Analysis of the carbon dioxide mole fraction variation and its transmission characteristics in Taiyuan

ZHANG Fengsheng,ZHU Lingyun,YAN Shiming,GAO Xing’ai and PEI Kunning

Shanxi Province Institute of Meteorological Sciences,Shanxi Provincial Meteorological Bureau,Taiyuan,China

ABSTRACT Based on the concentrations of CO2,PM2.5 and PM1.0,and conventional meteorological observation data from 2016 to 2018 at Taiyuan station,which belongs to the Shanxi greenhouse gas observation network,the CO2 concentration monthly and daily distribution characteristics,the weekend effect,and the variation characteristics on haze days and non-haze days,are analyzed.By using the Hybrid Single-Particle Lagrangian Integrated Trajectory model(backward trajectory model)and surface wind data,the transmission characteristics of atmospheric CO2 in Taiyuan are studied in various seasons.The results show that,in Taiyuan,the CO2 mole fraction in autumn and winter is higher than that in spring and summer, and on haze days is higher than that on non-haze days. The diurnal variation characteristic of CO2 mole fraction in each season is‘single peak and single valley’with the peak value around 0700 (hereafter refers to Beijing Time) and the valley value around 1600. The CO2 mole fraction on workdays is slightly higher than that on non-workdays and obviously different around 0800 of the early peak.Horizontal diffusion can reduce the CO2 mole fraction,while breezy weather is not beneficial to CO2 diffusion.The wind direction and speed in the upper levels are different from those near the surface,and the close air masses in the southwest-west-northwest sector raise the CO2 concentration in Taiyuan obviously.This indicates that the CO2 in Taiyuan is mainly contributed by local sources.

KEYWORDS Atmospheric carbon dioxide;mole fraction variation;transmission characteristics;Taiyuan

1. Introduction

Carbon dioxide (CO2) is a long-lived greenhouse gas.According to the WMO Greenhouse Gas Bulletin published on 22 November 2018,the global CO2content in 2017 reached its highest value since records began,up to 405.5 ± 0.1 ppm, which is 146% of the level before industrialization (before 1750) (WMO 2018). NOAA’s Annual Greenhouse Gas Index indicates that the radiation forcing of long-lived greenhouse gases increased by 41% from 1990 to 2017, and CO2accounted for about 82% of the increase (Butler and Montzka 2018).Human activities such as fossil fuel burning, cement production, and land-use variation are changing the concentrations of greenhouse gases at an unprecedented rate(Xu et al.2016).

Since 1959, the Mauna Loa observation station in the United States has carried out long-term observations of atmospheric CO2,accumulating observation data for nearly 60 years up to the present day (Keeling et al. 1976). In China, Waliguan station in Qinghai Province has been observing atmospheric greenhouse gases since 1994.Subsequently,observation stations such as Shangdianzi in Beijing, Lin’an in Zhejiang, and Longfengshan in Heilongjiang have been successively established.With the establishment of greenhouse gas observation stations in many places,related research has focused on the analysis of urban station data,which provides a scientific basis for the government to grasp the high-concentration state of atmospheric CO2in urban areas(Pu et al.2018;Gao et al.2015). According to the research report of the C40 Cities Climate Leadership Group, cities discharge 80% of the global anthropogenic greenhouse gases. Therefore, it is crucial to study the characteristics of atmospheric CO2mole fraction in urban areas as well as at background stations(Li et al.2013).

Shanxi Province is one of the important energy bases in China,with the second-largest coal output.Until 2017,the coal consumption in Shanxi accounted for about 84% of the disposable energy consumption, 24% points higher than the national average level (National Bureau of Statistics of China 2018).Carbon-intensive energy development has led to the rapid growth of greenhouse gas emissions in Shanxi Province.Understanding and mastering the regional characteristics of CO2concentration in Shanxi Province will provide important guidance to the effective implementation of local carbon reduction measures.As the leading province in monitoring the concentrations of environmental greenhouse gases in China, Shanxi constructed a greenhouse gas observation station network in 2012.At present, six stations, including Taiyuan, Datong, Linfen,Jincheng,Shuozhou,and Wutaishan,have been founded.Taiyuan,as the provincial capital,was the first to establish an observation station with high data integrity.The annual rate of increase in CO2concentration at Taiyuan site was 2.09%from 2016 to 2017,while it was 0.64%at Waliguan site in the same period.

Based on the online observational atmospheric CO2mole fraction at Taiyuan station during 2016-18,the characteristics of the monthly and daily variation in CO2mole fraction in Taiyuan are analyzed in this paper. Also, the concentrations on haze days and non-haze days, and on workdays and non-workdays are compared. Combined with the ground wind direction, wind speed, and the airmass transfer trajectory,the characteristics of CO2transfer in Taiyuan are preliminarily discussed.

Direct observation data of CO2in China are still very sparse. Thus, the importance of this study is to add new data over China. Furthermore, this research could help authorities to make decisions about measures to reduce fossil fuel consumption and implement climate change policies.

Figure 1.Location of Taiyuan station.

2. Site,data,and methods

2.1. Site

Taiyuan greenhouse gas observation station is located at Xiaodian meteorological station (37°44′N, 112°33?E; altitude: 775.4 m) to the south of Taiyuan City (8 km away)and adjacent to an urban expressway and the G5 Expressway(1 km away)(Figure 1).The sampling point is set on the tower 30 m above the ground,with no buildings and no obvious industrial emission sources nearby, but surrounded by rich vegetation.The main human activities are those related to traffic and the lives of local residents.Therefore,the observation data can represent the average atmospheric CO2mole fraction in Taiyuan to some extent.

2.2. Data and methods

The online atmospheric CO2mole fraction observation system adopts the G2301 high-precision CO2analyzer developed by Picarro, United States, which is based on wavelength scan cavity ring-down spectroscopy.The system is highly accurate, stable, has a rapid analysis speed,and is simple to operate. It uses the mixed standard gas calibrated by the Atmospheric Composition Observation and Service Center of the China Meteorological Administration. The calibration results can be traced to the international first-class standard gas maintained by the Central Calibration Laboratory of the WMO. The factory-reported precision of the instrument is 50 ppb for CO2in 5 min.Two standard gases,WH(high concentration of standard gases) and WL (low concentration of standard gases), are used to calibrate the measurements and a target gas is used to check the precision of the system routinely (Fang et al. 2015). The system is controlled by eight sample selection valves when it is running. In the 2-h cycle period,the valves rotate automatically and select to analyze the standard gases,target gas or sample air.The sample air is analyzed in 115 min,while the standard gases or target gas are analyzed in 5 min. The system uses an external standard method to calculate the sample air concentration. According to the linear correlation between the response signal of Picarro and the concentration of standard gas, the unknown sample air concentration is calculated through linear fitting by using the adjacent WH and WL.The target gas means that the standard gas with known concentration is alternately analyzed in the normal operation sequence.Then,by comparing the system analysis concentration value with the nominal value,the accuracy of the system's constant value is determined and the analysis data are quality controlled.In this paper,the average value of more accurate 5-min data is used as the original data, and abnormal values are eliminated according to the station records, including the data affected by system maintenance and human activities,such as the replacement of cold trap alcohol, the maintenance of calibration instruments, multiple people visiting, and garbage burning near the sampling port. About 80% of the effective data is retained, and the effective 5-min data are calculated to an hourly average,and then a daily average value. Based on the daily average, the seasonal average is obtained.

A haze day is defined according to the visibility,relative humidity,and the mass concentrations of PM2.5(particulate matter with a diameter of 2.5 μm or less)and PM1.0(particulate matter with a diameter of 1.0 μm or less),based on national standards in China. The mass concentrations of PM2.5and PM1.0are measured by 1405D and 1405 online aerosol mass concentration monitors produced by Thermo Scientific, United States, and the sampling point is set at 3 m above the ground.Conventional meteorological data,such as visibility,relative humidity,and wind(10 m),during the same period,are observed by automatic weather stations. The data used in this paper are all from the same observation station,with strict quality control.

The Hybrid Single-Particle Lagrangian Integrated Trajectory model developed by NOAA is utilized to analyze the influence of airmass transport on CO2concentration in different seasons(Draxler and Rolph 2003).The 72-h backward trajectory arriving in Taiyuan is calculated by using the GDAS database(ftp://www.arl.noaa.gov/pub/archives/gdas1/),with a spatial resolution of 1°×1°,provided by the National Centers for Environmental Prediction, United States, which is calculated at 0200, 0800, 1400, and 2000 every day.The Euclidean distance at 500 m between tracks is calculated through the TrajStat module of MeteoInfo.For all four seasons, the CO2concentration loadings of six cluster trajectories are selected for calculation.

3. Results and analysis

3.1. Characteristics of monthly variation in CO2 mole fraction and a comparison between haze days and non-haze days

Box plots are used to show the characteristics of the monthly variation in CO2mole fraction from 2016 to 2018 in Taiyuan(Figure 2(a))as well as its characteristics of variation on haze days and non-haze days in the same period (Figure 2(b)). Figure 2(a) shows that the CO2mole fraction in Taiyuan is high in autumn and winter, and low in spring and summer. The maximum value is in January and the minimum value is in August,which is related to emission sources in winter, vegetation photosynthesis in summer, meteorological conditions, and the change in vegetation cover. In autumn,the wind speed is low,the atmospheric stratification is stable, the atmospheric vertical movement ability is weak, and it is therefore difficult for CO2to be evacuated, meaning it gathers near the surface. In winter,photosynthesis is greatly weakened, reducing an important sink of the CO2.Meanwhile,the temperature is low,as well as the height of the boundary layer,and thus the occurrence frequency of inversion is increased and CO2diffuses slowly. Furthermore, the winter heating in the north increases the winter CO2source.Together,these factors result in high CO2mole fraction in autumn and winter.On the contrary,the wind speed in spring is high, the height of the boundary layer increases, and convection in the vertical direction is enhanced, which is convenient for the diffusion of CO2. In summer, the vegetation is lush, illumination is sufficient, temperatures are high, photosynthesis is strong, and CO2obtains an extra important sink.Together, the effect of these factors is to lower the CO2concentration in spring and summer.

Haze days and non-haze days are identified according to the visibility, relative humidity, and PM2.5and PM1.0concentration data.Figure 2(b)illustrates that the daily mean, maximum, minimum and each quantile of CO2mole fraction on haze days are greater than those on non-haze days.On haze days,the atmospheric stratification is stable and the vertical motion ability is weak, which makes it difficult for air pollutants accumulated in the surface layer to diffuse.

3.2. Characteristics of the daily variation in CO2 mole fraction and a comparison between workdays and non-workdays

The characteristics of the diurnal variation in CO2in each season in Taiyuan also indicate that the CO2concentrations in autumn and winter are higher than in spring and summer. The diurnal variation of CO2mole fraction in each season and the annual average all show a ‘single peak and single valley’ pattern (Figure 3(a)). After 1600, due to the weakening of the vertical transport process and the influence of plant respiration, the atmospheric CO2mole fraction gradually accumulates and increases, reaching a peak at around 0700.After that,photosynthesis gradually enhances due to sunrise, as well as atmospheric vertical movement,and the CO2mole fraction gradually decreases,reaching a minimum value at around 1600.Specifically,the peak value and valley value in each season also appear at different times,due to the different climatological characteristics. The peak value appears the earliest in summer,while the peak value in winter appears 3 h later than in summer because of the late sunrise. The valley value appears the earliest in winter and appears 1 h later than that in winter for the other three seasons.The amplitude in summer is the largest(39.5 ppm),while that in spring is the smallest(29.2 ppm),which may be related to the changes in temperature difference and solar radiation,which induce different rates of photosynthesis.

Figure 3(b)presents the characteristics of the daily variation in CO2mole fraction on workdays and nonworkdays. The two daily variations are similar, showing a ‘single peak and single valley’ pattern. The CO2mole fraction on workdays is slightly higher than that on nonworkdays, and obviously different at around 0800 of the early peak,which is related to the CO2of vehicle exhaust fumes (Liu 2018). Furthermore, the standard deviation of CO2mole fraction on workdays is higher than that on nonworkdays.The degree of dispersion of CO2mole fraction on workdays is larger,as well as the range of variation.

Figure 2.(a)Monthly variation of CO2 mole fraction and(b)a comparison between haze days and non-haze days,from 2016 to 2018 at Taiyuan station.In the boxes,dots represent the seasonal average value;the horizontal lines represent the median value;the top and bottom sides of the box represent the 75%and 25%quantiles,respectively;the top and bottom ends of the line segments represent the 95%and 5%quantiles,respectively;and the top and bottom dots outside the boxes represent the maximum and minimum values,respectively.The number of observations is 877.

Figure 3.(a)Diurnal variation of CO2 mole fraction in all seasons and(b)a comparison between workdays and non-workdays,from 2016 to 2018 at Taiyuan station.The number of observations is 234023.Error bars indicate confidence intervals of 95%.

3.3. Effect of wind on CO2 mole fraction

Wind direction and wind speed are important factors that affect the transmission capacity and horizontal diffusion effect of air pollutants(Zhang et al.2018).The CO2mole fraction and wind speed in spring, summer,autumn, and winter in 16 wind directions from 2016 to 2018 is arithmetically averaged, and rose diagrams of CO2mole fraction-wind speed and CO2mole fractionwind direction are drawn.As shown in Figure 4,the CO2mole fraction in autumn and winter is higher than that in spring and summer. The wind speed in autumn is low,but in spring it is high, which is consistent with the analysis of the seasonal variation in CO2mole fraction.The CO2mole fraction in the sector with higher wind speed is smaller.Conversely,the CO2mole fraction in the sector with lower wind speed is larger.Horizontal diffusion can reduce the CO2mole fraction in this region,while breezy weather is not beneficial to CO2diffusion.

The CO2mole fraction in the west-northwest direction is relatively small in spring,autumn,and winter,but larger than normal in summer,which may be due to the wind speed in the west-northwest direction is larger in summer than that in the other seasons.Besides,human activities in this direction are less.In the north direction is the urban area of Taiyuan,and the CO2mole fraction in this sector is higher than the seasonal average concentration for all seasons,indicating the important contribution of human activities to the CO2mole fraction.Combined with the wind frequency statistical results,the frequency of southerly wind in spring is the largest(8.97%), and the west-southwest wind frequency is the smallest (2.87%). As shown in Figure 4, the CO2mole fraction is not high in spring in the southerly wind, in contrast to the high CO2mole fraction in the west-southwest wind direction.High-frequency wind direction corresponds to low CO2mole fraction,while low-frequency wind direction corresponds to high CO2mole fraction,which is part of the reason why the CO2mole fraction at Taiyuan in spring is low. In summer, because of the strong photosynthesis of vegetation, the influence of wind speed and direction is weaker than in other seasons. In autumn, the northerly wind frequency is the largest,that is,the direction in the urban area is mostly upwind, and the station data are most affected by human activities. In winter, the combined west-northnorthwest wind frequency is 34.8%, but the CO2mole fraction in this sector is very low;meanwhile,in the eastnortheast-east-southeast sector, with high mole fraction, the sum of wind frequency is only 7.98%. In other words,most of the wind in winter is from the clean area,and the probability of serious pollution in the upwind direction is very low, which may be part of the reason why the difference in CO2mole fraction between winter and autumn in the Taiyuan area is not significant.

Figure 4.Rose diagram of CO2 mole fraction and wind speed for each season from 2016 to 2018 at Taiyuan station.

3.4. Effect of airmass transmission on CO2 mole fraction

The backward trajectory cluster analysis results for all four seasons are shown in Figure 5(a-d),respectively.It can be seen that air masses with a long transport distance move fast in each season, but with low CO2mole fraction.However, air masses with a short transport distance have a high CO2mole fraction.This indicates that an increase in CO2mole fraction in Taiyuan is affected by local sources.Especially in summer,the transport distance of the cluster 4 trajectory is the shortest, and the CO2mole fraction increases to the highest (about 9.0 ppm). This suggests,therefore, that the CO2in Taiyuan in summer is mainly contributed by local sources.

In each season, from the perspective of the direction of the high-concentration CO2air masses, the close air masses in the southwest-west-northwest sector raise the CO2mole fraction in Taiyuan obviously. Specifically, in spring and summer, they are from the southwest and west directions, respectively, and in both autumn and winter, they are from the northwest direction. Especially in winter, the air mass with high concentration (24.3 ppm higher than the average value) from the northwest direction accounts for the largest proportion, which is 24.35%.Also, this is the reason for the high CO2mole fraction in winter. In general, the concentrations corresponding to tracks 4 and 5 in spring are high, accounting for 46.2%, the concentrations corresponding to tracks 2 and 4 in summer are high, accounting for 26.27%,the concentrations corresponding to tracks 2, 4, and 6 in autumn are high, accounting for 60.36%, and the concentrations corresponding to tracks 1 and 6 in winter are high, accounting for 30.99%.

Figure 5.Cluster analysis of airmass backward trajectories in each season at Taiyuan station.

4. Conclusions and discussion

4.1. Conclusions

(1) Under the joint influence of vegetation cover,meteorological conditions, and artificial heating,the CO2mole fraction in Taiyuan in autumn and winter is higher than that in spring and summer.As haze days are mostly breezy weather,the CO2mole fraction is higher than that on non-haze days.

(2) Affected by the atmospheric vertical movement and sunshine, the characteristic of the diurnal variation in CO2mole fraction in each season in Taiyuan is a ‘single peak and single valley’ pattern, with the peak at around 0700 and the valley at around 1600.The difference in seasonal climate leads to slight differences in peak and valley values, their occurrence time,and the amplitude.The CO2mole fraction on workdays is slightly higher than that on nonworkdays.The difference is most obvious at around 0800 at the early traffic peak, which indicates that CO2mole fraction has a‘weekend effect’.

(3) Different wind directions and wind speeds will lead to different CO2mole fractions. Horizontal diffusion can reduce the CO2concentration,while breezy weather is not beneficial to CO2diffusion,and the CO2mole fraction in Taiyuan is mainly affected by local sources. Referring to the wind frequency data, it is found that most of the wind in spring and winter is from the clean area,and the probability of serious pollution in the upwind direction is very low, which leads to the low CO2mole fraction in spring, and the difference in CO2mole fraction between winter and autumn in Taiyuan area is not significant.

(4) The wind direction and speed in the upper levels are different from those near the surface, and the close air masses in the southwest-west-northwest sector raise the CO2mole fraction in Taiyuan obviously.This indicates that the CO2in Taiyuan is mainly contributed by local sources.In winter,the air mass with high concentration(24.3 ppm higher than the average value)from the northwest direction accounts for the largest proportion,and this is the reason for the high CO2mole fraction in winter.

4.2. Discussion

In Taiyuan, the concentration of CO2on haze days is significantly higher than that on non-haze days.Besides the weather conditions, this could also be related to the homology between greenhouse gases and the main air pollutants (Wei et al. 2019). The CO2mole fraction at the morning traffic peak on workdays is significantly different from that on non-workdays. So,should the CO2in vehicle exhaust fumes receive more concern? This will be studied in our future work, which will provide a scientific basis for the government to formulate climate change policies.

Acknowledgments

We express our gratefulness to the staffat Taiyuan station,who have contributed to the system maintenance at the stations.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Funding

This paper was supported by the Key Research and Development Project of Shanxi Province [grant number 201803D31220]and the General Program of Shanxi Provincial Meteorological Bureau [grant numbers SXKMSDW20205214 and SXKQNDW20205241].