Introduction

Air pollution in China has become an issue of concern not only within the Chinese community but also to the neighboring countries. This is because, the pollutants which originate in industrialized regions in China have been observed to affect other countries as a result of trans-boundary movement (Inomata et al. 2009; Jeong et al. 2011; Lee et al. 2013; Lu et al. 2016). China which is the fastest growing economy at the pace which has never been seen before is facing a serious air pollution problem (Feng et al. 2015; Li et al. 2017; Liu et al. 2013; Ran et al. 2009). Moreover, the air pollution problem seems to be universal as the report by World Health Organization (WHO) indicates that people living in cities are no longer safe, as 90% of the people live in cities which do not comply with the WHO air quality guideline (AQG) (WHO 2016). The report went further by revealing the sore fact that 10% of the annual deaths happen as a result of air pollution, while outdoor pollution on its own is responsible for the killing of around three million people each year. In an effort to combat the air pollution problems, the Chinese Ministry of Environmental Protection (CMEP) ratified National Ambient Air Quality Standard (NAAQS) in 2012. This new guideline aims at regulating the emission of six principal pollutants (i.e., NO2, SO2, CO, O3, PM2.5, and PM10) which are used to calculate the Air Quality Index (AQI) for cities. The set limit for daily (24 h average) concentration of PM10, PM2.5, SO2, NO2, CO, and 8 h average concentration of ozone are 150 μg/m3, 75 μg/m3, 150 μg/m3, 80 μg/m3, 4 mg/m3, and 160 μg/m3, respectively (GB3095 2012). It is worth to note that the standard set in NAAQS are 1.6~7.5 times higher than the standard which have been suggested by WHO for daily PM10, PM2.5, SO2, NO2, and 8 h average ozone of 50 μg/m3, 25 μg/m3, 20 μg/m3, 40 μg/m3, and 100 μg/m3, respectively.

Wuhan (29°58′~31°22′ N and 113°41′~115°05′ N), the capital city of Hubei province, is one of the megacities in central China. This city which is the hub of transportation and the major link of west and east China is reported to be polluted by a number of studies. For instance, Lyu et al. (2016) advocate Wuhan as one of the polluted cities because during their study period, they observed PM10 and PM2.5 frequently exceeding the NAAQS as a result of intensive biomass burning. The authors documented the average concentration of PM2.5 of 81.2 μg/m3 in summer and 85.3 μg/m3 in autumn. Wang et al. (2016) revealed the exceedance of annual average concentration of PM2.5, PM10, and NO2 stipulated in NAAQS by 256%, 192%, and 137%, respectively. Notably, monthly average concentration of PM2.5 was highest in December (winter month) and lowest in July (summer month) as a result of precipitation. Xu et al. (2017) reported the 3 years average ratio of PM2.5/PM10 of 0.62, the ratio which is higher than those found in most of the Chinese cities and town. The authors further revealed the existence of gradual increase of PM2.5/PM10 ratio at night time and they related this increase to stable atmospheric conditions which constrained vertical airflow. Based on the NAAQS, the concentration of PM2.5 and PM10 during the study period was above the daily acceptable limit by 40% and 27%, respectively. A follow-up study by Mbululo et al. (2017) reported that 83.05% of the days in 2015 were having PM2.5/PM10 ratio of more than 0.5, indicating the severity of PM2.5 in Wuhan atmosphere. The minimum and maximum ratio of organic carbon (OC)/elementary carbon (EC) were 3.5 and 24.2, respectively, indicating that the large portion of PM2.5 was composed of OC. A sampling campaign in spring of 2014 by Acciai et al. (2017) found average concentration of PM2.5 to be 95.53 μg/m3, which is higher than the acceptable limit set in NAAQS. Likewise, by using the Positive Matrix Factorization (PMF) model, the authors identified biomass burning as the largest (62.3%) contributor of PM2.5 in Wuhan, followed by metallurgical and steel industries (14%), crustal dust (12.5%), and dust and vehicle emission (10.4%).

Based on the available literature on air quality in Wuhan, the dynamics of atmospheric boundary layer (ABL) which is the main determinant of air quality (Hu et al. 2014; Wu et al. 2013) seems to have attracted limited attention to most of the researchers. Elsewhere, a study by Li et al. (2018) analyzed the characteristics of pollutants and the boundary layer structure by using meteorological instruments which were fixed at different heights (i.e., 10, 30, 50, 70, and 100 m) on a 100-m high tower. Nevertheless, the meteorological parameters (wind speed, air temperature, and relative humidity) captured at these altitudes cannot fully describe the dynamics of the ABL accurately. This is because the average height of ABL is far higher than this altitude. A number of studies (Tang et al. 2017; Wang et al. 2017) show the need for more detailed vertical study of meteorological parameters to fully understand the dynamics of ABL. This study therefore aims at describing the vertical structure of meteorological parameters (temperature, relative humidity, wind speed, and direction) at ABL of Wuhan city during pollution process by using 5 years data spanning from 2013 to 2017. Together with these high altitudes sounding data of ABL, this study also used daily ground observation data from meteorological station, air quality data, and air mass back trajectories.

Methodology

Data and sampling sites

Average daily air pollutant data (PM2.5, PM10, SO2, NO2, CO, and O3) from 2013 to 2017 from nine monitoring stations which are distributed all over Wuhan city were used in this study. These nine stations are administered by the Chinese Ministry of Environmental Protection (CMEP) and are used to determine the AQI of the city. Together with this, daily L-band radar data at 0700 LST with a vertical resolution of 10 m was also used to describe the vertical profile of different meteorological parameters (temperature, relative humidity, wind speed, and direction) up to 3000 m. The air quality data were obtained from Wuhan Environmental Protection Agency (WHEPA), while the L-band sounding data was obtained from Wuhan Meteorological Bureau (WMB). Daily average data of PM2.5 and AQI were used to plot the line graphs from which highly polluted and clean days could be defined. L-band sounding data were used to plot the vertical profile of the meteorological parameters.

Air quality index (AQI)

The overall AQI of the city is determined by the sub-AQI of the six air pollutants (PM10, PM2.5, SO2, NO2, CO, and O3) which are calculated by Eq. (1), as given in Technical Regulation on Ambient Air Quality Index (HJ633-2012, 2012). The overall AQI represents the maximum of the sub-AQI of all pollutants according to Eq. (2). Higher values of AQI are associated with “bad” while lower values are associated with “good” condition of air.

$$ IAQ{I}_p=\frac{I_{high}-{I}_{low}}{C_{high}-{C}_{low}}\left({C}_p-{C}_{low}\right)+{I}_{low} $$
(1)
$$ AQI=\max \left({I}_1,{I}_2,\dots, {I}_n\right) $$
(2)

where IAQIp is the air quality sub index for air pollutant p, Cp is the concentration of pollutant p, Clow is the concentration breakpoint that is ≤ Cp, Chigh is the concentration breakpoint that is ≥ Cp, Ilow is the index breakpoint corresponding to Clow, and Ihigh is the index breakpoint corresponding to Chigh. The daily AQIs were calculated by the 24-h average concentration of five air pollutants (PM10, PM2.5, SO2, NO2, and CO) and the daily average 8 h maximum concentration of O3. Table 1 summarizes the China air quality grading standard while the reference concentrations and index of pollutants can be found in HJ633-2012 2012.

Table 1 China air quality grading standards
Full size table

Air mass residence time

The movement of air parcel from one place to another acts as a transport agent of airborne particulate matter (PM). A study by Dimitriou (2015) considered the air mass residence time over specific source area to be linearly related to that area’s contribution to the receptor site. This study used 72-h air mass back trajectories approaching Wuhan at 100 m, 300 m, and 500 m to calculate and analyze airflow and diffusion by using Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model developed by the National Oceanic and Atmospheric Administration (NOAA) Air Resources Laboratory’s (ARL) of the USA. This model has a relatively complete transport, diffusion, and sedimentation model for handling a variety of meteorological element input fields, multiple physical processes, and different types of pollutant emission sources (Rolph et al. 2017; Stein et al. 2015). The TrajStat software which was developed based on Geographical Information System (GIS) by Wang et al. (2009) was used to cluster air mass back trajectories using Euclidean distance. The meteorological data input for running the software were from NCEP (National Centers for Environmental Prediction) fields obtained from NOAA which were available at every 3 h with 1°×1° spatial resolution.

Potential source contribution function (PSCF)

This study also used a PSCF method to identify the major contributing sources of PM2.5 in Wuhan city. The basis of PSCF method is that, if a receptor site is located at particular latitude and longitude, an air parcel back trajectory passing through that location indicates that the material from other sources can be transported through the trajectory to the receptor site (Dimitriou and Kassomenos 2014; Jeong et al. 2011; Polissar et al. 1999; Wang et al. 2009; Zeng and Hopke 1989). The formula, Eq. (3), defined the PSCF

$$ {\mathrm{PSCF}}_{(iJ)}=\frac{m_{(ij)}}{n_{(ij)}} $$
(3)

where n(ij) is the number of times that the trajectories passed through the cell (i, j) and m(ij) is the number of times that the pollutant concentrations were higher at the receptor site than the set criterion. In this study, the average daily acceptable limit of PM2.5 (75 μg/m3) stipulated in NAAQS was set as a criterion and utilized in PSCF calculation to identify the potential source areas (PSA) at distance which are more likely to have larger impact than the local sources. To reduce uncertainties resulted from the effect of simulation results of those grids with value of n(ij) that are too small, an empirical weighting coefficient, W(ij) (Eq. (4)) was multiplied by PSCF (Dimitriou and Kassomenos 2014; Jeong et al. 2011; Polissar et al. 1999; Wang et al. 2009). For the case of this study, navg is the average value (n(ij) > 0) of the number of all trajectory terminal points in grids.

$$ {\mathrm{W}}_{\left(i,j\right)}=\left\{\begin{array}{ll}1.0& \kern2.4em \left(3{n}_{avg}<{n}_{\left(i,j\right)}\right)\\ {}0.7& \left(1.5{n}_{avg}<{n}_{\left(i,j\right)}<3{n}_{avg}\right)\\ {}0.4& \kern0.24em \left({n}_{avg}<{n}_{\left(i,j\right)}<1.5{n}_{avg}\right)\\ {}0.2& \kern2.879999em {n}_{\left(i,j\right)}<{n}_{avg}\end{array}\right\} $$
(4)

Results and discussion

Analysis of pollution status

The results of six air pollutants (PM2.5, PM10, SO2, NO2, CO, and O3) from nine air monitoring stations for 5 years (2013~2017) in Wuhan show that they are above the daily and annual acceptable limits set by NAAQS except for SO2 and CO. Particulate matter (PM2.5 and PM10) pollution was serious during these 5 years. The ranking of years in the order of pollution status shows that 2013 was seriously polluted as 48% of its days were far above the PM2.5 daily average acceptable limit of 75 μg/m3, followed by 2014 (44%), 2015 (33%), 2016 (25%), and 2017 (20%). The same order is observed for PM10 as daily average acceptable limit of 150 μg/m3 in the year 2013 was exceeded by 30%, followed by 2014 (24%), 2015 (20%), 2016 (13%), and 2017 (8%). Table 2 shows the minimum and maximum concentration of PM2.5 and their standard deviation. Higher concentration is dominant in winter (December to February) and lower concentration in summer (June to August) months. Looking at the annual average acceptable limit, the concentration of PM10, PM2.5, and NO2 was also higher above the annual acceptable limit of 70 μg/m3, 35 μg/m3, and 40 μg/m3 respectively. The annual limit of PM2.5 in 2013 was 2.7 times higher than the average annual acceptable limit while for the other years, 2014, 2015, 2016, and 2017 were 2.3, 2, 1.6, and 1.5 times higher than the annual acceptable limit, respectively. For the case of PM10, it was above the annual average acceptable limit in 2013, 2014, 2015, 2016, and 2017 by 1.8, 1.7, 1.5, 1.3, and 1.3 times higher than the annual acceptable limit, respectively. The annual average concentration of NO2 was higher in 2013 by 1.5 times, 2014 by 1.3 times, and in 2015, 2016, and 2017 by 1.2 times the annual average acceptable limit. Corroborated results to this is reported by Wang et al. (2016) as they found the exceedance of annual average concentration of PM2.5, PM10, and NO2 of 2.6, 1.9, and 1.4 times higher than the NAAQS from March 2013 to February 2014, respectively.

Table 2 Monthly average concentration of PM2.5
Full size table

Moreover, the ratio of PM2.5/PM10 in 2013, 2014, 2015, 2016, and 2017 were 0.75, 0.71, 0.64, 0.63, and 0.61, respectively, indicating the significant contribution of PM2.5 to the total mass of PM10. In addition to this, the percentage of days having the PM2.5/PM10 ratio of above 0.5 in 2013, 2014, 2015, 2016, and 2017 were 83%, 89%, 78%, 69%, and 64%, respectively. Corroborated results to this have been reported in a study by Xu et al. (2017), as they reported PM2.5/PM10 ratio of 0.62 and Mbululo et al. (2017) who found percentage of days with PM2.5/PM10 ratio of above 0.5 of 83.05%. It is worth noting that PM2.5 which is the dominant pollutant in Wuhan has been associated with adverse health effect by a number of studies (Liu et al. 2013; Tao et al. 2012; Wan et al. 2012). Furthermore, during these 5-year period, there were 25 days which were recorded as “severe pollution” (AQI > 300) and out of them, 15 days were found in the year 2013 (Table S1). The highest AQI recorded during these 5 years was 500 on 6 May 2017, where the primary pollutant was PM10 with the concentration of 950 μg/m3 while the concentration of PM2.5 was only 168.6 μg/m3. This pollution event is considered as an isolated case as it was caused by a severe dust storm event which originated from the Gobi Desert during 2–7 May 2017 and masked the whole of southeast China, Korean peninsula, and Japan (Zhang et al. 2018).

The second highest AQI was 414 on 27 January 2013 where the primary pollutant was PM2.5 with the average concentration of 371 μg/m3 while the concentration of PM10 was 342 μg/m3. Likewise, the lowest concentration of PM2.5 on record was 6.3 μg/m3 on October 12, 2017 when the AQI was 33. Based on the above analysis, it indicates that the air quality in the year 2013 was worse as compared to the other years under study (Table S2) while the year 2017 can be considered as the cleanest year even though it has the highest AQI on record. Corroborated results to this have been reported by Wang et al. (2016). What is promising is that these statistical results suggest that the overall trend of air quality in Wuhan is improving even though the city is still polluted as the daily and annual acceptable limits set in NAAQS are still exceeded. This is thought to be the results of raising awareness of people and several government initiatives such as publishing the daily AQI for major cities and linking the AQI to performance evaluation of local governments (Chen et al. 2012). Note that all these measures aim at combating air pollution problem which is currently a common phenomenon in most of the cities in China.

Pollution process

In the case of this study, a pollution process is defined as a continuous accumulation of air pollutants (PM2.5) which results into increase of AQI value. During these 5-year period, a number of pollution processes have been identified, but for the case of this study, only one pollution process which happened at late November to early December of 2013, will be discussed since most cases look similar. It is worth to mention that out of the 15 days which were identified to have AQI of above 300 (severe pollution) in the year 2013, 8 days were found within these 2 months (November and December). Therefore, describing the pollution process during these months acts as a good representative of what is happening on the ABL during the pollution process. Figure 1 shows clear pollution process from November 28 to December 7, 2013 when the PM2.5 concentration increased from 46.2 to 312 μg/m3 and the AQI raised from 73 (good) to 362 (severe pollution). A closer look on the day before the beginning of the pollution process (November 27) shows that the condition of ABL was better, even though there was small inversion layer (0.5 °C/100 m) with the thickness of 80 m. Above the altitude of 90 m, there was no inversion layer and the wind speed at the ABL was very high, reaching 15 m/s (Fig. 2 (a)). This high wind speed which is observed throughout the ABL is a favorable condition for horizontal mixing of pollutants with clean air (Quan et al. 2013), as a result, the day recorded the PM2.5 concentration of 64.6 μg/m3 and the AQI of 87, indicating that the air quality was “good” (Table 1).

Fig. 1
figure 1

Daily average PM2.5 and AQI for Wuhan from November to December 2013 where the red line shows the daily average acceptable limit

Full size image
Fig. 2
figure 2figure 2

Vertical structure of the atmosphere on 27 November 2013, PM2.5 = 36.6 μg/m3 and AQI = 52 (a); 28 November 2013, PM2.5 = 88.6 μg/m3 and AQI = 117 (b); 29 January 2013, PM2.5 = 119.2 μg/m3 and AQI = 156 (c); 30 November 2013, PM2.5 = 173 μg/m3 and AQI = 223 (d); 1 December 2013, PM2.5 = 201 and AQI = 251 (e); 2 December 2013, PM2.5 = 239 and AQI = 289 (f); 3 December 2013, PM2.5 = 205 and AQI = 255 (g); 4 December 2013, PM2.5 = 309 and AQI = 359 (h); 5 December 2013, PM2.5 = 275 and AQI = 326 (i); 6 December 2013, PM2.5 = 310 and AQI = 360 (j); 7 December 2013, PM2.5 = 310 and AQI = 360 (k); 9 December 2013, PM2.5 = 310 and AQI = 360 (l) showing the profile of temperature (red line, °C), relative humidity (green line, %), wind velocity (blue line, m/s), and dominant wind direction (gray line) during the pollution process. The number on X-axis is the result after dividing the relative humidity value by 10 and the wind direction angle by 30, while temperature and wind speed remain the same

Full size image

A different scenario on the condition of the ABL began to be observed on November 28 (Fig. 2 (b)) as the strength of inversion layer was strengthened and the speed of wind was weakened significantly (Table 3). During this day, the strength of inversion layer increased to 3 °C/100 m and the wind speed decreased to 1 m/s. These observed conditions at the ABL are not favorable for pollutant dispersion as they limit horizontal and vertical mixing of air pollutants. Comparable results to this have been reported by Yassin et al. (2018) in Kuwait that whenever the mixing layer height decreases, the concentrations of air pollutants were observed to increase. Figure 2 (c and d) shows the condition of the ABL continued to worsen, where the strength of the inversion layer continued to build up (3.3~5 °C/100 m), and the thickness also continued to grow bigger (110~240 m). On 1 December, the condition of ABL was far worse than the previous days, as the wind was calm and the thickness of the inversion layer continued to increase (Fig. 2 (e)). This condition is thought to favor more accumulation of pollutants as horizontal and vertical mixing capacity of ABL was further weakened; therefore, pollutants were not distributed to other areas and got mixed with clean air. During this day, there was an increase of concentration of PM2.5 by 197% as compared to 3 previous days and the AQI of this day was 182, indicating the condition of the air is “moderately polluted.” There was slight decrease in the concentration of PM2.5 (2.8 μg/m3) and AQI on 2 December because on this day, the wind speed which reached 12 m/s was recorded at the altitude of 790 m (Fig. 2 (f)). Note that, on the previous day, the wind speed at ground was calm and at about 790 m, the wind speed was only 4 m/s. Even though the humidity on December 3 was the highest (99%), the concentration of PM2.5 and AQI increased to 178 μg/m3 and 228 (heavy pollution), respectively, as a result of calm wind and increased strength of inversion layer (Fig. 2 (g)).

Table 3 Ground meteorological variables during PM2.5 pollution process
Full size table

Severe pollution (AQI = 343) and PM2.5 concentration of 293 μg/m3 were recorded on December 4, as a result of pollutant accumulation from the previous days. On this day, the wind speed at the inversion layer which already limited vertical mixing was small, ranging between 1 and 4 m/s, the speed which is not higher enough to favor horizontal movement of air pollutant. Therefore, the condition on this day was not favorable for both vertical and horizontal mixing of air pollutants, as a result, more pollutants were allowed to accumulate (Fig. 2 (h)). The following 2 days (December 5 and 6) observed a decrease in PM2.5 concentration and AQI moderately due to increase in relative humidity near the ground (Fig. 2 (i)) and increase in wind speed which reached about 12 m/s at the altitude of 760 m (Fig. 2 (j)). December 7 was the peak of this pollution process as the recorded PM2.5 concentration was 312 μg/m3 and the AQI was 362 which indicate “severe pollution.” As compared to the previous day (December 6), the condition of ABL on this day (December 7) was further worsened (Fig. 2 (k)). For instance, on this day (previous day), ground wind speed was 1 m/s (2 m/s) and reached 2 m/s (12 m/s) at the altitude of 760 m, strength of inversion was 2.4 °C/100 m (1.3 °C/100 m), and the thickness of the inversion layer was 350 m (460 m) (Table 3). These general conditions of ABL are not favorable for pollutant dispersion as they have further weakened the horizontal and vertical mixing capacity of the ABL, as a result, more pollutants accumulate in Wuhan atmosphere. Corroborated results to this have been reported by Lee et al. (2013) as they found that the thickness of inversion layer was high during the polluted days as compared to periods before and after the polluted days. This is because an inversion layer has the tendency of trapping pollutants within the inversion layer, as a result, it limits the exchange of pollutants and clean air at free troposphere (Quan et al. 2013). Nevertheless, 2 days later (December 9), the air condition improved significantly; the concentration of PM2.5 and AQI dropped to 115 μg/m3 and 150 (“slightly polluted”), respectively as a result of improved ABL condition. On this day, the ground inversion layer which was observed only in the first 50 m was significantly weakened to 0.4 °C/100 m. Furthermore, the ABL was more humid than the previous day and the wind speed was high, reaching 16 m/s at the altitude of 850 m (Fig. 2 (l)). These observed conditions during this period are favorable for vertical and horizontal distribution of pollutants (Mbululo et al. 2017).

Analysis of the cluster mean back trajectory result for November and December 2013 identifies four major air masses, two among them contributed significant amount of air pollutants. The first important air mass is associated with local pollution which accounted for 58.06% of the total loading, while the second air mass which accounted for 29.03% and the third (8.39%), originated in Inner Mongolia and passed through Shaanxi, Shanxi, and Henan provinces before they arrived in Wuhan. The fourth air mass which accounted for 4.52% originated in southern Russia and passed through Inner Mongolia before it entered the most serious polluted area of Hebei and Henan province (Fig. 3 (a)). Since back trajectory results cannot stimulate the values of daily PM2.5 level caused by PSA (Jeong et al. 2011; Xin et al. 2016), further analysis was done by using PSCF. Figure 3 (b) shows the results of PSCF analysis, where red color represents high contribution level of PSA while blue color represents low contribution of PM2.5 concentration. As it can be seen on the map, the higher WPSCF value of above 1 was mainly from Wuhan, which indicates that the pollutants were mainly of local origin and WPSCF value of above 0.8 was mainly from Hunan and Henan provinces. These results compliment the results which have been reported on the ABL that during the pollution process, the vertical and horizontal mixing capacity was weak, which resulted into pollutant accumulation. It therefore suggests that the local air masses played a significant role in this observed pollution process.

Fig. 3
figure 3

Maps showing cluster mean back trajectories for November and December (a) and spatial distribution of WPSCF (b)

Full size image

Conclusion

Analysis of six air pollutants from 2013 to 2017 shows that PM10, PM2.5, NO2, and O3 exceeded the daily and annual acceptable limits set in NAAQS at different levels except SO2 and CO. During these 5 years, 2013, 2014, 2014, 2016, and 2017, PM2.5 was higher than the daily average acceptable limits for 175, 160, 121, 91, and 71 days, respectively. These results indicate that about a half of the years 2013 and 2014 and one third (4 months) of 2014 were polluted by PM2.5. Moreover, the annual average concentration of PM2.5 of the year 2013, 2014, 2015, 2016, and 2017 were 94.4 μg/m3, 80.4 μg/m3, 69.4 μg/m3, 57.5 μg/m3, and 53.2 μg/m3, respectively, the concentrations of which are higher than the annual acceptable limit of 35 μg/m3. The days which exceeded the daily average acceptable limit of PM10, NO2, and O3 during these 5 years were in the range of 30~111, 19~80, and 36~143 days, respectively. Based on the number of days which comply with NAAQS, the year 2017 is regarded as the cleanest year while the year 2013 is regarded as the most polluted year. During the pollution process, the condition of ABL was dominated by strong ground inversion layer and weak winds, the condition which weakens the horizontal and vertical transport mixing of air pollutants. As a result, pollutants were accumulated and restrained from mixing with clean air. Likewise, the cluster analysis of air mass back trajectory shows that the pollutants were mainly of local origins since higher value of WPSCF was mainly dominant in Wuhan. This therefore suggests that trans-boundary movement of air pollutants was not an important factor during this pollution process.