Characteristics of Air Pollutants and Greenhouse Gases at a Regional Background Station in Southwestern China

The characteristics of air pollutants and greenhouse gases at regional background sites are critical to assessing the impact of anthropogenic emissions on the atmospheric environment, ecosystems and climate change. However, observational studies are still scarce at such background sites. In this study, continuous hourly observations of air pollutants (O3, CO, SO2, NOx, PM2.5 and PM10) and greenhouse gases (CO2, CH4 and N2O) were performed for one year (from January 1 to December 31, 2017) at the Gongga Mountain background station (GGS; 101°97′E, 29°55′N; elevation: 3541 m) in southwestern China. The concentrations and variations of these air pollutants and greenhouse gases were determined, and the effect of transboundary atmospheric transport on the air pollution at the study site was investigated. The results showed that the average annual concentrations (mixing ratios) of the O3, CO, SO2, NO2, CO2, CH4, N2O, PM2.5 and PM10 were 74.7 ± 22.0 μg m, 0.3 ± 0.2 mg m, 0.5 ± 0.6 μg m, 1.7 ± 1.3 μg m, 406.1 ± 9.5 ppm, 1.941 ± 0.071 ppm, 324.5 ± 14.8 ppb, 6.5 ± 6.2 μg m and 10.6 ± 11.2 μg m, respectively. The concentrations (mixing ratios) of the abovementioned substances at the GGS are comparable to those at other background sites in China and around the world. The slight differences among concentrations at different sites may be mainly attributable to the impacts of anthropogenic emissions near the background sites and meteorological conditions. High values of O3 were observed in spring and summer, while SO2 and PM2.5 showed higher concentrations in summer than in autumn. Relatively high CO, NO2 and PM10 values were mostly observed in spring and winter. Relatively low CO2 concentrations were observed in summer due to the vigorous summertime photosynthesis of vegetation. The lowest concentrations for CH4 were recorded in summer, whereas the levels in the other three seasons were similar to each other; by contrast, the highest N2O concentrations were observed in summer due to enhanced microbial activity resulting from high ambient summer temperatures. A diurnal variation in O3 was observed, with early morning minima and afternoon maxima. CO and NO2 displayed higher concentrations in the daytime than in the nighttime. A slight increase in both PM2.5 and PM10 concentrations was also recorded in the daytime. These patterns were closely related to scattered anthropogenic emissions and regional atmospheric transport. Nevertheless, CO2 exhibited lower concentrations in the daytime than in the nighttime, although CH4 showed no obvious diurnal variation. The N2O concentration peaked between 10:00 and 12:00 (local time), which can be ascribed to the enhancement of microbial activity due to the increased soil temperature. The results based on the relationship between the wind and the concentrations of air pollutants and greenhouse gases were almost consistent with those based on the potential contribution source function. It appears that O3 and its precursors in parts of Inner * Corresponding author. E-mail address: jds@mail.iap.ac.cn Cheng et al., Aerosol and Air Quality Research, 19: 1007–1023, 2019 1008 Mongolia and Gansu, Ningxia, Sichuan, Chongqing and Hubei Provinces as well as adjacent areas of Hunan, Guizhou and Guangxi Provinces contributed to the increase in O3 at the study site. The potential source areas for CO and SO2 were similar and mainly distributed in India and Pakistan and areas of Inner Mongolia and Gansu and Guizhou Provinces in China. Potential source areas for NO2, PM2.5 and PM10 were found in neighboring countries of South Asia in addition to domestic regions, including Inner Mongolia, Gansu Province and the Cheng-Yu economic region. Furthermore, parts of Yunnan Province (China) as well as India and Pakistan were potential source areas for CO2, CH4 and N2O.


ABSTRACT
The characteristics of air pollutants and greenhouse gases at regional background sites are critical to assessing the impact of anthropogenic emissions on the atmospheric environment, ecosystems and climate change.However, observational studies are still scarce at such background sites.In this study, continuous hourly observations of air pollutants (O 3 , CO, SO 2 , NO x , PM 2.5 and PM 10 ) and greenhouse gases (CO 2 , CH 4 and N 2 O) were performed for one year (from January 1 to December 31, 2017) at the Gongga Mountain background station (GGS; 101°97′E, 29°55′N; elevation: 3541 m) in southwestern China.The concentrations and variations of these air pollutants and greenhouse gases were determined, and the effect of transboundary atmospheric transport on the air pollution at the study site was investigated.The results showed that the average annual concentrations (mixing ratios) of the O 3 , CO, SO 2 , NO 2 , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 were 74.7 ± 22.0 µg m -3 , 0.3 ± 0.2 mg m -3 , 0.5 ± 0.6 µg m -3 , 1.7 ± 1.3 µg m -3 , 406.1 ± 9.5 ppm, 1.941 ± 0.071 ppm, 324.5 ± 14.8 ppb, 6.5 ± 6.2 µg m -3 and 10.6 ± 11.2 µg m -3 , respectively.The concentrations (mixing ratios) of the abovementioned substances at the GGS are comparable to those at other background sites in China and around the world.The slight differences among concentrations at different sites may be mainly attributable to the impacts of anthropogenic emissions near the background sites and meteorological conditions.High values of O 3 were observed in spring and summer, while SO 2 and PM 2.5 showed higher concentrations in summer than in autumn.Relatively high CO, NO 2 and PM 10 values were mostly observed in spring and winter.Relatively low CO 2 concentrations were observed in summer due to the vigorous summertime photosynthesis of vegetation.The lowest concentrations for CH 4 were recorded in summer, whereas the levels in the other three seasons were similar to each other; by contrast, the highest N 2 O concentrations were observed in summer due to enhanced microbial activity resulting from high ambient summer temperatures.A diurnal variation in O 3 was observed, with early morning minima and afternoon maxima.CO and NO 2 displayed higher concentrations in the daytime than in the nighttime.A slight increase in both PM 2.5 and PM 10 concentrations was also recorded in the daytime.These patterns were closely related to scattered anthropogenic emissions and regional atmospheric transport.Nevertheless, CO 2 exhibited lower concentrations in the daytime than in the nighttime, although CH 4 showed no obvious diurnal variation.The N 2 O concentration peaked between 10:00 and 12:00 (local time), which can be ascribed to the enhancement of microbial activity due to the increased soil temperature.The results based on the relationship between the wind and the concentrations of air pollutants and greenhouse gases were almost consistent with those based on the potential contribution source function.It appears that O 3 and its precursors in parts of Inner

INTRODUCTION
Increased occurrence levels of air pollutants (ozone (O 3 ), carbon monoxide (CO), sulfur dioxide (SO 2 ), nitrogen oxides (NO x ), particulate matter with a diameter of 2.5 micrometers or less (PM 2.5 ) and particulate matter with a diameter of 10 micrometers or less (PM 10 )) and greenhouse gases (carbon dioxide (CO 2 ), methane (CH 4 ) and nitrous oxide (N 2 O)) caused by intensified anthropogenic emissions have adversely affected the atmospheric environment, ecosystems, climate change and human health (Orru et al., 2017;Wuebbles et al., 2017).Hence, there have been increasing concerns about air pollutants and greenhouse gases at local, regional and even global scales (Meng et al., 2009;Lin et al., 2011;Thunis et al., 2016).Ground-level O 3 is a powerful oxidant that can harm lung function and irritate the respiratory system; O 3 is also linked to premature deaths, heart attacks and other cardiopulmonary problems (www.epa.gov/ozone-pollution-and-your-patients-health;Weinhold, 2008).In addition, O 3 acts as a greenhouse gas (IPCC, 2001).As a precursor of O 3 , CO influences the oxidization of the atmosphere via interactions with hydroxyl radical (OH) (Gligorovski et al., 2015).Frequent occurrences of acid rain and smog are regional-scale environmental problems in China, and SO 2 and NO x play important roles in the formation of both problems (Ji et al., 2014;Seinfeld and Pandis, 2016).Additionally, NO x is a photochemical precursor resulting in the substantial enhancement of global background O 3 concentrations (Lin et al., 2014;Sun et al., 2016).CO 2 , CH 4 and N 2 O are the most important greenhouse gases (Watson et al., 1992) that can absorb infrared radiation emitted from the earth and partially reradiate this radiation back to the earth's surface (Seinfeld and Pandis, 2016).Therefore, to assess the impact of anthropogenic activities on the atmospheric environment, ecosystems, climate change and human health, it is necessary to conduct long-term continuous measurements of air pollutants (O 3 , CO, SO 2 , NO x , PM 2.5 and PM 10 ) and greenhouse gases (CO 2 , CH 4 and N 2 O).Nonetheless, colocated and simultaneous measurements of air pollutants and greenhouse gases at regional background sites are scarce, although a series of studies have been performed in regions with high anthropogenic emissions.
In contrast to studies in urban areas, studies on air pollutants and greenhouse gases at regional background sites not only provide valuable information on the influence of human activities on the atmospheric environment and global change but also are helpful for understanding the transboundary transport of air pollution at a regional scale.Regional background sites are affected by very limited local anthropogenic emissions; consequently, the mediumor long-range transport of air pollutants could be the main contributor to local air pollution.Therefore, given their critical importance, a number of studies on air pollutants and greenhouse gases have been carried out at several regional background sites of the Beijing-Tianjin-Hebei (BTH), Yangtze River Delta (YRD) and Pearl River Delta (PRD) regions in China (Chao et al., 2014;Pu et al., 2015;Wang et al., 2016).The results of these studies reflected the distinctive air pollution characteristics in the abovementioned regions in China and showed significant impacts of human activities on regional air quality.However, such studies at regional background sites in southwestern China are still scarce.The Gongga Mountain background station (GGS) is representative of the regional background in southwestern China (Fu et al., 2008;Zhang et al., 2012;Zhang et al., 2014;Li et al., 2017).Several field observations have been conducted at this station on volatile organic compounds, polar organic tracers in PM 2.5 and total particulate, reactive gaseous mercury and major chemical species of PM 10 in specific periods or months (Fu et al., 2008;Zhang et al., 2012;Zhang et al., 2014;Li et al., 2017).However, to the best of our knowledge, no long-term continuous measurement of air pollutants and greenhouse gases has been performed at any regional background sites in southwestern China.
In this study, we present observations of major air pollutants, including O 3 , CO, SO 2 , NO x , PM 2.5 and PM 10 , as well as greenhouse gases, such as CO 2 , CH 4 and N 2 O, at the GGS (101°97′E, 29°55′N; elevation: 3541 m) in southwestern China for the first time.The occurrence levels and temporal variations in these pollutants are discussed in detail, and potential contribution areas of the above substances are identified using the potential source contribution function (PSCF) method.

Sampling Site
As shown in Fig. 1, the GGS (101.97°E,29.55°N; elevation: 3541 m) was established in the Gongga Mountain Observation and Experimental Station of Alpine Ecosystem, which is located in the Hailuogou scenic area of the southeastern edge of the Qinghai-Tibetan Plateau.The Hailuogou scenic area is famous for its large unique glacier and forest park areas.Neither private vehicular transport nor industrial activities operate near the sampling site.This station is approximately 250 km from Chengdu, the capital of Sichuan Province.There are two major roads in the north and east, 500 m and 400 m away from the station, respectively.The study area is dominated by the Southeast Asian Monsoon, with an annual average temperature, relative humidity, wind speed, atmospheric pressure and visibility range of 2.1 ± 6.6°C, 82.5 ± 20.9%, 1.7 ± 0.9 m s -1 , 656.5 ± 25.0 hPa and 21.7 ± 24.8 km, respectively.In addition, higher precipitation mainly occurs in both summer and autumn.

Instruments and Measurement Data
The sampling campaign was conducted from January 1, 2017, to December 31, 2017.All instruments were deployed, operated and maintained by following the regulations and standard operating procedures defined by the Ministry of Ecology and Environment of the People's Republic of China (http://bz.mep.gov.cn/bzwb/dqhjbh/jcgfffbz/index_2.shtml).The precision, detection limits, and calibration methods of all analyzers/monitors for major air pollutants of interest, including O 3 , CO, SO 2 , NO x , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 , have been described elsewhere in detail (Ji et al., 2014;Christiansen et al., 2015;Cortus et al., 2015).Briefly, O 3 , CO, SO 2 , NO-NO 2 -NO x , CO 2 , CH 4 and N 2 O were observed using an ultraviolet photometric analyzer (Model 49i; Thermo Fisher Scientific (Thermo), USA), a gas filter correlation nondispersive infrared method analyzer (Model 48i TLE; Thermo, USA), a pulsedfluorescence analyzer (Model 43i-TLE; Thermo, USA), a chemiluminescence analyzer (Model 42i-TL; Thermo, USA), a cavity ring-down spectroscopy analyzer for CO 2 and CH 4 (Model G2301; Picarro, Inc., USA) and a gas filter correlation N 2 O analyzer (Model 320EU; Teledyne Technologies, USA), respectively.PM 2.5 and PM 10 were simultaneously monitored using a Tapered Element Oscillating Microbalance with a Filter Dynamics Measurement System (TEOM-FDMS; 1405-DF TEOM; Thermo, USA).The gases were calibrated daily by injecting a mixture of calibration gases (Scott-Marrin, Inc., CA, USA) and scrubbed ambient air.In addition, an internal catalytic converter was used to calibrate the CO blank.The TEOM-FDMS was calibrated with free-particle and standard filters (Thermo, USA).Meteorological parameters such as relative humidity (RH), wind direction (WD), wind speed (WS) and atmospheric temperature (T) were recorded via a colocated automatic meteorological station (Model AWS310; Vaisala, Finland).All data were processed using an Igor-based software (Wu et al., 2018).

Source Area Identification
The PSCF method has been extensively used in identifying source locations of atmospheric species (Lupu and Maenhaut, 2002).PSCF is defined as the probability that an air parcel with a concentration more than a specified threshold reaches the study site after having resided in a certain grid cell of the spatial domain of interest (Lupu and Maenhaut, 2002).In this study, the potential source areas are identified based on the PSCF method using a GIS-based tool, named Trajstat, which can perform a comprehensive investigation of the geographical distribution of atmospheric species origins (Wang et al., 2009).
Frequency distributions are a commonly used visualization tool used to display the number of observations within a given interval.In this study, histograms showing the normalized, lognormal and cumulative frequency distributions of O 3 , CO, SO 2 , NO, NO 2 , PM 2.5 and PM 10 concentrations and CO 2 , CH 4 and N 2 O mixing ratios were plotted and are presented in Fig. 3.The results show that all of the air pollutants of interest in this study possessed a unimodal bell-shaped lognormal distribution pattern during the entire sampling period.For example, O 3 concentrations ranging from 30 to 130 µg m -3 were dominant, and the distribution shape for O 3 was generally symmetrical, with the highest frequency at 90 µg m -3 , occupying more than 95% of the total sample data.It was reported that the frequency distribution of O 3 concentrations can be a good indicator for the studied type of sampling site because this distribution captures the impact of NO x emission sources mainly associated with combustion activities (Escudero et al., 2014).Based on the diurnal variations in both O 3 and NO (Section Seasonal and Diurnal Variations), in such a remote site, NO titration most likely seldom affects the O 3 frequency distribution, although other processes, such as deposition or scavenging, might play more important roles in removing O 3 from the troposphere (Monks et al., 2015).The frequency distributions of CO, SO 2 , NO 2 , PM 2.5 and PM 10 were more skewed to the right compared with the other distributions and featured a pronounced peak at approximately 0.25 mg m -3 , 0.22 µg m -3 , 1.0 µg m -3 , 4 µg m -3 and 6 µg m -3 , respectively, indicating the impact of relatively clean background continental air masses, while the long tails on the right side of the graph toward high concentration levels with a low cumulative distribution could indicate the minor influence of local rural pollution and/or urban pollution plumes.NO concentrations ranging from 0.1 to 1.0 µg m -3 were dominant, accounting for approximately 98% of the total sample data.Although the , mg m -3 , µg m -3 , µg m -3 , µg m -3 , ppm, ppm, ppb, µg m -3 and µg m -3 , respectively (in this manuscript, all units for the observed O NO dataset was only slightly more than half of that for the remaining pollutants, the right-skewed frequency distribution of NO was similar to the above trend, with a long tail on the right side.In addition, Fig. 3 clearly shows that the frequency distributions of the CO 2 , CH 4 and N 2 O mixing ratios were very narrow and were in the ranges of 380.1-435.0ppm, 1.802-2.232ppm and 300.0-379.9ppb, respectively; such patterns might indicate that high emissions of these gases had almost no impacts during the study period. As shown in Table 2, a comparative analysis was conducted between the observations in the present study and those reported earlier at background stations in China and worldwide.The annual mean O 3 concentration (74.7 ± 22.0 µg m -3 ) at the GGS was higher than the values observed at the Dinghushan (DHS) background site in southern China (24.6 ppb; Chao et al., 2014), the Jinsha (JS) background station in central China (24.6 ppb; Lin et al., 2011) and the Lin'an (LA) background station in the YRD of China (34.7 ppb; Xu et al., 2008); and it was similar to the values observed at the Shangdianzi (SDZ) background station in the BTH region of China (36.9 ppb; Lin et al., 2008) and the remote highland site of Dangxiong   Given that the GGS site is at a much higher latitude than the other sites, the contribution of stratospheric ozone intrusion at the GGS may be more significant than the stratosphere-totroposphere transport at most of the other stations discussed here (Mauzerall et al., 1996).The concentrations of NO 2 at the GGS were lower than those at the DHS, SDZ, and LA sites, which was consistent with the fact that the DHS, SDZ, and LA sites are in rapidly developing areas with much more intensive industrial emissions and a high increase in the number of motor vehicles.In contrast, the concentrations of SO 2 and CO observed at the GGS were lower than those observed at the DHS, SDZ, JS and LA stations, which also indicates the limited influence of anthropogenic activities at the GGS.The annual mean CO 2 concentration (406.1 ± 9.5 ppm) at the GGS was similar to that at the background sites of Mt.Dodaira, Japan (> 400 ppm in 2013); King's Park, Hong Kong, China (407.6 ppm in 2013); Korea (404.9 ppm in 2013); Mauna Loa (global background site; 406.53 ppm in 2017) and LA (404.7 ± 8.2 ppm, 405.6 ± 5.3 ppm and 407.0 ± 5.3 ppm for 2009, 2010 and 2011, respectively).However, the annual mean CO 2 concentration at the GGS was higher than that observed from September 2006 to August 2007 at the Waliguan (WLG; 383.5 ppm), SDZ (385.9 ppm), LA (387.8 ppm) and Longfengshan (LFS; 384.3 ppm) background sites.This trend suggests that CO 2 concentrations at the abovementioned background sites in China have increased significantly over the past decade, possibly due to the rapid economic development and extensive enhancement of energy consumption (http://www.stats.gov.cn/tjsj/ndsj/2017/indexch.htm).The average CH 4 concentrations at the GGS were higher than those observed at the Antarctica site (1.740-1.766ppm), the WLG background site (1.864 ppm) and the Shangri-La background site (1.861 ppm) as well as the global background concentration (1.798-1.824ppm) (Bian et al., 2016) but were similar to those at the SDZ (1.914 ppm), LA (1.965 ppm) and LFS (1.939 ppm) background sites.In addition, the monthly average N 2 O concentrations, which ranged from 321.3 to 329.8 ppb in the GGS, were similar to those at Zhongshan Station, East Antarctica, which varied from 320.5 to 324.8 ppb (Ye et al., 2016), but higher than those at the Xinglong (316.7 × 10 -9 , from 1995 to 2000) and WLG (314.9 × 10 -9 , from 1995 to 2000) background sites.
Compared to the observations at other background sites in the world, the PM 2.5 concentrations measured at the GGS site (6.5 ± 6.2 µg m -3 ) were slightly lower than or similar to those at the Portugal (9.4 µg m -3 ), Germany (10 µg m -3 ) and Scandinavia (6.6 µg m -3 ) background sites and were considerably lower than those recorded at the Switzerland (14.5 µg m -3 ) and Austria (19.7 µg m -3 ) background sites.The annual PM 10 concentration was almost the same as a previous result (10.8 µg m -3 ) from the Stockholm regional background site in Europe (Jonsson et al., 2013).Although most background monitoring sites are distant from urban areas, they might be affected by anthropogenic emissions to different extents.In addition, synoptic conditions play an important role in PM levels (Ji et al., 2012); for instance, differences in the precipitation intensity and atmospheric stagnation of various sites could affect the removal, transport and accumulation of particulate matter.Furthermore, the transboundary transport of air masses from regions of intense anthropogenic activities can lead to spatial variation in the levels of air pollutants at various background sites.
Overall, the air quality at the GGS site was better than that at the background sites located in the most populated and developed city clusters, such as the BTH, YRD and PRD regions of China.In addition to meteorological conditions and various sinks, the spatial variation in the levels of air pollutants in different regions could be highly affected by regional emission intensities of major air pollutants of interest (National Bureau of Statistics of the People's Republic of China, 2017).

Seasonal and Diurnal Variations
The average seasonal variations in O 3 , CO, SO 2 , N 2 O, PM 2.5 , PM 10 , NO 2 , CO 2 and CH 4 and the RH, T and WS values at the GGS are shown in Fig. 4. The highest values of O 3 were observed in spring, while the lowest values occurred in autumn.The maximum O 3 value in spring is consistent with the spring O 3 maximum frequently observed in the Northern Hemisphere (Monks, 2000;Vingarzan, 2004;Ran et al., 2014;Lin et al., 2015).Stronger ultraviolet (UV) radiation has been observed in spring at the GGS (Liu et al., 2017), which was favorable for O 3 production and further contributed to the highest O 3 concentrations in spring.The lower levels of CO from April to July were almost opposite to the seasonal variation in O 3 , which could be caused by the yield of O 3 from CO oxidation under stronger UV radiation (Seinfeld and Pandis, 2016).The higher CO 2 concentrations observed in April and May might originate from CO oxidation, which supports the above deduction to a degree.However, vigorous summertime photosynthesis resulted in a decline in CO 2 concentrations in the summer.The O 3 levels were relatively lower in late summer and autumn compared with those in spring.The high occurrence frequency of precipitation could scavenge more O 3 in late summer and autumn than the other two seasons.SO 2 and PM 2.5 showed the highest concentrations in summer and the lowest concentrations in autumn.As shown in Fig. 2, high concentration spikes of both SO 2 and PM 2.5 could be observed.Given that the lifetime of SO 2 is normally short under conditions of high RH and O 3 concentrations (Seinfeld and Pandis, 2016), it might be reasonable to infer that local emissions outweighed the regional transport of both pollutants, resulting in the increased concentrations of RH and O 3 at this background site.Note that pulse spikes of SO 2 and PM 2.5 accompanied the increase in CO and NO 2 concentrations.In addition, PM 10 did not increase as significantly as PM 2.5 during the sampling period.Hence, all these observations could lead to the hypothesis that the local air quality at such a remote site might be affected more by traffic emissions during summer than during other seasons.Because the Hailuogou scenic area is most attractive to tourists during summer, this difference could be ascribed to vehicular emissions from tourists.The seasonal pattern of CO was similar to that of NO 2 and PM 10 .The highest CO, NO 2 and PM 10 values were mostly observed in winter.The lowest CO 2 concentrations were observed in summer, which could be because vigorous summertime photosynthesis resulted in a decline in CO 2 mixing ratios.The lowest CH 4 concentrations were recorded in summer, while the CH 4 levels in the other three seasons were very similar to each other.Such a seasonality of the CH 4 mixing ratio in a clean-background environment is possibly related to OH radicals that are dependent on the seasonally varying intensity of ultraviolet radiation (Necki et al., 2003).In contrast, higher N 2 O concentrations were observed in summer than during the other seasons due to increases in the emission of N 2 O caused by enhanced microbial activity at higher ambient temperatures (Wang et al., 2018).
The average diurnal variations in O 3 , CO, SO 2 , NO x , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 as well as meteorological conditions, namely, RH, T and WS, at the GGS are shown in Fig. 5.The diurnal cycles of O 3 showed minimum and maximum values in the early morning and afternoon, respectively.During the nighttime, the ozone concentration decreased slowly, mainly due to the titration of NO and

RH (%) T (°C) WS m/s)
deposition processes (Seinfeld and Pandis, 2016); during the daytime, with the increase in solar radiation after sunrise, O 3 started increasing as more ozone was generated by photochemical reactions (Seinfeld and Pandis, 2016).Note that the daily amplitude of O 3 at the background site was lower than that at urban sites (Reddy et al., 2011).The diurnal variations in CO and NO 2 resulted in higher concentrations in the daytime and lower concentrations in the nighttime.The diurnal variations in NO and N 2 O resulted in obvious peaks at approximately 09:00 and 10:00, respectively.No obvious diurnal variation in SO 2 was observed.The diurnal variations in CO and NO 2 might be ascribed to two reasons.One reason could be that increasing anthropogenic activities in the daytime around the study site enhanced the primary emissions of these gases.The other reason may possibly be related to the evolution of the planetary boundary layer (PBL) and the transboundary transport of air pollutants.It is known that CO and NO 2 concentrations increase with the decay of the nocturnal boundary layer, consequently making the high concentrations of these pollutants over source areas migrate to downwind receptor sites and leading to enhancements in CO and NO 2 concentrations at the receptor sites during the daytime.The PM 2.5 and PM 10 concentrations did not show the obvious bimodal distribution that they show at urban sites (Wang et al., 2015).Similar to CO and NO 2 , a slight increase in the PM 2.5 and PM 10 concentrations was recorded in the daytime.Although the GGS is a regional background site, scattered anthropogenic emissions in the daytime could partly lead to the increase in the PM 2.5 and PM 10 concentrations; in addition, secondary aerosol formation via the atmospheric transformation of precursors from local sources or regional transport might contribute to higher concentrations of PM 2.5 and PM 10 during the daytime than during the nighttime.
The average diurnal variation in CO 2 is shown in Fig. 5.The diurnal cycle of CO 2 reflects a diurnal periodicity of atmospheric vertical mixing, the release and uptake of CO 2 by respiration, photosynthesis in the biosphere and human activities.In the daytime, photosynthesis is dominant in the net exchange between the biosphere and atmosphere, which could lead to the decline in CO 2 concentrations.During the nighttime, photosynthesis stops and vertical mixing is weak, while plant respiration continues.The CO 2 emitted by plant respiration accumulates near the ground.In addition, anthropogenic emissions also contributed slightly to the observed CO 2 , although the GGS site was distant from densely populated areas.The CH 4 mixing ratios showed a very weak diurnal variation, with minima during midday, which might be caused by the elevated PBL, as anthropogenic emissions of CH 4 are scarce in such a remote mountain area.N 2 O peaked at 10:00-12:00 and remained almost stable over the other time period.The N 2 O peak can probably be attributed to N 2 O emissions caused by enhanced microbial activity with the increase in soil temperatures.With the increase in solar radiation in the daytime, the soil temperature did not immediately increase with the atmospheric temperature due to the different heat capacities of the soil and atmosphere.With the increase in the soil temperature (generally 2 h later than that in the atmospheric temperature), the enhancement of microbial activity led to N 2 O peaks.In addition, N 2 O peaks did not occur with NO peaks, suggesting that they did not come from the same sources, i.e., vehicular emissions (Vojtíšek-Lom et al., 2018).Guizhou Provinces.Considering that SO 2 has similar emission sources as CO, the potential source areas for SO 2 overlapped with those of CO, but an intense SO 2 band caused by the emissions of power plants appeared in parts of Inner Mongolia and Gansu Province (Liu et al., 2015).The potential source areas for NO 2 , PM 2.5 and PM 10 were also found in neighboring countries in South Asia in addition to Inner Mongolia, Gansu Province and the Cheng-Yu economic region.High emissions of air pollutants in the abovementioned areas may result in the increase in NO 2 , PM 2.5 and PM 10 concentrations at the GGS site via transboundary transport.This result is almost consistent with that obtained by Qu et al. (2008), who found that anthropogenic sources in the Cheng-Yu economic region (Sichuan Basin), southeastern Yunnan Province and South Asian countries evidently influence Zhuzhang (in a mountainous rural area of southwestern China, 3583 m a.s.l.).For CO 2 , the potential source areas were focused on South Asian countries neighboring China and the adjacent areas of the Tibet Autonomous Region, Yunnan Province and Myanmar, which could be caused by high fuel consumption resulting in high CO 2 emissions (http://edgar.jrc.ec.europa.eu/overview.php?v=CO2andGH G1970-2016).The potential source areas of CH 4 were mainly distributed in Myanmar, India and Pakistan and parts of Yunnan Province, China.The potential source areas of N 2 O were recorded in the Cheng-Yu economic region, Yunnan Province, borders of the Tibet Autonomous Region of China and countries of South Asia; it is understandable that these areas with intense agricultural activities are the main sources of N 2 O and CH 4 (Davidson and Kanter, 2014;Saunois et al., 2016).

CONCLUSION
This report presents the first year-long (from January 1 to December 31, 2017) and real-time measurement study of O 3 , NO x , SO 2 , CO, CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 performed at the GGS background site in southwestern China.The frequency of occurrence and temporal variations of this pollution were discussed in detail, and potential contribution areas of the above substances were identified using PSCF.The conclusions of this study are as follows.
High O 3 and PM 2.5 concentrations that exceeded WHO thresholds were observed and were closely associated with the yield from CO oxidation and long-range transport and/or local emissions of PM.The O 3 , CO, SO 2 , NO x , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 concentrations presented typical lognormal patterns during the entire study period.The concentrations of these air pollutants at the GGS site were comparable to those at most background sites around the world.The slight differences in the levels of major air pollutants and greenhouse gases between background sites may be affected, to different extents, by anthropogenic emissions.
Obvious seasonal and diurnal variations were observed in the concentrations of O 3 , CO, NO x , SO 2 , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 .High values were observed for O 3 in spring and summer, and the diurnal variations were characterized by higher daytime than nighttime values.SO 2 and PM 2.5 reached their highest concentrations in summer and their lowest concentrations in autumn.Relatively high CO, NO 2 and PM 10 concentrations were mostly observed in spring and winter.CO and NO 2 exhibited higher concentrations in the daytime than in the nighttime.The vigorous summertime photosynthesis of vegetation resulted in lower CO 2 concentrations during summer than during the other seasons.The lowest concentrations for CH 4 were recorded in summer, whereas its levels in the other three seasons were similar to each other.High temperatures enhanced microbial activity, resulting in higher N 2 O concentrations during summer (compared to the other seasons) and in the daytime (compared to the nighttime).
Meteorological conditions significantly affect the background concentrations of the studied air pollutants.These pollutants can accumulate because of stagnant meteorological conditions and/or the contribution of midand long-range transport.The potential source areas for the air pollutants and greenhouse gases showed an obvious spatial distribution.The high potential source areas were distributed in parts of Inner Mongolia and Gansu, Ningxia, Sichuan, Chongqing and Hubei Provinces as well as adjacent areas of Hunan, Guizhou and Guangxi Provinces.High emissions in India and Pakistan also played an important role in increasing the CO, SO 2 , NO 2 , PM 2.5 and PM 10 concentrations.In addition, anthropogenic emissions from Inner Mongolia and Gansu Province contributed greatly to the enhancement of CO, SO 2 , PM 2.5 and PM 10 .Increases in the CO 2 concentration due to high fuel consumption were traceable primarily to South Asian countries neighboring China, and the adjacent areas of the Tibet Autonomous Region, Yunnan Province and Myanmar, whereas the enhancement of CH 4 and N 2 O concentrations in the Cheng-Yu economic region, Yunnan Province, the Tibet Autonomous Region of China, Myanmar, India and Pakistan were attributable to intense agricultural activities.
The results of this study contribute to a better understanding of how anthropogenic activities impact the occurrence of air pollution against regional backgrounds, which may improve the modeling of regional air quality in the future.
the mountainous rural area of southwestern China 3583 m from August 2004-February 2005 PM 10 ~10 µg m -3 Qu et al., 2009 Conversion factors between ppb and µg m -3and ppm and mg m -3 can be found in the following link: https://ukair.

Fig. 5 .
Fig. 5. Average diurnal variations in O 3 , CO, SO 2 , NO x , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 concentrations or mixing ratios (a) as well as RH, T and WS; (b) at the GGS station.

Fig. 6 .
Fig. 6.The dependent distributions between wind speed and direction and the concentrations of O 3 , CO, SO 2 , NO x , CO 2 , CH 4 , N 2 O, PM 2.5 and PM 10 at the GGS station.

Fig. 7 .
Fig. 7. Potential source areas for O 3 , CO, SO 2 , NO 2 , PM 2.5 and PM 10 concentrations as well as CO 2 , CH 4 and N 2 O mixing ratios during the study period.The color code denotes the PSCF probability.The location of the site is indicated by .

Table 1 .
Average concentrations and standard deviations (STDEVs) of O

Table 2 .
Comparison of the present study with previous measurements from background stations in China and around the world.