Refined Source Apportionment of Atmospheric PM2.5 in a Typical City in Northwest China

Samples were collected from 2017.12.18 to 2018.1.21 (winter), 2018.4.27 to 2018.5.25 (spring), 2018.7.25 to 2018.8.22 (summer), and 2018.10.31 to 2018.11.28 (autumn), respectively, at four stations, Fengdong New Town (FDNT), Chongwen Pagoda (CWP), Meteorological Bureau (MB) and Zhangjiashan Reservoir (ZJS) (Figure S1). Micro-flow samplers (Airmetrics, USA), which operated at a flow rate of 5 L/min, were used to collect PM2.5 samples. Each sample was collected over an interval of 24 h, nominally from 10:00 am (local time) to 10:00 am on the following day. The PM2.5 samples were collected using 47 mm diameter quartz filters (Whatman, QM/A, UK) and TeflonTM filter membranes (R2PJ047, Pall Life Sciences Ann Arbor, USA). A total of 1016 PM2.5 filter membrane samples was collected (508 each for quartz and Teflon). At the same time, we collected 32 blank filters to subtract background concentration for quartz samples and 16 blank filters for Teflon samples.

conducted (Cao et al., 2005(Cao et al., , 2012(Cao et al., , 2014Chen et al., 2016;Wang et al., 2014;Xu H.M., 48 et al., 2016;Xu Y., 2018;You et al., 2015), and the seriousness of air pollution in the 49 GZB has become all too evident (Huang et al., 2015;Li et al., 2017). It has been shown, 50 for example, that the PM 2.5 concentrations on hazy days were typically two times 51 higher compared with non-hazy days, and PM 2.5 often exceeds the National Air Quality 52 Standard value during haze events (Chen et al., 2016;Shen et al., 2009;You et al., 2015;53 Zhang et al., 2015;Zhang K.N., et al., 2019). The chemical components of the particles 54 were also found to differ for hazy versus non-hazy days, and those differences were 55 attributed to changes in the major sources and the effects of regional transport on 56 pollutants. 57 The Xixian New District (XXND) is located in the Fenwei Plain, and it is the 58 seventh national-level new district in China that was officially approved on January 6, 59 2014. The district is located between the main urban areas of Xi 'an and Xianyang, and 60 it is expected to become the main functional new district and ecological center of the 61 Xi'an metropolis. With the growing social, political and economic importance of 62 XXND, environmental issues, especially PM 2.5 pollution, are of great concern. 63 However, there has been little research on air pollution there, especially assessments 64 of the sources for PM 2.5 in the region. The development of effective measure for S1). The ZJS station is located in the north, far from urban areas and with no pollution October 31, 2018-November 29, 2018 as autumn. A micro-volume sampler, which 93 operated at a flow rate of 5 L/min, was used to collect the filter samples. Each sample 94 was pumped continuously for 24 h, and the monitoring period was from 10:00 am to 95 10:00 am of the next day. A total of 1016 PM 2.5 filter samples were collected, including 96 508 quartz filters and 508 Teflon™ filter membranes. 97 The mass concentrations of the PM 2.5 samples were determined gravimetrically 98 with the use of a micro-electronic balance (ME 5-F, Sartorius, Göttingen, Germany). A 99 thermal/optical carbon analyzer (Model 2001, Atmoslytic Inc., Calabasas, CA, USA) 100 was used to measure the organic and elemental carbon components (OC and EC, 101 respectively). An ion chromatograph (IC,Metrohm 940,Swiss Metrohm Group,102 Switzerland) was used to quantitatively analyze water-soluble ion concentrations (Cl -, 103 NO 3-, SO 4 2-, Na + , NH 4 + , K + , Mg 2+ , Ca 2+ ). Eighteen elements (Na, Mg, Al, Si, S, Cl, K,    In summer, however, the sulfate concentrations exceeded those of nitrate (Table 1). 175 The highest concentrations of Cland K + (3.6 and 1.4 μg m -3 , respectively) were 176 likely due to the burning of large quantities of coal and biomass in winter for heating, 177 whereas ions that typically are associated with dust (Mg 2+ , Ca 2+ ) were relatively high 178 in spring (0.1 μg m -3 , 1.1 μg m -3 ).    farmlands) in the dust-related source were 11.4%, 8.0% and 3.9% respectively. The In fact, secondary pollutants (nitrate, sulfate, and organic matter) accounted for a 282 large percentage (41.5%) of the PM 2.5 mass (Figure 3). In addition to gas-to-particle 283 conversion of precursor gases from coal combustion, some of the secondary species 284 were likely produced from NH 3 emitted from farmlands and livestock and from VOCs 285 that originated from numerous widespread sources. These findings are consistent with 286 the WRF-CHEM model results presented above which showed that regional transport 287 accounted for 60.8% of PM 2.5 during the high-pollution events in the winter. Dust 288 sources were not included in the current WRF-CHEM modelling, however, and the 289 contribution from regional sources likely would have been lower than 60.8% if we had 290 been able to include that dust component of the PM 2.5 . In a study in Beijing, for instance, 291 most of the dust was found to be associated with local emissions (Han et al., 2007). 292 The percentage of PM 2.5 mass from secondary aerosols obtained from the receptor 293 models (41.5%) can explain much of the regional contribution to the pollution aerosol 294 over the annual scale. fraction of PM 2.5 mass at XXND was considerably higher (Table 3). This is can be 299 explained by large number of construction sites for buildings and roads in the new 300 district; indeed, the dust faction of the PM 2.5 mass was as high as 23.3%. Industrial 301 sources also showed a stronger impact in XXND compared with Xi'an in that period; 302 this can be explained by several large industrial sources in XXND, including 303 petrochemical plants, methanol plants, etc. The proportion of PM 2.5 mass assigned to 304 motor vehicles in XXND was about half that at Xi'an, which is probably because 305 XXND is a new and developing urban area, and the number of motor vehicles in use is 306 lower than in Xi'an. Biomass burning in XXND was 2 to 3% percent higher than in

A C C E P T E D M A N U S C R I P T
2 Fig. 1

A C C E P T E D M A N U S C R I P T
3 Fig. 2

A C C E P T E D M A N U S C R I P T
5 Zn 0.2 ± 0.16 0.12 ± 0.09 0.08 ± 0.04 0.14 ± 0.1 0.13 ± 0.05 1 n.d. stands for < minimum detection limit.

A C C E P T E D M A N U S C R I P T
6 Table 2. Contributions of local emissions and regional transport to the main air pollutants in the Xixian New District. See Table S3 for detailed experimental design details. The interaction term represents the interactions between background, local emissions and regional transport.

A C C E P T E D M A N U S C R I P T
7