern China: overview of the recent

mea-surements at a rural site Geophys Res Lett 28, 2373e2376

Wiedensohler, A., Cheng, Y.F., Nowak, A., Wehner, B., Achtert, P., Berghof, M.,

Birmili, W., Wu, Z.J., Hu, M., Zhu, T., Takegawa, N., Kita, K., Kondo, Y., Lou, S.R.,

Hofzumahaus, A., Holland, F., Wahner, A., Gunthe, S.S., Rose, D., Su, H., P€oschl, U.,

2009 Rapid aerosol particle growth and increase of cloud condensation nucleus

activity by secondary aerosol formation and condensation: a case study for

regional air pollution in northeastern China J Geophys Res 114, D00G08.

http://dx.doi.org/10.1029/2008JD010884

Wiedensohler, A., Birmili, W., Nowak, A., Sonntag, A., Weinhold, K., Merkel, M.,

Wehner, B., Tuch, T., Pfeifer, S., Fiebig, M., Fj€araa, A.M., Asmi, E., Sellegri, K.,

Depuy, R., Venzac, H., Villani, P., Laj, P., Aalto, P., Ogren, J.A., Swietlicki, E.,

Williams, P., Roldin, P., Quincey, P., Hüglin, C., Fierz-Schmidhauser, R., Gysel, M., Weingartner, E., Riccobono, F., Santos, S., Grüning, C., Faloon, K., Beddows, D., Harrison, R., Monahan, C., Jennings, S.G., O'Dowd, C.D., Marinoni, A., Horn, H.G., Keck, L., Jiang, J., Scheckman, J., McMurry, P.H., Deng, Z., Zhao, C.S., Moerman, M., Henzing, B., de Leeuw, G., L€oschau, G., Bastian, S., 2012 Mobility particle size spectrometers: harmonization of technical standards and data structure to facilitate high quality long-term observations of atmospheric par-ticle number size distributions Atmos Meas Technol 5, 657e685 http:// dx.doi.org/10.5194/amt-5-657-2012 , 2012.

WMO-No782, 2005 Aerodrome Reports and Forecasts: a User's Handbook to the Codes

Zhang, X.Y., Wang, Y.Q., Niu, T., Zhang, X.C., Gong, S.L., Zhang, Y.M., Sun, J.Y., 2012 Atmospheric aerosol compositions in China: spatial/temporal variability, chemical signature, regional haze distribution and comparisons with global aerosols Atmos Chem Phys 12, 779e799 http://dx.doi.org/10.5194/acp-12-779-2012 , 2012.

Zhang, L., Sun, J.Y., Shen, X.J., Zhang, Y.M., Che, H., Ma, Q.L., Zhang, Y.W., Zhang, X.Y., Ogren, J.A., 2015 Observations of relative humidity effects on aerosol light scattering in the Yangtze River Delta of China Atmos Chem Phys 15, 8439e8454 http://dx.doi.org/10.5194/acp-15-8439-2015

Zhang, R.H., Li, Q., Zhang, R.N., 2014 Meteorological conditions for the persistent severe fog and haze event over eastern China in January 2013 Sci China Earth Sci 57, 26e35 http://dx.doi.org/10.1007/s11430-013-4774-3

X.J Shen et al / Atmospheric Environment 120 (2015) 307e316 316