Environ. Res. Lett. 13 (2018) 114002 https://doi.org/10.1088/1748-9326/aae492

LETTER

Evidence of heterogeneous HONO formation from aerosols and theregional photochemical impact of this HONO source

Xingcheng Lu1 , YuhangWang2,5, Jianfeng Li2, Lu Shen3 and JimmyCHFung1,4,5

1 Division of Environment and Sustainability, HongKongUniversity of Science andTechnology,HongKong SAR, People’s Republic ofChina

2 School of Earth andAtmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332,United States of America3 School of Engineering andApplied Sciences,HarvardUniversity, Cambridge,MA 02138,United States of America4 Department ofMathematics, HongKongUniversity of Science andTechnology,HongKong SAR, People’s Republic of China5 Authors towhomany correspondence should be addressed.

E-mail: ywang@eas.gatech.edu andmajfung@ust.hk

Keywords: 1D framework, aerosol, unknownHONO, regional impact, heterogeneous conversion

Supplementarymaterial for this article is available online

AbstractLargemissing daytimeHONO sources have been reported bymany previous studies around theworld. PossibleHONO sources include ground heterogeneous conversion, aerosol heterogeneousformation, soil emission, and photochemical production. In this study, a consistent 1D frameworkbased on regional and 3D chemical transportmodels (CTMs)was used to analyze the unknowndaytimeHONO sources in 14 cases worldwide.We assume that the source ofHONO fromaerosols isthroughNO2 hydrolysis (not including its oxidation products) and that non-localmixing effect isnegligible. Assuming all themissing unknownHONOsource is from the ground, it would imply aNO2-to-HONOground heterogeneous conversion exceeding 100% in daytime, which is unphysical.In contrast, a strongR2 reaching up to 0.92 is found between the unknownHONO sources and theproducts of aerosol wet surface area and short-wave radiation. Because the largest unknowndaytimeHONO sources are found inChina due to high concentrations of aerosols andNO2, we derive anoptimizedNO2 uptake coefficient on the basis of thesemeasurements. The 3DCTMsimulationssuggest that in some regions of central, eastern, and southwestern (e.g. SiChuan province)China, theaerosolHONO source has the greatest effects on ozone (>10 ppbv) andOH (>200%) inwinter. InJanuary, the simulated particulate sulfate level over these three regions increases by 6–10 μgm−3 afterincluding the aerosol-HONO source, which helps reduce the previousmodel underestimation ofsulfate production inwinter. Additionalmeasurement studies that target the daytimeHONO sourceswill be essential to a better understanding of themechanisms and resulting effects on atmosphericoxidants.

1. Introduction

Ozone (O3) is one of the most notorious ambientphotochemical pollutants worldwide. High concen-trations of this pollutant can adversely affect humanhealth and cause substantial reductions in crop yields(Avnery et al 2011, Lu et al 2016). O3 formation isinfluenced by the photochemical reactions of ROx(OH+HO2+RO2) and NOx in the atmosphere.Many recent studies have investigated the role of theROx cycle in O3 formation (Liu et al 2012a, Lu

et al 2017). Known atmospheric sources ofOHradicalsinclude (but are not limited to) reactions betweenwater vapor and electronically excited oxygen atoms(O1D), the photodissociation of some types of volatileorganic compounds (Jaeglé et al 2011), and thephotolysis of nitrous acid (HONO) (Liu et al 2012a).According to Lu et al (2017), HONO can contribute toover 30%of theO3 production rate during the daytimeat a site of central China inWuhanCity.

Many field campaigns have analyzed the ambientlevel, diurnal profile, and sources ofHONObecause of

OPEN ACCESS

RECEIVED

26November 2017

REVISED

25 September 2018

ACCEPTED FOR PUBLICATION

27 September 2018

PUBLISHED

24October 2018

Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2018TheAuthor(s). Published by IOPPublishing Ltd

https://doi.org/10.1088/1748-9326/aae492https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0003-2787-7016https://orcid.org/0000-0003-2787-7016mailto:ywang@eas.gatech.edumailto:majfung@ust.hkhttps://doi.org/10.1088/1748-9326/aae492http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/aae492&domain=pdf&date_stamp=2018-10-24http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/aae492&domain=pdf&date_stamp=2018-10-24http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0

its importance to atmospheric OH radical formation(Su et al 2008, Liu et al 2012a, Lee et al 2016). Short-wave radiation (SWR) induced photolysis has beenacknowledged as the major sink of HONO during thedaytime. However, the daytime source of HONOremains unknown based on current knowledge andcannot merely be explained by the gaseous reactionbetween atmospheric NO andOH. As a result, the per-formance of the 3D chemical transport model (CTM)simulations of OH radicals and O3 might be influ-enced by the missing HONO source (Zhanget al 2012). Hence, the missing daytime HONO sourceis highly significant andmust be understood to achievebetter 3D CTM simulations of O3. Several mechan-isms have been proposed to explain the missing day-time HONO source. Su et al (2011) proposed thatnitrites in low-pH fertilized soilmay be a strong sourceof ambient HONO (Su et al 2011), whereas Vanden-Boer et al (2015) reported that surface acid displace-ment processes were an important potential HONOdaytime source in both urban and vegetated regions.Zhou et al (2011) found that surface nitrate loadingcorrelated with HONO flux in a forest environment.Furthermore, in the ambient environment, the het-erogeneous conversion of NO2 to HONO on the sootsurface was found to increase dramatically in responseto solar radiation (Monge et al 2010).

Increasing attention has recently been directed tothe heterogeneous conversion of NO2 to HONO onaerosol surfaces. Tong et al (2016) reported that theRH (relative humidity) and PM2.5 (atmospheric parti-culate matter with a diameter

Modern-Era Retrospective analysis for Research andApplications, Version 2 (MERRA-2) for cases withinChina and outside China, respectively. We used AODdata from the moderate resolution imaging spectro-radiometer (MODIS) AOD 550 nm product. In short,the observation and model data used in this studyincludes: HONO data, aerosol surface area data, Jvalues (HONO), SWR value (simulated), dry deposi-tion velocities for HONO and NO2 (simulated,section 2.3), and vertical diffusion coefficient (simu-lated, section 2.3).

2.2.Model configurationWRF v3.2 was used to simulate the meteorology fieldfor the CMAQ model. WRF has a domain resolutionof 27 km, and the domain extent is shown in figure 1(blue dash line). Final Operational Global Analysis(FNL)Datawere used to drive theWRFmodel. CMAQ(v5.0.1) also has a domain resolution of 27 km, and itsdomain extent is plotted as a red line in figure 1. TheCMAQ domain covers most of China and someneighboring countries, including Korea and Japan.CB05 and ISORROPIA were respectively selected asthe gas-phase chemistry scheme and inorganic aerosolscheme for the simulation. The in-line photolysismodule, which considers aerosol, ozone, and cloudattenuation effects, was selected for the chemicalphotolysis rate calculation. The dry deposition velocityin the CMAQ model is calculated using the M3Drymodule (Pleim and Byun 2004). The MIX emissioninventory was used as the anthropogenic emissioninput for this simulation (Li et al 2017). More detailsabout the WRF and CMAQ schemes settings havebeen reported by Lu et al (2015).

GEOS-Chem v10.01 was used in this work. Thismodel was driven by the GEOS-FPmet fields and has aworldwide domain extent with a spatial resolution of2×2.5 degree. The GEOS-Chem model is run withfull gaseous chemistry and an updated isoprenescheme (Mao et al 2013), and the simulated aerosol

species included sulfate, nitrate, ammonium, blackcarbon (BC), sea salts, natural dust, and primaryorganic carbon. A non-local scheme (Lin and McEl-roy 2010) and a relaxed Arakawa-Schubert scheme(Suarez et al 2008) were selected for vertical mixingand convection, respectively, in the simulation. Themodel-based dry deposition velocity calculation wasbased on the Wesely resistance-in-series model(Wesely 1989). More details of the dry deposition cal-culations in GEOS-Chem and CMAQ can be found inthe supplementalmaterial.

2.3. UnknownHONOsource estimationWe proposed an analytic 1D framework based on theequations shown below to analyze the unknownHONO source. The proposed equations consider thegaseous HONO source (NO+OH), daytime HONOsink, NO2 heterogeneous conversion to HONO fromthe ground and an unknown source:

¶¶

= -¶¶

- + ¢

-¶¶

==

⎧⎨⎪⎪

⎩⎪⎪

( )

c

tK

c

zK c S

Kc

zF

, 1z c

zz

2

2

0

where c is the HONO concentration, t is the time,Kz isthe vertical eddy diffusion coefficient (table S2), Kc isthe HONO chemical reaction coefficient (s−1), F is thevertical HONO flux from the ground and S′ is thecombination of the unknown HONO source and thesource from the reaction of NO and OH (moleculecm−3 s−1). The pseudo-steady state is assumed,(¶¶

c

t=0), and hence we selected for analysis the period

when the daytime HONO concentration reached asteady value (the period for each case is listed intable 1). The daytime HONO photochemical lifetimeis on the order of 10–20 min. Therefore, when daytimeHONO concentrations do not change, it suggests thatmixing processes reached a steady state with HONOchemical sources and sinks. We assumed that theunknown HONO source would not change with the

Table 1.References analyzed for theHONOunknown source in this work.

No. References Location Time period Aerosol surface area Note

1 Liu et al (2014) Beijing, China 2007.08 Observation Average Case2 Liu et al (2014) Beijing, China 2007.08 Observation Heavy PollutionCase3 Hou et al (2016) Beijing, China 2014.02/03 CMAQ a CleanCase4 Hou et al (2016) Beijing, China 2014.02 CMAQ a Heavy PollutionCase5 Zhao et al (2015) Shanghai, China 2010.06 CMAQ6 Yue et al (2015) Jiangmen, China 2013.10 CMAQ a

7 Su et al (2008) Guangzhou, China 2004.10 Observation8 Acker et al (2006) Rome, Italy 2001.05/06 GEOS-Chem9 Michoud et al (2014) Paris, France 2009.07 GEOS-Chem a Summer case10 Michoud et al (2014) Paris, France 2010.01/02 GEOS-Chem a Winter case11 Li et al (2010) MexicoCity,Mexico 2006.03 GEOS-Chem12 VandenBoer et al (2014) California, USA 2010.05/06 Observation13 VandenBoer et al (2013) Colorado,USA 2011.02 GEOS-Chem14 Gall et al (2016) Dallas, USA 2011.06 Observation

a The aerosol wet surface areawas corrected by the ratio of the observed PM2.5 concentration over the simulated PM2.5 concentration.

3

Environ. Res. Lett. 13 (2018) 114002

height near the surface, allowing us to derive ananalytical solution for equation (1):

= + ( )X FK K

S

K, 2

c z c

where X is the observed HONO concentration sub-tracted the HONO formed by NO+OH. Otherunknown gas phase HONO formation reaction is notconsidered here. Kc is calculated based on the HONOphotolysis rate and the reaction rate between HONOand OH, which determine the HONO lifetime duringthe daytime. F is the net HONO ground flux and iscalculated as the upwardHONO flux subtracted by thedownward HONO flux (dry deposition). S is theunknown HONO source (molecule cm−3 s−1), whichis not the ground heterogeneous conversion or thegaseous phase formation. Zhang et al (2016) evaluatedseveral PBL schemes in a 1D column model toinvestigate the boundary-layer vertical gradients ofreactive nitrogen oxides using the DISCOVER-AQ2011 campaign and reported that the ACM2 scheme isthe best and that it can reproduce the profile of thedeep turbulent mixing. Hence, in this work, ACM2scheme was used to calculate the vertical diffusioncoefficients for the 1D analytic framework. Non-localmixing is not considered here, which may cause biasesin the computedHONObudget calculation.However,the differences among the different WRF boundarylayer mixing schemes found by Zhang et al (2016) isconsiderably less than the uncertainties in unknownHONOaerosols.

In this analytic 1D framework, we assumed that asthe NO2molecules reached the ground via dry deposi-tion, part of them would be converted to HONO andreleased to the atmosphere immediately (upwardHONO flux). According to Liu et al (2014), weassumed that the ground heterogeneous conversion ofNO2 and dry deposition of HONO were the majornighttime HONO source and sink, respectively. The

NO2 ground conversion ratio ( f ) during the nighttimecould then be calculated by equation (3):

´ = ´ ´[ ] [ ] ( )V f VHONO NO . 3HONO 2 NO2

The dry deposition velocities of HONO andNO2 werecalculated using the CMAQ (within China) andGEOS-Chem3Dmodels. TheNO2 ground conversionratio ( f ) from equation (3) was then used to calculatethe upward HONO flux during the daytime. Hence,the F term in equation (2) is calculated by subtractingthe right-hand side by the left-hand side ofequation (3). The conversion ratio ( f ) for each casecan be found in table S2 in the supplemental material.The parameter, ( f ), ismainly dependent on the surfacecharacteristics, e.g. surface area, surface type. Hence, itis reasonable to keep the value of ( f ) constant whenthe sampling location did not change. Some of otherworks also applied the dry deposition rate of NO2 tocalculate the NO2 ground conversion rate during thenighttime (Li et al 2012, Liu et al 2014).

After acquiring the unknown HONO source S(molecule cm−3 s−1) from equation (1), the NO2uptake coefficient for aerosol could then be calculatedusing equations (4) and (5) listed below:

= ´ [ ] ( )S k NO , 4a 2

å gw= +=

-⎡⎣⎢⎢

⎤⎦⎥⎥ ( )k

r

DA

4, 5a

i

npi

gi

1

1

where ka is the first-order NO2 uptake coefficient foraerosol (s−1), Ai and rp

i are the aerosol surface area(μm2 cm−3) and particle radius (μm) for the ith sizebin, respectively.ω is theNO2meanmolecular velocity(m s−1), Dg is the gas molecular diffusion coefficient(m2 s−1) and γ is theNO2 uptake coefficient.

We discuss the correlation between the unknownHONO source and aerosol wet surface area insection 3. TheNO2 uptake coefficient, calculated usingequation (5), was implemented into the CMAQmodel

Figure 1.HONOdata sampling locations andWRF-CMAQmodel domain setting. GreenTriangle represents the location of thestudied site; blue and red lines represent the domain extents ofWRF andCMAQsimulations, respectively.

4

Environ. Res. Lett. 13 (2018) 114002

and its effects on O3 and OH concentrations in Chinawere investigated.

3. Results and discussion

3.1.HeterogeneousHONO formation fromaerosolThe calculated unknown HONO concentration/source and time periods we selected for the study areshown in table 2. The calculated unknown HONOconcentrations ranged from 7.17 × 108 (No. 13 forColorado case) to 4.07 × 1010 molecules cm−3 (No.2for 2007 Beijing Heavy Pollution Case), and theunknown HONO sources ranged from 8.70 × 105

(No.13 for Colorado case) to 3.48 × 107 moleculescm−3 s−1 (No.4 for 2014 Beijing Heavy PollutionCase). The calculated unknown HONO concentra-tions were generally higher in China than in othercountries. The unknown HONO sources in Beijing(cases No.1, 2 and 4), Jiangmen (No.6) and Guangz-hou (No.7 for Xinken site) all exceeded 107 moleculescm−3 s−1.

Figure 2(a) depicts the R2 between the calculatedunknownHONO source and aerosol wet surface area.A total of 14 points were used for the correlation ana-lysis, and each point represents an individual case lis-ted in table 1. The R2 between these two parametersreached 0.87. Some studies have reported that SWRcould enhance heterogeneous HONO formation byaerosol (Liu et al 2014, Lu et al 2017). Accordingly, thethe R2 value between the unknownHONO source andaerosol wet surface area increased to 0.92 after multi-plying by SWR (figure 2(b)). These results provide theevidence to support that the aerosol surfaces comprisean important medium for HONO formation. The OHradical concentrations for Jiangmen (No.6), Shanghai(No.5) and Rome (No.8) cases were calculated by the3D CTM. A sensitivity test was performed by adding

50% more to the OH concentrations to these threecases and the changes of the calculated unknownHONO source are 2.9% (No.6 for Jiangmen case),1.5% (No.8 for Rome case) and 3.2% (No.5 for Shang-hai case) respectively, when compared to the originalvalues. This shows that HONO production from thereaction of NO+OH is limited when compared tothe unknown HONO source and the uncertaintycaused by the OH radical computation in these threecases would not influence the results. Since the effectof multiple scattering by aerosols is not included, theattenuation effect of SWR may be overestimated.However, the biases do not lead to false correlationsbetween the unknown HONO source and aerosolsince the bias in SWR estimate is monotonic withaerosol AOD.

NO2 is themajor gaseous precursor of bothHONOand aerosol formation, and this might drive the strongcorrelation between the aerosol surface area and theunknown HONO source. As shown in figure 2(c), theR2 between NO2 and the unknown HONO source wasonly 0.62, much lower than the R2 between theunknown HONO source and aerosol wet surface area.Multiplication by SWR decreased the R2 value betweenthe unknown HONO source and NO2 to 0.19. There-fore, figures 2(c) and (d) indicate that the strong R2

between the unknown HONO source and the aerosolsurface area was not driven by NO2. However, the R

2

between the unknown HONO source and the productof aerosol wet surface area×NO2×SWR droppeddown to 0.78. Except the cases of California, Colorado,Beijing 2007 and Shanghai, the NO2 data in the otherdatasets were measured by the chemiluminescencemethod, whichmight be influenced by theNOy species(e.g. PAN). We further applied the observed seasonalratios of coincident NO2 measurements of by the che-miluminescence instrument to the more selective pho-tolytic instrument reported by Zhang et al (2018) tocorrect the NO2 data and found that the R

2 (0.87)increased after the correction (figure 2(f)). Xu et al(2013) reported that this ratio varied in different sea-sons and locations. Hence, the applied correction ratiosby Zhang et al (2018)may not be representative for thedifferent campaigns in this study. By using a consistentNO2 dataset, the R

2 between the unknown HONOsource and the product of SWR×NO2×aerosol sur-face area was the highest when compared to other cor-relations in Liu et al (2014). Therefore, figures 2(e), (f)implies that the decrease of R2 after adding NO2 mightbe caused by the inconsistent uncertainties in NO2measurements in different campaigns. The decrease ofthe R2 value may reflect some of the approximationsused in this study. However, we believe that the changemay be more numerical than physical. Panel (b) offigure 2 shows much more evenly distributed datapoints without NO2 than panel (f) with NO2. In panel(f), there are effectively only 7 data points because thelow NO2×SWR×aerosol wet surface area datapoints are essentially lumped together. The large

Table 2.Calculated unknownHONOconcentrations/sources forthe selected campaigns.

Time

UnknownHONO

concentrationaUnknown

HONOsourceb

1 14:00–16:00 1.78×1010 2.00×107

2 15:00–17:00 4.07×1010 2.84×107

3 15:00–16:00 1.25×1010 9.38×106

4 13:00–14:00 4.00×1010 3.48×107

5 14:00–15:00 5.76×109 6.82×106

6 10:00–14:00 2.39×1010 1.70×107

7 12:00–14:00 1.98×1010 2.50×107

8 14:00–15:00 3.25×109 6.88×106

9 14:00–16:00 2.09×109 3.07×106

10 15:00–17:00 7.49×109 1.75×106

11 14:00–16:00 3.86×109 4.24×106

12 16:00–17:00 1.70×109 1.90×106

13 11:00–13:00 7.17×108 8.70×105

14 9:00–11:00 5.19×109 7.93×106

a Values are shown inmolecules cm−3.b Values are shown inmolecules cm−3 s−1.

5

Environ. Res. Lett. 13 (2018) 114002

deviations of the three data points (in red, black, andblue) from the regression line explain most of thedecrease of theR2 value. Recent studies pointed towardsmuch faster photolysis of aerosol nitrate as a source ofHONO (Ye et al 2016, Ye et al 2017). Our analysis can-not rule out that this pathway contributes significantlyto daytime HONO formation. Given the relativelysmall amount of nitrate in aerosols, the supply of aero-sol nitrate through the reaction of OH andNO2 in day-time essentially balances out nitrate photolysis andtherefore the production of HONO in aerosols has asimilar daytime cycle as OH.Modeling analysis of sum-mer data at Wangdu, China (e.g. Liu et al 2017, Tanet al 2017) suggests that if aerosol nitrate photolysis rateis ten times higher than that of gas-phase HNO3 photo-lysis, the pathway will provide

equation (2) and set the unknown HONO as directlyequal to the ground-generated HONO. The groundNO2-to-HONO conversion ratio, ( f ), was then calcu-lated by substituting all other parameters into thismodified equation. The new f values for different casesare shown in table 3. Based on our calculation, theNO2-to-HONO conversion ratios of all cases, exceptBeijing 2007 Heavy Pollution (99%), Mexico City(37%) and Colorado (87%), exceeded 100% if theaerosol heterogeneous formation was not considered.In other words, these results suggest the non-physicalconversion of one NO2 molecule to more than oneHONO molecule on the ground. This result indicatesthat NO2 ground heterogeneous conversion cannot bethe only major unknown HONO source under mostcircumstances. We note that it is assumed that soilnitrate is not a source of atmosphericHNO2.

The calculated unknown HONO concentrationsat Chinese sites were large and possibly derived fromsoil emissions. Through laboratory experiments,Oswald et al (2013) found that the ammonia-oxidizingbacteria in soil can directly emit HONO and reportedHONO emission flux values from different types ofsoil. Notably, differences in emission were great, evenfor the same type of soil, as demonstrated by the>500% difference between two pasture soil samplestaken from different locations. Sörgel et al (2015)reported that acidic soil might not promote HONOemission and stated that the mechanisms of soil emis-sion remained under debate. Given this uncertainty,the HONO emission flux for related soil types couldnot be applied to equation (1) when estimating theexact HONO budget from soil emissions based oncurrent knowledge.

Hence, we discuss herein the relative importanceof soil emissions compared to the aerosol hetero-geneous source. Among the seven sites in China, theBeijing and Shanghai sites are located in urban areas,far from farmlands, whereas the Guangzhou (No.7 for

Xinken site) and Jiangmen (No.6) sites are in ruralareas. The urban areas mainly comprise concrete sur-faces, with a small proportion of tree and grass cover-age. Hence, soil emission is unlikely to be a majorHONO source in urban areas. However, it remainsinteresting to compare the relative amounts of HONOgenerated by aerosol sources and estimates of soilHONO flux. Because the sites did not have homo-geneous land-use types, we calculated the averageHONO emission fluxes of pasture, woody savannah,and grassland according to Oswald et al (2013) andadded the optimum value (∼5.8×1010 moleculescm−2 s−1 at 25 °C) to the second term of equation (2).The HONO soil emission flux depends on the soilwater content and can decrease to zero when the soilwater content is near zero or exceeds 40%. For the soilflux, we used the peak value (optimum) for each soiltype at an approximate soil water content of 15%.After adding the potential soil emission flux, the ratiosof aerosol-generated HONO (the third term inequation (2)) versus observed HONO are 56.5%,80.5%, 34.1%, 69.9%, 53.2%, and 60.9% for the Beij-ing 2007 average and heavy cases (No.1 and 2), Beijing2014 clean and heavy cases (No.3 and 4), Guangzhou(No.7), and Jiangmen cases (No.6), respectively. Theoptimum soil emission could sustain the daytimeHONO budget in the Shanghai case (No.5), for whichthe aerosol concentration was relatively low. At theoptimum HONO soil flux, however, HONO from anaerosol source still accounted formore than 60% of thetotal HONO budget for both the Beijing heavy pollu-tion cases. This further demonstrates that aerosol is animportant HONO source in China in urban sites withhigh aerosol loading. Earlier soil studies found thatN2Ocorrelated positively with soil temperature (Dobbie andSmith 2001, Schindlbacher et al 2004). In a soil model,Parton et al (2001) reported that at temperatures below10 °C, the parameter used to characterize the nitrifica-tion process decreased dramatically.Oswald et al (2013)also showed that the soil HONO emissions woulddecrease with decreasing temperature. In winter, lowsoil temperatures prohibit soil HONO emissions, andaerosol becomes amore importantmedium for hetero-geneousHONOformation (e.g.No.4 case).

3.2. NO2 reactive uptake coefficientAs introduced in section 1, the NO2 reactive uptakecoefficient can be calculated using equations (4) and(5) once the unknown HONO source has beenacquired. Because high daytime HONO sources havebeen along with high aerosol loading are found inChina (table 2), the NO2 reactive uptake coefficientsfor the first seven cases were calculated and used tostudy the regional effect onO3 inChina (in section3.3).As shown in table 4, the calculated γ values range from1.0×10−4 (Beijing 2014 heavy pollution case) to7.9×10−4 (Shanghai 2010 case). The magnitudes ofour calculated γ values are equivalent to those reported

Table 3.NO2-to-HONOground conversion ratioswithout consideration ofthe aerosol source.

Conversion ratio

1 118%

2 99%

3 320%

4 290%

5 232%

6 250%

7 195%

8 208%

9 100%

10 195%

11 37%

12 131%

13 87%

14 450%

7

Environ. Res. Lett. 13 (2018) 114002

by Colussi et al (2013) and those used by Liu et al(2014) and Wong et al (2012) for 1D/3D modelsimulations. As introduced in section 1, some aerosolcomponents (e.g. dicarboxylic acid anions) canenhanceHONO formation. Nie et al (2015) observed ahigh NO2/HONO ratio in the mixed plumes resultingfrom biomass burning and fossil fuel emission andconcluded that this mixed aerosol could promoteHONO formation. Other studies reported that theNO2-to-HONO conversion ratio could be influencedby mineral (e.g. Al2O3) (Romanias et al 2013) andmetal (e.g. Fe2+) (Kebede 2015) compositions.According to an observation study, the metal compo-sitions of aerosols in China exhibit large daily varia-tions (Okuda et al 2004). Recently, Han et al (2016 and2017) reported that the NO2 uptake coefficient andHONO yield would not change with light intensity onthe humid acid; however, once the benzophenone wasadded, the HONO yield increased linearly with thelight intensity. Hence, differences in γ values may beattributable to differences in aerosol componentsduring different periods and at various locations.Another factor that influences the uptake coefficientcalculation is the interference of the chemilumines-cence measurements of NO2 as discussed insection 3.1. More field experiments are needed toverify the dominant factors that affect the magnitudeof the reactive NO2 uptake coefficient in an ambientenvironment.

Liu et al (2014) assumed a linear dependence of γon SWR. Using their method, we calculated the γ/SWR (γ′), as shown in the second column of table 4.Our γ′ values for the seven cases were on the order of10−7 to 10−6. In some modeling studies, such as thatconducted by Zhang et al (2012), the reactive NO2uptake coefficient was set to equal the magnitude of γ′calculated in this work, without considering thephoto-enhancement effect. Using such a para-meterization, the NO2 uptake coefficient during day-time could be underestimated (Liu et al 2014).Accordingly, the aerosol contribution to hetero-geneous HONO formation is negligible in some 3Dmodeling studies. Because the exact soil HONO fluxcould not be identified, the γ′ calculated in this workharbors some uncertainty.

3.3. Effect onO3 spatial distributionThe formation of HONO from heterogeneous NO2conversion on aerosol and ground surfaces wasimplemented into the CMAQ model. The effects ofaerosols on regional-scale ambient HONO, OH andO3 concentrations were also investigated. The follow-ing simulations were set for this comparison: (1)HONO formation mechanisms, including gaseousformation in CB05, tailpipe emission (0.8% of totalNOx emission), and ground heterogeneous formation( f set as 0.1) and (2) the mechanism listed in the firstset of simulations, together with the aerosol hetero-geneous HONO formation based on equation (4). Weused the same CMAQ domain shown in figure 1 forthis sensitivity comparison. January, April, July, andOctober 2015 were selected for the simulation, repre-senting four different seasons. To determine hetero-geneous HONO formation on the aerosol surface, weparameterized the reactive NO2 uptake coefficient byγ=γ′×SWR. As shown in table 4, the γ′ valuediffered for each case. Given the short lifetime ofHONO for our analysis periods, HONO sourcesshould be balanced by its chemical loss. We thereforeobtain the optimal γ′ value byminimizing the squareddifference between HONO sources and chemical sinkfor all data points in table 1. The value of the optimizedγ′ is 5.0×10−7.

The difference between the CMAQ baseline andupdated HONO heterogeneous formation scheme isshown in figure 3. Following the update, during winterdaytime (9:00 am–17:00 pm), a simulated HONO dif-ference generally greater than 0.5 ppbv mainlyappeared in central, southwestern (SiChuan Basin)and eastern China (Yangtze River Delta). The differ-ence in OH was also centered in this region andreached approximately 200% after updating thescheme. The O3 differences in some of areas of thisregion reached 10 ppbv. In summer, the differences inHONO, O3, and OH in the northern plain and easterncoast of China were relatively larger than those inother places. The simulated spatial difference in July inthis study was similar to that in August as reported byLiu et al (2014). The differences in the spatial patternsof these three species between April (spring) andOcto-ber (autumn) were similar. In the northern centralplain of China, the differences for O3 andHONOwerelarger during October than during April. When anadequateHONO soil emissionmodule is included, thesimulation differences of HONO, O3, OH are expec-ted to become larger.

The simulation over China yielded a much largerdifference in winter than in summer, which can beattributed largely to two factors. (1) The aerosol con-centration over the northern plain of China is higherin winter than in summer because of weaker, stablemeteorology conditions (Zou et al 2017). (2) As notedby Liu et al (2014), the OH source generated by O(1D)+H2O is much smaller in winter than in sum-mer because of the low RH in winter. Accordingly,

Table 4.CalculatedNO2 reactiveuptake coefficients (γ) for the sevencases inChina.

γ γ/SWR

1 7.6×10–4 2.2×10–6

2 1.9×10–4 8.8×10–7

3 6.2×10–4 1.8×10–6

4 1.0×10–4 4.2×10–7

5 7.9×10–4 1.5×10–6

6 2.3×10–4 8.2×10–7

7 2.1×10–4 4.5×10−7

8

Environ. Res. Lett. 13 (2018) 114002

heterogeneous HONO formation from aerosol plays amore important role with respect to the radical budgetand the O3 ambient level during winter. Based on thisresult, the O3 concentration would decrease if the par-ticulate matter level decreases, especially during win-ter. However, the O3 production rate would increaseas the aerosol optical depth decreases, in contrast tothe effect of aerosol on the formation of HONO (Luet al 2017). The HO2 uptake by aerosol has not yetbeen included in the CMAQmodel. If this mechanismis active, however, the HO2 concentration wouldincrease as the aerosol concentration decreases, whichwould promote O3 formation (Liu et al 2012a). Hence,in the future, it is important to investigate the com-bined effects of aerosol onO3 formation inChina.

Two important factors are needed in future studiesto improve the 3D CTM simulations of the ambientHONO concentrations in China: (1) the effects ofaerosol components on the coefficient of NO2 hetero-geneous uptake and (2) the important factors deter-mining HONO soil emissions. Liu et al (2012b)investigated satellite products and reported a relativelyhigh concentration of glyoxal (discarboxylic acid pre-cursor) in central China. Colussi et al (2013) reportedthat the carboxylate anions can work as a catalyst topromote the disproportionation of NO2 on water andHONO formation. This may have driven the hetero-geneous HONO generation in the northern centralplain of China. Yabushita et al (2009) also reportedthat the NO2 uptake coefficient can be enhanced by

Figure 3.Monthly average changes inHONO (ppbv), O3(ppbv) andOH (%) levels during 9:00 am–17:00 pm in January, April, JulyandOctober 2015 (1–4 represent the cases in different seasons and (a)–(c) represent the changes ofHONO,O3 andOH).

9

Environ. Res. Lett. 13 (2018) 114002

several order of magnitude by the disproportionationof gas phase NO2 on the air/water interface. Moreexperimental studies are needed to quantify the effectof the aerosol component on the production ofHONO. Although Oswald et al (2013) reported theHONO emissions of different soil types, large emis-sion differences were observed even for single soiltype, and the factors that determine this process willrequire further identification in the future.

Several recent studies found that current mechan-isms in 3D CTMs could not explain the observed sul-fate production rates during winter in north andcentral China (Wang et al 2014, Zheng et al 2015). Thefollowing reactions could oxidize sulfur(IV) and pro-duce particulate sulfate (R2–R4 are aqueous phasereactions):

+ ¾ ¾ ++

( )SO OH H SO HO , R12H O O

2 4 22 2

+ +⟶ ( )SO O H SO O , R22 3H O

2 4 22

+ + +- - + ( )HSO O SO H O , R33 3 42 2+ + +- - + ( )HSO H O SO H H O. R43 2 2 42 2

As shown in figure 3, the OH and O3 concentrationsincreased substantially after the heterogeneousHONOproduction in Januarywas added.H2O2wouldincrease as a result of increases in OH and HO2.According to R1–R4, increases in OH and O3 couldpromote the formation of -SO ,4 2 which might partlyexplain themissing sulfate formation in the 3DCTMs.Spatially enhanced sulfate production in January isdemonstrated in figure S2 in the supplemental mat-erial. Sulfate production increased by approximately6–8 μg m−3 in the northern central plain and by10 μg m−3 or more in the Sichuan Basin. Czader et al(2013) noted that the CMAQ underestimated thedaytime OH and HO2 concentrations of a polluted airmass. In other words, the missing daytime OH couldexplain part of the missing particulate sulfate in the3DCTMs.

3.4. ConclusionIn this study, a consistent 1D framework based onregional and 3DCTMswas used to analyze the unknowndaytime HONO sources in 14 cases worldwide. Weassume that the source of HONO from aerosols isthrough NO2 hydrolysis (not including its oxidationproducts) and that non-local mixing effect is negligible.We found that the daytime unknown HONO sourcecorrelated stronglywith the aerosolwet surface area. If allunknown HONO were derived from NO2 groundheterogeneous formation, the NO2-to-HONO conver-sion fraction would exceed 100% inmost cases, which isunphysical. After adding the optimized soil emission,heterogeneous HONO formation from aerosol stillaccounted for more than 60% in the two urban heavypollution cases. Furthermore, low soil temperatures inwinter could prohibit the emission of HONO from soils.Hence, our results indicate that aerosol should play animportant role in daytime HONO heterogeneous

formation, especially on days affected by heavy pollutionand winter temperatures. This result contrasts with thegeneral assumption that aerosols are insignificant withregard to daytime HONO formation. The NO2 uptakecoefficient was estimated for seven cases in China,leading us to propose that variations in the uptakecoefficients depend on the aerosol composition. Accord-ing to the CMAQ simulation results, the aerosolgenerated HONO can contribute up to 10 ppbv of O3 incentral China. Further investigations of the effects ofdifferent aerosol components on HONO production inChina arewarranted, because such efforts could improveO3 andPM2.5 simulations basedon3DCTMs.

Acknowledgments

We thank the CalNex 2010 campaign for making thedata public. We also thank the constructive commentsfrom the anonymous reviewers. We appreciate thehelp from Prof Alexis Lau (HKUST) for the CMAQmodel consultation. This work was supported by theResearch Grants Council of Hong Kong Government(Project No. T24/504/17). Xingcheng Lu was sup-ported by the Overseas Research Award from theHong Kong University of Science and Technology.Yuhang Wang was supported by the AtmosphericChemistry Program of the US National ScienceFoundation.

ORCID iDs

Xingcheng Lu https://orcid.org/0000-0002-0962-9855Lu Shen https://orcid.org/0000-0003-2787-7016

References

Acker K, FeboA, Trick S, PerrinoC, Bruno P,Wiesen P,Möller D,WieprechtW,Auel R andGiustoM2006Nitrous acid in theurban area of RomeAtmos. Environ. 40 3123–33

Avnery S,Mauzerall D L, Liu J andHorowitz LW2011Global cropyield reductions due to surface ozone exposure: I. Year 2000crop production losses and economic damageAtmos.Environ. 45 2284–96

Cai R, YangD, FuY,WangX, Li X,MaY,Hao J, Zheng J and Jiang J2017Aerosol surface area concentration: a governing factorin newparticle formation in BeijingAtmos. Chem. Phys. 1712327–40

Colussi A J, Enami S, Yabushita A,HoffmannMR, LiuW-G,MishraH andGoddardWA III 2013Tropospheric aerosol asa reactive intermediate FaradayDiscuss. 165 407–20

Czader BH, Li X andRappenglueck B 2013CMAQmodeling andanalysis of radicals, radical precursors, and chemicaltransformations J. Geophys. Res.: Atmos. 118 11376–87

Dobbie K and SmithK 2001The effects of temperature, water‐filledpore space and land use onN2O emissions from animperfectly drained gleysol Eur. J. Soil Sci. 52 667–73

Gall E T,GriffinR J, Steiner A L,Dibb J, Scheuer E, Gong L,Rutter A P, Cevik BK,KimS and Lefer B 2016 Evaluation ofnitrous acid sources and sinks in urban outflowAtmos.Environ. 127 272–82

HanC, YangW,WuQ, YangH andXueX 2016Heterogeneousphotochemical conversion ofNO2 toHONOon the humic

10

Environ. Res. Lett. 13 (2018) 114002

https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0002-0962-9855https://orcid.org/0000-0003-2787-7016https://orcid.org/0000-0003-2787-7016https://orcid.org/0000-0003-2787-7016https://orcid.org/0000-0003-2787-7016https://doi.org/10.1016/j.atmosenv.2006.01.028https://doi.org/10.1016/j.atmosenv.2006.01.028https://doi.org/10.1016/j.atmosenv.2006.01.028https://doi.org/10.1016/j.atmosenv.2010.11.045https://doi.org/10.1016/j.atmosenv.2010.11.045https://doi.org/10.1016/j.atmosenv.2010.11.045https://doi.org/10.1039/c3fd00040khttps://doi.org/10.1039/c3fd00040khttps://doi.org/10.1039/c3fd00040khttps://doi.org/10.1002/jgrd.50807https://doi.org/10.1002/jgrd.50807https://doi.org/10.1002/jgrd.50807https://doi.org/10.1046/j.1365-2389.2001.00395.xhttps://doi.org/10.1046/j.1365-2389.2001.00395.xhttps://doi.org/10.1046/j.1365-2389.2001.00395.xhttps://doi.org/10.1016/j.atmosenv.2015.12.044https://doi.org/10.1016/j.atmosenv.2015.12.044https://doi.org/10.1016/j.atmosenv.2015.12.044

acid surface under simulated sunlight Environ. Sci. Technol.50 5017–23

HanC, YangW, YangH andXueX 2017 Enhanced photochemicalconversion ofNO2 toHONOonhumic acids in the presenceof benzophenone Environ. Pollut. 231 979–86

HoK,Ho S, Lee S, KawamuraK, Zou S, Cao J andXuH2011Summer andwinter variations of dicarboxylic acids, fattyacids and benzoic acid in PM2.5 in Pearl delta river region,ChinaAtmos. Chem. Phys. 11 2197–208

HoK, Lee S,Ho S SH,KawamuraK, Tachibana E, Cheng Y andZhuT2010Dicarboxylic acids, ketocarboxylic acids,α‐dicarbonyls, fatty acids, and benzoic acid in urban aerosolscollected during the 2006Campaign of air quality research inBeijing (CAREBeijing‐2006) J. Geophys. Res.: Atmos. 11519312

Hou S, Tong S, GeMandAn J 2016Comparison of atmosphericnitrous acid during severe haze and clean periods in Beijing,ChinaAtmos. Environ. 124 199–206

Jaeglé L, JacobD J, BruneWHandWennberg PO2011ChemistryofHOx radicals in the upper troposphereAtmos. Environ. 35469–89

KawamuraK,OkuzawaK, Aggarwal S, IrieH, Kanaya Y andWang Z2013Determination of gaseous and particulate carbonyls(glycolaldehyde, hydroxyacetone, glyoxal,methylglyoxal,nonanal and decanal) in the atmosphere atMt. TaiAtmos.Chem. Phys. 13 5369–80

KebedeMA2015Nitrous acid formation on environmentalsurfaces containing titanium and ironPhDThesis IndianaUniversity

Lee J,Whalley L,HeardD, StoneD,Dunmore R,Hamilton J,YoungD, Allan J, Laufs S andKleffmann J 2016Detailedbudget analysis ofHONO in central London reveals amissingdaytime sourceAtmos. Chem. Phys. 16 2747–64

LiG, LeiW, ZavalaM,Volkamer R,Dusanter S, Stevens P andMolina L 2010 Impacts ofHONO sources on thephotochemistry inMexicoCity during theMCMA-2006/MILAGOCampaignAtmos. Chem. Phys. 10 6551–67

LiM, ZhangQ,Kurokawa J-I,Woo J-H,HeK, LuZ,Ohara T,Song Y, StreetsDG andCarmichael GR 2017MIX: amosaicAsian anthropogenic emission inventory under theinternational collaboration framework of theMICS-Asia andHTAPAtmos. Chem. Phys. 17 935–63

Lin J-T andMcElroyMB2010 Impacts of boundary layermixingonpollutant vertical profiles in the lower troposphere: implicationsto satellite remote sensingAtmos. Environ.44 1726–39

LiuX et al 2017High levels of daytimemolecular chlorine and nitrylchloride at a rural site on theNorthChina plain Environ. Sci.Technol. 51 9588–95

Liu Z et al 2012a Summertime photochemistry duringCAREBeijing-2007: ROx budgets andO3 formationAtmos.Chem. Phys. 12 7737–52

Liu Z,Wang Y,Costabile F, AmorosoA, ZhaoC,Huey LG,Stickel R, Liao J andZhuT 2014 Evidence of aerosols as amedia for rapid daytimeHONOproduction over ChinaEnviron. Sci. Technol. 48 14386–91

Liu Z,Wang Y,VrekoussisM, Richter A,Wittrock F, Burrows J P,ShaoM,ChangCC, Liu SC andWangH2012b Exploring themissing source of glyoxal (CHOCHO) over ChinaGeophys.Res. Lett. 39 L10812

LuK, Rohrer F,Holland F, FuchsH, BohnB, Brauers T, ChangC,Häseler R,HuMandKita K 2012Observation andmodellingofOHandHO2 concentrations in the Pearl river delta 2006: amissingOH source in aVOC rich atmosphereAtmos. Chem.Phys. 12 1541–69

LuX, ChenN,Wang Y,CaoW, ZhuB, YaoT, Fung J CHandLauAKH2017Radical budget and ozone chemistry duringautumn in the atmosphere of anUrban site inCentral ChinaJ. Geophys. Res.: Atmos. 122 3672–85

LuX, Fung J CHandWuD2015Modelingwet deposition of acidsubstances over the PRD region inChinaAtmos. Environ. 122819–28

LuX, YaoT, Fung J CHand LinC 2016 Estimation of health andeconomic costs of air pollution over the Pearl river deltaregion inChina Sci. Total Environ. 566 134–43

Mao J, Paulot F, JacobD J, CohenRC, Crounse JD,Wennberg PO,Keller CA,HudmanRC, BarkleyMP andHorowitz LW2013Ozone and organic nitrates over the easternUnitedStates: sensitivity to isoprene chemistry J. Geophys. Res.:Atmos. 118 11256–68

MichoudV, ColombA, BorbonA,Miet K, BeekmannM,CamredonM,Aumont B, Perrier S, Zapf P and SiourG 2014Study of the unknownHONOdaytime source at a Europeansuburban site during theMEGAPOLI summer andwinterfield campaignsAtmos. Chem. Phys. 14 2805–22

MongeME,D’AnnaB,Mazri L, Giroir-Fendler A, AmmannM,DonaldsonD andGeorgeC 2010 Light changes theatmospheric reactivity of sootProc. Natl Acad.Sci. USA 1076605–9

NieW,Ding A, Xie Y, XuZ,MaoH,KerminenV-M, Zheng L,Qi X,HuangX andYangX-Q 2015 Influence of biomass burningplumes onHONOchemistry in easternChinaAtmos. Chem.Phys. 15 1147–59

OkudaT, Kato J,Mori J, TenmokuM, Suda Y, Tanaka S,HeK,MaY, Yang F andYuX 2004Daily concentrations of tracemetals in aerosols in Beijing, China, determined by usinginductively coupled plasmamass spectrometry equippedwith laser ablation analysis, and source identification ofaerosols Sci. Total Environ. 330 145–58

Oswald R, Behrendt T, ErmelM,WuD, SuH,Cheng Y,Breuninger C,MoravekA,Mougin E andDelonC 2013HONOemissions from soil bacteria as amajor source ofatmospheric reactive nitrogen Science 341 1233–5

PartonW,Holland E,Del Grosso S,HartmanM,Martin R,Mosier A,OjimaD and SchimelD 2001GeneralizedmodelforNOx andN2O emissions from soils J. Geophys. Res.:Atmos. 106 17403–19

Pleim J E andByunDW2004Application of a new land-surface, drydeposition, and PBLmodel in themodels-3 communitymulti-scale air quality (CMAQ)model systemAir PollutionModeling and Its Application XIV (Berlin: Springer)pp 297–305 ( https://doi.org/10.1007/0-306-47460-3_30)

RomaniasMN, Bedjanian Y, Zaras AM,Andrade-Eiroa A,Shahla R,Dagaut P and Philippidis A 2013Mineral oxideschange the atmospheric reactivity of soot: NO2 uptake underdark andUV irradiation conditions J. Phys. Chem.A 11712897–911

Schindlbacher A, Zechmeister‐Boltenstern S andButterbach‐Bahl K2004 Effects of soilmoisture and temperature onNO,NO2,andN2O emissions fromEuropean forest soils J. Geophys.Res.: Atmos. 109D17302

Stemmler K,NdourM, Elshorbany Y, Kleffmann J, D’annaB,GeorgeC, BohnB andAmmannM2007 Light inducedconversion of nitrogen dioxide into nitrous acid onsubmicron humic acid aerosolAtmos. Chem. Phys. 7 4237–48

SuH,Cheng Y,Oswald R, Behrendt T, Trebs I,Meixner F X,AndreaeMO,Cheng P, Zhang Y and PöschlU 2011 Soilnitrite as a source of atmosphericHONOandOHradicalsScience 333 1616–8

SuH,Cheng Y F, ShaoM,GaoDF, YuZY, Zeng LM, Slanina J,Zhang YH andWiedensohler A 2008Nitrous acid (HONO)and its daytime sources at a rural site during the 2004 PRIDE‐PRDexperiment inChina J. Geophys. Res.: Atmos. 113D14312

SuarezM J, RieneckerM, Todling R, Bacmeister J, Takacs L, LiuH,GuW, SienkiewiczM,Koster R andGelaro R 2008TheGEOS-5Data Assimilation System-Documentation ofVersions 5.0. 1, 5.1. 0, and 5.2. 0 (https://ntrs.nasa.gov/search.jsp?R=20120011955)

SörgelM, Trebs I,WuD andHeldA 2015A comparison ofmeasuredHONOuptake and releasewith calculated sourcestrengths in a heterogeneous forest environmentAtmos.Chem. Phys. 15 9237–51

11

Environ. Res. Lett. 13 (2018) 114002

https://doi.org/10.1021/acs.est.5b05101https://doi.org/10.1021/acs.est.5b05101https://doi.org/10.1021/acs.est.5b05101https://doi.org/10.1016/j.envpol.2017.08.107https://doi.org/10.1016/j.envpol.2017.08.107https://doi.org/10.1016/j.envpol.2017.08.107https://doi.org/10.5194/acp-11-2197-2011https://doi.org/10.5194/acp-11-2197-2011https://doi.org/10.5194/acp-11-2197-2011https://doi.org/10.1029/2009JD013304https://doi.org/10.1029/2009JD013304https://doi.org/10.1016/j.atmosenv.2015.06.023https://doi.org/10.1016/j.atmosenv.2015.06.023https://doi.org/10.1016/j.atmosenv.2015.06.023https://doi.org/10.1016/S1352-2310(00)00376-9https://doi.org/10.1016/S1352-2310(00)00376-9https://doi.org/10.1016/S1352-2310(00)00376-9https://doi.org/10.1016/S1352-2310(00)00376-9https://doi.org/10.5194/acp-13-5369-2013https://doi.org/10.5194/acp-13-5369-2013https://doi.org/10.5194/acp-13-5369-2013https://doi.org/10.5194/acp-16-2747-2016https://doi.org/10.5194/acp-16-2747-2016https://doi.org/10.5194/acp-16-2747-2016https://doi.org/10.5194/acp-10-6551-2010https://doi.org/10.5194/acp-10-6551-2010https://doi.org/10.5194/acp-10-6551-2010https://doi.org/10.5194/acp-17-935-2017https://doi.org/10.5194/acp-17-935-2017https://doi.org/10.5194/acp-17-935-2017https://doi.org/10.1016/j.atmosenv.2010.02.009https://doi.org/10.1016/j.atmosenv.2010.02.009https://doi.org/10.1016/j.atmosenv.2010.02.009https://doi.org/10.1021/acs.est.7b03039https://doi.org/10.1021/acs.est.7b03039https://doi.org/10.1021/acs.est.7b03039https://doi.org/10.5194/acp-12-7737-2012https://doi.org/10.5194/acp-12-7737-2012https://doi.org/10.5194/acp-12-7737-2012https://doi.org/10.1021/es504163zhttps://doi.org/10.1021/es504163zhttps://doi.org/10.1021/es504163zhttps://doi.org/10.1029/2012GL051645https://doi.org/10.5194/acp-12-1541-2012https://doi.org/10.5194/acp-12-1541-2012https://doi.org/10.5194/acp-12-1541-2012https://doi.org/10.1002/2016JD025676https://doi.org/10.1002/2016JD025676https://doi.org/10.1002/2016JD025676https://doi.org/10.1016/j.atmosenv.2015.09.035https://doi.org/10.1016/j.atmosenv.2015.09.035https://doi.org/10.1016/j.atmosenv.2015.09.035https://doi.org/10.1016/j.atmosenv.2015.09.035https://doi.org/10.1016/j.scitotenv.2016.05.060https://doi.org/10.1016/j.scitotenv.2016.05.060https://doi.org/10.1016/j.scitotenv.2016.05.060https://doi.org/10.1002/jgrd.50817https://doi.org/10.1002/jgrd.50817https://doi.org/10.1002/jgrd.50817https://doi.org/10.5194/acp-14-2805-2014https://doi.org/10.5194/acp-14-2805-2014https://doi.org/10.5194/acp-14-2805-2014https://doi.org/10.1073/pnas.0908341107https://doi.org/10.1073/pnas.0908341107https://doi.org/10.1073/pnas.0908341107https://doi.org/10.1073/pnas.0908341107https://doi.org/10.5194/acp-15-1147-2015https://doi.org/10.5194/acp-15-1147-2015https://doi.org/10.5194/acp-15-1147-2015https://doi.org/10.1016/j.scitotenv.2004.04.010https://doi.org/10.1016/j.scitotenv.2004.04.010https://doi.org/10.1016/j.scitotenv.2004.04.010https://doi.org/10.1126/science.1242266https://doi.org/10.1126/science.1242266https://doi.org/10.1126/science.1242266https://doi.org/10.1029/2001JD900101https://doi.org/10.1029/2001JD900101https://doi.org/10.1029/2001JD900101http:// https://doi.org/10.1007/0-306-47460-3_30https://doi.org/10.1021/jp407914fhttps://doi.org/10.1021/jp407914fhttps://doi.org/10.1021/jp407914fhttps://doi.org/10.1021/jp407914fhttps://doi.org/10.1029/2004JD004590https://doi.org/10.5194/acp-7-4237-2007https://doi.org/10.5194/acp-7-4237-2007https://doi.org/10.5194/acp-7-4237-2007https://doi.org/10.1126/science.1207687https://doi.org/10.1126/science.1207687https://doi.org/10.1126/science.1207687https://doi.org/10.1029/2007JD009060https://doi.org/10.1029/2007JD009060http://arXiv.org/abs/https://ntrs.nasa.gov/search.jsp?R=20120011955http://arXiv.org/abs/https://ntrs.nasa.gov/search.jsp?R=20120011955https://doi.org/10.5194/acp-15-9237-2015https://doi.org/10.5194/acp-15-9237-2015https://doi.org/10.5194/acp-15-9237-2015

TanZ et al 2017Radical chemistry at a rural site (Wangdu) in theNorthChina plain: observation andmodel calculations ofOH,HO 2 andRO 2 radicalsAtmos. Chem. Phys. 17 663–90

TongS,HouS,ZhangY,ChuB,LiuY,HeH,ZhaoPandGeM2016Exploring thenitrous acid (HONO) formationmechanism inwinterBeijing: direct emissions andheterogeneousproduction inurbanandsuburbanareasFaradayDiscuss.189213–30

VandenBoer T,MarkovicM, Sanders J, RenX, Pusede S, Browne E,CohenR, Zhang L, Thomas J and BruneW2014 Evidence fora nitrous acid (HONO) reservoir at the ground surface inBakersfield, CA, during CalNex 2010 J. Geophys. Res.: Atmos.119 9093–106

VandenBoer TC, Brown S S,Murphy JG, KeeneWC, YoungC J,PszennyA, KimS,WarnekeC, Gouw JA andMaben J R 2013Understanding the role of the ground surface inHONOvertical structure: high resolution vertical profiles duringNACHTT‐11 J. Geophys. Res.: Atmos. 118 10155–71

VandenBoer TC, YoungC J, Talukdar RK,MarkovicMZ,Brown S S, Roberts JM andMurphy JG 2015Nocturnal lossand daytime source of nitrous acid through reactive uptakeand displacementNat. Geosci. 8 55–60

Wang J,WangS, Jiang J,DingA,ZhengM,ZhaoB,WongDC,ZhouW,ZhengGandWangL2014 Impactof aerosol–meteorologyinteractionsonfineparticlepollutionduringChina’s severehazeepisode in January2013Environ.Res. Lett.9094002

WeselyM1989 Parameterization of surface resistances to gaseousdry deposition in regional-scale numericalmodelsAtmos.Environ. 23 1293–304

WongK,TsaiC,LeferB,HamanC,GrossbergN,BruneW,RenX,LukeWandStutz J 2012DaytimeHONOvertical gradientsduringSHARP2009 inHouston,TXAtmos.Chem.Phys.12635–52

XuZ,Wang T, Xue L, Louie P, LukC,Gao J,Wang S, Chai F andWangW2013 Evaluating the uncertainties of thermalcatalytic conversion inmeasuring atmospheric nitrogendioxide at four differently polluted sites inChinaAtmos.Environ. 76 221–6

Yabushita A, Enami S, Sakamoto Y, KawasakiM,HoffmannMRandColussi A J 2009Anion-catalyzeddissolution ofNO2 on aqueousmicrodroplets J. Phys. Chem.A 113 4844–8

YeC et al 2016Rapid cycling of reactive nitrogen in themarineboundary layerNature 532 489–91

YeC, ZhangN,GaoHandZhouX 2017 Photolysis of particulatenitrate as a source ofHONOandNOxEnviron. Sci. Technol.51 6849–56

YueD, Zhong L, Shen J, Zhang T, Xie Z, Zeng L andDongH2015Effect of atmosphericHNO2 andO3 onOHradical formationduring daytime in the Pearl river delta regionChina Sci. Paper10 1387–91 (in Chinese)

ZhangR, SarwarG, Fung JC, LauAK andZhang Y 2012 Examiningthe impact of nitrous acid chemistry on ozone and PMoverthe Pearl RiverDelta RegionAdv.Meteorol. 2012 140932

ZhangR,Wang Y, Smeltzer C,HangQ, KoshakW, BoersmaKF andCelarier E 2018Reconciling the differences betweenOMI-based and EPAAQS in situNO2 trendsAtmos.Meas. Tech. 113955–67

Zhang Y et al 2016 Large vertical gradient of reactive nitrogen oxidesin the boundary layer:modeling analysis ofDISCOVER‐AQ2011 observations J. Geophys. Res.: Atmos. 121 1922–34

ZhaoH,Wang S,WangW, Liu R andZhouB 2015 Investigation ofground-level ozone and high-pollution episodes in amegacity of EasternChina PLoSOne 10 e0131878

Zheng B, ZhangQ, Zhang Y,HeK,WangK, ZhengG,Duan F,MaY andKimoto T 2015Heterogeneous chemistry: amechanismmissing in currentmodels to explain secondaryinorganic aerosol formation during the January 2013 hazeepisode inNorthChinaAtmos. Chem. Phys. 15 2031–49

ZhouX, ZhangN, TerAvestM, TangD,Hou J, Bertman S,AlaghmandM, Shepson PB, CarrollMA andGriffith S 2011Nitric acid photolysis on forest canopy surface as a source fortropospheric nitrous acidNat. Geosci. 4 440–3

ZouY,WangY, Zhang Y andKoo J-H 2017Arctic sea ice, Eurasiasnow, and extremewinter haze inChina Sci. Adv. 3 e1602751

12

Environ. Res. Lett. 13 (2018) 114002

https://doi.org/10.5194/acp-17-663-2017https://doi.org/10.5194/acp-17-663-2017https://doi.org/10.5194/acp-17-663-2017https://doi.org/10.1039/C5FD00163Chttps://doi.org/10.1039/C5FD00163Chttps://doi.org/10.1039/C5FD00163Chttps://doi.org/10.1002/2013JD020971https://doi.org/10.1002/2013JD020971https://doi.org/10.1002/2013JD020971https://doi.org/10.1002/jgrd.50721https://doi.org/10.1002/jgrd.50721https://doi.org/10.1002/jgrd.50721https://doi.org/10.1038/ngeo2298https://doi.org/10.1038/ngeo2298https://doi.org/10.1038/ngeo2298https://doi.org/10.1088/1748-9326/9/9/094002https://doi.org/10.1016/0004-6981(89)90153-4https://doi.org/10.1016/0004-6981(89)90153-4https://doi.org/10.1016/0004-6981(89)90153-4https://doi.org/10.5194/acp-12-635-2012https://doi.org/10.5194/acp-12-635-2012https://doi.org/10.5194/acp-12-635-2012https://doi.org/10.5194/acp-12-635-2012https://doi.org/10.1021/jp900685fhttps://doi.org/10.1021/jp900685fhttps://doi.org/10.1021/jp900685fhttps://doi.org/10.1038/nature17195https://doi.org/10.1038/nature17195https://doi.org/10.1038/nature17195https://doi.org/10.1021/acs.est.7b00387https://doi.org/10.1021/acs.est.7b00387https://doi.org/10.1021/acs.est.7b00387https://doi.org/10.1155/2012/140932https://doi.org/10.5194/amt-11-3955-2018https://doi.org/10.5194/amt-11-3955-2018https://doi.org/10.5194/amt-11-3955-2018https://doi.org/10.5194/amt-11-3955-2018https://doi.org/10.1002/2015JD024203https://doi.org/10.1002/2015JD024203https://doi.org/10.1002/2015JD024203https://doi.org/10.1371/journal.pone.0131878https://doi.org/10.5194/acp-15-2031-2015https://doi.org/10.5194/acp-15-2031-2015https://doi.org/10.5194/acp-15-2031-2015https://doi.org/10.1038/ngeo1164https://doi.org/10.1038/ngeo1164https://doi.org/10.1038/ngeo1164https://doi.org/10.1126/sciadv.1602751

1. Introduction2. Data and methods2.1. Observational and model data2.2. Model configuration2.3. Unknown HONO source estimation

3. Results and discussion3.1. Heterogeneous HONO formation from aerosol3.2. NO2 reactive uptake coefficient3.3. Effect on O3 spatial distribution3.4. Conclusion

AcknowledgmentsReferences