Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region

JES-00836; No of Pages 13 J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/jes

3Q1

Weiwei Chen1,⁎, Daniel Q. Tong2,3 , Mo Dan4 , ShiChun Zhang1 , XueLei Zhang1 , YuePeng Pan5

6 7 8 9 10

O

1. Key Laboratory of Wetland Ecology and Environment,Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun 130102, China 2. Cooperative Institute for Climate & Satellites, University of Maryland, College Park, MD 20740, USA 3. Center for Spatial Information Science and Systems, George Mason University, Fairfax, VA 22030, USA 4. Beijing Municipal Institute of Labor Protection, Beijing 100054, China 5. State key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China

R O

5

P

4

F

2

Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region

1

D

11

AR TIC LE I NFO

ABSTR ACT

15

Article history:

This study presents the mass concentrations of PM2.5, O3, SO2 and NOx at one urban, one 20

16

Received 24 September 2015

suburban and two rural locations in the Changchun region from September 25 to October 27 21

17

Revised 5 March 2016

2013. Major chemical components of PM2.5 at the four sites were daily sampled and 22

18

Accepted 18 March 2016

analyzed. Most of daily concentrations of SO2 (7–82 μg/m3), O3 (27–171 μg/m3) and NOx 23

19

Available online xxxx

(14–213 μg/m3) were below the limits of the National Ambient Air Quality Standard (NAAQS) 24

41

Keywords:

Higher PM2.5 concentrations (~150 μg/m3) were measured during the pre-harvest and 26

42

Aerosol

harvest at the urban site, while PM2.5 concentrations significantly increased from 250 to 27

43

Air quality

400 μg m−3 at suburban and rural sites with widespread biomass burning. At all sites, PM2.5 28

44

Agriculture

45

Biomass burning

46

PM2.5

47

PMF

C

T

E

14 13

components were dominated by organic carbon (OC) and followed by soluble component 29

R

+ − sulfate (SO2− 4 ), ammonium (NH4) and nitrate (NO3). Compared with rural sites, urban site 30

had a higher mineral contribution and lower potassium (K+ and K) contribution to PM2.5. 31 Severe atmospheric haze events that occurred from October 21 to 23 were attributed to 32

N C O

48

strong source emissions (e.g., biomass burning) and unfavorable air diffusion conditions. 33 Furthermore, coal burning originating from winter heating supply beginning on October 18 34 increased the atmospheric pollutant emissions. For entire crop harvest period, the Positive 35 Matrix Factorization (PMF) analysis indicated five important emission contributors in the 36 Changchun region, as follows: secondary aerosol (39%), biomass burning (20%), supply 37 © 2016 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences.

54 53

Introduction

55

Fine particulate matter (PM2.5), ozone (O3), sulfur dioxide (SO2) and nitrogen oxide (NOx) have been generally recognized as important atmospheric pollutants that most significantly

56 57

38

heating (18%), soil/road dust (14%) and traffic (9%).

U

39 40 51 50 49 52

R

E

in China. However, PM2.5 concentrations (143–168 μg/m3) were 2-fold higher than NAAQS. 25

Published by Elsevier B.V.

affect human health and visibility (Heal et al., 2012). In China, increasing public concern about air quality and haze events requires accurate emission source quantifications, detailed chemical components and an in-depth understanding of the transport and photochemical reaction processes of

⁎ Corresponding author. E-mail: [email protected] (Weiwei Chen).

http://dx.doi.org/10.1016/j.jes.2016.03.031 1001-0742/© 2016 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

58 59 60 61 62

2

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

123 Q7

1. Materials and methods

145 144

1.1. Study region

146

P

R O

O

F

particulate matter production (Holmén et al., 2008). Additionally, coal burning for the heating supply in cities and biomass burning in most villages increase the amount of atmospheric pollutants during the wintertime in northeastern China (Cao et al., 2012a, 2012b). Thus, the possibility of haze events may greatly increase with the combination of crop harvesting and winter heating. This study presents the atmospheric concentrations of fine particulate matter (PM2.5) and gaseous pollutants (i.e., O3, SO2 and NOx) from September 25 to October 27, 2013, in the Changchun region, Jilin Province, northeastern China. Using portable real-time PM2.5 analyzers, we measured the hourly PM2.5 concentration at one urban, one suburban and two rural locations in the region. Gaseous pollutant concentrations were determined by real-time gas analyzers established at the suburban site. Daily sampling of PM2.5, except for rain days, and the major chemical component analysis of PM2.5 at the four sites were also conducted. Our primary objectives were to gain insight into the atmospheric pollutants during the harvest season and to determine the cause of the severe haze events in northeastern China.

D

atmospheric pollutants to develop suitable and sustainable mitigation strategies (Chan and Yao, 2008; Cheng et al., 2013; 65 Pui et al., 2014). However, the complexity of chemical 66 components in atmospheric PM2.5, spatio-temporal distribu67 tions as well as heterogeneous reactions with gases remains 68 Q2 unclear in different regions (R. Zhang et al., 2012). Therefore, 69 long-term and systemic studies are required to elucidate the 70 air pollution levels, possible emission sources and transport 71 processes at specified regions. 72 The frequency of regional atmospheric haze events in73 creased rapidly along with the fast-paced urbanization and 74 increasing energy consumption over the past years (Kan et al., 75 2012). Previous studies have shown that haze primarily occurs 76 in four regions: the Bejing-Tianjin-Hebei zone, the Yangtze 77 Delta, the Pearl River Delta and the Sichuan Province (e.g., Cao 78 et al., 2012a, 2012b; X.Y. Zhang et al., 2012; Wang et al., 2013a, Q6 Q5 Q4 79 Q3 2013b; Li et al., 2014; Ji et al., 2012); thus, a considerable 80 number of studies has been ongoing in these regions. In 81 addition, atmospheric haze problems have increased in 82 northeastern China under the national five-year development 83 plan in the region; thus, this area may become the fifth severe 84 haze region. An extreme haze event from October 20 to 85 October 23 2013, affected the public over an expansive area 86 (> 1 million km2), nearly covering the entire region and 87 showing an ultra-high PM2.5 concentration in some locations 88 (> 1000 μg/m3) (Xinhua News Agency, 2013). However, air 89 quality studies in this region have been limited to a few cities 90 and to basic chemical component analysis (e.g., Han et al., 91 2010; Huang et al., 2011). More information about emission 92 patterns and the chemical composition of the atmospheric 93 pollutants is necessary in this region due to the importance of 94 identifying emission sources and verifying an air quality 95 model (Heal et al., 2012). Furthermore, studies on the issues 96 involving emission inventory, the mechanism of haze events 97 and a system of forecasting and warning by numerical 98 simulation in northeastern China have been lacking. 99 Agricultural emission is crucial to regional air quality in 100 northeastern China due to intense activity and the vast 101 farmlands, which occupy approximately 30% of the total 102 land area in northeastern China (China Agricultural Yearbook, 103 2012). Particulate matter and gaseous pollutants emitted from 104 agricultural operations may initially change the air quality in 105 a rural area and then affect adjacent towns/cities by diffusion. 106 The air quality in agricultural regions is closely linked to the 107 local weather, soil properties, crop types and field operations 108 as well as the living habits in village (Hinz and 109 Tamoschat-Depolt, 2007). The primary particles mainly orig110 inate from operations, such as soil tillage, fertilization, 111 application of chemical substances, crop cutting and process112 ing, straw burning and animal feed, while ammonia, biogenic 113 VOC, and other chemical substances from these operations 114 could generate secondary fine particles by homogeneous or 115 heterogeneous reactions (Aneja et al., 2009). In many agricul116 tural regions in China, spring plowing, fall harvesting and, 117 especially biomass burning are responsible for regional haze 118 events (Zhang et al., 2010; Qin and Xie, 2011; Liu et al., 2014). 119 Similarly, most straw residues are open burned to be 120 incorporated into the soil for farming season and to reduce 121 the cost of recycling in northeastern China. Mechanical 122 harvest and land preparation after burning also contribute to

C

E

R

R

O

C

N

U

124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143

This study was conducted in the Changchun region (43°05′N– 43°05′N, 124°18′E–127°05′E, 250–350 m a.s.l), which is the capital of Jilin Province and the natural geographical center in northeastern China (Bureau of Statistics of Jilin, 2007). The local climate is characterized as temperate continental monsoon climate. The mean annual temperature is 4.8°C, with a mean January temperature of −15.1°C and a July mean of 23.1°C. The annual precipitation is 522–615 mm, with more than 60% falling in the summer from June to August. Changchun city is an industrial city, having the largest automobile manufacture enterprise in China, whereas there are also vast farmlands around the city in the rural areas. Maize is the dominant crop, accounting for 58% of total crop areas in Jilin province. Generally, maize is planted in early May and harvested in October. During the crop harvest, three stages are divided by different operations. The first stage is pre-harvest (i.e., farm machinery preparation), the second stage is the reaping of maize with small straw burning (i.e., harvest) and the third stage is characterized by strong straw burning and small land preparation after harvest (i.e., post-harvest). Straw burning is primarily implemented in the second half of October. In urban areas, wintertime heating normally begins October 15, and the heat is formally supplied around October 25 in the Changchun region.

147

1.2. Experimental design

171

To elucidate the local air pollution characteristics, the possible emission sources and the role of agriculture in the Changchun region, four sampling sites (i.e., one urban, one suburban and two rural) and four major atmospheric pollutants (i.e., PM2.5, O3, NOx and SO2) were monitored. As shown in Fig. 1, the urban site is located in the Liaoyang residential

172

T

64

E

63

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

148 149 150 Q8 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170

173 174 175 176 177

3

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

support frame) at the urban, suburban and rural sites, 210 respectively. 211

1.3. Real-time measurements of gaseous pollutants and PM2.5 212 213

1.4. Sampling and chemical component analysis of PM2.5

242

Atmospheric PM2.5 was sampled every one or two days using a filter-based gravimetric sampling method. Portable samplers (Model Omni, BGI Inc., USA) were used with a flow-rate of 5 L/min on a 24-hr basis. There were two types of Whatman filters (i.e., Teflon and Quartz, 47 mm) that were applied to collect atmospheric particles. The mass concentrations of PM2.5 were calculated by dividing the weight increase in the filter by the standard sampling air volume, which was converted using the actual air volumes and the periodic air temperature and pressure. These filters were weighted on an electronic microbalance with a precision of 0.01 mg (Model XS105DU, Mettler Toledo Inc., Switzerland). Before and after filter sampling, these filters were stored in a dessicator at 20–25°C and 35%–45% in relative humidity for 48 hr. Subsequently, the sampled filters were stored in a refrigerator at 4°C until chemical component analysis. Teflon filters were used to measure the ionic speciation, + + including anions (i.e., F−, Cl−, NO−3 and SO2− 4 ), cations (Na , NH4, + 2+ 2+ K , Mg and Ca ), and inorganic elements (i.e., Al, Si, Ca, Fe, Mg, K, Mn, Ni, Cu, Zn, As, Se, Sr, Ba, Cd, Cr, Nd, and Pb). The concentrations of anions and cations were determined by ion chromatography (ICS-1000, Dionex Inc., USA). The eluent used for anion was a 3.5-mmol/L Na2CO3/1.0-mmol/L NaHCO3 solution, whereas a 20-mmol/L methane sulfonic acid (MSA)

243

R O

O

F

Atmospheric PM2.5 concentrations were determined in real-time using a Model 8520 DUSTTRAK™ Aerosol Monitor (TSI Incorporated, Shoreview, MN, USA), which uses light scattering technology (MacIntosh et al., 2002). An aerosol sample is drawn into the sensing chamber in a continuous stream, and particles in the aerosol stream scatter light in all directions. A laser beam collects some of the scattered light and focuses it onto a photodetector. The detection circuitry converts the light into a voltage, and finally, this voltage is proportional to the amount of light scattered, which is proportional to the mass concentration of the aerosol. The portability and low-power-supply requirement of this monitoring method ensures that the PM2.5 determination could be performed simultaneously at the four sites, especially at rural sites without good infrastructure. The atmospheric concentrations of O3, NOx and SO2 were detected using a Model 49C O3 analyzer, a Model 42CTL NO and NO2 analyzer and a Model 43CTL SO2 analyzer, respectively (Ji et al., 2012). The NOx concentration is the sum of NO and NO2. These analyzers were manufactured by Thermo Environmental Instruments (TEI), Inc. The detection limit and precision of Model 49C were 1 ppbv. The Model 42CTL had a detection limit of 0.4 ppbv with a precision of 0.05 ppbv. The Model 43CTL exhibited a detection limit of 0.1 ppbv and a precision of 1 ppbv. Meteorological data were obtained from the Changchun Meteorological Bureau, which includes daily precipitation, air temperature, relative humidity, visibility, wind speed and direction.

183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 Q9 206 207 208 209

C

E

182

R

181

R

180

N C O

179

communities (43°48′38′′N, 125°14′51′′E) in the center of Changchun city, which is primarily characterized by a dense population. The source of atmospheric pollutants at this site might be greatly derived from daily life, vehicle operation, and road or construction dust, as well as the nearby automobile manufacturers (~7 km). The suburban site is situated at the Northeast Institute of Geography and Agro-ecology (IGA, 43°59′41′′N, 125°24′00′′E), Chinese Academy of Sciences, and is approximately 25 km apart from the selected urban site. The IGA site is on the edge of Changchun city and this district is constructed from farmland since 2010. The potential emission sources for atmospheric pollutants include low-density factories, such as a refuse processing plant, coal burning from scattered dwellings and constructing buildings. The first rural site (i.e., rural_1) is a black-soil protection station (44°12′29′′N, 125°34′04′′E) that is attached to the IGA. It is approximately 50 km away from the urban site. The second rural site (i.e., rural_2) is in a village (44°51′27′′N, 126°24′45′′E) that is 150 km from the urban site. Agriculture-related emissions of atmospheric pollutants might drive the local air quality at this site. We simultaneously measured the atmospheric PM2.5 concentration every 5 min at four sites in from September 25 to October 27 2013. At the suburban site, the concentrations of gaseous pollutants (i.e., O3, NOx and SO2) were also detected with high-frequency signals in five-second intervals during the crop harvest. Furthermore, atmospheric PM2.5 was sampled daily, except for rainy days, at the four sites, and these samples were sent to a lab for chemical component analysis. The heights of sampling and real-time measurements are approximately 12 m (i.e., the roof of a two-floor building), 6 m (i.e., the roof of a one-floor building) and 3 m (i.e., the top of a

U

178

T

E

D

P

Fig. 1 – Locations of the sampling sites in the Changchun region, Jilin Province, China. The capitals of three provinces (Liaoning, Jilin and Heilongjiang) in northeast China are labeled in gray circles, i.e., Shenyang (SY), Changchun (CC) and Haerbin (HEB). The gray triangles indicate the four sampling sites distributed in an urban, a suburban and two rural sites.

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241

244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265

4

280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301

305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323

C

1.6. Data analysis

331

Daily O3 concentrations were calculated by the average of a maximum of 8 hr on the given days. Hourly atmospheric pollutants, wind direction and wind speed were collected to analyze their relationships. The significance of the differences in the PM2.5 concentrations among the three stages for the specified site and among the different sites was investigated using the paired-samples t test. Linear or non-linear regression analysis was used to evaluate the relationships between atmospheric pollutants, meteorological data and chemical species. All of the statistical procedures were performed using the software packages SigmaPlot 10.0 (SPSS Inc., Chicago, USA) and R-packages (e.g., base function and open-air project).

332

F

279

E

278

R

277

R

276

O

275

C

274

N

273

U

272

O

The version of apportionment model (EPA PMF 3.0, Norris et al., 2014) is used to analyze the potential emission sources of airborne PM2.5 during the harvest season in the Changchun region. The chemical species at the four sites were assembled as input data, and weak species or bad species were modulated by considering the input data statistics. The data uncertainty was calculated by taking into account of analytical error and sampling error. PMF results can be greatly influenced by species with a high proportion of data lower than the minimum detection limits, thus each variable must be individually evaluated according to the number of cases that its value is lower than the detection limit. In PMF 3.0 the variables can be characterized as “bad”, “weak” or “strong”. A variable is defined to be weak if its S/N is between 0.2 and 2. If S/N < 0.2, then a variable is defined as bad. In this study, Ni, Cu, As, Sr, Cd, Cr and Nd were set as “bad” and thus were exclude from the analysis. The PM2.5 concentration was set as total variable. A major consideration in searching for the PMF solution is to find the best number of factors to fit the dataset. It should be noted that the choice of factors in PMF is always a

271

R O

304

270

P

1.5. Source apportionment

269

324

D

303

268

compromise because PMF is a descriptive model and there is no objective criterion to choose the ideal solution. We selected the numbers of factors (e.g., 3–10) and executed the base model runs. After a considerable amount of testing and by considering the actual emitters on regional scale, the main factors were finally determined with the reasonable emissions based on the PMF results.

2. Results 2.1. Meteorological conditions during harvest season

325 326 327 328 329 330

333 334 335 336 337 338 339 340 341 342 343

345 344 346

There is a significant decrease in air visibility during the harvest season (y = 12–0.6 ×, r2 < 0.84, p < 0.001) from 2004 to 2013 based on the meteorological data during this season. The temporal trend suggests the gradual degradation of air quality in the Changchun region, with the worst year being 2013 (7.1 km). In October 2013, the air temperature decreased to the freezing point, and three rainfall events occurred on October 1, 13 and 24 (Fig. 2a, b). A southwesterly wind with a low speed below 3 m/sec prevailed during the harvest season (Fig. 2b). The air visibility generally fluctuated from 5 km to 10 km for the entire harvest season. However, the visibility significantly decreased from October 17 to 24 (< 5 km), and the lowest value occurred around October 22 (<1 km).

347

2.2. Temporal variations of gas pollutants and PM2.5

361

2.2.1. Daily variations

362

The daily atmospheric concentrations of SO2, O3 (average of a maximum of 8 hr) and NOx at the suburban site ranged from 7 to 82 μg/m3, from 27 to 171 μg/m3 and from 1 to 116 μg/m3 during the sampling period (i.e., September 24 to October 31), respectively (Fig. 2c). The temporal coefficients of variation (CV) of SO2, O3 and NOx were 65%, 49% and 58% at the suburban site, respectively. Daily average mass concentrations were 26 μg/m3 for SO2, 71 μg/m3 for O3 and 74 μg/m3 for NOx during the harvest season. The daily SO2 concentration fluctuated over a narrow range below 50 μg/m3 until October 23, and then increased to the peak at the end of October. The O3 concentrations exhibited large fluctuations and high concentrations at the first half of the month and exhibited small variations at the second half of the month. The daily

363

T

302

solution was used for the cation eluent. The ion chromatography method had a detection limit of 0.05 mg/L and an uncertainty of ±10% for all of the ions. For the inorganic elements, the Teflon filter was extracted for half an hour by 6 ml HNO3 and 2 ml HCL in a microwave laboratory system with 1400 W in power, 170°C in maximal temperature and 20 bar in ultimate pressure before determination. Then, the extracted solution was injected into an inductively coupled plasma-atomic emission spectrometer (ICP-AESIRIS, Intrepid II, Thermo Electron, USA) to obtain the element concentrations. The precision and bias of the element concentrations of the ICP-AESIRIS method were typically less than 10%. The concentration of mineral dust in PM2.5 was calculated by summing the content of the oxides of Al, Si, Ca, Fe, Mg and K, (i.e., 1.89 Al + 2.14 Si + 1.40 Ca + 1.43 Fe + 1.66 Mg + 1.21 K) (Hueglin et al., 2005). The Si concentration was estimated according to the average ratio of Si/Al (3.6) in the earth's crust (Hueglin et al., 2005) because Si was not determined the by ICP-AESIRIS method in this study. Quartz filters were used to determine the particulate EC and OC concentrations using a thermal-optical carbon aerosol analyzer (Sunset-OCEC RT-4, Sunset Lab Inc., Tigard, OR, USA) (Aneja et al., 2006). This method is based on the thermal desorption/oxidation of particulate carbon to CO2, which is then reduced to methane and is subsequently measured using a flame-ionization detector. The analysis sequence was initialized in a nonoxidizing atmosphere (helium) with a 10-sec purge followed by four temperature ramps to a maximum of 900°C. A cooling blower was then used, and the temperature dropped to 600°C before oxygen was added. The temperature was held at this point until the transmittance or reflectance returned to the initial point before the sample was heated. This point determines the distinction between OC and EC, so that all of the carbon measured up to this point is OC, whereas all of the carbon measured after this point is EC. The total carbon is equal to EC plus OC. The precision is 0.19 at 1 μg of carbon and 0.01 at 10–72 μg of carbon.

267

E

266

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

348 349 350 351 352 353 354 355 356 357 358 359 360

364 365 366 367 368 369 370 371 372 373 374 375 376

5

N C O

R

R

E

C

T

E

D

P

R O

O

F

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

U

Fig. 2 – Temporal variations in air temperature, relative humidity (a), wind speed, wind direction and air visibility (b), major gas pollutants (O3, SO2 and NOx) concentrations (c), atmospheric PM2.5 concentrations measured by the online portable Dustrak 8520 II analyzer (d) and those measured by manual Teflon or Quartz films (e) sampling method at the urban, suburban and rural sites. The gas pollutants were only determined at the suburban site by established online gas analyzers. Each PM2.5 concentration point represents the daily mean value at the site. The hatched background marks three continuous rainfall events in October 1, 13 and 24, 2013, respectively. PM2.5: particulate matter.

377 378 379 380 381 382 383

pattern of NOx concentrations was opposite to that of O3, showing high NOx levels with low O3 levels (Fig. 2c). The real-time and filter-based measurements showed a significant increasing trend of atmospheric PM2.5 concentrations at the four sites with the advancement of the harvesting process (Fig. 2d, e). The range of PM2.5 values at the urban site was between 26 and 408 μg/m3 with a CV of 58%. The suburban

site had a similar PM2.5 level range (28–373 μg/m3) and CV (58%) to the urban site. However, there were larger variations of the PM2.5 concentrations (11–503 μg/m3) and a higher CV value of 85% at the rural sites compared with the urban and suburban sites. The daily mean PM2.5 concentrations were 148 ± 15, 142 ± 14, 168 ± 22 and 146 ± 23 μg/m3 at the urban, suburban, rural_1 and rural_2 sites, respectively (Table 1).

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

384 385 386 387 388 389 390

6

t1:15 t1:16 t1:17 t1:19 t1:18

All sites

Rural_2

Rural_1

Urban

t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14

Suburban

196 ± 28 (46–381) 162 ± 30 (65–488) 204 ± 53 (34–590) 150 ± 28 (45–270) 167 ± 13 (34–590) 148 ± 15 (25–408) 143 ± 14 (28–373) 168 ± 22 (18–503) 146 ± 23 (11–499) 151 ± 6 (11–499)

U

N

5.6 ± 1.2 (0.2–12.8) 9.4 ± 2.2 (0.8–31.4) 10.8 ± 3.2 (0.5–41.0) 7.1 ± 2.0 (0.6–18.0) 8.2 ± 1.2 (0.2–41.0)

C

O

R

2.3 ± 0.5 (0.3–5.9) 2.2 ± 0.5 (0.3–6.5) 3.3 ± 0.9 (0.4–10.3) 2.8 ± 0.3 (0.8–6.2) 2.7 ± 0.3 (0.3–10.3) 4.0 ± 0.5a (1.3–6.8) 1.6 ± 0.3b (0.1–3.5) 0.2 ± 0.1c (0–1.5) 0.4 ± 0.1c (0.1–0.9) 1.6 ± 0.9 (0–6.8)

R

E

391

The SO2 concentrations had a typical bimodal pattern and showed two peaks appearing at 9:00 and 20:00 (Fig. 3a). The diurnal variation of O3 was unimodal, with the highest value occurring at 14:00–15:00 local time, while this period was the lowest for the NOx concentrations (Fig. 3b). Significantly higher NOx concentrations were observed in the nighttime than in the daytime. The diurnal variation of the PM2.5 concentrations was consistent at the urban, suburban and rural sites (Fig. 3c). The PM2.5 pattern showed that the PM2.5 concentrations were lowest at 13:00–14:00, rapidly increased to the peak at 19:00–23:00, and gradually decreased until 13:00–14:00 the following day. Furthermore, an increase in the PM2.5 concentration was also determined distinctly from 7:00–8:00. Compared with the urban and suburban sites, significantly higher PM2.5 concentrations occurred from 16:00 to 22:00, while they were lower for other periods.

392

2.2.3. Relationships of atmospheric pollutants and wind direction and speed

408

High NOx concentrations primarily occurred with a low wind speed of < 2 m/sec (Fig. 4a). The SO2 concentrations displayed significantly higher levels with southwesterly winds, regardless of the wind speed (Fig. 4b). However, O3 concentrations greatly depended on the southwesterly wind and a strong speed of >6 m/sec (Fig. 4c). In addition, a higher ratio of SO2 and NOx was calculated with i southwest and northwest wind directions with wind speeds of > 4 m/sec (Fig. 4d). The relationships among PM2.5 concentrations and wind direction and speed were similar at the four sites (Fig. 4e–h). A PM2.5 level above 150 μg/m3 was distributed by a wind speed below 4 m/sec accompanied by easterly wind. At the urban and suburban sites, slightly lower ranges of PM2.5 concentrations (100–150 μg/m3) were observed in southwesterly wind. Moreover, southwesterly winds with high wind speeds (>8 m/sec) could also cause high PM2.5 concentrations (150–200 μg/m3) (Fig. 4f).

410

E

D

P

R O

O

F

2.2.2. Diurnal variations

T

C

The figures outside the parentheses are the daily mean values over the entire sampling period and the standard error, and those inside the parentheses are their ranges. PM2.5_O and PM2.5_F are the atmospheric PM2.5 concentrations measured by the online portable Dustrak 8520 II analyzer and manual Teflon or Quartz films, respectively. Mineral, TE, OC and EC refer to is the mineral dust, trace elements, organic carbon and elemental carbon in atmospheric PM2.5 samples, respectively. The different superscript case letter indicates a significant difference in mineral dust and Ca2+ at p < 0.01. PM2.5: particulate matter.

6.7 ± 0.8 (3.0–11.4) 8.6 ± 1.6 (3.1–23.1) 11.4 ± 3.2 (1.5–40.6) 7.6 ± 1.6 (1.7–13.1) 8.6 ± 1.0 (1.5–40.6) 66.0 ± 12.0 (25.3–195.4) 78.9 ± 16.8 (27.5–263.9) 110.0 ± 28.9 (13.2–332.1) 76.9 ± 17.4 (15.2–144.2) 83.0 ± 9.5 (13.2–332.1) 1.3 ± 0.3 (0.5–3.9) 1.4 ± 0.4 (0.1–5.0) 0.2 ± 0.1 (0–0.4) 1.8 ± 0.5 (0.4–3.9) 1.2 ± 0.3 (0.1–5.0) 65.2 ± 8.5a (26.2–111.6) 12.7 ± 2.7b (2.9–36.1) 5.5 ± 1.0c (1.7–13.0) 7.6 ± 0.9c (3.6–11.2) 22.8 ± 14.2 (1.7–111.6) 4.3 ± 1.1 (0.9–10.7) 5.2 ± 1.2 (0.6–14.2) 9.6 ± 3.0 (0.4–38.8) 8.3 ± 1.6 (2.9–17.3) 6.9 ± 1.3 (0.4–38.8) 13.3 ± 3.1 (1.6–37.0) 16.3 ± 4.3 (2.6–60.6) 11.5 ± 3.5 (0–47.5) 6.4 ± 1.5 (2.5–15.6) 11.9 ± 2.1 (0–60.6) 9.6 ± 1.7 (0–19.3) 14.3 ± 3.3 (1.8–45.9) 11.5 ± 3.5 (0–47.5) 6.4 ± 1.5 (2.5–15.6) 10.4 ± 1.7 (0–47.5) 0.3 ± 0.1 (0.1–0.4) 0.3 ± 0.1 (0.1–0.5) 0.1 ± 0.1 (0–0.3) 0.1 ± 0.1 (0–0.3) 0.2 ± 0.1 (0–0.5)

EC (μg/m3) OC (μg/m3) TE (μg/m−3) Mineral (μg/m3) Cl− (μg/m3) NO−3 (μg/m3) SO2− 4 (μg/m3) Na+ (μg/m3) K+ (μg/m3) Ca2+ (μg/m3) NH+4 (μg/m3)

PM2.5_F (μg/m3) PM2.5_O (μg/m3) Site t1:4

t1:3

t1:1 t1:2

Table 1 – Atmospheric PM2.5 concentrations and major chemical components in the sampled PM2.5 from urban, suburban and rural sites in the Changchun region during the crop harvest from September 25 to October 27 2013.

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

Fig. 3 – Diurnal variations of atmospheric pollutant concentrations of SO2 (a), O3 and NOx (b), PM2.5 (c) at the sampled sites during crop harvest season. PM2.5: particulate matter.

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

393 394 395 396 397 398 399 400 401 402 403 404 405 406 407

409 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425

7

E

D

P

R O

O

F

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454

E

R

429

Detailed information of the PM2.5 concentrations and the corresponding chemical components is summarized in Table 1. Based on the filter method, daily average PM2.5 concentration at the four sites during the harvest season was 167 ± 13 μg/m3, in which OC was the most abundant chemical component, with daily averaged mass concentrations ranging from 66 to 110 μg/m3 at the investigated sites (Table 1). The mineral dust (6–65 μg/m3), nitrate ions (6–16 μg/m3), sulfate ions (6–14 μg/m3), ammonium ions (7–11 μg/m3) and potassium ions (2–3 μg/m3) mainly composed the remaining parts. For the entire harvest season, significant differences were only found in mineral dust and calcium ions among the urban, suburban and rural sites (Table 1). However, the differences in the total PM2.5 concentrations and its major components were distinct at the three stages (i.e., pre-harvest, harvest and post-harvest) (Fig. 5). At the pre-harvest stage, the PM2.5 and OC levels were in the order from highest to lowest at the urban, suburban and rural sites. As harvesting progressed, the PM2.5 and OC concentrations increased at the rural sites, approaching the suburban level but remaining lower than that at the urban site. The post-harvest stage significantly promoted the PM2.5, OC and NO−3 concentrations at all of the sites, with the rural site having the highest levels, followed by the suburban and urban sites. The ratios of OC in the PM2.5 samples gradually increased in the three stages. Additionally, the Si concentrations in PM2.5 samples at the urban site remained high throughout the harvest season (Fig. 5).

R

428

2.2.4. Chemical components in atmospheric PM2.5 at different stages

Airborne PM2.5 at all of the sites was well correlated to OC, + + EC, NO−3, SO2− 4 , NH4, and K (Table 2). At the urban site, the 2+ + contributions of OC, Si, NO−3, SO2− to 4 , EC, NH4 and Ca atmospheric PM2.5 were 33%, 12%, 6%, 5%, 3% and 2% during the crop harvest, respectively (Fig. 6). The PM2.5 at the suburban site was primarily composed of 45% OC, 9% SO2− 4 ,

N C O

427

U

426

C

T

Fig. 4 – Relationships of atmospheric pollutant concentrations, wind direction and wind speed during the crop harvest, including NOx (a), SO2 (b), O3 (c) and the ratio of SO2 and NOx (d), and PM2.5 at the four sampled sites (e, f, g and h). PM2.5: particulate matter.

Fig. 5 – Differences in atmospheric PM2.5 concentrations and major chemical components in PM2.5 from the urban, suburban and rural sites at three stages. The stage_1, stage_2 and stage_3 are divided manually by rainfall events, representing pre-harvest (i.e., before October 1), harvest (i.e., between October 2 and October 13) and post-harvest (i.e., from October 14 to October 27). PM2.5: particulate matter.

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

455 456 457 458 459 460

8

Table 2 – Pearson correlation coefficients among the major species of PM2.5 at the urban, suburban and rural sites.

464

1 0.55 0.93 0.98 0.59 0.75 0.68 0.90 0.83 0.84

1 0.73 0.51 0.53 0.49 0.44 0.45 0.42 0.58

1 0.89 0.49 0.65 0.56 0.80 0.73 0.83

1 0.65 0.84 0.79 0.96 0.85 0.90

1 0.99 0.75 0.96 0.94 −0.44 −0.44 −0.35 0.86 0.92 0.93

1 0.72 0.96 0.94 −0.50 −0.50 −0.37 0.87 0.98 0.96

1 0.87 0.63 −0.15 −0.31 −0.20 0.58 0.68 0.65

1 0.91 −0.40 −0.50 −0.36 0.79 0.92 0.91

1 −0.45 −0.51 −0.39 0.76 0.94 0.90

1 0.97 0.85 0.97 0.96 0.17 0.11 0.02 0.95 0.95 0.98

1 0.90 0.98 0.90 0.04 0.10 −0.07 0.88 0.96 0.97

1 0.94 0.77 −0.01 0.05 −0.17 0.74 0.84 0.82

1 0.93 0.13 0.13 −0.03 0.89 0.97 0.96

Ca2+

Al

Ca

K

OC

EC

1 0.76 0.83 0.70 0.66 0.64

1 0.97 0.90 0.79 0.81

1 0.88 0.76 0.75

1 0.84 0.90

1 0.80

1

1 0.73 0.63 −0.52 −0.56 −0.60

1 0.88 −0.26 −0.52 −0.55

1 −0.15 −0.39 −0.38

1 0.86 0.84

1 0.96

1

1 0.76 0.18 0.11 0.12

1 0.07 0.01 0.02

1 0.95 0.90

1 0.98

1

R O

P

T

E

D

O

K+

F

1 0.94 0.61 0.88 0.97 0.77 0.89 0.86 0.96 0.86 0.90

R

1 0.31 0.22 0.16 0.95 0.96 0.93

8% NO−3, 5% EC, 4% Si and 3% Cl−. Compared with the urban and suburban sites, the rural sites had a higher OC ratio (49%), followed by the other species in order, i.e., NO−3 (9%), SO2− 4 (6%), NH+4 (5%), EC (5%) and Cl− (4%).

O

463

Rural PM2.5 NO−3 SO2− 4 NH+4 K+ Ca2+ Al Ca K OC EC

NH+4

1 0.45 0.76 0.10 0.16 0.15

2.2.5. Source apportionment of atmospheric PM2.5

465

Based on the PMF results, we identified five primary emission 466 sources and their contributions to atmospheric PM2.5 in the 467 Changchun region during harvest season (Fig. 7). The sources 468

C

462

Suburban PM2.5 NO−3 SO2− 4 NH+4 K+ Ca2+ Al Ca K OC EC

SO2− 4

N

461

Urban PM2.5 NO−3 SO2− 4 NH+4 + K Ca2+ Al Ca K OC EC

NO−3

U

t2:5 t2:6 t2:7 t2:8 t2:9 t2:10 t2:11 t2:12 t2:13 t2:14 t2:15 t2:16 t2:17 t2:18 t2:19 t2:20 t2:21 t2:22 t2:23 t2:24 t2:25 t2:26 t2:27 t2:28 t2:29 t2:30 t2:31 t2:32 t2:33 t2:34 t2:35 t2:36 t2:37 t2:38 t2:39 t2:40 t2:41 t2:42

PM2.5

C

t2:4

E

t2:3 t2:2

R

t2:1

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

Fig. 6 – Contributions of major chemical components to atmospheric PM2.5 at urban, suburban and rural sites for the entire crop harvest period. The contribution is calculated by the ratio of the mass concentration of individual components to the total PM2.5 concentration. PM2.5: particulate matter. Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

9

T

E

D

P

R O

O

F

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

472 473 474 475 476 477 478 479 480

R

471

include secondary aerosols (39.1%), biomass burning (20.0%), supply heating (17.9%), road/soil dust (13.6%) and traffic (9.3%). The significant (r2 = 0.87, p < 0.01) correlation between the predicted and observed PM2.5 concentration indicates that this is a reasonable conclusion. The secondary aerosols occurred along with a high loading of NO−3, NH+4, OC and EC. The biomass burning portion was characterized by an obviously high concentration of K+ and K. The heating supply mainly contributed high concentrations of SO2− 4 . Road and soil dust provided the primary mineral species of Al, Ca, Fe and Mg. Vehicle use might release certain amounts of Zn, which is widely used in automobile oil.

N C O

470

U

469

R

E

C

Fig. 7 – Source profiles from PMF (Positive Matrix Factorization) analysis during the entire sampling period using data from the four sites. Percentage of species mass (%) is the concentration of a species contained in one factor divided by the total factor concentration (i.e., each compound sums vertically to 100%). The total contribution of each factor to the atmospheric PM2.5 mass concentration is presented in parentheses, respectively. PM2.5: particulate matter.

482 481

3. Discussion

483

3.1. Air pollution levels during the harvest season in the Changchun region

484 485 486 487 488 489 490

This study reveals the severe air pollution and haze events and critical atmospheric pollutant of PM2.5 during the harvest season (Table 1 and Fig. 2). The daily average mass concentration of the maximum pollutant limits of the NAAQS provided by the Ministry of Environmental Protection of People's Republic of China (MEP) are 160, 100, 150 and 75 μg/m3

491

for O3, NOx, SO2 and PM2.5, respectively (MEP, 2012), which are much higher than the guidelines provided by the World Health Origination (WHO) (i.e., O3: 100 μg/m3, NO2: 40 μg/m3; SO2: 20 μg/m3; PM2.5: 25 μg/m3) (WHO, 2005). Judging by the NAAQS standards of China, most of the SO2 and O3 concentrations (i.e., >95%) during harvest season meet the criterion, while one-fourth days of the days exceeded the allowed NOx level. However, the proportions out of the PM2.5 limits were 91%, 85% and 62% at the urban, suburban and rural sites, respectively. Approximately one-third of the days had PM2.5 concentrations that were 2-fold higher than the standard (i.e., 150 μg/m3) at all of the sites. The mean PM2.5 concentration (308–382 μg/m3) from October 20 to 23 at all of the sites is four and five times higher than the PM2.5 limits in NAAQS, indicating the extremely severe and continuous haze weather. This level of daily PM2.5 concentrations (18–503 μg/m3) is comparable to the reported values in the Tongyu rural site (Zhang et al., 2007), the Longfengshan rural site (R. Zhang et al., 2012) and Shengyang city (Han et al., 2010) in northeastern China, as well as other large cities, such as Beijing (Ji et al., 2012), Guangzhou (Chang et al., 2013) and Xi'an (Cao et al., 2012a, 2012b).

510

3.2. Major emission sources of gaseous pollutants

511

492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 Q10 507 508 509 Q11

Our results showed that the local emission sources were 512 dominant for NOx, transportation affected O3 and the 513

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

10

524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544

F

523

O

522

R O

521

C

520

E

519

R

518

P

3.3. Potential emission sources of airborne PM2.5

517

Fig. 8 – Spatial distribution of MODIS fire points during sampling period (i.e., October) in northeastern China.

D

546

516

PM2.5 levels for the second haze event on October 23 (Cao et al., 2012a, 2012b). In addition, the lower gaseous pollutant levels (i.e., SO2, NOx and O3) possibly indicate that these precursors were used to form secondary aerosols during the haze events (Fig. 2c). Along with the strong emissions, the low speed (<2 m/sec) and high relative humidity (>80%) were not conducive to air diffusion and transportation, which aggravated air pollution on October 23. The diurnal PM2.5 concentrations revealed that severe air pollution began at 16:00 and remained at a high level during the nighttime. This pattern was most similar to the combination of strong PM2.5 emissions and weak vertical diffusion in a low planetary boundary layer (PBL) height at night (Fan et al., 2005). In addition to the rural diffusion, urban traffic-related PM2.5 release and smoke from outdoor street barbecues most likely contribute to the longer duration of urban or suburban air pollution in a day (Huang et al., 2010). The morning traffic rush also causes the PM2.5 to increase in the urban site, and cooking smoke using straw burning might account for the morning's PM2.5 increase. Carbonaceous species dominated the principal component of airborne PM2.5 due to the 30%–70% contributions during the observed period. The EC predominately originates from incomplete combustions, whereas OC arises from primary and secondary (through gas-to-particle conversion) sources (Castro et al., 1999). The good correlation of OC and EC indicates that they might be emitted by similar emission sources. The petrol cars, biogenic burning and coal burning have been reported as the most significant emission sources of OC (Chow, 1995). At the urban site, high OC and EC contents before harvest were most similar to from automotive exhaust around residential areas and adjacent to “The First Automobile works”, which is one of the largest automobile enterprises in China. Outdoor street barbecues were also identified as an important source of OC and EC for urban areas. At the rural

T

545

combination of local sources and transportation contributed the SO2 (Fig. 4a–c). A high concentration with weak wind generally indicates that the local emission sources are primary; otherwise, the source is transportation from upstream areas. Gaseous pollutants were only determined at the suburban site, indicating the possible dispersion path of urban air pollutants to the downstream suburban site under southwesterly wind. In urban areas, the NOx and VOC levels generated by fuel combustion are more conducive to O3 production by photochemical reactions (Heal et al., 2012). In addition, high O3 levels often occurred with lower NOx values, and this phenomenon was distinct in the diurnal dynamics due to the NOx consumption in photochemical reactions (Fig. 3a). The NOx level was possibly linked to the local motor vehicle traffic flow on an inter-city highway and a city beltline that surrounded the suburban site as well as the farm machinery during the harvest season. After sunset, NOx originating from traffic, power plant and coal burning accumulated and reached a maximum at approximately 21:00 local time and slightly increased during the morning rush hour (i.e., 7:00). High SO2 concentrations in the southwest direction might be caused by adjacent point emissions, such as a waste treatment plant, and SO2 may be obtained from an urban industrial release, such as power plant. Furthermore, the high ratio of SO2/NOx distributed over the west-related directions supports the transportation of industrial point source emissions from these directions (Wang et al., 2002). The winter heat supply since October 25 most likely significantly promoted the SO2 emissions by the increase of coal burning. The double peaks (i.e., approximately 9:00 and 21:00) of the SO2 concentrations in a day appear to encompass the time of heat supply because the largest contributor of SO2 is fossil burning.

515

E

514

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

547

U

N

C

O

R

Airborne PM2.5 at each site was primarily contributed by local 548 sources based on the distribution of the concentrations in wind 549 speed (Fig. 4e–h). In the Changchun region, urban areas can be 550 considered a complex point source composed of traffic, 551 Q12 industrial process, daily life, power plant and coal burning 552 (Han et al., 2010), thus maintaining the highest PM2.5 concen553 tration with continuous PM2.5 releases from sources in the 554 pre-harvest stage. In contrast, rural areas showed a gradual 555 increase in the strong area sources of PM2.5 with the enhance556 ment of agricultural burning, maintaining the highest PM2.5 557 levels at the end of harvest (Ichoku et al., 2012). Influenced by 558 urban and rural emissions, the suburban site had higher PM2.5 559 values in the pre-harvest stage than the rural site but the lowest 560 values in harvest stage and post-harvest with the increase of 561 coal burning in the urban site and straw burning in the 562 rural site. Rainfall events significantly improve air quality by 563 clearing away air particles with wet deposition and reducing 564 straw-burning operations under wet conditions as well as road 565 and soil dust emission. Subsequently, the PM2.5 concentration 566 rapidly increased and reached the maximum around October 10 567 and 23 before the next rainfalls. These PM2.5 concentrations at 568 the rural sites were driven by intense and expansive straw 569 burning pre-rainfall, which was also supported by more fire 570 points from MODIS in the period (Fig. 8). At the urban site, PM2.5 571 releases from coal burning for heat supply would increase the

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605

11

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665

F

618

Local chemical and temporal profiles induced by agricultural burning are of great importance in air quality models because they provide key parameters (Tong et al., 2012). In this study, the particle chemical profile of crop residue burning was not directly determined under a controlled burning condition; therefore, we performed an indirect estimation by comparing the increased PM2.5 and the components in the post-harvest stage with pre-harvest stage at a more distant rural site. We assume that atmospheric PM2.5 at this rural site was primarily influenced by agricultural burning during the concentrated-burning stage, and the pre-harvest stage acted as the ambient treatment. Based on the straw burning-induced chemical profile of PM2.5 (Fig. 9), straw burning released large amounts of OC (~58%), and formed secondary aerosol-related NO−3, SO2− and NH−4 in the aging 4 process. Furthermore, as the tracer of biomass burning, K+ accounted for 2% of the composition. Although these composition ratios were comparable to previous studies, the amounts of farm machinery in the field might also significantly contribute to our estimation method. Similar to the chemical profile, the increased hourly PM2.5 concentrations were calculated using on-line data at a more distant rural site to represent the agricultural burning diurnal

O

617

R O

616

P

615

666 667 668 669 670 671

3.4. Agricultural burning-induced chemical profile and diurnal 672 profile of PM2.5 673

D

614

E

613

T

612

C

611

E

610

R

609

should include all types of motor vehicles in the urban and suburban areas and the farm machinery used in vast rural areas. However, the weakness in our source apportionment is that the sample numbers were insufficient to analyze source apportionments for each site, which might create some uncertainties in the contributions.

R

608

site, along with an increase of the biomass burning and the use of farm machinery, daily average OC concentrations increased from 19 to 208 μg/m3 in the three stages. The ratios of OC and EC at all of the sites varied from 5.5 to 17.2, with an average of 9.7 ± 2.2, indicating remarkable primary OC from biomass burning. Based on the OC/EC minimum ratio method (Turpin and Huntzicker, 1995), the contributions of SOC to OC were estimated as 37%, 32% and 40% at the urban, suburban and rural sites, respectively. In addition, as the tracer of biomass burning, rapid increases of K+ and K at the rural sites confirmed their emission in agricultural operations. These high non-point pollutant emissions would increase the OC and K+ concentration in the urban and suburban sites. + The high loading of NO−3, SO2− 4 , NH4 represents the basic characteristic of nitrate and sulfate, which are formed during photochemical reaction-induced secondary pollution (Cao et al., 2005). NOx (NO2 + NO) is an important precursor of nitrate ion. The major emission sources of NO2 include vehicle exhaust, heat supply, and biomass burning. Compared with pre-harvest period, a higher NO2/NOx value (0.70 ± 0.04) during harvest period suggests the significant increases in major sources. Besides heat supply and crop residue burning, a large number of agricultural machinery is used during harvest period, indicating their possible contribution to NO2 emission due to more automobile exhaust pollution. The power plant and heat supply in October also contribute the sulfate precursor. In addition, previous studies have reported that particles from biogenic burning would become finer and aged, and a high content of NO−3 and SO2+ 4 was found in aged particles (Hays et al., 2005). This trend was consistent with the increasing NO−3 and SO2+ 4 concentrations at the post-harvest stage, supporting the importance of biomass burning to secondary aerosol formation. Furthermore, we found that the crustal elements remained stable and had a good association with airborne PM2.5 at the urban site (Table 2 and Fig. 5). This high level was primarily caused by re-suspension of traffic dust and road construction around residential quarters under the surge of construction in Changchun city. On a regional scale in Changchun, our initial analysis of the PMF model showed that secondary aerosols were the largest contributor (i.e., two-fifth) to PM2.5 according to the relationship of PM2.5 and the chemical components. Although the precursors of secondary aerosols are primarily from automobile exhaust, industry emission and biomass burning, we could not identify each of their each of their ratios; thus, they were summed in this study. Biomass burning was the important emitter (i.e., one-fifth) in harvest season and represented the role of agricultural activity roles in regional air quality, especially in the second half of October. The third emission source was the heat supply associated only with high SO2− 4 , which may primarily represent coal combustion from the thermal power plant in Changchun city. Road and soil dust was estimated to approximately one-seventh of the total emission sources and was composed of automobile-induced re-suspend dust and agricultural-induced soil release. Because Zn is one of the basic characteristics of motor vehicle pollution due to its use in lubricants, brake parts and tires, we obtained its direct contribution (approximately one-tenth) to the regional sources; its indirect contribution to the formation of secondary aerosols was not involved in this part. The vehicle sources in this period

N C O

607

U

606

Fig. 9 – Chemical profile of agricultural burning-induced PM2.5 in the Changchun region. The increased PM2.5 and components by agricultural burning were estimated by the subtraction of average PM2.5 concentration in the post-harvest stage and the pre-harvest stage at the rural site. The ratio of each component to the total PM2.5 concentration was calculated based on the difference of mass concentrations. PM2.5: particulate matter.

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695

12

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

732

Acknowledgments

741 740

F

October along with the straw burning and winter heating supply, especially during days under poor diffusion conditions. The chemical and diurnal profiles estimated for agricultural burning in this study might provide additional methods to improve air quality simulations during the harvest season. In summary, therefore, attention from scientists and policy makers should be enhanced to prevent regional complex air pollution in northeastern China.

4. Conclusions

715

In the Changchun region, our results numerically confirmed the severe and continuous air pollution during harvest season (i.e., October). The critical pollutant is PM2.5, which was dominated by carbonaceous species, secondary aerosol-related ions and potassium in both urban and rural areas. Regional PM2.5 pollution levels gradually increased with the progression of agricultural activities. In a day, the PM2.5 accumulation under a low boundary layer and under diffusion conditions often causes more serious pollution in the first half of the night. According to these characteristics, automobile exhaust, road dust, coal burning and outdoor barbeques were identified as important sources of the high PM2.5 concentrations in urban areas, while straw burning and the use of farm machinery apparently release organic carbon and soil dust as area pollution sources. Typical haze events occurred in three days before rainfall due to the strong burning operations. It should be noted that extreme haze would occur during this time at the urban site in late

707 708 709 710 711

716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731

C

706

E

705

R

704

R

703

O

702

C

701

N

700

U

699

737 738 739

R O

O

74 6

P

714 713

698

736

REFERENCES

Aneja, V.P., Wang, B., Tong, D.Q., Kimball, H., Steger, J., 2006. Characterization of major chemical components of fine particulate matter in North Carolina. J. Air Waste Manage. Assoc. 56 (8), 1099–1107. Aneja, V.P., Schlesinger, W.H., Erisman, J.W., 2009. Effects of agriculture upon the air quality and climate: research, policy, and regulations. Environ. Sci. Technol. 43 (12), 4234–4240. Bureau of Statistics of Jilin, 2007. Jilin Statistical yearbook 2012. China Statistic Press, Beijing. Cao, J.J., Wu, F., Chow, J.C., 2005. Characterization and source apportionment of atmospheric organic and elemental carbon during fall and winter of 2003 in Xi'an, China. Atmos. Chem. Phys. 5, 3127–3137. Cao, J.J., Shen, Z.X., Chow, J.C., Watson, J.G., Lee, S.C., Tie, X.X., et al., 2012a. Winter and summer PM2.5 chemical compositions in fourteen Chinese cities. J. Air Waste Manage. 62 (10), 1214–1226. Cao, J.J., Wang, Q.Y., Chow, J.C., Watson, J.G., Tie, X.X., Shen, Z.X., et al., 2012b. Impacts of aerosol compositions on visibility impairment in Xi'an, China. Atmos. Environ. 59, 559–566. Castro, L.M., Pio, C.A., Harrison, R.M., Smith, D.J.T., 1999. Carbonaceous aerosol in urban and rural European atmospheres: estimation of secondary organic carbon concentrations. Atmos. Environ. 33, 2771–2781. Chan, C.K., Yao, X., 2008. Air pollution in mega cities in China. Atmos. Environ. 42, 1–42. Chang, S.Y., Chou, C.K., Liu, S., Zhang, Y.H., 2013. The characteristics of PM2.5 and its chemical compositions between different prevailing wind patterns in Guangzhou. Aerosol Air Qual. Res. 13 (4), 1373–1383. Cheng, Z., Jiang, J., Fajardo, O., Wang, S., Hao, J., 2013. Characteristics and health impacts of particulate matter pollution in China (2001–2011). Atmos. Environ. 65, 186–194. China Agricultural Yearbook. 2012. China Agriculture Press. Chow, J.C., 1995. Measurement methods to determine compliance with ambient air quality standards for suspended particles. J. Air Waste Manage. Assoc. 45 (5), 320–382. Fan, J., Zhang, R., Li, G., Nielsen-Gammon, J., Li, Z., 2005. Simulations of fine particulate matter (PM2.5) in Houston, Texas. J. Geophys. Res. 110, D16203. http://dx.doi.org/10.1029/ 2005JD005805. Han, B., Kong, S.F., Bai, Z.P., Du, G., Bi, T., Li, X., et al., 2010. Characterization of elemental species in PM2.5 samples collected in four cities of Northeast China. Water Air Soil Pollut. 209 (1), 15–28.

T

712

profile (Fig. 10). Based on the observed diurnal profile, a high PM2.5 concentration occur from 16:00–22:00. However, this does not imply that most of the straw be burned during this time because high levels were also caused by the shortening of the PBL height. Using the Weather Research and Forecasting model, we simulated the PBL heights in the Changchun region from October 14 to 23. Then, we adjusted the emission diurnal profile, showing the largest burning activities from 14:00–20:00. This diurnal pattern of biomass burning emissions is clearly different from the parameters provided by the EPA but is coincident with the local conventional practice according to the local farmers. The reason for the high PM2.5 concentration during the nighttime was most likely to the large accumulation of burning smoke under low PBL height. Therefore, these parameter patterns for the chemical and diurnal profile of biomass burning could improve the accuracy of load forecasting in local or even regional air quality.

697

735

742

D

696

734

This study was financially supported by the National Natural Science Foundation of China (Nos. 41205106, 41275158). We also thank the staff of the sampling sites for their support in the field experiments and for providing agricultural information.

E

Fig. 10 – Diurnal profile of the observed agricultural burning-induced PM2.5 concentration (Obs-conc.), PBL-adjusted agricultural burning emission (PBL-b.e.) and EPA-provided corn residue burning emission style (EPA-b.e). PM2.5: particulate matter; PBL: planetary boundary layer.

733

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

743 744 745

747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792

13

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

D

P

R O

O

F

Pui, D.Y., Chen, S., Zuo, Z., 2014. PM2.5 in China: measurements, sources, visibility and health effects, and mitigation. Particuology 13 (2), 1–26. Qin, Y., Xie, S., 2011. Historical estimation of carbonaceous aerosol emissions from biomass open burning in China for the period 1990–2005. Environ. Pollut. 159, 3316–3323. Tong, D.Q., Lee, P., Saylor, R.D., 2012. New directions: the need to develop process-based emission forecasting models. Atmos. Environ. 47, 560–561. Turpin, B.J., Huntzicker, J.J., 1995. Identification of secondary organic aerosol episodes and quantitation of primary and secondary organic aerosol concentrations during SCAQS. Atmos. Environ. 29, 3527–3544. Wang, T., Cheung, T.F., Li, Y.S., Yu, X.M., Blake, D.R., 2002. Emission characteristics of CO, NOx, SO2 and indications of biomass burning observed at a rural site in eastern China. J. Geophys. Res. 107, D12. http://dx.doi.org/10.1029/ 2001JD000724. Wang, J., Hu, Z., Chen, Y., Chen, Z., Xu, S., 2013a. Contamination characteristics and possible sources of PM10 and PM2.5 in different functional areas of Shanghai, China. Atmos. Environ. 68, 221–229. Wang, Q.Y., Cao, J.J., Shen, Z.X., Tao, J., Xiao, S., Luo, L., et al., 2013b. Chemical characteristics of PM2.5 during dust storms and air pollution events in Chengdu, China. Particuology 11 (1), 70–77. WHO, 2005. WHO air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulphur dioxide. http://whqlibdoc. who.int/hq/2006/WHO_SDE_PHE_OEH_06.02_eng.pdf. Xinhua News Agency, 2013. http://www.webcitation.org/ 6KZoMVk7n. Zhang, Y.X., Shao, M., Zhang, Y.H., Zeng, L.M., He, L.Y., Zhu, B., et al., 2007. Source profiles of particulate organic matters emitted from cereal straw burnings. J. Environ. Sci. 19, 167–175. Zhang, G., Li, J., Li, X., Xu, Y., Guo, L., Tang, J., et al., 2010. Impact of anthropogenic emissions and open biomass burning on regional carbonaceous aerosols in South China. Environ. Pollut. 158, 3392–3400. Zhang, R., Tao, J., Ho, K.F., Shen, Z., Wang, G., Cao, J., et al., 2012a. Characterization of atmospheric organic and elemental carbon of PM2.5 in a typical semi-arid area of northeastern China. Aerosol Air Qual. Res. 12 (5), 792–802. Zhang, X.Y., Wang, Y.Q., Niu, T., Zhang, X.C., Gong, S.L., Zhang, Y.M., et al., 2012b. Atmospheric aerosol compositions in China: spatial/temporal variability, chemical signature, regional haze distribution and comparisons with global aerosols. Atmos. Chem. Phys. 12, 779–799.

E

T

C

E

R

R

N C O

888

Hays, M.D., Fine, P.M., Geron, C.D., Kleeman, M.J., Gullett, B.K., 2005. Open burning of agricultural biomass: physical and chemical properties of particle-phase emissions. Atmos. Environ. 39, 6747–6764. Heal, M.R., Kumar, P., Harrison, R.M., 2012. Particles, air quality, policy and health. Chem. Soc. Rev. 41 (19), 6606–6630. Hinz, T., Tamoschat-Depolt, K., 2007. Particulate Matter in and from Agriculture. Special Issue 308. Landbauforschung Völkenrode. Holmén, B.A., Miller, D.R., Hiscox, A.L., Yang, W., Wang, J., Sammis, T.W., Bottoms, R., 2008. Near-source particulate emissions and plume dynamics from agricultural field operations. J. Atmos. Chem. 59 (2), 117–134. Huang, L.K., Wang, K., Yuan, C.S., Wang, G.Z., 2010. Study on the seasonal variation and source apportionment of PM10 in Harbin, China. Aerosol Air Qual. Res. 10 (1), 86–93. Huang, L., Yuan, C.S., Wang, G., Wang, K., 2011. Chemical characteristics and source apportionment of PM10 during a brown haze episode in Harbin, China. Particuology 9 (1), 32–38. Hueglin, C., Gehrig, R., Baltensperger, U., Gysel, M., Monn, C., Vonmont, H., 2005. Chemical characterisation of PM2.5, PM10 and coarse particles at urban, near-city and rural sites in Switzerland. Atmos. Environ. 39, 637–651. Ichoku, C., Kahn, R., Chin, M., 2012. Satellite contributions to the quantitative characterization of biomass burning for climate modelling. Atmos. Res. 111, 1–28. Ji, D.S., Wang, Y.S., Wang, L.L., 2012. Analysis of heavy pollution episodes in selected cities of northern China. Atmos. Environ. 50, 338–348. Kan, H., Chen, R., Tong, S., 2012. Ambient air pollution, climate change, and population health in China. Environ. Int. 42 (1), 10–19. Li, L., Qian, J., Qu, C.Q., Zhou, Y.X., Guo, C., Guo, Y.M., 2014. Spatial and temporal analysis of air pollution index and its timescale-dependent relationship with meteorological factors in Guangzhou, China, 2001–2011. Environ. Pollut. 190, 75–81. Liu, T., Zhang, Y., Xu, Y., Lin, H., Xu, X., Luo, Y., 2014. The effects of dust-haze on mortality are modified by seasons and individual characteristics in Guangzhou, China. Environ. Pollut. 187, 116–123. MacIntosh, D.L., Williams, P.L., Yanosky, J.D., 2002. A comparison of two direct-reading aerosol monitors with the federal reference method for PM2.5 in indoor air. Atmos. Environ. 36, 107–113. Ministry of Environmental Protection of People's Republic of China, 2012o. Ambient air quality standards. http://www. zzemc.cn/em_aw/Content/GB3095-2012.pdf. Norris, G., Duvall, R., Brown, S., Bai, S., 2014. EPA Positive Matrix Factorization (PMF) 5.0 Fundamental and User Guide. U.S. Environmental Protection Agency.

U

793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839

Please cite this article as: Chen, W., et al., Typical atmospheric haze during crop harvest season in northeastern China: A case in the Changchun region, J. Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.031

840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887