Five-year Record of Black Carbon Concentrations in Urban Wanzhou, Sichuan Basin, China

The atmospheric fine particle black carbon (BC) was measured from June 2013 till February 2018 in Wanzhou District, the second largest metropolitan area in Chongqing Municipality, China, which is located in the eastern Sichuan Basin. The average daily concentration ranged from 0.5 to 10.4 μg m, with a mean of 4.4 ± 2.2 μg m, and the annual mean displayed a significantly decreasing trend, from 5.3 μg m in 2013 to 3.7 μg m in 2017. The frequency distribution of the average daily concentrations during the years 2014–2017 skewed toward lower values (compared to the rest of the study period), with over 60% falling between 1 and 5 μg m. The BC exhibited a seasonal pattern, with the highest concentrations being reported during winter, followed by spring and fall, and the lowest ones during summer, as well as a double-peaked diel variability all year round, with a morning peak occurring between 07:00 and 09:00 and an evening peak between 19:00 and 21:00. Furthermore, the concentration was negatively correlated with the planetary boundary layer (PBL) height and influenced by the strong scavenging effect of precipitation. An inverse relationship between the concentration and the wind speed was observed when the latter was below 2.0 m s. The wind direction also affected the concentration, with easterly and southeasterly winds accompanying higher levels of BC regardless of the season.


INTRODUCTION
Black carbon (BC), primarily originating from incomplete combustion of fossil fuel and biomass burning, is one of the important components in atmospheric aerosol, and has strong absorption of light (Jacobson, 2001;Ramanathan and Carmichael, 2008;Chow et al., 2009;Tian et al., 2018). In recent years, BC has attracted extensive attention because of its important effects on climate (Jacobson, 2002;Chung and Seinfeld, 2005;Schmidt, 2011), environment (Ramachandran and Rajesh, 2007;Huang et al., 2012) and human health (Kopp and Mauzerall, 2010;Janssen et al., 2011;Farzad et al., 2020). According to researches on global climate change, absorption of BC accounts for 80-90% of aerosols' contribution to global warming (Chung and Seinfeld, 2005;Kopp and Mauzerall, 2010;Rai et al., 2019). BC aerosol is considered to be the second largest contributor to anthropogenic radiative forcing after CO 2 (Jacobson, 2002;Panicker et al., 2013). In addition to enhancing the absorption of solar radiation, BC in clouds and aerosols also accelerates cloud evaporation rate, thus changing the duration of cloud existence (Quaas, 2011;Dalirian et al., 2018). BC can heat the air to alter the stability and vertical movement of the regional atmosphere, thus having an important impact on large-scale air circulation and water circulation with implication on crop yield reduction (Chameides et al., 1999;Menon et al., 2002;Richardson et al., 2018). BC is mainly present in fine particles in the urban environment, and can be inhaled into the human body and cause adverse health effects on the respiratory system (Sahu et al., 2011;Wang et al., 2011;Jung et al., 2017).
Research on BC has particular importance in China because the country has been regarded as the largest emitter of BC in the world (Novakov et al., 2003;Bond et al., 2007;Ramanathan and Carmichael, 2008;Zhou et al., 2018). BC emissions in China roughly accounted for one fourth of the global anthropogenic sources (Street et al., 2001). Numerous studies on BC have been conducted in Chinese megacities, such as Beijing Ji et al., 2016;Ji et al., 2017Ji et al., , 2019bLiu et al., 2019), Shanghai (Feng et al., 2014;Peng et al., 2019), Guangzhou Sun et al., 2020), Nanjing (Zhuang et al., 2014;Xiao et al., 2016), Xi'an (Cao et al., 2009;Wang et al., 2015a;Zhang et al., 2015;Zhang et al., 2018). Collectively, these studies produced valuable data on the distribution of BC in China, and also provided results on BC's optical property and radiative forcing. However, measurements from aforementioned studies were mostly short-term, and BC data of southwest China remains scarce.
Chongqing Municipality of southwestern China is known for its position at the hinterland of the Three Gorges Reservoir (TGR) on the Yangtze River. Wanzhou, situated in the eastern margin of the Sichuan Basin, is a district located 230 km away from downtown Chongqing. Wanzhou covers an area of 3457 km 2 and has a registered population of approximately 1.8 million at the end of 2017 (Introduction to Wanzhou, http://www.wz.gov.cn/zjwz/wzjj/202003/t202 00331_6849138.html). In recent years, the acceleration of urbanization and energy consumption growth in Wanzhou together increased the amount of air pollutant emission, threatening local air quality. In 2013, Chinese government issued Air Pollution Prevention and Control Action Plan (2013-2017) with aggressive measures starting in October 2013. Since then, Chongqing Municipal Government has adopted a series of corresponding measures to improve the city's air quality. Studying BC, a product of fossil fuel combustion and biomass burning, can shed light on the effectiveness of these emission control measures. However, to date no publicized data is available for BC aerosol in Wanzhou. In this study, we conducted continuous measurement of BC at an urban site in Wanzhou from June 1, 2013, to February 28, 2018. The purposes of this study are: 1) to evaluate the temporal variability of BC in Wanzhou, 2) to explore the correlation between BC and meteorological factors, and 3) to investigate the possible emission sources of BC.

Site Description
BC mass concentration was measured on the rooftop of the Experimental Building (30.79°N, 108.37°E) on Chongqing Three Georges University campus, at a height about 27 m above the ground (Fig. 1). Surrounding the site are residences, restaurants, and shops. At about 150 m east of the measurement site lies a major road, Shalong Road. There is no major stationary pollution source within 1 km around the sampling site. This site represents a mixed residential, traffic, and commercial environments of urban area in Wanzhou.

Black Carbon Measurement
BC concentration was monitored using a Thermo Carusso Model 5012 Multi-Angle Absorption Photometer (MAAP). The MAAP was equipped with an additional impactor system removing particles with aerodynamic diameter larger than 2.5 µm. The sampling flow of the instrument was maintained at 16.7 L min -1 at ambient temperature and pressure. Briefly, the instrument collected aerosol particles on a glass-fiber tape and measured light attenuation caused by the sampled particles. More detailed information about the BC analysis protocol is provided in the study of Kanaya et al. (2013). The time of day (hour) referred to throughout this paper is the starting time of each hourly measurement at local time. The study period was June 1, 2013, to February 28, 2018. Yearly sampling campaigns were conducted during the following periods: June 2013 to February 28, 2014;March 1, 2014, to February 28, 2015March 1, 2015, to February 29, 2016March 1, 2016, to February 28, 2017andMarch 1, 2017, to February 28, 2018.

Meteorological Data
Meteorological parameters (wind speed and direction) were collected simultaneously with BC monitoring using a miniature weather instrument (WS500-UMB; Lufft Corp., Germany). No direct measurement was available for planetary boundary layer (PBL) heights and precipitation during the period of BC measurement; we used the closest grid value of the reanalysis PBL height and precipitation to the Wanzhou as the real PBL height and precipitation. The PBL heights and precipitation during the study period were calculated by the U.S. National Oceanic and Atmospheric Administration (NOAA) READY archived meteorological data (http://www.arl.noaa.gov/ready/hysplit4.html). The program used the archived dataset GDAS (1°, 3-h, global) based on the average value of Coordinated Universal Time (UTC), and the time series of calculated daily PBL heights were obtained. All UTC values were converted to local time.

BC Concentration and Variability
Daily average BC concentration from June 1, 2013, to February 28, 2018, in Wanzhou ranged from 0.5 to 10.4 µg m -3 , with a mean level of 4.4 ± 2.2 µg m -3 , implying a dramatic day-to-day variation (Fig. 2). The annual average concentration of BC is shown in Table 1. There was a decreasing trend of the annual mean BC concentration, which dropped from 5.3 µg m -3 in 2013 to 3.7 µg m -3 in 2017 (30% reduction). Overall, at the 95% confidence level, the annual mean BC decreased by 0.33 µg m -3 year -1 from 2013 to 2017. The decline in BC concentrations is closely associated with various policy interventions taken at Chongqing. As part of mitigation implementation to control traffic pollution, Chongqing eliminated 0.26 million yellow-label cars and outdated vehicles between 2013 and 2017, while adopting Chinese national Stage V standard for diesel and gasoline vehicles and eliminating the sale and use of substandard oil. Additionally, the government had designated prohibited areas (nearly 4000 m 2 ) for high emission non-road mobile machinery (NRMM) and strengthened the supervision of NRMM usage in these areas. The total energy consumption in Chongqing accounted for 72.5 and 84.5 million tons of standard coal equivalent (SCE) in 2013 and 2017, respectively. However, the proportion of coal consumption in total energy consumption dropped from 63.2% in 2013 to 53.8% in 2017. Coal has been gradually replaced by natural gas and other cleaner energy in Chongqing (Chongqing environment statement in 2017, http://sthjj.cq.gov.cn/hjzl_249/hjzkgb/20 1912/P020191228643991023980.pdf). At the same time, Chongqing had closed some heavily polluting industries, had renovated coal-fired boilers with clean energy, and had promoted the ultra-low emission transformation of coalfired power plants. Additional mitigation measures were applied for curbing domestic pollution. For example, it had reformed the lampblack emission from catering business and canteens of government and public institutions, and had intensified law enforcement on illegal activities such as burning straw, garbage, and barbecuing in the open air. It could be inferred that BC concentration reduction over the past few years in Wanzhou might be related to the implementation of these governmental mitigation measures.    pollution in Wanzhou was worse. Overall, BC concentration in Wanzhou was moderate compared to major cities in China; however, taking in consideration of Wanzhou's city size, population, and economy, its BC contamination was concerning.

Seasonal Variations of BC Concentrations
Monthly and seasonal trends of BC during 2013-2017 were evident (Figs. 4 and 5). The lowest monthly average BC concentrations occurred in July or August, while the highest concentrations occurred in December or January. Monthly BC concentration varied by nearly fourfold, from a minimum of 2.3 ± 0.5 µg m -3 (August 2017) to the maximum of 8.9 ± 1.6 µg m -3 (January 2014). As shown in Fig. 4, a decreasing trend for BC was clear in all months; October and January are the main months that contributed to the decrease in annual average BC concentration. The decreasing trends in October and January are significant at the 95% confidence level. The peaks decreased at approximately 0.89 and 0.97 µg m -3 year -1 in October and January, respectively. Seasonal BC concentration was highest in winter, followed by fall and spring, with summer showing the lowest concentration. Similar seasonal variation of BC concentrations can be found in other Chinese sites Feng et al., 2014;Ji et al., 2016;Chen et al., 2019).
The seasonality of BC concentrations is mainly affected by meteorological factors and anthropogenic activities. Precipitation was usually abundant during summer (Fig. 7)   and is an important mechanism that scavenges air pollutants, resulting in the lowest BC concentrations occurring in summer. At the same time, the PBL height was higher in summer which facilitated the dispersal of BC. Additionally, anthropogenic BC emission from fossil fuel over China was lower in summer while higher in winter (Zhuang et al., 2014;Zhang et al., 2009), also contributing to low BC concentrations in summer. High BC concentrations in fall might be due to the increased biomass burning such as burning of leftover crop residues (Huang et al., 2018). For spring, diffusion conditions were limited compared to those in summer, and could thus result in higher BC concentration than summer.

Diel Variations of BC Concentrations
The average diel BC mass concentrations measured in four seasons are shown in Fig. 6. In all seasons, a bimodal distribution of BC concentration was marked with a morning peak at 07:00-09:00 and an evening peak at 19:00-21:00. Meanwhile, the lowest BC concentration occurred between 14:00-15:00 for all seasons. This diel profile for BC is similar to those observed at different locations (Sahu et al., 2011;Chen et al., 2014;Feng et al., 2014). The morning peak results from morning rush hour traffic, suggesting an important role of vehicle emissions. In addition, atmospheric inversion in the morning can also lead to the high BC concentration. After sunrise, the increase in solar radiation promotes the development of daytime mixing layer, which facilitates the dispersion and dilution of pollutants, subsequently lowering BC concentration from morning hours. The BC evening peak could be attributed to afternoon traffic emissions and cooking, combined with low mixing heights. After midnight, due to the reduction of human activities and vehicle emissions, BC concentration decreased slightly, but remained at high levels. The BC concentrations at nighttime (00:00-06:00) maintained at 2.9-3.4 µg m -3 in summer, 3.7-4.3 µg m -3 in spring and fall, 5.5-6.2 µg m -3 in winter, respectively. This may be attributed to the formation of a nocturnal boundary layer and stagnant air flow which favor the accumulation of pollutants.
The specific timing of BC's diel variability was distinct among the seasons. The morning peak of BC concentration in winter appeared at 09:00, which was 1-2 hours later than other seasons. It could be associated with the inversion layer and the human activity regularity. During winter morning, the inversion phenomenon occurs more frequently and longer in mountain area (Wang et al., 2015b), which inhibits vertical diffusion and facilitates the accumulation of BC. Meanwhile, human activities were later than other seasons, leading to a delay in BC emission. In addition, the highest ratio between maximum and minimum of diel BC concentrations was found in summer (1.9), and the lowest in winter (1.3), reflecting the different diel variation of meteorological conditions in different seasons.

Relationships to Meteorological Conditions
PBL height, precipitation and wind speed are three important meteorological parameters affecting BC concentration. Monthly average BC mass concentration together with meteorological parameters in Wanzhou are plotted in Fig. 7. The PBL height determines the volume through which surface-emitted pollutants can be diluted and reflects boundary layer turbulence (Stull, 1988). It presented a strong seasonal variation in Wanzhou. The highest seasonal PBL height was observed in summer (465 m), which was 1.6 times higher than those measured in fall (289 m) and winter (297 m). A significant negative correlation coefficient (r = -0.58) was found between BC concentration and PBL height, as shown in Fig. 7(a). As mentioned in Section 3.2, the average BC concentrations were higher in winter (6.0 µg m -3 ) than in fall (4.3 µg m -3 ). However, the PBL height was almost the same in winter and fall. This result implies that PBL height was not the main reason for the higher BC concentration in winter.
Precipitation is an important mechanism of washing out particles as well as BC. As shown in Fig. 7(b), most of the precipitation occurred between June and October in Wanzhou, accounting for 59-66% of the annual precipitation (1198 mm). There was a significant negative correlation (r = -0.74) between precipitation and BC concentration. However, a poor correlation between BC and rainfall was found at Xi'an (r = -0.35), in northwestern China (Cao et al., 2009) and at Beijing (r = -0.27) in northern China . This distinct feature may be due to the more frequent precipitation in southwest of China compared to the northwest (Dai et al., 2014). Among the meteorological parameters, wind speed is an important factor determining BC concentration Park et al., 2015). Fig. 7(c) shows the monthly wind speed from 2013 to 2017. The wind speed is relatively low, with monthly averaged values generally ranged from 0.7 to 1.1 m s -1 . BC was observed to be negatively correlated with the wind speed (r = -0.64). This result is similar to the findings of many previous studies (Bhat et al., 2017;Kucbel et al., 2017).
The hourly wind speeds and corresponding BC concentrations are plotted in Fig. 8. The calm conditions (wind speeds < 0.5 m s -1 ) occurred frequently in Wanzhou, accounting for 22% of the entire measurement period, during which the highest average BC concentration (5.5 µg m -3 ) appeared. Wind speed between 0.5 and 1 m s -1 was dominant (51%), and the corresponding average BC concentration was 4.4 µg m -3 . When wind speed reached 1.5-2.0 m s -1 , the average BC concentration dropped to 3.0 µg m -3 (45% reduction). A clear inverse relationship between BC concentration and wind speed was found for wind speed lower than 2.0 m s -1 . However, when the wind speed was higher than 2.0 m s -1 , no clear relation between wind speed and BC concentrations could be seen. Similar relationship between BC and wind speed have also been reported for other urban locations globally, such as Helsinki, Finland (Pakkanen et al., 2000), and Urumqi, China (Li et al., 2012). The contribution of hourly BC from different wind directions is illustrated by wind rose plots in Fig. 9. The prevailing winds in Wanzhou were mainly from eastern and southeastern directions in different seasons. The main road, Shalong Road, is located to the east of the sampling site, so it can be speculated that the higher BC concentration from this direction may be affected by traffic emissions. Since the local industrial area lies in southeastern Wanzhou, relatively high BC concentration from this direction was likely due to the influence from industrial emissions. In addition, it is noteworthy that frequencies of wind directions between 180° to 270° increased in fall and winter, and BC concentration observed by these directions was considerably high. Patches of farmlands lie to the west of the sampling site; higher BC concentration from these directions could be attributed to the burning of crop residues on these farmlands.
The conditional probability function (CPF) was used to identify the likely locations of local emission sources affecting concentrations of BC in sampling site. BC data obtained at wind speeds < 1.0 m s -1 were excluded in the CPF analysis due to the isotropic behavior of wind vane under low wind speeds. Fig. 10 shows the CPF concentration plots of BC when the threshold was set at the upper 25 th percentile of the concentrations. The emission sources are likely located in the directions that have high conditional probability values. There was a clear indication of higher probability for these concentrations from the southeast, i.e., in the direction of industrial emissions. Additionally, there was high probability to the west, corresponding to the direction where some farmlands lie. Combining Fig. 10 with Fig. 9 produces an interesting feature: The highest concentration occurred under very low wind speed conditions from all wind directions, particularly the east, corresponding to the direction where the main road lies. Thus, it is suggestive that BC concentration was mainly affected by traffic emissions at low wind speed (< 1.0 m s -1 ), while BC concentration was affected by transport from industrial emissions and biomass burning at high wind speed (> 1.0 m s -1 ).

CONCLUSIONS
In this study, BC was measured in Wanzhou metropolitan area from June 2013 till February 2018 in order to investigate its temporal variability and potential sources. The average daily concentration ranged from 0.5 to 10.4 µg m -3 , with a mean of 4.4 ± 2.2 µg m -3 , and the average annual concentration exhibited a declining trend, decreasing from 5.3 µg m -3 in 2013 to 3.7 µg m -3 in 2017. Although these values qualify as moderate in China, BC contamination is nevertheless an issue when considering the size, population, and economy of Wanzhou. The seasonality of the BC-specifically, the occurrence of the minimum and maximum concentrations during summer and winter, respectively-can be attributed to changes in the emission sources and variability in the ambient meteorological conditions, especially the low precipitation and shallow PBL height during winter. The concentrations peaked during the morning between 07:00 and 09:00 and again during the evening between 19:00 and 21:00, likely as a result of anthropogenic activities, including traffic, and dropped to their lowest daily values between 14:00 and 15:00, which can be ascribed to the development of the PBL.
The BC concentration was found to be negatively correlated with the PBL height (r = -0.58), rainfall (r = -0.74), and wind speed (r = -0.64). Furthermore, winds blowing from the east and southeast, the most frequent directions, were accompanied by higher concentrations during all of the seasons, and CPF analysis indicated a high probability of sources to the southeast, where industry is located, and to   Winter 25% upper BC level (wind speed > 1.0m/s ) the west, where farmlands lie. Thus, at low wind speeds (< 1.0 m s -1 ), traffic emissions may have been the primary factor influencing the BC concentration, whereas at high wind speeds (> 1.0 m s -1 ), industrial emissions and biomass burning potentially played the largest roles.