Sensitivity of Nitrate Aerosol Production to Vehicular Emissions in an Urban Street

This study investigated the sensitivity of nitrate aerosols to vehicular emissions in urban streets using a coupled computational fluid dynamics (CFD)–chemistry model. Nitrate concentrations were highest at the street surface level following NH3 emissions from vehicles, indicating that ammonium nitrate formation occurs under NH3-limited conditions in street canyons. Sensitivity simulations revealed that the nitrate concentration has no clear relationship with the NOx emission rate, showing nitrate changes of only 2% across among 16 time differences in NOx emissions. NOx emissions show a conflicting effect on nitrate production via decreasing O3 and increasing NO2 concentrations under a volatile organic compound (VOC)-limited regime for O3 production. The sensitivity simulations also show that nitrate aerosol is proportional to vehicular VOC and NH3 emissions in the street canyon. Changes of VOC emissions affect the nitrate aerosol and HNO3 concentrations through changes in the O3 concentration under a VOC-limited regime for O3 production. Nitrate aerosol concentration is influenced by vehicular NH3 emissions, which produce ammonium nitrate effectively under an NH3-limited regime for nitrate production. This research suggests that, when vehicular emissions are dominant in winter, the control of vehicular VOC and NH3 emissions might be a more effective way to degrade PM2.5 problems than the control of NOx.


Introduction
Nitrate aerosol is a fine particulate matter (PM 2.5 ) component produced from the reaction of gas-phase nitrate (nitric acid; HNO 3 ) and ammonia (NH 3 ).During a haze event, nitrate aerosol often contributes to the total observed particle mass by as much or more than the organic fraction across East Asia [1][2][3][4].As severe haze events have increased across East Asia [5], nitrate contributions to PM 2.5 mass have also increased in polluted urban areas [1,6].
The production of nitrate aerosol in urban areas is affected by vehicular emissions such as nitrogen oxides (NO x = NO + NO 2 ) and NH 3 [7][8][9].These vehicular emissions are highly concentrated and are transported by turbulence from the complex geometry of buildings, causing steep gradients of pollutant concentrations [10][11][12][13].Nitrate aerosol chemistry is highly nonlinear and follows the concentrations of precursor gases and humidity, so the formation and distribution of nitrate aerosol are not spatially uniform [14,15].Most modeling studies on nitrate aerosols are based on regional or global air-quality models that have clear limitations due to their coarse resolution [16][17][18].Therefore, to accurately investigate nitrate formation in urban areas, fine-scale simulations that can conserve highly concentrated emission plumes and turbulence are necessary.
Meanwhile, policies regarding vehicular nitrate control focus on reduction of NO x (e.g., through banning diesel vehicles) [19,20], based on studies of regional or global air-quality modeling.However, studies from observation campaigns have often reported that other vehicle emissions are The model also includes the production of HNO 3 via heterogeneous chemistry between aerosols and gases following Jacob (2000) [38].The reactions that contribute to HNO 3 production by heterogeneous chemistry can be written as: The uptake coefficients, γ, of R2 and R3 were 10 −4 and 0.1, respectively, following Jacob (2000) [38].For R4, γ was set to 0.01 as suggested by Zhang et al. (2012) and Walker et al. (2012) [39,40].N 2 O 5 is essential for night-time nitrate aerosol chemistry; the production and loss reactions of N 2 O 5 can be written as: The simulation of carbonaceous aerosols follows the GEOS-Chem model [35].The primary carbonaceous aerosol follows the passive tracer without any chemical reactions.However, the model resolves primary Black Carbon (BC) and Organic carbon (OC) with a hydrophobic and a hydrophilic fraction for each (i.e., making four aerosol types) for deposition processes.All sources emit hydrophobic aerosols that then become hydrophilic with an e-folding time of 1.2 days, as per Cooke et al. (1999) [41].Although secondary organic aerosol (SOA) chemistry is not considered in the model, we treat boundary inflow and the transport of SOAs.However, the model does not account for either dust or sea-salt aerosol.
The dry deposition of gases and aerosols was simulated with a standard big-leaf resistance-in-series model [42].The model accounts for the dry deposition of 46 species, including aerosols.The CFD model has no atmospheric physics module that simulates hydrometeors (e.g., clouds and rain); thus, wet deposition was not calculated, following Kim et al. (2012) [25].

Simulation Set-Up
We assumed a street canyon selected for a simulation located in SMR, one of the most polluted cities, to investigate nitrate formation in urban streets.The domain size was 120 × 80 × 100 m in the x, y, and z directions, respectively.The grid intervals in all directions were 2 m, and the building height was 20 m with unified aspect ratios for all street canyons.Figure 1 shows the detailed structure of the simulation domain.
First, we set a control run (CNTL hereafter) with standard emissions.We estimated the pollutant emissions from average traffic volume in 2017 obtained from the Traffic Monitoring System (http: //www.road.re.kr), which provides traffic statistics in SMR.On that basis, the daily traffic volume for each street was assumed to be 15,130 vehicles day −1 .The monthly averaged diurnal variation in traffic volume was also obtained from the Traffic Monitoring System.Vehicular emissions were computed using the emission factor from the Clean Air Policy Support System (CAPPS) emission inventory [43] and calculated by multiplying the mean ratios of vehicle sizes in Korea.The emission factors following vehicle size and the ratios of vehicle size are summarized in Tables 1 and 2, respectively.The calculated averaged emissions per vehicle were 0.10 g km −1 , 0.15 g km −1 , 0.015 g km −1 , and 0.011 g km −1 for NO x , CO, VOC, and NH 3 , respectively.NO x emissions were separated into NO and NO 2 emissions at a 10:1 ratio by volume [44].Total VOC emissions were further speciated using the method developed by EMEP/EEA (2016) [45].Table 3 lists the emission rate and ratios of speciated VOC in the CNTL simulation.All vehicular emissions were emitted at z = 1 m from 4 m wide area sources located at the center of the streets.Other emissions in the model domain were not considered in the simulation.Previous studies have reported that the largest proportion of emissions in SMR is vehicular emissions [43,46,47].
aerosols.The CFD model has no atmospheric physics module that simulates hydrometeors (e.g., clouds and rain); thus, wet deposition was not calculated, following Kim et al. (2012) [25].

Simulation Set-Up
We assumed a street canyon selected for a simulation located in SMR, one of the most polluted cities, to investigate nitrate formation in urban streets.The domain size was 120 × 80 × 100 m in the x, y, and z directions, respectively.The grid intervals in all directions were 2 m, and the building height was 20 m with unified aspect ratios for all street canyons.Figure 1 shows the detailed structure of the simulation domain.First, we set a control run (CNTL hereafter) with standard emissions.We estimated the pollutant emissions from average traffic volume in 2017 obtained from the Traffic Monitoring System (http://www.road.re.kr), which provides traffic statistics in SMR.On that basis, the daily traffic volume for each street was assumed to be 15,130 vehicles day −1 .The monthly averaged diurnal variation in traffic volume was also obtained from the Traffic Monitoring System.Vehicular emissions were computed using the emission factor from the Clean Air Policy Support System Table 1.Emission estimates from vehicles used in the coupled CFD-chemistry model simulations for the CNTL simulation.The text discusses details about the species emission estimates.
[g km −1  For meteorological conditions, we used values for the SMR in winter that provided favorable conditions for nitrate formation [48].To simulate diurnal changes in buoyancy and their effects on transport and chemical reaction rates in the model, we used the hourly temperature and relative humidity obtained from the Seoul station of the Korea Meteorological Administration (http://web.kma.go.kr/eng/).Table 4 shows the hourly temperature and relative humidity used in this study.The wind speed on the rooftop was assumed to be the observed seasonal mean value for SMR, which is 2.4 m s −1 .The wind direction is westerly (most frequently observed in winter) and is perpendicular to the street canyon [49].The ambient wind speed and direction were fixed during a one-day simulation.The following vertical profiles of the wind, turbulent kinetic energy (TKE), and TKE dissipation rate were imposed: Here, u * , z 0 , and κ represent the friction velocity, roughness length (=0.05 m), and von Karman constant (=0.4), respectively; C µ is an empirical constant (=0.0845), and θ is the wind direction.The surface and top boundary pressures in the model were assumed to be 1013.15hPa and 993.72 hPa, respectively.For the rooftop boundary conditions of the species, we used a reanalysis dataset from the Monitoring Atmospheric Composition and Climate (MACC) project in SMR with 6 h diurnal variations [50].In addition, we used the boundary conditions of the species from the GEOS-Chem simulation with 6 h diurnal variations at SMR in winter for the species that were not provided in the MACC reanalysis dataset [51].The initial conditions of the species were assumed as the concentrations of the boundary conditions in the first step.We conducted 48 h model simulations for each case: the first 24 h for the model spin-up, and the results from the last 24 h were used.The chemical and dynamical time steps were 1 min and 1 s, respectively.
Sensitivity simulations were conducted to examine the effect of emissions on nitrate aerosols in urban streets.Twelve sensitivity simulations were set by changing the vehicular emissions of NO x , VOC, and NH 3 .Each simulation was conducted with different emissions by multiplying the original emission by 0.25, 0.5, 2.0, and 4.0 for each species.We named the sensitivity simulations "species name" × "multiplying factor."For example, simulations named NO x × 0.25, NO x × 0.5, NO x × 2, and NO x × 4 indicate multiplying the vehicular NO x by 0.25, 0.5, 2.0, and 4.0, respectively, while other emissions were fixed.

Model Validation
The coupled CFD-chemistry model in this study was thoroughly validated for the flow and dispersion of passive tracers in street canyons by comparing the results from this model with those from a wind tunnel, an idealized numerical study, and fluid experiments [25,27,52].Park et al. (2015) found good agreements when comparing the model with wind tunnel data and experimental data by implementing improved wall functions for the momentum and thermodynamic energy equations in the CFD model to more accurately represent the effects of solid-wall boundaries [27].
The coupled CFD model also evaluated the dispersion of reactive pollutants compared with idealized simulations and field campaigns [25].Kim et al. (2012) applied the coupled CFD-chemistry model to simulations using the same building configuration as in Baker et al. (2004) [25,53].Their results showed that the concentrations of NO x and O 3 have a pattern and magnitude consistent with the simulated concentrations by Baker et al. (2004) under steady-state O 3 -NO-NO 2 photochemistry.Kim et al. (2012) reproduced reactive pollutants on Dongfeng Middle Street, Guangzhou, China, using a full tropospheric NO x -O x -VOC chemistry scheme and compared the results to a field campaign by Xie et al. (2003) [13].The coupled model, with the full photochemical mechanism, also successfully captured the time variation in the observed CO concentrations for both upwind and downwind sites in the Dongfeng Street canyon.However, the coupled model overestimated NO x concentrations compared to observations by Xie et al. (2003) due to estimating excessive NO emissions from traffic volume, implying the necessity of utilizing an accurate emissions inventory [13].
The coupled CFD model has also been used to evaluate the dispersion of reactive aerosol in street canyons [28].Kim et al. (2019) evaluated the composition of PM 1 in summer and winter in a street canyon by comparison with the field campaign in Elche, Spain, by Yubero et al. (2015) [28,54].The model generally captured seasonal variations of PM 1 in the street canyon.We evaluated the seasonal variation in nitrate concentration by comparing our model results to those of Yubero et al. (2015) [54].Four simulations were conducted to represent the four seasons (spring, summer, autumn, and winter) in Elche, Spain.The street was approximately 7 m wide and surrounded by buildings that were approximately 25 m in height.The domain size was 20 m × 40 m × 50 m, and the number of grid points was 42 × 82 × 52.The meteorological conditions used were the observed seasonal mean values during the campaign periods.Pollutant emissions were estimated from traffic volume obtained from the Elche Traffic Office [54].Vehicular emissions were computed using Spain's emission rates in EMEP/EEA (2016).The detailed model configuration generally followed that of Kim et al. (2019) [28].
Figure 2 shows the observed and simulated nitrate concentrations by season.The observed nitrate concentrations were highest in winter and lowest in summer due to the thermal evaporation of nitrate aerosols, showing 1.3 and 0.3 µg m −3 , respectively.The observed nitrate concentrations in spring and autumn were 0.4 and 0.5 µg m −3 , respectively, falling between those in winter and summer and indicating the dominant effect of temperature on nitrate aerosol.The simulated nitrate concentrations were 1.0, 0.8, 1.0, and 1.4 µg m −3 in spring, summer, autumn, and winter, respectively, showing clear seasonal variation.The simulated concentrations reproduce the observed seasonal variation in nitrate concentration.However, the simulated nitrate concentration in summer was twice that of the observed magnitude, indicating the weaker thermal evaporation of ammonium nitrate predicted by the model.The model captured the magnitude of the observed nitrate in winter located within the standard deviation of the observed nitrate concentration.These results indicate that the model successfully calculated the nitrate aerosol in cold environments.

Results
Before investigating the production and sensitivity of nitrate aerosol, we checked the precursor gases of nitrate aerosol and oxidant concentrations that affect nitrate aerosol chemistry.Figure 3a indicates the meridionally averaged NOx concentration in the domain.The NOx concentration is highest at the street surface level, indicating that vehicular NOx emission is trapped by the canyon vortex in the street canyon.The NOx concentration reached 100 ppbv, ten times higher that outside the canyon, showing a steep gradient of NOx concentration in the street canyon.Note that concentrations in the three street canyons have slightly different values and dispersion patterns owing to different vortex patterns under non-infinite consecutive 3-D street canyons following the dispersion rates of TKE [26]. Figure 3b shows the spatial distribution of O3 concentration.The O3 concentration was lowest at the surface, showing a negative correlation with the NOx concentration.The O3 concentrations outside and inside the street canyon were 38 and 11 ppbv, respectively, suggesting NOx titration in the street canyon.This distribution of O3 and NOx fits the general dispersion pattern in the street canyon reported in previous studies [25,27].Figure 3c displays the HNO3 concentration, which reached 3.8 ppbv in the street canyon.The HNO3 concentration in the street canyon was higher than that outside the street canyon, indicating the oxidation of HNO3 from vehicular NOx.However, the HNO3 concentration at the surface was the lowest, even though the

Results
Before investigating the production and sensitivity of nitrate aerosol, we checked the precursor gases of nitrate aerosol and oxidant concentrations that affect nitrate aerosol chemistry.Figure 3a indicates the meridionally averaged NO x concentration in the domain.The NO x concentration is highest at the street surface level, indicating that vehicular NO x emission is trapped by the canyon vortex in the street canyon.The NO x concentration reached 100 ppbv, ten times higher that outside the canyon, showing a steep gradient of NO x concentration in the street canyon.Note that concentrations in the three street canyons have slightly different values and dispersion patterns owing to different vortex patterns under non-infinite consecutive 3-D street canyons following the dispersion rates of TKE [26]. Figure 3b shows the spatial distribution of O 3 concentration.The O 3 concentration was lowest at the surface, showing a negative correlation with the NO x concentration.The O 3 concentrations outside and inside the street canyon were 38 and 11 ppbv, respectively, suggesting NO x titration in the street canyon.This distribution of O 3 and NO x fits the general dispersion pattern in the street canyon reported in previous studies [25,27].Figure 3c displays the HNO 3 concentration, which reached 3.8 ppbv in the street canyon.The HNO 3 concentration in the street canyon was higher than that outside the street canyon, indicating the oxidation of HNO 3 from vehicular NO x .However, the HNO 3 concentration at the surface was the lowest, even though the NO x concentration was highest at the surface.The low level of HNO 3 at the surface was caused by low O 3 concentrations at the surface under VOC-limited conditions, suppressing the production of OH and HNO 3 .Figure 3d displays the NH 3 concentrations; the daily averaged NH 3 concentration reached 2.3 ppbv, with the highest value at the surface.The dispersion pattern of NH 3 was similar to that of NO x , indicating the high impact of vehicular emissions.Figure 4a shows the nitrate concentrations in the street canyon, which are higher than those outside the canyon, showing values of up to 11.4 μg m −3 .The high concentration of nitrate aerosol is due to the high HNO3 concentration from vehicular NOx in the street canyon.However, the spatial pattern of nitrate aerosol in the street canyon differs from that of HNO3, because NH3 is a precursor gas of ammonium nitrate.The nitrate aerosol was highest at the surface following vehicular NH3.These high correlations between NH3 and nitrate aerosol indicate that ammonium nitrate forms Figure 4a shows the nitrate concentrations in the street canyon, which are higher than those outside the canyon, showing values of up to 11.4 µg m −3 .The high concentration of nitrate aerosol is due to the high HNO 3 concentration from vehicular NO x in the street canyon.However, the spatial pattern of nitrate aerosol in the street canyon differs from that of HNO 3 , because NH 3 is a precursor gas of ammonium nitrate.The nitrate aerosol was highest at the surface following vehicular NH 3 .These high correlations between NH 3 and nitrate aerosol indicate that ammonium nitrate forms under NH 3 -limited conditions in the street canyon.Figure 4b shows the ammonium concentration in the street canyon; the spatial distribution of ammonium aerosol is similar to that of nitrate, implying that most ammonium aerosols combine with nitrate aerosols in winter.The maximum concentration of ammonium aerosol was 4.7 µg m −3 , and the spatial patterns of ammonium also indicate a low concentration of ammonium sulfate in winter.Note that the vehicular emission rate of SO 2 was very low, implying that ammonium sulfate inside the street canyon might also be low [55].The sum of the ammonium and nitrate concentrations was 16.1 µg m −3 , higher than the air quality guidelines set by the World Health Organization (WHO, 10 µg m −3 ) and 46% of the WHO Interim Target-1 (35 µg m −3 ), indicating the hazardous effect of ammonium nitrate aerosols on pedestrians [56].Considering that nitrate chemistry is highly nonlinear, this cannot be resolved and may lead to uncertainty in regional models due to their coarser resolution.
Atmosphere 2019, 10, x FOR PEER REVIEW 9 of 20 the street canyon; the spatial distribution of ammonium aerosol is similar to that of nitrate, implying that most ammonium aerosols combine with nitrate aerosols in winter.The maximum concentration of ammonium aerosol was 4.7 μg m −3 , and the spatial patterns of ammonium also indicate a low concentration of ammonium sulfate in winter.Note that the vehicular emission rate of SO2 was very low, implying that ammonium sulfate inside the street canyon might also be low [55].The sum of the ammonium and nitrate concentrations was 16.1 μg m −3 , higher than the air quality guidelines set by the World Health Organization (WHO, 10 μg m −3 ) and 46% of the WHO Interim Target-1 (35 μg m −3 ), indicating the hazardous effect of ammonium nitrate aerosols on pedestrians [56].Considering that nitrate chemistry is highly nonlinear, this cannot be resolved and may lead to uncertainty in regional models due to their coarser resolution.We investigated the sensitivity of nitrate aerosol production to NOx emission.Figure 5a shows the average nitrate concentration in the street canyon (i.e., below 20 m) by following vehicular NOx emission changes to investigate the sensitivity of nitrate aerosol production to the vehicular NOx emission rate.Surprisingly, the nitrate concentration did not show a clear relationship with the NOx emission rate.Nevertheless, the change in nitrate concentrations was at most 2% compared to the standard simulation, and the average nitrate concentration was highest in the CNTL and lowest in the NOx × 0.25 simulations, indicating the nonlinearity of the nitrate aerosol production to the NOx emission rate.These results contradict the conventional belief that high NOx emissions from vehicles can cause nitrate aerosol air quality problems.
Figure 5b,d display the average HNO3, O3, and NO2 concentrations, respectively, in the street canyon according to the sensitivity simulations.The HNO3 concentration in the street canyon shows changes similar to those of nitrate aerosol, indicating that the changes in nitrate aerosol in the sensitivity simulations are closely related to the changes in HNO3 (Figure 5b).The O3 concentration decreases as the NOx emissions increase because of NOx titration (Figure 5d).Note that O3 formation falls under a VOC-limited regime in the street canyon due to vehicular emissions.A low We investigated the sensitivity of nitrate aerosol production to NO x emission.Figure 5a shows the average nitrate concentration in the street canyon (i.e., below 20 m) by following vehicular NO x emission changes to investigate the sensitivity of nitrate aerosol production to the vehicular NO x emission rate.Surprisingly, the nitrate concentration did not show a clear relationship with the NO x emission rate.Nevertheless, the change in nitrate concentrations was at most 2% compared to the standard simulation, and the average nitrate concentration was highest in the CNTL and lowest in the NO x × 0.25 simulations, indicating the nonlinearity of the nitrate aerosol production to the NO x emission rate.These results contradict the conventional belief that high NO x emissions from vehicles can cause nitrate aerosol air quality problems.
Figure 5b,d display the average HNO 3 , O 3 , and NO 2 concentrations, respectively, in the street canyon according to the sensitivity simulations.The HNO 3 concentration in the street canyon shows changes similar to those of nitrate aerosol, indicating that the changes in nitrate aerosol in the sensitivity simulations are closely related to the changes in HNO 3 (Figure 5b).The O 3 concentration decreases as the NO x emissions increase because of NO x titration (Figure 5d).Note that O 3 formation falls under a VOC-limited regime in the street canyon due to vehicular emissions.A low concentration of O 3 prevents the conversion of NO 2 to HNO 3 through either photochemical production during the daytime due to inhibited OH production and heterogeneous nitrate production at night.The NO 2 (the precursor gas of HNO 3 ) , concentration in the street canyon is proportional to the NO x emissions because it affects the direct NO 2 emissions from vehicles and the reduction in photodissociation of NO 2 under low O 3 concentrations (Figure 5c).High NO 2 creates suitable conditions for HNO 3 formation, compensating for the effect of decreased O 3 on HNO 3 .Thus, HNO 3 and nitrate aerosols have no clear relationship with NOx emissions and only undergo small changes.These results imply that NO x emission controls cannot improve PM 2.5 levels in urban street conditions.
Atmosphere 2019, 10, x FOR PEER REVIEW 10 of 20 reduction in photodissociation of NO2 under low O3 concentrations (Figure 5c).High NO2 creates suitable conditions for HNO3 formation, compensating for the effect of decreased O3 on HNO3.Thus, HNO3 and nitrate aerosols have no clear relationship with NOx emissions and only undergo small changes.These results imply that NOx emission controls cannot improve PM2.5 levels in urban street conditions.In addition, we estimated the sensitivity of nitrate aerosol production to VOC emissions.Figure 6a shows the average concentration of nitrate aerosol in the street canyon following vehicular VOC emission changes; nitrate concentrations are proportional to VOC emissions.The average nitrate concentration in the VOC × 0.25 simulation was 8% lower than that of the CNTL simulation, indicating the higher sensitivity of VOC emissions to nitrate aerosols compared with that of NOx emissions.The nitrate aerosols in VOC × 0 only showed a 12% difference with those of the CNTL simulation, which implies the large impact of the boundary condition on nitrate formation.The changes in HNO3 concentration follow the changes in nitrate concentration, implying that the former is caused by the latter (Figure 6b).Reducing VOC emissions drives a decrease in both NO2 and O3 concentrations, creating unfavorable conditions for HNO3 production, in contrast to the effect of NOx emissions (Figure 6c,d).These results are consistent with a previous box modeling study, which suggests that increases in VOC emissions induce nitrate production via O3 increases under a VOC-limited regime for O3 production [57].Considering that megacities are generally under a VOC-limited regime [58,59], VOC emission control can improve both O3 control and PM2.5 control in urban street canyons.In addition, we estimated the sensitivity of nitrate aerosol production to VOC emissions.Figure 6a shows the average concentration of nitrate aerosol in the street canyon following vehicular VOC emission changes; nitrate concentrations are proportional to VOC emissions.The average nitrate concentration in the VOC × 0.25 simulation was 8% lower than that of the CNTL simulation, indicating the higher sensitivity of VOC emissions to nitrate aerosols compared with that of NO x emissions.The nitrate aerosols in VOC × 0 only showed a 12% difference with those of the CNTL simulation, which implies the large impact of the boundary condition on nitrate formation.The changes in HNO 3 concentration follow the changes in nitrate concentration, implying that the former is caused by the latter (Figure 6b).Reducing VOC emissions drives a decrease in both NO 2 and O 3 concentrations, creating unfavorable conditions for HNO 3 production, in contrast to the effect of NO x emissions (Figure 6c,d).These results are consistent with a previous box modeling study, which suggests that increases in VOC emissions induce nitrate production via O 3 increases under a VOC-limited regime for O 3 production [57].Considering that megacities are generally under a VOC-limited regime [58,59], VOC emission control can improve both O 3 control and PM 2.5 control in urban street canyons.Finally, we investigated the sensitivity of NH3 emissions to nitrate production.Figure 7a shows the average concentration of nitrate aerosols in the street canyon following vehicular NH3 emission changes.The concentration of nitrate aerosol was considerably influenced by the NH3 emission changes; the nitrate concentration in the NH3 × 4 simulation was 42% higher than that in the standard simulation, and the nitrate concentration in NH3 × 0.25 was 85% of that in the standard simulation.These results indicate that the sensitivity of NH3 emissions to nitration is much higher than that of NOx and slightly higher than that of the VOC emissions.The HNO3 concentrations are inversely proportional to NH3 emissions, indicating that higher NH3 emissions induce a higher conversion rate of HNO3 to nitrate aerosol (Figure 7b).These results suggest that the production of ammonium nitrate is reduced by the low concentration of NH3 under an NH3-limited regime for nitrate production.Studies based on both modeling and observed campaigns have reported that nitrate formation occurs under an NH3-limited regime in East Asian megacities, including SMR [21,60].The nitrate aerosols in the NH3 × 0 simulation were 19% lower than those with the CNTL simulation, indicating that ammonium and NH3 concentrations from the boundary also have an important role in nitrate formation in the urban street canyon.These results indicate that the control of NH3 emissions might be the most effective way to degrade PM2.5 problems where vehicular emissions are dominant in winter.The regulation of vehicle emissions is mostly focused on the control of NOx emissions; considering the present findings, we should instead focus on controlling VOC and NH3.
Though we used the coupled CFD-chemistry model to investigate the sensitivity of nitrate aerosols from vehicular emissions under complex geometry, our simulation still has limitations.Sea-salt aerosol significantly impacts the formation of nitrate aerosols via heterogeneous reactions when interacting with trace gases on the surface of sea-salt aerosol [61].This process drives the efficient production of nitrates under an NH3-limited environment.This model does not account for the effect of sea-salt aerosol on nitrate aerosol production.Therefore, the simulation might Finally, we investigated the sensitivity of NH 3 emissions to nitrate production.Figure 7a shows the average concentration of nitrate aerosols in the street canyon following vehicular NH 3 emission changes.The concentration of nitrate aerosol was considerably influenced by the NH 3 emission changes; the nitrate concentration in the NH 3 × 4 simulation was 42% higher than that in the standard simulation, and the nitrate concentration in NH 3 × 0.25 was 85% of that in the standard simulation.These results indicate that the sensitivity of NH 3 emissions to nitration is much higher than that of NO x and slightly higher than that of the VOC emissions.The HNO 3 concentrations are inversely proportional to NH 3 emissions, indicating that higher NH 3 emissions induce a higher conversion rate of HNO 3 to nitrate aerosol (Figure 7b).These results suggest that the production of ammonium nitrate is reduced by the low concentration of NH 3 under an NH 3 -limited regime for nitrate production.Studies based on both modeling and observed campaigns have reported that nitrate formation occurs under an NH 3 -limited regime in East Asian megacities, including SMR [21,60].The nitrate aerosols in the NH 3 × 0 simulation were 19% lower than those with the CNTL simulation, indicating that ammonium and NH 3 concentrations from the boundary also have an important role in nitrate formation in the urban street canyon.These results indicate that the control of NH 3 emissions might be the most effective way to degrade PM 2.5 problems where vehicular emissions are dominant in winter.The regulation of vehicle emissions is mostly focused on the control of NO x emissions; considering the present findings, we should instead focus on controlling VOC and NH 3 .
Though we used the coupled CFD-chemistry model to investigate the sensitivity of nitrate aerosols from vehicular emissions under complex geometry, our simulation still has limitations.Sea-salt aerosol significantly impacts the formation of nitrate aerosols via heterogeneous reactions when interacting with trace gases on the surface of sea-salt aerosol [61].This process drives the efficient production of nitrates under an NH 3 -limited environment.This model does not account for the effect of sea-salt aerosol on nitrate aerosol production.Therefore, the simulation might underestimate the nitrate formation of heterogeneous chemistry.Moreover, we only considered the effect of vehicular emissions; NH 3 and VOCs emissions from heating or biogenic emissions might affect the sensitivity of nitrate formation in the street canyon.The absence of these emissions in the model domain might create uncertainty in the nitrate aerosol calculation in this simulation.
underestimate the nitrate formation of heterogeneous chemistry.Moreover, we only considered the effect of vehicular emissions; NH3 and VOCs emissions from heating or biogenic emissions might affect the sensitivity of nitrate formation in the street canyon.The absence of these emissions in the model domain might create uncertainty in the nitrate aerosol calculation in this simulation.

Model Sensitivity to Geometry and Speciation of VOC Emissions
We examined the sensitivity of the model to the canyon by conducting sensitivity model simulations in which we changed the street canyon aspect ratios (the ratio of building height to street width) of the street canyon to 0.5 and 2.0.The conditions of the sensitivity simulations were identical to those of the CTNL simulation except that the height of the buildings, 10 m and 40 m, respectively, indicating canyon aspect ratios of 0.5 and 2.0 (Figure 8).We named the sensitivity simulations for different aspect ratios "species name" × "multiplying factor" _A "aspect ratio" (e.g.CNTL_A2.0 and NOx × 2_A0.5).Figures 9 and 10 indicate the meridionally averaged NOx, O3, HNO3, and nitrate aerosol concentrations in the CNTL_A0.5 and CNTL_A 2.0 simulations.The nitrate aerosol and their precursors showed similar distributions as those of the CNTL simulations despite the difference in their aspect ratios.The NOx concentration was highest at the surface and is an order of magnitude higher than that outside the canyon, indicating the trapping of vehicular emissions due to the strong canyon vortex in the street canyon (Figures 9a and 10a).The spatial distribution of NOx indicates that concentrated vehicular emissions drive the NOx titration of the O3 concentration at the surface under a low VOC emissions condition in both cases (Figures 9b and 10b).The HNO3 concentrations also show similar distribution to those of the CTNL simulation, indicating the suppressing of the production of OH and HNO3 (Figures 9c and 10c).Figures 9d and 10d show the nitrate concentrations in the street canyon for different canyon aspect ratios.The averaged nitrate concentrations in the street canyon (<10 m and <40 m, respectively) show 5.7 and 6.2 μg m −3 in the CNTL_A0.5 and CNTL_A 2.0 simulations, respectively, which are 9%, and 19% higher than those of the CTNL simulation.These differences are mainly due to the complex canyon vortex driving for the building geometry [27].Despite the dispersion patterns differing according to canyon geometry, the mechanism of nitrate aerosol formation was consistent with that of CNTL, suggesting high nitrate formation from vehicular emissions in the street canyon.

Model Sensitivity to Geometry and Speciation of VOC Emissions
We examined the sensitivity of the model to the canyon geometry by conducting sensitivity model simulations in which we changed the street canyon aspect ratios (the ratio of building height to street width) of the street canyon to 0.5 and 2.0.The conditions of the sensitivity simulations were identical to those of the CTNL simulation except that the height of the buildings, 10 m and 40 m, respectively, indicating canyon aspect ratios of 0.5 and 2.0 (Figure 8).We named the sensitivity simulations for different aspect ratios "species name" × "multiplying factor" _A "aspect ratio" (e.g., CNTL_A2.0 and NO x × 2_A0.5).Figures 9 and 10 indicate the meridionally averaged NO x , O 3 , HNO 3 , and nitrate aerosol concentrations in the CNTL_A0.5 and CNTL_A 2.0 simulations.The nitrate aerosol and their precursors showed similar distributions as those of the CNTL simulations despite the difference in their aspect ratios.The NO x concentration was highest at the surface and is an order of magnitude higher than that outside the canyon, indicating the trapping of vehicular emissions due to the strong canyon vortex in the street canyon (Figures 9a and 10a).The spatial distribution of NO x indicates that concentrated vehicular emissions drive the NO x titration of the O 3 concentration at the surface under a low VOC emissions condition in both cases (Figures 9b and 10b).The HNO 3 concentrations also show similar distribution to those of the CTNL simulation, indicating the suppressing of the production of OH and HNO 3 (Figures 9c and 10c).Figures 9d and 10d show the nitrate concentrations in the street canyon for different canyon aspect ratios.The averaged nitrate concentrations in the street canyon (<10 m and <40 m, respectively) show 5.7 and 6.2 µg m −3 in the CNTL_A0.5 and CNTL_A 2.0 simulations, respectively, which are 9%, and 19% higher than those of the CTNL simulation.These differences are mainly due to the complex canyon vortex driving for the building geometry [27].Despite the dispersion patterns differing according to canyon geometry, the mechanism of nitrate aerosol formation was consistent with that of CNTL, suggesting high nitrate formation from vehicular emissions in the street canyon.We tested the NOx, VOC, and NH3 emission sensitivity to nitration formation for different canyon aspect ratios (0.5 and 2.0).Figures 11 and 12 show the averaged nitrate concentration in the street canyons (<10 m and <40 m, respectively) by following the vehicular NOx, VOC, and NH3 emission changes for the different canyon aspect ratios of 0.5 and 2.0, respectively.The sensitivity of the nitrate formation following NOx, VOC, and NH3 emission changes is also similar to those for the canyon aspect ratio of unity.The nitrate concentration changes show no clear relationship with the NOx emission rate in either case, which is related to the nitrate precursor changes, particularly in the daytime as we mentioned (not shown).The maximum concentrations occurred in NOx × 0.5_A0.5 and NOx × 0.5_A2.0,which differed slightly with the simulations for the canyon aspect ratio of unity.Nevertheless, the difference in the nitrate concentrations between the simulations was only 2%.The sensitivity of nitrate formation following VOC and NH3 emissions also follows consistent results with those for the aspect ratio of unity.Enhanced (reduced) VOC emissions drive an increase (decrease) in the nitrate aerosol concentration affecting the O3 concentration under a VOC limited regime for O3 production.The nitrate concentration in the street canyon is proportional to the vehicular NH3 emission, indicating that NH3 limits the condition of nitrate formation in both cases.These results indicate that the sensitivity of nitrate formation following emission changes in the We tested the NO x , VOC, and NH 3 emission sensitivity to nitration formation for different canyon aspect ratios (0.5 and 2.0).Figures 11 and 12 show the averaged nitrate concentration in the street canyons (<10 m and <40 m, respectively) by following the vehicular NO x , VOC, and NH 3 emission changes for the different canyon aspect ratios of 0.5 and 2.0, respectively.The sensitivity of the nitrate formation following NO x , VOC, and NH 3 emission changes is also similar to those for the canyon aspect ratio of unity.The nitrate concentration changes show no clear relationship with the NO x emission rate in either case, which is related to the nitrate precursor changes, particularly in the daytime as we mentioned (not shown).The maximum concentrations occurred in NO x × 0.5_A0.5 and NO x × 0.5_A2.0,which differed slightly with the simulations for the canyon aspect ratio of unity.Nevertheless, the difference in the nitrate concentrations between the simulations was only 2%.The sensitivity of nitrate formation following VOC and NH 3 emissions also follows consistent results with those for the aspect ratio of unity.Enhanced (reduced) VOC emissions drive an increase (decrease) in the nitrate aerosol concentration affecting the O 3 concentration under a VOC limited regime for O 3 production.The nitrate concentration in the street canyon is proportional to the vehicular NH 3 emission, indicating that NH 3 limits the condition of nitrate formation in both cases.These results indicate that the sensitivity of nitrate formation following emission changes in the street canyon is consistent, regardless of the building aspect ratio, due to concentrated emissions from vehicles and the canyon vortex.
Atmosphere 2019, 10, x FOR PEER REVIEW 15 of 20 street canyon is consistent, regardless of the building aspect ratio, due to concentrated emissions from vehicles and the canyon vortex.The VOC speciation of vehicular emissions might change the model sensitivity on nitrate formation by affecting the ozone and OH production, considering all mechanisms are explained under VOC limited regime for O3 production.Therefore, we checked the sensitivity of VOC speciation of emissions on nitrate aerosol formation using the different VOC chemical speciation used by Kim et al. (2006) [62].Table 5 summarizes the emission rates and the ratio of speciated VOC with the method of Kim et al. (2006).Figure 13 indicates the averaged nitrate concentration in the street canyon (i.e., below 20 m) by following vehicular NOx, VOC, and NH3 emission changes with the VOC speciation of Kim et al. (2006).Similar to other sensitivity simulations, NOx emission changes did not affect the nitrate formation (due to the conflicting effects of NO2 and O3) even though we changed the VOC speciation (Figure 13a).The sensitivity of VOC concentration to nitrate formation also shows a similar relationship to that with EMEP/EEA speciation.However, the sensitivity was slightly lower than the nitrate concentration in CNTL, showing only a 6% difference between VOC × 0.25 and CNTL (Figure 13b).These results show that the reduction of vehicular VOC emission is a more effective way to regulate nitrate problems in urban streets than NOx emissions under different VOC speciations.The VOC speciation of vehicular emissions might change the model sensitivity on nitrate formation by affecting the ozone and OH production, considering all mechanisms are explained under a VOC limited regime for O3 production.Therefore, we checked the sensitivity of VOC speciation of emissions on nitrate aerosol formation using the different VOC chemical speciation used by Kim et al. (2006) [62].Table 5 summarizes the emission rates and the ratio of speciated VOC with the method of Kim et al. (2006).Figure 13 indicates the averaged nitrate concentration in the street canyon (i.e., below 20 m) by following vehicular NOx, VOC, and NH3 emission changes with the VOC speciation of Kim et al. (2006).Similar to other sensitivity simulations, NOx emission changes did not affect the nitrate formation (due to the conflicting effects of NO2 and O3) even though we changed the VOC speciation (Figure 13a).The sensitivity of VOC concentration to nitrate formation also shows a similar relationship to that with EMEP/EEA speciation.However, the sensitivity was slightly lower than the nitrate concentration in CNTL, showing only a 6% difference between VOC × 0.25 and CNTL (Figure 13b).These results show that the reduction of vehicular VOC emission is a more effective way to regulate nitrate problems in urban streets than NOx emissions under different VOC speciations.The VOC speciation of vehicular emissions might change the model sensitivity on nitrate formation by affecting the ozone and OH production, considering all mechanisms are explained under a VOC limited regime for O 3 production.Therefore, we checked the sensitivity of VOC speciation of emissions on nitrate aerosol formation using the different VOC chemical speciation used by Kim et al. (2006) [62].Table 5 summarizes the emission rates and the ratio of speciated VOC with the method of Kim et al. (2006).Figure 13 indicates the averaged nitrate concentration in the street canyon (i.e., below 20 m) by following vehicular NO x , VOC, and NH 3 emission changes with the VOC speciation of Kim et al. (2006).Similar to other sensitivity simulations, NO x emission changes did not affect the nitrate formation (due to the conflicting effects of NO 2 and O 3 ) even though we changed the VOC speciation (Figure 13a).The sensitivity of VOC concentration to nitrate formation also shows a similar relationship to that with EMEP/EEA speciation.However, the sensitivity was slightly lower than the nitrate concentration in CNTL, showing only a 6% difference between VOC × 0.25 and CNTL (Figure 13b).These results show that the reduction of vehicular VOC emission is a more effective way to regulate nitrate problems in urban streets than NO x emissions under different VOC speciations.

Conclusions
Nitrate contributions to PM2.5 mass have increased in polluted urban areas, with an increasing number of severe haze events in East Asia.This study investigates the favorable conditions for the production of nitrate aerosols in urban streets using a coupled CFD-chemistry model.It was found that the nitrate concentrations in street canyons are higher than those outside the canyons due to the high HNO3 concentrations from vehicular NOx in these canyons.However, the spatial pattern of nitrate aerosols in street canyons differs from that of HNO3 due to NH3, thus indicating that ammonium nitrate formation occurs under NH3-limited conditions in street canyons.
Sensitivity simulations indicate that nitrate concentration does not show a clear relationship with the NOx emission rate, with nitrate changes of only 2% across among 16 time differences in NOx emissions.The HNO3 concentration in street canyons changes in a similar manner to that of nitrate aerosols, indicating that the changes in nitrate aerosols in the sensitivity simulations are closely related to HNO3 changes.An increase in the NOx emissions induces a decrease in O3 and an increase in NO2 under a VOC-limited regime for O3 production.These changes in O3 and NO2 have a conflicting effect on the HNO3 production in urban streets.Therefore, HNO3 and nitrate aerosols have no linear relationship with NOx emissions and only undergo small changes.The sensitivity simulations were conducted by varying the vehicular VOC emissions to investigate their effect on nitrate production.The results show that nitrate concentrations are proportional to VOC emissions.Nitrate was decreased by 9% in the VOC × 0.25 simulation, indicating a relatively high sensitivity compared to that of NOx.Decreased VOC emissions drive a decrease in both NO2 and O3 concentrations, creating unfavorable conditions for HNO3 production, unlike the effect of changes in NOx emissions.The nitrate aerosol concentration is considerably influenced by NH3 emissions, which show a higher sensitivity to nitrate production than do NOx and VOC emissions in urban streets.The nitrate concentration is proportional to NH3 emissions with the additional production of ammonium nitrate under an NH3-limited regime for nitrate production.This research implies that, where vehicular emissions are dominant in winter, the control of vehicular VOC and NH3 emissions might be a more effective way to degrade PM2.5 problems than controlling NOx.

Conclusions
Nitrate contributions to PM 2.5 mass have increased in polluted urban areas, with an increasing number of severe haze events in East Asia.This study investigates the favorable conditions for the production of nitrate aerosols in urban streets using a coupled CFD-chemistry model.It was found that the nitrate concentrations in street canyons are higher than those outside the canyons due to the high HNO 3 concentrations from vehicular NO x in these canyons.However, the spatial pattern of nitrate aerosols in street canyons differs from that of HNO 3 due to NH 3 , thus indicating that ammonium nitrate formation occurs under NH 3 -limited conditions in street canyons.
Sensitivity simulations indicate that nitrate concentration does not show a clear relationship with the NO x emission rate, with nitrate changes of only 2% across among 16 time differences in NO x emissions.The HNO 3 concentration in street canyons changes in a similar manner to that of nitrate aerosols, indicating that the changes in nitrate aerosols in the sensitivity simulations are closely related to HNO 3 changes.An increase in the NO x emissions induces a decrease in O 3 and an increase in NO 2 under a VOC-limited regime for O 3 production.These changes in O 3 and NO 2 have a conflicting effect on the HNO 3 production in urban streets.Therefore, HNO 3 and nitrate aerosols have no linear relationship with NO x emissions and only undergo small changes.The sensitivity simulations were conducted by varying the vehicular VOC emissions to investigate their effect on nitrate production.The results show that nitrate concentrations are proportional to VOC emissions.Nitrate was decreased by 9% in the VOC × 0.25 simulation, indicating a relatively high sensitivity compared to that of NO x .Decreased VOC emissions drive a decrease in both NO 2 and O 3 concentrations, creating unfavorable conditions for HNO 3 production, unlike the effect of changes in NO x emissions.The nitrate aerosol concentration is considerably influenced by NH 3 emissions, which show a higher sensitivity to nitrate production than do NO x and VOC emissions in urban streets.The nitrate concentration is proportional to NH 3 emissions with the additional production of ammonium nitrate under an NH 3 -limited regime for nitrate production.This research implies that, where vehicular emissions are dominant in winter, the control of vehicular VOC and NH 3 emissions might be a more effective way to degrade PM 2.5 problems than controlling NO x .
We checked the model sensitivity by changing the model's building geometry and VOC speciation.The sensitivity of nitrate formation by following emissions changes acts in a similar direction as CNTL simulation despite changing the building geometry and speciation of VOC emissions.The sensitivity simulations revealed that our results about the sensitivity of nitrate production to emission changes are robust.

Figure 1 .
Figure 1.Schematic diagram of the coupled computational fluid dynamics (CFD)-chemistry simulation domain for the control run (CNTL) simulation.

Figure 1 .
Figure 1.Schematic diagram of the coupled computational fluid dynamics (CFD)-chemistry simulation domain for the control run (CNTL) simulation.

Atmosphere 2019 , 20 Figure 2 .
Figure 2. Comparisons of nitrate concentrations in the spring, summer, autumn, and winter cases between this work and the previous results of Yubero et al. (2015) (see Figure 1 in Yubero et al.).The values are averaged for the analysis period at the sample station.The black solid line is the observed concentrations and the red dashed line is the simulated concentrations.The error bars indicate the standard deviations of the observed nitrate aerosol in each season.

Figure 2 .
Figure 2. Comparisons of nitrate concentrations in the spring, summer, autumn, and winter cases between this work and the previous results of Yubero et al. (2015) (see Figure 1 in Yubero et al.).The values are averaged for the analysis period at the sample station.The black solid line is the observed concentrations and the red dashed line is the simulated concentrations.The error bars indicate the standard deviations of the observed nitrate aerosol in each season.
Atmosphere 2019, 10, x FOR PEER REVIEW 8 of 20 reached 2.3 ppbv, with the highest value at the surface.The dispersion pattern of NH3 was similar to that of NOx, indicating the high impact of vehicular emissions.

Figure 4 .
Figure 4. Distribution of the daily average concentrations of (a) nitrate and (b) ammonium (µ g m −3 ) in the CNTL simulation.

Figure 4 .
Figure 4. Distribution of the daily average concentrations of (a) nitrate and (b) ammonium (µg m −3 ) in the CNTL simulation.

Figure 5 .
Figure 5. Average (a) nitrate, (b) HNO 3 , (c) NO 2 , and (d) O 3 concentrations in the street canyon (i.e., below 20 m) following a change in the emission rate of vehicular NO x .

Figure 7 .
Figure 7. Average (a) nitrate and (b) HNO3 concentrations in the street canyon (i.e., below 20 m) following a change in the emission rate of vehicular NH3.

Figure 7 .
Figure 7. Average (a) nitrate and (b) HNO 3 concentrations in the street canyon (i.e., below 20 m) following a change in the emission rate of vehicular NH 3 .

Figure 8 .
Figure 8. Schematic diagrams of the coupled sensitivity simulation domain for the canyon aspect ratios (a) 0.5 and (b) 2.0.

Figure 8 . 20 Figure 8 .
Figure 8. Schematic diagrams of the coupled sensitivity simulation domain for the canyon aspect ratios (a) 0.5 and (b) 2.0.

Figure 11 .
Figure 11.The average nitrate concentration in the street canyon following changes in the emission rate of vehicular (a) NOx, (b) VOC, and (c) NH3 for the 0.5 canyon aspect ratio; units are µ g m −3 .

Figure 12 .
Figure 12.The same as Figure 11 but for the 2.0 canyon aspect ratio; units are µ g m −3 .

Figure 11 .
Figure 11.The average nitrate concentration in the street canyon following changes in the emission rate of vehicular (a) NO x , (b) VOC, and (c) NH 3 for the 0.5 canyon aspect ratio; units are µg m −3 .

Figure 11 .
Figure 11.The average nitrate concentration in the street canyon following changes in the emission rate of vehicular (a) NOx, (b) VOC, and (c) NH3 for the 0.5 canyon aspect ratio; units are µ g m −3 .

Figure 12 .
Figure 12.The same as Figure 11 but for the 2.0 canyon aspect ratio; units are µ g m −3 .

Figure 12 .
Figure 12.The same as Figure 11 but for the 2.0 canyon aspect ratio; units are µg m −3 .

Figure 13 .
Figure 13.Average nitrate concentration in the street canyon following changes in the emission rate of vehicular (a) NOx and (b) VOC with the speciation method of VOC emission used by Kim et al. (2006); units are µ g m −3 .

Figure 13 .
Figure 13.Average nitrate concentration in the street canyon following changes in the emission rate of vehicular (a) NO x and (b) VOC with the speciation method of VOC emission used by Kim et al. (2006); units are µg m −3 .

Table 2 .
Ratios of vehicles following size and fuel type used in the CNTL simulation obtained from the Traffic Monitoring System.

Table 3 .
Emission rates per vehicle and ratios of speciated volatile organic compound (VOC) following the method developed by EMEP/EEA (2016).

Table 4 .
Diurnal variations of the observed hourly surface temperature and relative humidity used in this model.

Table 5 .
Kim et al. (2006)r vehicle and ratios of speciated VOC obtained by following the method developed byKim et al. (2006).

Table 5 .
Kim et al. (2006)r vehicle and ratios of speciated VOC obtained by following the method developed byKim et al. (2006).

Table 5 .
Kim et al. (2006)r vehicle and ratios of speciated VOC obtained by following the method developed byKim et al. (2006).