Development of a high temporal–spatial resolution vehicle emission inventory based on NRT traffic data and its impact on air pollution in Beijing – Part 2: Impact of vehicle emission on urban air quality

18 In a companion paper (Jing et al., 2015), a high temporal–spatial resolution vehicle emission 19 inventory (HTSVE) for 2013 in Beijing has been established based on near real time (NRT) traffic 20 data and bottom up methodology. In this study, based on the sensitivity analysis method of switching 21 on/off pollutant emissions in the Chinese air quality forecasting model CUACE, a modeling study 22 was carried out to evaluate the contributions of vehicle emission to the air pollution in Beijing main 23 urban areas in the periods of summer (July) and winter (December) 2013. Generally, CUACE model 24 had good performance of pollutants concentration simulation. The model simulation has been 25 improved by using HTSVE. The vehicle emission contribution (VEC) to ambient pollutant 26 concentrations not only changes with seasons but also changes over moment. The mean VEC, affected 27 by regional pollutant transports significantly, is 55.4 and 48.5 % for NO2, while 5.4 and 10.5 % for 28 PM2.5 in July and December 2013, respectively. Regardless of regional transports, relative vehicle 29 emission contribution (RVEC) to NO2 is 59.2 and 57.8 % in July and December 2013, while 8.7 and 30 13.9 % for PM2.5. The RVEC to PM2.5 is lower than PM2.5 contribution rate for vehicle emission in 31 total emission, which may be caused by easily dry deposition of PM2.5 from vehicle emission in near32

November to 30 th November) was conducted to reduce the effect of chemical initial and boundary 23 conditions. 24 Two real simulations which based on default emission of CUACE and the improved emission with 25 high temporal-spatial resolution vehicle emission (hereafter refer to HTSVE) are carried out to 26 evaluate the accuracy of pollutant concentrations simulated by CUACE and analyze the influence of 27 HTSVE on Beijing air quality, and hereafter refer to SIM1 and SIM2 respectively . The contribution 1 rate to ambient pollution level ( or source apportionment) based on air quality numerical model 2 includes source sensitivity simulations using the brute force method (also referred as zero-out method) 3 or the decoupled direct method (DDM), air pollution tagged method, and the adjoint method (An et  4 al., 2015; Burr and Zhang, 2011;Zhang et al., 2015). With comprehensible physical and chemical 5 process, adjoint method has a significant advantage in source apportionment compared to sensitivity 6 simulations or tagged method. However, the development of adjoint model is facing a challenge due 7 to complicated mathematics and a large amount of data processing and programming, which results 8 in less available regional scale air quality ajoint model. At  Beijing (Huang et al., 2015)  This study focus on vehicle source and its influence. HTSVE based on NRT traffic data was used to 3 replace the vehicle emission in CUACE emission module to analyze its effects on air quality 4 simulation. The detailed description of high temporal-spatial resolution vehicle emission and 5 comparison with vehicle emission in CUACE emission module were presented in part 1. The 6 contribution of major species from vehicle emission is presented in Table 3. The vehicle emission of 7 NO, NO2 and HC from HTSVE is higher, while CO and PM2.5 is lower than that from CUACE.  However, low correlation (0.21 and 0.12 for SIM1 and SIM2 respectively) in July reflects the 1 complexity of air quality numerical simulation. Simulated PM2.5 daily mean concentration is basically 2 consistent with observed value. Minor difference of PM2.5 concentration is observed between SIM1 3 and SIM2 due to less vehicle emission change (Table 3). Based on temporal correlation analysis, 4 SIM2 improves PM2.5 time trends slightly, with correlation coefficients of 0.75 and 0.77 in two 5 periods for SIM1, 0.76 and 0.78 for SIM2. Compared with SIM1, the RMSE of PM2.5 daily mean 6 concentration has slightly decrease for SIM2. It is obviously that simulated PM2.5 concentration is 7 more accurate than simulated NO2 concentration in July, similar phenomena was found in previous basically consistent with sites observation (Fig. 3). The mean wind in Beijing urban region is the 27 southwest wind in July, and drives local pollutant transports from southwest to northeast. The high 28 NO2 concentration is located in northeastern city, while two high PM2.5 concentration regions appear 1 in west and center city ( Fig. 3a and b). The spatial distribution of NO2 is different from that of PM2.5 2 because of emission sources distribution difference with one high emission area inner 5 th ring road 3 for NO2 and two high emission areas in west 6 th ring road and inner 3 rd ring road for PM2.5 (Fig. 4). 4 High concentrations present in high emissions or its downwind. The mean concentrations of NO2 and 5 PM2.5 are 29.8 and 91.3 μg m -3 in July. Beijing urban region is dominated by northwest wind in 6 December, and pollutant concentration distribution is obviously different from that in July. NO2 7 concentration is high in southeast city, and gradually decreases outward (Fig. 3c). High PM2.5 8 concentration is mostly located in west and southeast city (Fig. 3d). It is found that significant 9 difference presents in NO2 distribution between July and December while slightly difference for 10 PM2.5 due to the combined effect of wind fields and emission distributions. The mean concentrations 11 of NO2 and PM2.5 are 42.8 and 136.4 μg m -3 in December respectively. 12

The effect of vehicle emission on urban air quality 13
VEC on ambient pollutant concentration is analyzed through comparison simulation with and without 14 vehicle emission (SIM2 and SIM3 respectively). Probability density function (PDF) is a good way to 15 describe the total representation. The PDF of instantaneous VEC in two periods is shown in Fig. 5. 16 The maximum frequencies of VEC to NO2 in July and December are appeared in 55-60 % and 50-17 55 % respectively. The frequencies of VEC to NO2 from 15 to 60 % in December are larger than that 18 in July (Fig. 5a), which indicates large contribution presents in summer while small contribution 19 presents in winter. Based on one-way analysis of variance, the difference of VEC to NO2 in summer 20 and winter is significant. This may relates to seasonal differences of meteorological condition and 21 pollutant emission. In summer, high temperature and strong solar radiation lead to strong atmosphere 22 oxidation ability, and therefore it is easy to convert from NO to NO2, which results in large 23 contribution to NO2 concentration. Meanwhile, the high rate of NO2 emission from vehicle (Table 3)  24 is another reason for large contribution to ambient NO2 concentration in summer. The VEC to PM2.5 25 is considerably lower than that to NO2. The maximum frequencies of VEC to PM2.5 in July and 26 December are appeared in 0-5 % and 5-10 % respectively. Different from NO2, the mean VEC to 27 PM2.5 in summer is smaller than that in winter, with a significant difference from one-way analysis 1 of variance. Relative humidity in summer is larger than that in winter, and high relative humidity is 2 conductive to gas-particle conversion processes of other emission sources (Yao et al., 2014), which 3 may be one of the reason for small VEC to PM2.5 in summer. The strong turbulence mixing in summer 4 makes rapidly vertical exchange and transport of pollutant in boundary layer, and finally results in 5 small VEC to PM2.5 in summer. Wind field variation is another reason for seasonal change of VEC to 6 PM2.5, which will be investigated in the following part. 7 As the local transports of pollutants, the VEC in Beijing city depends on wind field and spatial 8 distribution of vehicle emission. Wind dependency map of VEC to NO2 and PM2.5 are shown in Fig.  9 6. High VEC to NO2 in July is appeared in south wind with 3-4 m s -1 , while north wind with 6-7 m 10 s -1 for that in December. Due to the difference of lifetime between NO2 and PM2.5, the wind 11 dependency map to PM2.5 is quite different from that to NO2. High VEC to PM2.5 in July and 12 December appeared in north wind due to many vehicle emission of particle matter in northeast city 13 (Jing et al., 2015). The dominant wind is southwest wind in July and northwest in December (Fig. 3), 14 which brings a small VEC to PM2.5 in summer. Significant regional transport which is analyzed in 15 next section is one of the reason for relative small VEC to PM2.5 in summer. 16  2007), but also changes with time. Time series of regional mean VEC is 49. [8][9][10][11][12][13][14][15][16][17][18][19] 60.0 % to ambient NO2 concentration in July, with a mean contribution rate of 55.4 %. In December, 20 regional mean contribution on NO2 concentration decreases to 28.5-57.9 % at different days, with a 21 mean contribution rate of 48.5 %. VEC to ambient PM2.5 concentration is less than 10.3 and 13.6 % 22 at different times, with mean contribution rate of 5.4 and 10.5 % in July and December respectively. 23 The change of VEC to PM2.5 between July and December is most caused by meteorological condition 24 in two periods. With different lift time of PM2.5 and NO2, PM2.5 concentration is more affected by 25 regional transports, while NO2 concentration is more affected by local emissions. Therefore the 26 contribution with time variation for PM2.5 is different from that for NO2. Except for wind field, 27 pollution level is an important factor to VEC. It is obviously that low VEC presents in serious 28 pollution, while high VEC presents in low pollution concentration level, especially for NO2 (Fig. 8). 1 The absolute contribution of vehicle emission increases in severe pollution mostly because of adverse 2 dispersion condition. However, pollutant regional transport is enhanced in severe pollution, which 3 results in negatively correlation between VEC and pollution concentration level. The VEC has a 4 significant spatial variation, previous study pointed that PM2 As can be seen from Table 4 In this study, the rates of NO2 and PM2.5 from vehicle emission in total emission takes account for 27 55.1 and 22.3 % in July and 53.9 and 20.6 % in December (Table 3) of total emission. Because of the 28 effect of pollutant regional transports, the contribution rate of vehicle emission on ambient pollutant 1 concentration is lower than the rate of vehicle emission in total emissions. The difference between 2 these two rates became significantly larger with more contribution of outside emission, which implies 3 the importance of weather condition. In order to avoid the effect of weather situation on analysis 4 results, the relative contribution of vehicle emission on pollutant concentrations is analyzed in 5 following section. 6 The chemical components of PM2.5 represents the characteristics of emission source and complexity 7 chemical processes of pollutant in atmosphere. Based on sensitivity test, the VECs of BC, OC and NI 8 are large, while relative small for SF, and AM ( Table 5). The VECs of BC and OC in December are 9 approximately twice of that in July. Seasonal changes for the rates of BC and OC from vehicle 10 emission in total emission are inapparent which indicates that it is not the reason for seasonal change 11 of VECs. Beijing is controlled by southerly wind dominantly, which results in significant regional

Relative contribution of vehicle emission 18
Air pollution in Beijing is attributed not only from local emissions but also from regional transports. concentrations. Pollutant regional transport depends on atmospheric circulation and regional emission 24 characteristics. By comparing pollutant concentrations between SIM2 and SIM4, local emissions in 25 Beijing contributes 93.6 % and 62.6 % to NO2 and PM2.5 concentrations in July, and 83.8 % and 76.1 % 26 to NO2 and PM2.5 concentrations in December, which have a profound effect on RVEC. 27 Figure 10 depicts the spatial distribution of RVEC to NO2 and PM2.5 in July and December, and 1 similar distribution is found in two periods. The RVEC to NO2 is large in southeast and northeast 2 main urban areas, while small in west main urban areas. Time series of regional mean RVEC to NO2 3 in main urban areas range from 52.3 to 63.4 %, and 49.4 to 61.2 %, with the mean of 59.2 and 57.8 % 4 in July and December respectively. Different from NO2, the RVEC to PM2.5 is large in northeast of 5 main urban areas in two periods. Time series of regional mean RVEC to PM2.5 range from 5.7 to 11.3 % 6 and 9.9 to 16.1 %, with the mean of 8.7 and 13.9 % in July and December respectively. The differences 7 of RVECs to NO2 and PM2.5 in July and December are significant based on one-way analysis of 8 variance. The spatial distribution of RVEC are tremendously affected by vehicle emissions, as they 9 are mostly consistent with the rate of vehicle emission in total emission (Fig. 4) concentrations on the ground level. Regardless of regional transports, the contribution of vehicle 21 emission to ambient PM2.5 concentration is substantial lower than the rate of vehicle emission in total 22 emission in this study. Our finding is seemingly in conflict with Cheng et al. (2013), but may be more 23 reasonable for following reasons. Different from elevated emission, PM2.5 from vehicle emission in 24 near-surface layer easily descends to the ground or is absorbed by vegetation, which leads to low 25 contribution rate to PM2.5 concentration. Secondary aerosol generated by photochemical reaction is 26 different for different sector emissions. The VEC to SF is low in Beijing (Table 5), which indirectly 27 causes low VEC to PM2.5. Furthermore, pollutant regional transport and the background concentration 28 may result in lower VEC to PM2.5 than the rate of emission. 1

Conclusion 2
Air quality simulation has been improved by using HTSVE. In summer (July), high NO2 3 concentration was located in the northeastern part of city, while two high PM2.5 concentration regions 4 appeared in west and center of the city. In winter (December), NO2 concentration was high in 5 southeast city, then gradually decreased outward, while high PM2.5 concentration was mostly located 6 in west and southeast part of city. The VEC in Beijing city depends on wind field, spatial distribution 7 of vehicle emission and air pollution level. High VEC to NO2 in July appeared along with south wind 8 and low pollution concentration level, while north wind and low pollution concentration level for that 9 in December. High VEC to PM2.5 in July and December appeared along with north wind and low 10 pollution concentration level. 11 Seasonal change of VEC was observed in this study. The mean VECs to NO2 were 55.4 and 48.5 %, 12 while the mean VECs to PM2.5 were 5.4 and 10.5 % in July and December respectively. Regional had an important effect on RVEC. Regardless of regional transports, the RVEC to NO2 was large in 18 the southeast and northeast main urban areas, and northeast main urban areas for PM2.5. The mean 19 RVECs to NO2 were 59.2 and 57.8 %, while the mean RVECs to PM2.5 were 8.7 and 13.9 % in July 20 and December respectively. The RVEC to PM2.5 was lower than PM2.5 contribution rate for vehicle 21 emission, which was caused by easily dry deposition of PM2.5 from vehicle emission in near-surface 22

layer. 23
Acknowledgments 24 Stockwell, W. R., Middleton, P., Chang, J. S., and Tang, X.: The second generation regional acid 12 deposition model chemical mechanism for regional quality modeling, J.    Beijing, white arrows represent near-surface mean wind field. 4