Analysis of SO2 Pollution Changes of Beijing-Tianjin-Hebei Region over China Based on OMI Observations from 2006 to 2017

Sulfur dioxide (SO2) in the planetary boundary layer (PBL) as a kind of gaseous pollutant has a strong effect regarding atmospheric environment, air quality, and climate change. As one of the most polluted regions in China, air quality in Beijing-Tianjin-Hebei (BTH) region has attracted more attention. +is paper aims to study the characteristics of SO2 distribution and variation over BTH. Spatial and temporal variations for a long term (2006–2017) over BTH derived fromOMI PBL SO2 products were discussed. +e temporal trends confirm that the SO2 loading falls from average 0.88DU to 0.16DU in the past 12 years. Two ascending fluctuations in 2007 and 2011 appeared to be closely related to the economic stimulus of each five-year plan (FYP). +e spatial analysis indicates an imbalanced spatial distribution pattern, with higher SO2 level in the southern BTH and lower in the northern. +is is a result of both natural and human factors. Meanwhile, the SO2 concentration demonstrates a decreasing trend with 14.92%, 28.57%, and 27.43% compared with 2006, during the events of 2008 Olympic Games, 2014 Asia-Pacific Economic Cooperation (APEC) summit, and 2015Military Parade, respectively.+e improvement indicates that the direct effect is attributed to a series of long-term and short-term control measures, which have been implemented by the government. +e findings of this study are desirable to assist local policy makers in the BTH for drawing up control strategies regarding the mitigation of environmental pollution in the future.


Introduction
Sulfur dioxide (SO 2 ) is a short-lived gas primarily produced by volcanoes, power plants, refineries, metal smelting, and burning of fossil fuels.When SO 2 remains near the Earth's surface, it is toxic, causes acid rain, and degrades air quality.It forms sulfate aerosols that can alter cloud reflectivity and precipitation in the free troposphere [1,2].As a kind of important atmospheric pollutants, SO 2 critically affects the global environment,climate change, and public health.SO 2 has become one of the popular research topics in the past decades, to examine its changes over some of the world's most polluted regions [3][4][5][6][7][8][9][10].
China, with its incredible economic growth has been the focus of many studies during the previous decade because of its increasing sulfur dioxide emissions' contribution to the Earth's atmosphere [11][12][13][14][15][16]. e sources of SO 2 are both natural (volcanic) and anthropogenic emissions.Natural emissions include intentional biomass burnings and volcanic eruptions.Anthropogenic emissions are mainly due to fossil fuel burning (e.g., coal and oil), which accounts for more than 75% of global emissions [3].Anthropogenic SO 2 emissions are predominantly in or slightly above the planetary boundary layer (PBL), impacting on regional variations of aerosol types [17].
Satellite measurements of trace gases have been widely used and been an essential way to provide global, consistent observations for detecting, monitoring, and quantifying the SO 2 .e first space-based quantitative data on SO 2 mass of the El Chichon volcanic eruption in 1983 were obtained from Total Ozone Mapping Spectrometer (TOMS) on board Nimbus 7 [18].Subsequently, anthropogenic SO 2 sources from power plants in eastern Europe [19,20] and smelters in Peru and Russia [21] were demonstrated through detection of SO 2 emissions using Global Ozone Monitoring Experiment (GOME) measurements on the Earth Research Satellite 2 (ERS-2).e tropospheric SO 2 were detected by the SCanning Imaging Absorption spectroMeter for Atmospheric CHartographY (SCIAMACHY) on board the ENVISAT [22] and the Global Ozone Monitoring Experiment-2 (GOME-2) instrument on MetOp-A [23].
e Ozone Monitoring Instrument (OMI) on NASA's Aura spacecraft enables to provide daily, nearly global measurements of ozone columns and aerosols, and the trace gases with the highest spatial resolution and the longest data record currently available [24,25].
OMI data have been applied to assess the e ect of pollutant transmission, analyze pollutant source contribution, evaluate pollutant emission inventory, observe regional pollution changes, and quantify the reduction of power plant emissions [14,16,17,[26][27][28], due to the higher spatial and temporal resolution.ree major air pollutants (NO 2 , SO 2 , and CO) in China before, during, and after the Olympic Games from Aura's Ozone Monitoring Instrument (OMI) and Terra's Measurements of Pollutants in the Troposphere (MOPITT) instrument have been measured [29].Annual emissions by sector and fuel types calculated from satellite data show an increasing trend of SO 2 during 1996-2008 and decreasing thereafter in China [16].Substantial changes in SO 2 emissions in the northern China for the period 2005-2008 were analyzed [14].e spatiotemporal variation of SO 2 concentration during 2005-2008 over China from the planetary boundary layer (PBL) SO 2 column concentration retrieved from OMI has been analyzed [28].Long-term SO 2 pollution changes over China or region have been observed through OMI observations [29].A long-term trend of NO 2 and SO 2 levels (2005-2014) of the Henan province in China has been retrieved from the OMI [30].In the past decades, China has adopted di erent policies for air quality control consistently, such as carbon reduction, energy saving, and other measures to reduce SO 2 emissions [31,32].e latest ndings represent that large reductions in SO 2 are bene ting from the e ective control policies in China [33,34].
In this study, we analyze trend variation and distribution in SO 2 concentrations over the Beijing-Tianjin-Hebei (BTH) region observed by the OMI between January 2006 and December 2017.Figure 1 shows the multiyear average spatial distribution map of SO 2 based on OMI data (2006-2017) over China, which clearly shows the hotspots of SO 2 in the North China Plain.Meanwhile, it is noted that the BTH region is one of the most polluted regions in China.Accordingly, tight emission control arrangement of this area always adopted, the SO 2 emission of Hebei rank 3rd in 2013 went down to 5th with 17.4% rate of decline among all provinces in China, and Beijing and Tianjin declined 24.1% and 17.2%, respectively.Regional SO 2 time evolution and spatial distribution are discussed in the following.e datasets and monitoring area are presented in Section 2. e analysis and associated ndings are described in Section 3. Finally, the main conclusions are summarized in Section 4. e Ozone Monitoring Instrument (OMI) is a sun-synchronous polar orbiting Dutch/Finnish sensor on the AURA satellite launched on 15 July 2004.e science goals of OMI are directly related to these questions and focus on (1) measuring the ozone layer and its destroying trace gases BrO and OClO, (2) tropospheric pollution by ozone, nitrogen dioxide, tropospheric aerosols, SO 2 , and formaldehyde, and (3) detection of species important for climate change such as aerosols, clouds, and ozone.e OMI measures the radiation backscattered by the Earth's atmosphere and surface over the entire wavelength range from 270 to 500 nm, with a spectral resolution of about 0.5 nm, and high spatial resolution (13 × 24 km 2 ), and daily global coverage [24].OMI data have 4 processing grade products: Level-0, Level-1, Level-2, and Level-3.For this study, we used the PBL SO 2 vertical column density from the Level-3 0.25 × 0.25 degree gridded OMI/AURA SO 2 data product.e data used here with the time span of January 1, 2006 to December 31, 2017, were obtained from Giovanni interface (http://giovanni.gsfc.nasa.gov/giovanni/), derived from the NASA Goddard Earth Sciences Data Active Archive Center (GES DISC; http://disc.sci.gsfc.nasa.gov)[35,36].ese Level-3 products have been widely used [15,34,37,38].

Method Description
e original OMI PBL SO   Advances in Meteorology product employed the band residual di erence (BRD) algorithm [39].But this product has a high noise level and systematic artifacts that required empirical corrections [5,6].A new operational OMI PBL SO 2 product produced with the principal component analysis (PCA) algorithm was released [40,41].
Validation of these two algorithmic products has been analyzed [42].Daily satellite observations were retrieved with the given longitude and latitude for China and BTH region, as shown in Figure 2, to gain insight into the distribution of SO 2 columns in the BTH region.e data were gridded onto monthly 0.25 × 0.25 elds and then onto seasonal and yearly maps.e meteorological conditions used for analysis are from NCEP reanalysis data (https://www.esrl.noaa.gov/psd/).

Study Area.
Our study area focuses on the Beijing-Tianjin-Hebei (BTH) region (Beijing, Tianjin, and Hebei integration), which is the most polluted industrialized regions in China (Figure 1).e BTH region is located in the northwest part of the North China Plain as shown in Figure 2 (36 °05′-42 °37′N, 113 °11′-119 °45′E), with a total area of 216,000 km 2 and more than 110 million residential population.
As one of the most economically vibrant regions in China, the BTH region covers only 2.3% of the Chinese territory but generates over 10% of the total national gross domestic product (GDP) in 2016 (National Bureau of Statistics of China (NBSC)) [43].As the main high-tech and heavy industry base of China, there are mainly the automotive industry, electronic industry, machinery industry, iron industry, and steel industry.e SO 2 map (Figure 2) shows hotspots associated with the major coal-red power plants and industrial activities.Figure 3 reveals that high sulfur coal-red power plants are the major contributors to the SO 2 concentrations over the BTH region [44,45].OMIderived spatial distribution shows generally good agreement with these main anthropogenic emission sources from burning sulfur-contaminated fossil fuels as Figure 3. e last decade has seen frequent occurrences of severe air pollution episodes (haze), and the high SO 2 loading observed certainly contributed to PM 2.5 problems, especially in winter.Atmospheric environment quality has attracted more and more attention related to air pollution prevention and control policy-making.

Results and Discussion
e temporal and spatial variations of SO 2 concentration in the region of Beijing-Tianjin-Hebei during the period of 2006-2017 were analyzed based on the satellite OMI data to characterize the variation of SO 2 columns.e brief period of emission growth can probably be attributed to the government stimulus for resurgence of economy in response to the global nancial crisis of 2007-2008.Subsequently, there was a sharp decrease with 37.48% reduction over the 7-year period during 2011-2017.rough the seasonal variation tendency (Figures 4(b)-4(e)), the increase in 2011 mostly comes from summer, accompanied by the industrial production slowdown in the latter half of 2011.

Overall Temporal
As an industrialized and populated region, the levels of BTH air pollution are determined by population density, economic activity, type of power generation and fuel used, and regulatory policies [47].Here, we examine the SO 2 changes with the regional SO 2 emission data derived from national statistics (http://www.stats.gov.cn/tjsj/).It is interesting to nd out the yearly emission trend that agrees well with the trend based on OMI retrieved SO 2 as mentioned earlier (Figure 5), along with the estimated SO 2 emissions data from power plants.Some mismatch is expected considering the di erent observation means and uncertainties with satellite and ground measurements.
Two uctuations occur in the long trends; therefore, we broke down the data into two periods, 2006-2010 and 2011-2017, to verify the coherence with precursors' investigation [30,48] as to the changes in SO 2 loadings.Figure 6 shows the 5-year mean SO 2 concentration distribution maps over the BTH region of China's 11th ve-year plan (2006-2010) and second 7 years mean SO 2 meets China's 12th veyear plan (2011-2015) and China's new 13th ve-year plan (2016-2017), respectively.e dramatic decrease in SO 2 loading (Figure 6) well illustrates the achievements and improvements due to a series of air pollution control policies.e average SO 2 was 0.87 DU for 2006-2010 (Figure 6(a)).e average concentration decreased to 0.72 DU for 2011-2015 and 0.23 DU for 2016-2017.
According to the above statistics (Figure 5), it is more clearly shown that the SO 2 loading over the BTH peaked in 2007, and then presented overall decreasing trend.is can be attributed to the emission control measures taken by the government.e sharp decline from 2007 to 2010 is closely related to the installation of ue gas desulfurization (FGD) and follow-up e ects of strict pollution reduction measures implemented before the 2008 Beijing Olympic Games.e results also show that atmospheric SO 2 loadings in BTH have drastically decreased by 17.24% (2011-2015) and 75.29% (2016-2017) related to the average quantity in 11th ve-year plan (2006-2010), due to more stricter emission reduction targets, such as ue gas desulfurization control of enterprises, new energy to replace polluting energy, 50% of privately owned vehicles were banned through an odd and even number system, and switching from coal to natural gas for heating.And it is noted that the SO 2 loadings have been with a short-lived upswing in the early of 11th and 12th ve-year plan that may be caused by the government's economy stimulus.In addition, the SO 2 in the PBL has short lifetimes during the warm season, on the short time scale of days to months, and meteorology plays an important role in regional air pollution [6].e pollutants in the atmosphere have a dilution e ect.e meteorological data of 2007, the highest loading over the past twelve years, were chosen as an example to interpret the seasonal di erences (Figure 9).In summer, near-surface temperature is higher and air convection is stronger, besides relative humidity and amount of precipitation reaches the annual maximum.All these aforementioned basics can speed up the di usion of atmospheric pollutants, which have given birth to short lifetimes of SO 2 loading.On the contrary, the temperature, relative humidity, and precipitation in winter go against the transformation from SO 2 to sulfate, making the SO 2 with the longest lifetime in winter [48].

Spatial Distribution Pattern.
In Figure 10, the spatial distribution of the average SO 2 column in the Beijing-Tianjin-Hebei region from 2006 to 2017 is presented.As shown in Figure 10, the high SO 2 concentration in BTH is signi cantly distributed in the southwest and eastern regions.
As for the thirteen cities, multiyear average SO 2 clearly presents the spatial distribution discrepancy.We set up two borderlines with SO 2 column amount (1.0 DU and 0.5 DU) to identify hotspots over BTH (Figure 11).e SO 2 concentrations of Handan, Xingtai, and Shijiazhuang in the southwest of BTH are more than 1.0 DU.Cities with SO 2 concentrations exceeding 0.5 DU include Tangshan, Tianjin, and Qinhuangdao located in the coastal beach areas of Bohai Bay, Hengshui, and Cangzhou adjacent to the Shandong province, as well as Baoding and Langfang around Beijing. e SO 2 loading of Zhangjiakou and Chengde in the north BTH region cities are lower than the other eleven cities less than 0.5 DU.
e characteristic of spatial distribution over BTH can be interpreted from both natural factors and human activities.As shown in Figure 12, the terrain declines semicircularly from the northwest to the southeast over the BTH region due to the mountains and plains landforms.Figure 13 also shows the landscapes in this region, plateaus, mountains, and hills account for 54% of the entire BTH region centering on the area of northwest.Surface pressures gradually decrease from eastern to the northwest over the mountains (Figure 14).e central and southeast plains account for 46% of the region.
Forests and grasslands are representative natural ecological landscapes in the region and account for 38% of the study area, and they are mainly distributed in the Yanshan Mountains, the Taihang Mountains, and the northwest edge of the Inner Mongolia Plateau [49].We also found that relative humidity is di erent in this region due to the terrain and surface coverage (Figure 9).e whole area is mostly above 30%, especially the northwest is with higher values above 50% in winter, and wind also plays a major part in contributing to the SO 2 spatial distribution.Wind direction and wind speeds determine the pollutant transmission path and di usion velocity.e southern part of BTH underlies the leeward area which makes the SO 2 di cult to di usion and dilution, leading to the SO 2 accumulation.However, the area of northern BTH located in the upwind accompanied with high wind speeds in favor of SO 2 di usion as shown in Figure 14.From the point of human factors, coal, petrication, motor vehicles, and iron and steel industrial emissions are the major source of the BTH region, which has a close relationship with the spatial distribution of population.e spatial pattern of population over BTH exhibits that the southern plain area is more likely to cause human activities variance than the northern region.e southern region is so at and densely populated that human activities and industrial emissions greatly in uence the environment.
Besides natural factors discussed above, anthropogenic factors also take up signi cant role, contributing to heavy pollution.e BTH region is adjacent to Inner Mongolia, Shanxi, Shandong, and Henan, which is the top ve major pollution emission provinces in China (Figure 12), where gathering the most heavily emission sources in China is shown in Figure 3.As shown in Figure 15, we calculate the annual SO 2 columns for each city of the BTH region.Handan, Xingtai, and Shijiazhuang as the emission hotspots will be served for further analysis.At rst, the high SO 2 loadings in the southwest of BTH around the province of

Advances in Meteorology
Shanxi has nearly a hundred of coal-red power plants that help contribute to the higher value.Secondly, the BTH region has the largest iron and steel industrial scale in China, among that Beijing, Tianjin, Shijiazhuang, Handan, and Tangshan occupies the majority.Meanwhile, these cities have been the most polluted in BTH region due to unfavourable factors, such as meteorological and geographic conditions and population density.Tangshan, the iron and steel producer of Hebei province, is the second most polluted city because of its high smoke SO 2 emission.Although population and enterprises of Beijing are high, the highest green coverage rate and strict pollution control measures make the pollution level relatively low in the BTH region.Zhangjiakou and Chengde had a low SO 2 column with the least amount of population density, industrial enterprises as well as the terrain advantage.e amplitude of each urban curves variation is able to reveal the degree of being in uenced by natural and anthropogenic sources.As shown in Figure 16, the SO 2 average concentration for the period of Olympic Games decreased signi cantly to 14.92% and 7.76% compared to the neighboring years; the average SO 2 concentration in the APEC conference was signi cantly lower than that of the same periods in past ten years, and it declined 28.57% compared to the average of November in other years between 2006 and 2015; and the SO 2 concentration in September of the 2015 Military Parade, with 27.43% decline, reached the lowest value compared to ten years before.Standard deviations can re ect that more clearly.Besides, the standard deviations verify the previous analysis of seasonal characteristic and brief ascent in 2011.
During the three events, air pollution control policies have been reinforced.To achieve the goal of "green Olympic Games" [55], China has released a series of air pollution control policies to improve air quality.
e government has implemented a series of long-term pollution reduction measures, such as coal-red power plant in Beijing to install the desulfurization equipment, part of closure of small power plants in this area near Beijing during the Olympic Games, and about 94% of the small coal-red boilers to use clean energy transformation.e government also implemented some short-term strategies; for example, from July 1st to September 20th of 2008, the vehicles with exhaust emissions that failed to meet the European No. 1 standard were all-day forbidden on the roads; from July 20th to September 20th, the odd/even license plate number rule was applied on personal vehicles in Beijing [12]; power generation facilities were run only 30% of the equipment to stop all construction activities; some heavy-polluting factories were closed during the Olympic Games; and some heavily polluting  Advances in Meteorology companies around Beijing city were closed [29,31].During the APEC meeting and the Military Parade, the government also implemented a series of measures to ensure the safety and environmental protection measures, and these measures are similar even more and stricter than the 2008 Olympic Games, as with Beijing, the surrounding six provinces also have taken similar measures [27,56].During the period from August 20th to September 3rd of 2015, Beijing adopted to strengthen urban transportation management to strictly limit the motor vehicle population [57].Advances in Meteorology events to con rm the e ectiveness of environmental management measures.e relative changes clearly demonstrate that these measures were e ective in reducing SO 2 concentration during the periods of those a airs taken place.

Conclusions
In this study, the past   and temporal trends in SO 2 emission over the Beijing-Tianjin-Hebei region.e SO 2 loading distribution has close correlation with their emission sources.Spatiotemporal variation characteristics over BTH can more clearly reflect the natural and anthropogenic emission sources, to provide references to air pollution prevention and control.e main conclusions are as follows: (1) e temporal changes (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)  (3) Spatial distribution of BTH is also characterized, which is interesting to find that progressive changes from the northern to the southern regions are associated with the terrain and surface coverage, besides industrial pattern and population distribution.
In the cities of BTH, we found that rapid SO 2 reductions generally correlate well with sharp reductions in industrial activity.(4) China made great progress in pollution control with implementation of a series of major policies.e concentration of SO 2 during 2008 Olympic Games, 2014 APEC, and 2015 Military Parade have been compared to identical months in other years, and the result shows that concentration of SO 2 column was relatively low in these episodes because of reinforced emission controls.ese findings demonstrate that SO 2 concentration over BTH do not follow simple linear trends, but instead reflect a repercussion of environmental measures and political economic activities.In terms of temporal changes, the decreasing trend has been observed since 2011, mainly due to the government efforts to restrain emissions from the power and industrial sectors.On the other hand, spatial characteristics are regarding with natural and anthropogenic factors.One of the important goals in this work is to evaluate the effectiveness of a series of policy to control the pollution problems, and three typical time quantum status reflects that the national environmental pollution control measures have taken an evident effect.ese findings can provide a basis for the development of environmental management measures during the Winter Olympics in 2022, also for the national environmental pollution prevention and control as a reference.
e results clearly illustrate effectiveness of central governmental policies regarding emission mitigation of SO 2 and aid in important policy implications for the future reduction action plans to provide better air quality of China.
e SO 2 emission data have been publicly released from http://www.stats.gov.cn/tjsj/.e data used to support the findings of this study are included within the article.

Figure 2 :
Figure 2: Location of the study area in this study.e right image represents the region of Beijing-Tianjin-Hebei.

Figure 3 :Figure 4 :
Figure 3: Geographic distribution of the major SO 2 sources in China.

4
Advances in Meteorology2007 with an upward of 14.96% compared to 2006, re ecting the total SO 2 emission in China had substantially increased from 2000 to 2007.e phenomenon has been largely driven by expansion in manufacturing industries and fueled by coal for the Chinese economic growth[14].And then, an obvious downward trend appeared from 2007 to 2010, and the decline was 12.92% in 2007-2008 and 13.23% in 2008-2009, respectively.Afterwards from 2009 to 2010, the downward trend slowed down with the rate of decline 0.68%.e decrease was mainly due to China's 11th FYP requiring power plants to install FGD devices.Nonetheless, there was a temporary rebound from 2010 to 2011 with 19.19% growth rate.

Figure 5 :
Figure 5: Annual average SO 2 column (DU) based on OMI (black circle dots) and SO 2 emission statistics (blue asterisks) in (a) Beijing, (b) Tianjin, and (c) Hebei province (statistical data for 2017 have not been published).

10
Relevant conclusion comes out that any air pollution activities by human beings in the entire BTH region should be controlled with the most rigorous management.Beijing as China's political and administrative center has strict pollution control measures and high execution e ciency.Meanwhile, the environment policies to a certain extent a ect the surrounding cities of Beijing.We focused on several a airs, such as the 2008 Olympic Games, 2014 APEC (Asia-Paci c Economic Cooperation), and 2015 Victory Day Military Parade, by means of analyzing the concentration of SO 2 during the

Figure 15 :
Figure 15: Annual mean curve of SO 2 column concentration of each city in the BTH region, 2006-2017.

Figure 16 :
Figure 16: Monthly average SO 2 of August (a), September (b), and November (c) from 2006 to 2017.e black solid squares show relative changes based on 2006.e error bars express standard deviation.e red solid squares represent the months when events occur.
Data used in this study over the BTH are based on OMI SO 2 products with Dobson Units (DU, 1 DU 2.69 × 10 16 molecules/cm 2 ).

Table 1 :
Relevant information for study areas (data based on 2016).
[41,46]s in MeteorologyBTH from 2006 to 2017.e SO 2 loading has decreased over the recent years without clear regularity, which is in line with the study results by others[41,46].e plot describes the irregular upward and downward rule.It is necessary to identify the speci c period for the pollution attenuation, to a certain extent, related to governmental actions.It was found that SO 2 peak in

6
Advances in Meteorology mean distribution has an obvious seasonal change sequence as winter (December, January, and February)>autumn (September, October, and November)>spring (March, April, and May)>summer (June, July, and August).e average SO 2 loading of four seasons (spring, summer, autumn, and winter) was 0.59 DU, 0.52 DU, 0.73 DU, and 0.89 DU, respectively.Figure8also shows seasonal variation by calculating the monthly SO 2 columns from 2006 to 2017.SO 2 levels are peak in winter and minimum in summer and they are mainly attributed to the di erence in pollution di usion, which is caused by anthropogenic activities and meteorological conditions.Coal heating is a major source accounting for high SO 2 levels in winter.e SO 2 rapid drawdown in 2016 and 2017 represents the fact that coal heating had been replaced by natural gas is successful in reducing sulfur dioxide emissions.
over the BTH region exhibit the upward and downward trend consistent with the national trend in China.According to the dipping and heaving, we find that the gridded SO 2 data can be divided into two phases: the first (2006-2010) and second (2011-2017) period, to observe changes in SO 2 loadings presenting the Chinese government actions and policies and accomplishments in addressing air pollution.eSO 2 loadings have drastically decreased by more than 30% from the 2006-2010 period to the 2011-2017 period.SO 2 peaked in 2007 and the secondary peak was in 2011, to a certain extent, referring to the economic policy stimulus in the early of each FYP.Meanwhile, OMI observations show generally good agreement with independent SO 2 emission.(2)e annual cycles of SO 2 show a pronounced seasonal pattern, with the highest values occurring in winter and the lowest values in summer.is seasonal variation can be explained mainly by the seasonality of emission strengths, lifetimes of these pollutants, and meteorological factors.