Seasonal characterization of submicron aerosol chemical composition and organic aerosol ...

19 Pages·2016·3.54 MB·English

Atmos.Chem.Phys.,16,5171–5189,2016
www.atmos-chem-phys.net/16/5171/2016/
doi:10.5194/acp-16-5171-2016
©Author(s)2016.CCAttribution3.0License.
Seasonal characterization of submicron aerosol chemical
composition and organic aerosol sources in the
southeastern United States: Atlanta, Georgia,
and Look Rock, Tennessee
SriHapsariBudisulistiorini1,KarstenBaumann2,EricS.Edgerton2,SolomonT.Bairai3,StephenMueller4,
StephanieL.Shaw5,EladioM.Knipping6,AvramGold1,andJasonD.Surratt1
1DepartmentofEnvironmentalSciencesandEngineering,GillingsSchoolofGlobalPublicHealth,
TheUniversityofNorthCarolinaatChapelHill,ChapelHill,NC,USA
2AtmosphericResearch&Analysis,Inc.,Cary,NC,USA
3Battelle,Pueblo,CO,USA
4Ensafe,Nashville,TN,USA
5ElectricPowerResearchInstitute,PaloAlto,CA,USA
6ElectricPowerResearchInstitute,Washington,DC,USA
Correspondenceto:J.D.Surratt([email protected])
Received:29July2015–PublishedinAtmos.Chem.Phys.Discuss.:20August2015
Revised:13April2016–Accepted:17April2016–Published:26April2016
Abstract. A year-long near-real-time characterization of observedthroughouttheyearatbothsitesandcontributedup
non-refractorysubmicronaerosol(NR-PM )wasconducted to66%oftotalOAmass.HOAwasobservedduringtheen-
1
atanurban(Atlanta,Georgia,in2012)andrural(LookRock, tireyearonlyattheurbansite(onaverage21%ofOAmass).
Tennessee, in 2013) site in the southeastern US using the BBOA (15–33% of OA mass) was observed during winter
Aerodyne Aerosol Chemical Speciation Monitor (ACSM) andfall,likelydominatedbylocalresidentialwoodburning
collocatedwithestablishedair-monitoringnetworkmeasure- emission. Although SV-OOA contributes quite significantly
ments. Seasonal variations in organic aerosol (OA) and in- (∼27%),itwasobservedonlyattheurbansiteduringcolder
organicaerosolspeciesareattributedtometeorologicalcon- seasons. IEPOX-OA was a major component (27–41%) of
ditions as well as anthropogenic and biogenic emissions in OAatbothsites,particularlyinspringandsummer.Anion
thisregion.ThehighestconcentrationsofNR-PM wereob- fragmentatm/z75iswellcorrelatedwiththem/z82ionas-
1
served during winter and fall seasons at the urban site and sociated with the aerosol mass spectrum of IEPOX-derived
duringspringandsummerattheruralsite.Acrossallseasons secondaryorganicaerosol(SOA).Thecontributionof91Fac
andatbothsites,NR-PM wascomposedlargelyofOA(up tothetotalOAmasswassignificant(onaverage22%ofOA
1
to 76%) and sulfate (up to 31%). Six distinct OA sources mass)attheruralsiteonlyduringwarmermonths.Compar-
were resolved by positive matrix factorization applied to ison of 91Fac OA time series with SOA tracers measured
the ACSM organic mass spectral data collected from the fromfiltersamplescollectedatLookRocksuggeststhatiso-
two sites over the 1 year of near-continuous measurements prene oxidation through a pathway other than IEPOX SOA
at each site: hydrocarbon-like OA (HOA), biomass burn- chemistry may contribute to its formation. Other biogenic
ing OA (BBOA), semi-volatile oxygenated OA (SV-OOA), sources could also contribute to 91Fac, but there remains a
low-volatility oxygenated OA (LV-OOA), isoprene-derived need to resolve the exact source of this factor based on its
epoxydiols (IEPOX) OA (IEPOX-OA) and 91Fac (a factor significantcontributiontoruralOAmass.
dominated by a distinct ion at m/z91 fragment ion previ-
ously observed in biogenic influenced areas). LV-OOA was
PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion.
5172 S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition
1 Introduction Table1.SeasonalclassificationofmeasurementsatJSTandLRK
isbasedondirectionofangleoftheEarthtothesunandtheangle
Characterizationofthechemicalcompositionofatmospheric ofthesunlightasithitstheEarth.
fineaerosolisimportant,becauseofitsadversehumanhealth
effects (Pope III and Dockery, 2006) and possible impacts JST LRK
ontheEarth’sclimatesystem(Forsteretal.,2007).Aerosol Winter 22/12/2011–19/03/2012 18/01/2013–19/03/2013
with aerodynamic diameters ≤1µm (PM1) plays a signif- Spring 20/03/2012–19/06/2012 20/03/2013–31/05/2013
icant role in scattering and/or absorbing solar radiation as Summer 20/06/2012–21/09/2012 01/06/2013–21/09/2013∗
well as cloud formation (IPCC, 2013). Long-term regional Fall 22/09/2012–20/12/2012 22/09/2013–20/12/2013
characterizationsofambientPM arerequiredtounderstand ∗MeasurementsinsummerattheLRKsiteincludedinthe2013SOAScampaignfrom
1
1Juneto17July2013.
its sources, formation and aging mechanisms, as well as at-
mosphericlifetime.Thisinformationwillleadtomoreaccu-
rately constrained air quality models for making regulatory
decisionstomitigatethepotentialadverseimpactsofPM .
1
Over the past decade, online aerosol mass spectrometry but was limited by 1-month measurement periods at each
(AMS) has been used to extensively characterize ambient site(Xuetal.,2015a).Non-fossilcarbonderivedfrommod-
non-refractory (NR)-PM (Zhang et al., 2007; Jimenez et ern sources (e.g., biogenic) is reported to account for 50%
1
al., 2009; Ng et al., 2010; Crippa et al., 2014); however, of carbon at two urban sites and 70–100% of carbon at
prior studies were limited by short measurement periods 10 near-urban or remote sites in the US (Schichtel et al.,
(weeks to a several months) because the need for intensive 2008).Additionally,isoprene-derivedSOAwasrecentlyob-
instrumentmaintenancerequiredthecontinuouson-sitepres- served to contribute substantially to SOA in downtown At-
ence of skilled personnel in order to generate high-quality lanta during summer (Budisulistiorini et al., 2013; Xu et
data. The Aerodyne Aerosol Chemical Speciation Monitor al., 2015a, b). The isoprene-derived SOA was attributed to
(ACSM) based on the AMS technology has been modified the heterogeneous chemistry of isomeric isoprene epoxydi-
to allow for long-term operation with less maintenance (Ng ols (IEPOX), known oxidation products of isoprene under
et al., 2011b). The ACSM has been recently used for long- both low- (Paulot et al., 2009) and high-NO (Jacobs et al.,
termNR-PM measurements(Petitetal.,2015;Ripolletal., 2014) conditions, in the presence of acidic sulfate aerosol
1
2015; Parworth et al., 201; Zhang et al., 2015) and shown (Budisulistiorinietal.,2013).
tobedurableanddataarecomparabletodatacollectedfrom Biogenic hydrocarbons and their oxidation products are
existingfineaerosolmonitoringnetworks(Budisulistioriniet major contributors to ambient fine aerosol in rural areas
al.,2014). whereanthropogenicsourcesarelow(Budisulistiorinietal.,
Worldwide studies have shown that tropospheric PM 2015).Insummer2001,thefractionofnon-fossilcarbonwas
1
mass is dominated by organic aerosol (OA; Zhang et al., reportedtovaryfrom66to80%oftotalcarbonatLookRock
2007; Jimenez et al., 2009). OA consists of aerosol di- (LRK), Great Smoky Mountains National Park (GSMNP),
rectly emitted into the atmosphere, primary organic aerosol Tennessee (TN), indicating the likely importance of photo-
(POA) and aerosol formed from atmospheric oxidation chemical oxidation of biogenic VOCs (BVOCs; Tanner et
of volatile organic compounds (VOCs), secondary organic al.,2004a).Sulfatedidnotshowsignificantdiurnalvariabil-
aerosol (SOA). POA sources include fossil fuel combus- ity at LRK, TN, suggesting that local meteorological con-
tionfromvehicles,powergeneration,andresidentialburning ditions are less influential in determining concentrations of
(cooking and heating) as well as forest fires (Kanakidou et long-lived species (Tanner et al., 2005). SOA is a predomi-
al., 2005). Contribution of hydrocarbon-like OA (HOA) as- nantcomponentofPM massduringsummerandearlyfall
2.5
sociatedwithPOAtourbanOAmassmaybesignificantdur- but POA is more dominant in the late fall (Ke et al., 2007),
ingmorningtraffic,whileoxygenatedOA(OOA)associated suggesting that the LRK site is influenced by biogenic and
withSOAexceedsPOAatmiddayorintheafternoon(Zhang anthropogenicemissions.
etal.,2005).SOAhasbeenobservedtocontributeupwards Wepresenta2-yearstudycomparingnear-real-timechem-
of 90% to the total OA mass (Docherty et al., 2008), indi- icalcharacterizationsofNR-PM collectedfor1-yearatthe
1
cating the critical role of photochemical processes in SOA urbanJeffersonStreet(JST)siteindowntownAtlanta,Geor-
formation. gia (GA), and a subsequent year at the rural LRK site lo-
StudiesinAtlanta,Georgia,havecharacterizedthechemi- catedintheGSMNP,TN.NR-PM wassampled,chemically
1
cal components of ambient aerosol collected during differ- characterized and quantified over a 2-year period, spanning
ent seasons (Lee et al., 2002; Kim et al., 2003; Butler et 2012–2013 using the ACSM. OA sources were seasonally
al., 2003); however, they were limited by low time or mass analyzedbypositivematrixfactorization(PMF).OAfactors
resolution. A recent study reported characterization of am- resolved by PMF were compared with collocated data col-
bient NR-PM by high-resolution time-of-flight AMS (HR- lected from both air-monitoring sites in order to associate
1
ToF-AMS)frommultiplesitesinGeorgia,includingAtlanta, themwithspecifictypesofOAsources.
Atmos.Chem.Phys.,16,5171–5189,2016 www.atmos-chem-phys.net/16/5171/2016/
S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition 5173
Figure1.AnnualtemporalvariationsofOAandinorganicspecies Figure2.AnnualtemporalvariationsofOAandinorganicspecies
(µgm−3)measuredattheJeffersonStreet(JST)site,Atlanta,Geor- (µgm−3) measured at the Look Rock (LRK) site, Great Smoky
◦
gia,in2012.Includedintheplotsareambienttemperature( C)and Mountains, Tennessee, in 2013. Included in the plots are ambi-
relative humidity (RH, in %) measured by SEARCH network, as ent temperature and RH (%) measured by Tennessee Valley Au-
wellaspHandliquidwatercontent(LWC,inmolL−1)estimated thority (TVA), as well as pH and LWC (molL−1) estimated by
byISORROPIA-II. ISORROPIA-II.
2 Methods
wind of urban areas, such as Knoxville and Maryville, TN,
2.1 Fineaerosolsamplinganddataanalysis andsmallfarmswithanimalgrazingareas.Coal-firedpower
plants Kingston and Bull Run are located within 50–60km
Real-time continuous chemical measurements were con- northwestofLRKsite(TennesseeValleyAuthority,2015).In
ducted during 2012 at a downtown urban site (JST) in At- summer, up-slope flow carries pollutants emitted in the val-
lanta,GA,andduring2013atarural/forestedsite(LRK)in ley during early morning to the LRK site by mid-morning,
GSMNP, TN, respectively. Analysis of data obtained from and in the evening down-slope flow accompanies a shift of
measurementsatJSTandLRKwasclassifiedbyseason(Ta- wind direction to the south and east that could isolate the
ble1),whichwasabletocapturechangesinmeteorology,in sitefromfreshprimaryemissionsfromthevalleyandallows
particular ambient temperature, at JST in 2012 and LRK in agedsecondaryspeciestoaccumulate(Tanneretal.,2005).
2013asillustratedinFigs.1and2.Theperiodwiththecold- Ambient NR-PM was analyzed using the Aerodyne
1
esttemperaturesisclassifiedasthewinterseason,andwhen ACSMinasimilarmanneratbothsites.DetailsofNR-PM
1
the temperature rises, the period is classified as the spring sampling at the JST and LRK sites have been described in
season.Summerseasonissignifiedbyconstanthightemper- Budisulistiorini et al. (2013, 2015). Briefly, the ACSM was
atureattheJSTandLRKsites.Whentemperaturedecreases operatedwithasamplingflowrateof3Lmin−1,resultingin
aftersummer,thisperiodiscategorizedasthefallseason. aresidencetimeof<2sforPM inthesamplingline.The
2.5
Organicandinorganicspeciescharacterizationsduringthe aerodynamiclensmountedontheACSMinletcontinuously
2013 Southern Oxidant Aerosol Study (SOAS; Budisulis- samplesPM fromthebypassPM samplingline(Ngetal.,
1 2.5
tiorini et al., 2015) were included in analysis of the sum- 2011b). Particle-laden air was dried using a 50-tube Nafion
mer season at the LRK site in this study. Detailed descrip- dryer(PermaPurePD-50T-24SS)inwhichadry-airsystem
tions of both sites have been published (Budisulistiorini et delivered7Lmin−1 ofdrysheathairtokeepthesampleair
al., 2013, 2015). Briefly, the JST site is one of several re- relativehumidity(RH)wellbelow10%,preventingconden-
searchsitesoftheSoutheasternAerosolResearchandChar- sationwithinthesamplinglinethatcouldadverselyaffectthe
acterization (SEARCH) network. The JST site is located in collectionefficiency(CE)ofPM andclogtheACSMsam-
1
a mixed industrial-residential area about 4.2km northwest plinginlet.TheACSMwastunedforionizerandelectronic
of downtown Atlanta and within approximately 200m of a offsetandcalibratedforionizationefficiencyon-site(fiveto
busmaintenanceyardandseveralwarehousefacilitiestothe seven times) throughout each year of sampling at each site.
south and southwest (Hansen et al., 2003; Solomon et al., Masscalculationofaerosolconstituentsisdescribedindetail
2003), and within 53km of a coal-fired power plant (Plant elsewhere(Ngetal.,2011b).Atbothsites,aCEvalueof0.5
Bowen; Edgerton et al., 2006). The LRK site is located on forallspecieswasusedbasedonevaluationofcomposition-
a ridge top on the northwestern edge of the GSMNP down- dependent CE as described in Budisulistiorini et al. (2013,
www.atmos-chem-phys.net/16/5171/2016/ Atmos.Chem.Phys.,16,5171–5189,2016
5174 S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition
2015). We estimated dry density of ambient PM based on 0.28,suggestingthattheerrorsareoverestimated(Ulbrichet
1
average particle composition for each season, and the as- al.,2009).However,theerrorvaluesaredeemedappropriate
sumption of organic, inorganic and elemental carbon (EC) sinceQ/Q isconsistentlylessthanunity,regardlessofthe
exp
densities are 1.4 (Hallquist et al., 2009), 1.77 (Turpin and numberoffactorsandthedatasets.
Lim,2001),and1.77gcm−3(Parketal.,2004),respectively.
TheestimateddryaerosoldensitiesatboththeJSTandLRK 2.3 EstimationofaerosolaciditybyISORROPIA
sites are 1.55gcm−3 on average (Table S1 in the Supple-
The thermodynamic model, ISORROPIA-II, in forward
ment), which is about 13% less than the density of 1.75 at
mode(FountoukisandNenes,2007;Nenesetal.,1999),was
JST(Budisulistiorinietal.,2014)andsimilartothedensity
of1.52gcm−3 atLRK(Budisulistiorinietal.,2015)during used to estimate aerosol pH. Inputs for the model include
aerosol-phase sulfate, nitrate, and ammonium as µmolm−3,
summer.IfaCEof1wasappliedtoJSTandLRKdatasets,
the estimated aerosol density is <1gcm−3, which is much measured by the ACSM under ambient conditions. In addi-
lower than the suggested organic of 1.4gcm−3 (Hallquist tion, RH and temperature obtained from the SEARCH net-
et al., 2009) and inorganic aerosol density of 1.77gcm−3 workandtheNationalParkService(NPS)forJSTandLRK
sites, respectively, were used as inputs. Inputs of gas-phase
(Crossetal.,2007).Therefore,weappliedaCEvalueof0.5
ammonia for the JST site were obtained from SEARCH
toallseasonaldatasets.
and for the LRK site, from the Ammonia Monitoring Net-
work(AMoN,TN01/GreatSmokyMountainsNationalPark
2.2 OrganicaerosolcharacterizationbyPMF
– Look Rock). ISORROPIA-II predicted particle hydro-
niumionconcentrationpervolumeofair(H+,µgm−3)and
Details of PMF analysis of the organic mass fraction have
aerosol liquid water content (LWC, molL−1). Calculation
been describedpreviously (Lanz etal., 2007; Ulbrichet al.,
of aerosol pH follows that of Eq. (1) in Budisulistiorini et
2009;Zhang,2011).ThePMF2algorithm(PaateroandTap-
al.(2015).
per,1994)wasusedinrobustmodeviaPMFEvaluationTool
panel (PET v2.04) using the methods outlined in Ulbrich
et al. (2009) and Zhang et al. (2011). Only the mass range
3 Results
m/z12–120 was utilized for PMF because no organic frag-
mentionsarepossibleatm/z<12andlowtransmissionef-
Seasonally averaged NR-PM was typically higher at JST
1
ficiency for ions with m/z>120 (Ng et al., 2011b), which in 2012 (6–13µgm−3) compared to LRK in 2013 (5–
resultsinlowsignal-to-noiseratiosaswellaspossibleinter- 8µgm−3), especially during colder seasons (fall and win-
ferencefromnaphthalenecalibrantatm/z128.
ter; Table 2). However, during warmer seasons (spring and
PMF analysis of year-long data collected from JST and
summer) the average NR-PM concentrations were similar
1
LRKyieldedsimilarfactorsolutionsasthoseobtainedfrom
at both sites. The highest average seasonal concentration of
seasonal data, but showed additional factor splitting that NR-PM atJSTwasobservedduringthefall(12.5µgm−3),
1
made solid identification of unique factors difficult. There-
whereasthesummerseasonyieldedthehighestaverageNR-
fore,wepresentresultsfromPMFanalysisperformedsepa- PM concentrationattheLRKsite(8.4µgm−3).Thesepat-
1
ratelyforwinter,spring,summerandfallseasonsfortheJST
ternscorrespondtoOAandsulfateseasonaltrends,suggest-
and LRK sites. Solutions were chosen based on the qual-
ingtheimportantrolesofthesespeciestototalNR-PM mass
1
ity of PMF fits as well as interpretability when compared
aturbanandruralsitesacrossthesoutheasternUS(Tanneret
to reference mass spectra (Ng et al., 2011a; Robinson et
al.,2015;Xuetal.,2015a).
al.,2011)andindependentgas-andparticle-phasemeasure-
ments (Budisulistiorini et al., 2013, 2015). For each analy- 3.1 Submicronaerosolchemicalcomposition
sis,uncertaintyofselectedfactorsolutionswasinvestigated
with different seeds (seed parameter varied from 0 to 100, AttheLRKsite,averageOAloadingsincreasedfromspring
in steps of 5), FPEAK parameters, and 100 bootstrapping (3.2µgm−3)tosummer(5.3µgm−3),andthendecreasedin
runs. PMF analysis of each season is detailed in Figs. S1– fall(2.8µgm−3),whichislikelyrelatedtoBVOCemissions
S24intheSupplementandcorrelationsofselectedPMFfac- thatdependonleafsurfacearea,solarradiationandambient
torswithexternaltracersandreferencemassspectraarepro- temperature (Fig. 2; Guenther et al., 2006). A different pat-
vided in Tables S2–S3. Q/Q from PMF analysis of JST tern was observed at the urban site (Fig. 1), where average
exp
data for all four seasons is 2.2–2.9, indicating that the er- OA loadings were highest during the fall (8.2µgm−3) fol-
rorsaresomewhatunderestimated(Ulbrichetal.,2009).This lowedbywinter(7.2µgm−3),suggestingcontributionsfrom
couldbeduetosomemissingdatapointsandthelackofdis- biomass-burning-relatedOAandnon-biogenicsources.High
tincttimeseriesduringnighttimeduetoatmosphericstability concentrationofOAinfallisslightlylowerthanACSMmea-
andlimitationofACSMmeasurements(nothigh-resolution), surementinfall2011(Budisulistiorinietal.,2013),butcon-
suchasobservedbyGuhaetal.(2015).Q/Q fromPMF sistentwithHR-ToF-AMSmeasurementsinNovember2012
exp
analysis of LRK data for all four seasons is between 0.15– (Xu et al., 2015a; Table S4), suggesting the role of meteo-
Atmos.Chem.Phys.,16,5171–5189,2016 www.atmos-chem-phys.net/16/5171/2016/
S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition 5175
Table2.Seasonalaveragedmassconcentrationsofnon-refractoryPM1(NR-PM1)inµgm−3,percentcontributionsoforganicandinorganic
speciesmeasuredbytheAerodyneACSMandPMFfactorsresolvedfromAtlanta,GA(JSTsite),andLookRock,TN(LRKsite),during
2012and2013.
Winter Spring Summer Fall
JST LRK JST LRK JST LRK JST LRK
NR-PM1 10.50±7.32 4.77±3.32 6.19±2.85 5.59±3.47 8.78±4.46 8.39±4.44 12.47±6.72 4.55±2.55
OA 69.0% 50.2% 75.9% 57.8% 70.0% 63.4% 65.9% 62.1%
SO2− 13.4% 30.6% 12.0% 26.9% 17.4% 24.5% 15.8% 21.6%
4
−
NO 9.3% 9.2% 5.6% 6.1% 4.5% 3.8% 9.3% 7.2%
3
+
NH 7.9% 9.9% 6.3% 9.0% 7.9% 8.2% 8.6% 9.1%
4
−
Cl 0.3% 0.1% 0.2% 0.1% 0.1% 0.1% 0.3% 0.0%
OAspeciation
HOA 24% n.a. 20% n.a. 18% n.a. 20% n.a.
BBOA 19% 33% n.a. n.a. n.a. n.a. 15% n.a.
SV-OOA 26% n.a. n.a. n.a. n.a. n.a. 28% n.a.
LV-OOA 30% 66% 43% 37% 44% 36% 37% 63%
91Fac n.a. n.a. n.a. 21% n.a. 23% n.a. 9%
IEPOX-OA n.a. n.a. 37% 41% 38% 40% n.a. 27%
Residuals 1% 1% 0% 1% 0% 1% 0% 1%
n.a.denotesvaluesnotavailableorresolvedfromPMFanalysis.PMFanalysisyieldedsomeresidualsofunresolvedOAmassthatmakeuptheremainingpercentageofOA
factors.
rology. Average OA contributions to NR-PM were higher chlorideloadingswerelow(<0.1µgm−3),indicatingthatit
1
inspringandsummeratJSTandLRK,suggestingthatbio- is not a significant contributor to inorganic aerosol mass in
genicSOAplaysasignificantroleduringtheseperiods.OA this region. The increasing average contributions from the
characterizationisfurtherdiscussedinSect.3.2. sum of sulfate, ammonium and nitrate in winter and fall at
Averagesulfateconcentrationswerehighestinsummerfor JSTsuggeststheimportantroleofinorganicsinNR-PM ,in
1
LRK(2.1µgm−3)andfallforJST(∼2µgm−3;Fig.3).This accordwithobservationsinothermajorurbanareas(Sunet
suggests that sulfate may contribute to enhanced SOA for- al.,2011;Petitetal.,2015).
mation in this region (Lin et al., 2013a; Xu et al., 2015b; ThelowestseasonalaveragepHwasobservedinsummer
Budisulistiorini et al., 2015). Changes in sulfate concentra- (1.45) for JST (Fig. 3) and in fall (1.53) for LRK (Fig. 3).
tion at LRK were mainly affected by changes in SO emis- On the other hand, the highest seasonal average pH was
2
sions from electrical-generating units in the region (Tanner 2.01 for JST and 1.81 for LRK, which were observed dur-
etal.,2015).AtJST,sulfatemeasurementsarelowerbutstill ingwinter.Overall,seasonalaerosolpHwas1.5–2.0atboth
within a standard deviation of those measured by HR-ToF- sites,indicatingthatNR-PM inthesoutheasternUSisacidic
1
AMSinMayandJuly2012inAtlanta(Xuetal.,2015a).SO yearround.ThisisconsistentwitharecentstudybyGuoet
2
emissions from coal-fired power plants nearby Atlanta con- al.(2015).Nodirectcorrelation(r2<0.1)wasobservedbe-
tributedtospatialvariabilityofsulfateconcentration(Peltier tween aerosol pH and OA at both sites. However, this does
etal.,2007).TheaveragecontributionofsulfatetoNR-PM not necessarily rule out the potential role of aerosol acidity
1
loading was quite significant throughout the year, ranging in enhancing SOA formation in light of laboratory studies,
from12to17%atJSTandfrom21to31%atLRK(Table2). demonstratingasignificantpHeffect(Gaoetal.,2004;Sur-
Average concentrations of ammonium and nitrate were <1 ratt et al., 2007; Lin et al., 2013b). Uncertainty of aerosol
JST and <0.5µgm−3 at LRK. The average ammonium and acidityestimationbyISORROPIA-IIbyomissionoforganic
nitratecontributiontoseasonalaverageNR-PM loadingsis sulfateasinput(Linetal.,2014)couldleadtounderpredic-
1
small compared to OA and sulfate (Table 2). Both ammo- tion of aerosol acidity and the observed lack of correlation
niumandnitrateshowedsimilartrendsattheJSTsite,where with OA. Seasonal averages of LWC were highest during
they were highest during colder seasons (winter and fall), summer at both JST (33.97molL−1 of aerosol) and LRK
whileshowingnosignificantfluctuationsduringtheduration (38.17molL−1) sites. It should be noted that the possible
ofthestudyatLRK.Thisobservationisconsistentwithpre- LWC contributions from OA are not included because the
viousstudies(Tanneretal.,2004b;Olszynaetal.,2005)re- organic hygroscopicity parameter estimated from observed
portingthataveragecontributionsofammoniumandnitrate cloudcondensationnucleiactivitiesofOA(Guoetal.,2015)
are not significant for rural PM . Average non-refractory was not available in this study. Studies have suggested that
1
www.atmos-chem-phys.net/16/5171/2016/ Atmos.Chem.Phys.,16,5171–5189,2016
5176 S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition
4 3.0
-3OA (µg m) 1842 2--3SO (µg m)4 321 --3NO (µg m)3 21..00
0 0 0.0
wtr spr smr fall wtr spr smr fall wtr spr smr fall
3.0 0.20 3.0
+-3H (µg m)4 21..00 --3Cl (µg m) 000...110505 pH 21..00
N
0.0 0.00 0.0
wtr spr smr fall wtr spr smr fall wtr spr smr fall
Figure 3. Seasonal averages of OA, inorganic species and pH from JST (solid squares) and LRK (open triangles). Error bars show ±1
standarddeviation.Seasonsareclassifiedintowinter(wtr),spring(spr),summer(smr)andfall.
reactiveuptakedecreaseswithenhancedRH(Nguyenetal., (e.g. deposition) could influence diurnal patterns observed
2014; Gaston et al., 2014); however, some isoprene-derived hereforthePMFfactors.
SOAtracerswereelevatedbyhighRH(Zhangetal.,2011).
Although organic water fraction in total LWC was found to 3.2.1 Winter
besignificant,Guoetal.(2015)suggestedthatpHprediction
using ISORROPIA-II based on inorganic ions alone gave a PMF analysis of winter OA yielded a four-factor solution
reasonableestimate.ThelackofcorrelationbetweenOAand at JST (Figs. 4a and 5a) and a two-factor solution at LRK
pH as well as LWC indicates that pH and LWC might not (Figs.6aand7a).HOA,BBOA,SV-OOAandlow-volatility
be limiting factors in OA production in this region, consis- oxygenated OA (LV-OOA) factors (Ng et al., 2011a) were
tent with previous studies in Georgia and Alabama (Xu et resolvedfromtheJSTdataset,whereasonlytheBBOAand
al., 2015a) and Tennessee (Budisulistiorini et al., 2015). It LV-OOA factors were resolved from the LRK data set. In-
shouldbenotedthatthisstudydidnotincludethecontribu- creasingthenumberoffactorsinPMFanalysisofLRKdata
tion of organic water into pH estimation, which could con- resultedinsplittingfactorsthatsharesimilaritieswithBBOA
tributetotherelationshipbetweenpHandOA. factor. Thus, we selected a two-factor solution (p=2) for
LRKinwinter.
The temporal variation of the HOA factor correlates well
3.2 OAcharacterizations
(r2>0.7) with black carbon (BC), carbon monoxide (CO)
and reactive nitrogen species (NO ; Table S2). Moreover,
y
The mass spectra and time series of OA factors resolved itsdiurnalvariation(Fig.9)showedamorningpeak,consis-
from PMF analysis at JST in 2012 are provided in Figs. 4 tentwithanexpectedcontributionfromvehicularemissions
and 5, respectively, and at LRK in 2013 are provided in (Zhangetal.,2007).
Figs. 6 and 7, respectively. More PMF factors were re- TheBBOAfactorconcentrationincreasedduringthenight
solved from JST OA than from LRK OA, which could be and decreased during the day at JST (Fig. 9), which could
due to a larger number of OA source types in urban areas. berelatedtoresidentialandnon-residentialwoodburningas
Each factor had a distinctive time trend throughout 2012 wellasPBLdynamics.BBOAattheLRKsitealsoshowed
(Fig. 5) at JST and 2013 at LRK (Fig. 7). OA measured a large nighttime peak with a gradual decrease during the
at JST in 2012 and LRK in 2013 was composed primar- day (Fig. 10). The large peak appears to result from a short
ilyoflow-volatilityoxygenatedOA(LV-OOA)andIEPOX- period of intense biomass burning that occurred during 15–
derivedOAfactor(IEPOX-OA).ConcentrationsofLV-OOA 18 March 2013. Since a source for this event could not be
and IEPOX-OA at both sites were 1.9 and 1.6µgm−3 on identified, we do not report it specifically in this study. The
average,respectively(Fig.8).Hydrocarbon-likeOA(HOA) timeseriesofBBOAshowedlowtomoderatecorrelation(r2
andsemi-volatileoxygenatedOA(SV-OOA)concentrations 0.4–0.5atJSTandr2 0.2–0.4atLRK)withBC,suggesting
variedbetween1and2µgm−3 atJSTandbiomassburning thatitislikelyinfluencedbysomelocalsources(e.g.,fires).
OA (BBOA) was ∼1µgm−3 at both sites. A biogenically BBOA mass spectra from JST and LRK were highly corre-
influencedfactor(91Fac)wasobservedonlyatLRKandac- lated (r2∼0.7), indicating similarity of the sources. Com-
countedfor∼1µgm−3.Duetoalackofmeasurements,the parisonoftheBBOAmassspectrawithreferencemassspec-
potential role of planetary boundary layer (PBL) height to tra showed correlation with other OOA factors (Tables S2
diurnal variation of PMF factors was not accounted for in and S3), a known caveat in resolution of BBOA based on
thisstudy.However,diurnalPBLdynamicsorlossprocesses unitmassresolution(UMR)datasuchasthosefromACSM
Atmos.Chem.Phys.,16,5171–5189,2016 www.atmos-chem-phys.net/16/5171/2016/
S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition 5177
measurements(Woodetal.,2010).ThesimilarityofBBOA
and OOA factor mass spectra could indicate aging of the
BBOAfactor,whichwasobservedtohaveenhancedsignals
(a) 0.12
HOA atm/z18,29and44ionsandlowsignalsatm/z60and73
0.08
ions (Bougiatioti et al., 2014). However, the BBOA factor
0.04
observed at JST and LRK displayed an enhanced signal at
0.00
al0.12 BBOA m/z44 ion but retained signals at m/z60 and 73 ions, sug-
gn0.08 gestingthatitwasnotasoxidizedastheagedBBOAfactor.
si0.04
of 0.00 LV-OOA is characterized by a high fraction of total ion
on 0.16 SV-OOA intensity at m/z44 (f44) resulting from high oxygen con-
acti00..1028 x 5 tent (Ng et al., 2011a) and is the most abundant OA type
Fr00..0040 at both JST and LRK (Table 2). Maxima around midnight
0.20 LV-OOA at JST (Fig. 9) and in the mid-afternoon at LRK (Fig. 10)
0.15 x 5
0.10 were not significant, indicating that LV-OOA concentration
0.05
0.00 is relatively constant throughout the day in this region. LV-
(b) OOA has been shown to correlate with non-volatile sec-
0.12 HOA
ondaryspecies,suchassulfate(Jimenezetal.,2009).Weak
0.08
0.04 correlation(r2<0.2)betweenLV-OOAandsulfatemightbe
al 0.00 due to a complex oxidation process, as previously observed
n
Sig 0.12 IEPOX-OA in urban ambient aerosol (Sun et al. 2011a). On the other
action of 000...000840 x 5 hsiatnesd,shthoewmedassstrsopnegctcroarlrecloamtiopnari(sro2n∼o1f,LFVig-O. OS2A5)f,ropmossbiboltyh
Fr 0.20 LV-OOA suggestingsimilarsourcesofLV-OOAatthesesites.
0.15
x 5 SV-OOA, which was observed only at JST, showed an
0.10
0.05 f smaller than that of LV-OOA (Fig. 4a), indicating that
44
0.00
the factor is less oxidized and thus semi-volatile (Ng et al.,
(c) 0.10
0.08 HOA 2011a).ThetemporalvariationofSV-OOAwasmoderately
0.06 correlated (r2∼0.4) with nitrate (Table S2) while the mass
0.04
0.02 spectrumwaswellcorrelatedwithpreviouslyresolved82Fac
al 0.00
n
g 0.16 IEPOX-OA and IEPOX-OA factors (82Fac and IEPOX-OA are equiv-
n of Si 00..1028 x 5 alent and are characterized by a prominent ion at m/z82;
o 0.04 Robinson et al., 2011; Budisulistiorini et al., 2013, 2015).
Fracti 00..0200 LV-OOA Since isoprene emission is expected to be negligible during
0.15 winter season, SV-OOA might not relate to IEPOX-derived
0.10 x 5
SOA. The diurnal profile of SV-OOA showed an increase
0.05
0.00 in the evening and decrease in the morning, similar to the
(d) 0.10 BBOA profile and HOA factors. Moreover, it tracked well
0.08 HOA −
0.06 withthediurnalprofileofNO .Thissuggestsapossiblein-
3
0.04 fluence of nitrate-radical chemistry on nighttime SOA for-
0.02
0.00 mationduringwinter(Xuetal.,2015b;Rollinsetal.,2012).
0.12 BBOA
al0.08
n
g0.04
of si0.00 3.2.2 Spring
n 0.15 SV-OOA
o
cti0.10 x 5
a0.05 PMF analysis of spring OA resulted in a three-factor solu-
Fr0.00
tion (i.e., HOA, LV-OOA, and IEPOX-OA) for the JST site
0.20 LV-OOA
0.15 (Figs.4band5b)andathree-factorsolution(i.e.,LV-OOA,
0.10 x 5
0.05 91Fac,andIEPOX-OA)fortheLRKsite(Figs.6band7b).
0.00
IncreasingthenumberoffactorsinPMFanalysisofJSTre-
20 40 60 80 100 120
sultedinsplittingcomponents,andthus,SV-OOAwasnotre-
m/z
solvedinspring.ThelackoftheSV-OOAfactormightresult
Figure 4. Mass spectra of PMF factors resolved from (a) winter, fromevaporationofsemi-volatilespeciesinwarmerperiods
(b)spring,(c)summerand(d)fallOAmeasuredatJSTin2012. and/ortheinabilityoftheACSMtopickuponthevariability
ofafactorwithlowconcentration.Similarly,asplittingcom-
ponent was observed in PMF analysis of LRK data p=4.
www.atmos-chem-phys.net/16/5171/2016/ Atmos.Chem.Phys.,16,5171–5189,2016
5178 S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition
40
30 Winter Spring Summer Fall HOA
20
10
0
6 BBOA
-3m) 42
µg 0
c. ( 12 IEPOX-OA
n 8
o
c 4
ss 0
Ma 16 SV-OOA
12
8
4
0
12 LV-OOA
8
4
0
1/1/2012 1/3/2012 1/5/2012 1/7/2012 1/9/2012 1/11/2012
Date and time (local)
Figure5.AnnualtemporalvariationofPMFfactorsresolvedfromOAmeasuredatJSTin2012.
Thus, BBOA and/or HOA were not resolved from LRK in constant concentration until the evening suggests that this
spring. factorisdrivenbyphotooxidationofisoprene(Budisulistior-
TheaverageconcentrationofHOAinAtlantawaslowerin ini et al., 2013). At JST, the diurnal pattern of IEPOX-OA
spring(0.7µgm−3)thaninwinter(1.7µgm−3),whichcould followedthatoftotalOA,whereitslightlydecreasedduring
beinfluencedbydilution–fromariseofthePBL–andevap- thedaybeforeitincreasedagainintheevening.Thisdiurnal
oration of POA during warmer conditions (Robinson et al., patternisdifferentfrompreviousobservationsatJSTduring
2007).Althoughitsconcentrationdecreases,thediurnalpat- summer2011(Budisulistiorinietal.,2013),butquitesimilar
tern of HOA was consistent from winter to spring (Fig. 9) toisopreneOAfromMay2012reportedbyXuetal.(2015a),
and correlation with primary species was strong (r2∼0.6, suggestinginfluenceofyear-to-yearchangesinmeteorology,
TableS2). such as precipitation and solar radiation (Table S1). Never-
AverageLV-OOAconcentrationatJSTalsowasthelowest theless,themassspectraofIEPOX-OAatJSTandLRKare
in spring (1.4µgm−3), which might be attributed to warm- tightlycorrelated(r2∼1),indicativeofsimilarcomposition.
ing temperatures that elevate the PBL and enhance atmo- 91Fac was resolved only at the LRK site and accounted
sphericmixing.DiurnalvariationofLRKLV-OOA(Fig.10) for 0.7–1.2µgm−3. 91Fac has been attributed to various
showedasmalldiurnalmaximumintheafternoon,whereas sources: monoterpenes-derived SOA (Budisulistiorini et al.,
no variation was observed for JST LV-OOA (Fig. 9). LRK 2015; Boyd et al., 2015), biogenic SOA (Chen et al., 2015)
LV-OOAshowedmoderatecorrelationwithsulfate(r2>0.4, and aged BBOA (Robinson et al., 2011). However, a recent
Table S3), suggesting influence of sulfate at this site dur- field study identified ions at m/z18, 29 and 44 as mark-
ingspring(Tanneretal.,2015).Althoughnocorrelationwas ers for aged BBOA but not m/z91 ion (Bougiatioti et al.,
foundforJSTLV-OOAvs.sulfate,comparisonofmassspec- 2014). Since BBOA was not resolved from OA measure-
tra revealed the same strong correlation (r2∼1, Fig. S25) ments in spring, aging of BBOA seems unlikely to be the
betweenJSTandLRKLV-OOAfactorsobservedinwinter, source of 91Fac in this study, although it cannot be conclu-
suggestingpossiblesimilarsourcesoveraregionalscale. sively ruled out. The lack of 91Fac at the JST site suggests
TheIEPOX-OAfactor,attributedtoIEPOXheterogeneous that its sources may be limited to emissions and chemical
chemistry(Budisulistiorinietal.,2013;Linetal.,2012),was processes in forested and/or rural areas. 91Fac will be fur-
resolvedfromdatasetsatbothJSTandLRK.Itwasthesec- therdiscussedinSect.4.2.
ond most abundant OA type after LV-OOA at JST, but the
mostabundantOAcomponentatLRK(Table2).Theaverage 3.2.3 Summer
IEPOX-OAconcentrationwasslightlyhigheratLRKthanat
JST,whichisexpectedduetoabundantemissionsofisoprene PMF analysis of summer OA resolved the same factors as
attheforestedsite.DiurnalpatternsofIEPOX-OAarediffer- springatbothsites:HOA,LV-OOAandIEPOX-OAfactors
ent at JST and LRK. At LRK, IEPOX-OA has insignificant at JST (Figs. 4c and 5c), and LV-OOA, 91Fac and IEPOX-
diurnal variability, which is likely influenced by small vari- OA factors at LRK (Figs. 6c and 7c). Average HOA mass
ability of sulfate as previously observed at this site (Tanner concentration at JST increased in summer to ∼1.1µgm−3
etal.,2005).However,asmallincreaseintheafternoonand (Fig. 8). Temporal variation of HOA was well correlated
(r2∼0.6) with BC, CO and NO (Table S2) and the diur-
x
Atmos.Chem.Phys.,16,5171–5189,2016 www.atmos-chem-phys.net/16/5171/2016/
S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition 5179
(a) 0.10 blesS2–S3).ThediurnalprofileofLRKLV-OOAshoweda
al 0.08 BBOA localmaximuminmid-afternoon(Fig.10)andhasamoder-
n 0.06
n of sig 000...000420 iantegctohrarteslautlifoante(rp2la∼ys0a.4r)owleithinthLeVs-uOlfOaAte(inTasbulmeSm3e)r,sautgLgResKt-.
ctio 00..2250 LV-OOA ComparisonofJSTandLRKLV-OOAmassspectrarevealed
Fra 00..1150 x 5 astrongcorrelation(r2=0.94),possiblysuggestingsimilar
0.05 sourcesbetweentwosites.
0.00
Averageconcentrationofthe91FacOAatLRKwashigher
(b)
0.15 91Fac insummerthanspring,whichindicatestheroleofmeteorol-
0.10 x 5 ogy – an increasing temperature from ∼13◦C in spring to
0.05
al 0.00 ∼21◦C in summer (Table S1). The relative contribution of
n
sig 0.15 IEPOX-OA 91FactototalOAincreasedatLRK(Table2)anditsdiurnal
n of 0.10 x 5 profile showed a local maximum around noon. A moderate
actio 00..0050 correlation of 91Fac with nitrate (r2∼0.5, Table S3) sug-
Fr 0.25 LV-OOA geststhatthefactorismoderatelyoxidized.
0.20
0.15 x 5 AverageconcentrationofIEPOX-OAatJSTandLRKin-
0.10
0.05 creased during summer. At LRK, the average concentration
0.00
ofIEPOX-OAreachedamaximuminsummer,butitsrelative
(c)
0.15 91Fac contributiontototalOAmasswaslowerduetotheincreas-
0.10 x 5 ingconcentrationof91Fac.ConcentrationsofIEPOX-OAat
0.05 bothsitesarecomparable(Fig.8),suggestingthatinsummer
al 0.00
sign 0.15 IEPOX-OA thisfactormaybecomespatiallyhomogeneousinthesouth-
on of 00..1005 x 5 edausctteerdndUuSri.nSgindcifefemreenatsuyreeamrse,nmtsetaetoJrSoTloganicdalLcRhKanwgeesremciognh-t
acti 0.00 play a role in site-to-site comparison. At LRK, IEPOX-OA
Fr 0.20 LV-OOA showedasmallincreasearoundnoon,whileatJSTtherewas
0.15 x 5
0.10 alocalmaximuminthemid-afternoon,suggestinganinflux
0.05
0.00 of IEPOX-OA likely transported from surrounding forested
(d) 0.12 areas. The time series of IEPOX-OA was moderately cor-
0.08 91Fac related with nitrate (r2∼0.4) at JST, and at LRK, stronger
X 2
0.04 correlations(r2>0.5)withsulfateandnitratewereobserved,
al 0.00 suggestingthatthisfactorismoderatelyoxidized.
n
sig 0.15 IEPOX-OA
of 0.10 x 5
n
o 0.05 3.2.4 Fall
acti 0.00
Fr 0.20 LV-OOA
0.15
0.10 x 5 At JST, PMF analysis of fall OA resulted in a four-factor
0.05 solution (i.e., HOA, BBOA, SV-OOA and LV-OOA), while
0.00
at LRK a three-factor solution was resolved (i.e., LV-OOA,
20 30 40 50 60 70 80 90 100 110 120
m/z 91FacandIEPOX-OA).Increasingthenumberoffactorsin
PMFanalysisofJSTfalldataresultedinfactorsplitting,and
Figure 6. Mass spectra of PMF factors resolved from (a) winter,
thus, the IEPOX-OA factor was not resolved from this data
(b)spring,(c)summerand(d)fallOAmeasuredatLRKin2013.
set. Similarly, we could not resolve the BBOA factor from
LRKfalldatabecausetheanalysisresultedinsplittingcom-
ponents.
nal pattern was similar to that of spring (Fig. 9). Similar to TheconcentrationofJSTHOAincreased toalevelcom-
spring,SV-OOAwasnotresolvedinsummer,whichcouldbe parabletothatinwinter(Fig.5),whichmightbeinfluenced
attributedtorapidevaporationofsemi-volatilespeciesunder by meteorology – a low ambient temperature and less solar
highambienttemperatures(TableS1). radiation–infallandwinter.Thecorrelationofthetimese-
AverageLV-OOAconcentrationsatbothsitesincreasedin ries of HOA with BC, CO and NO (r2>0.7) was similar
x
summer;however,theproportionalcontributiondecreasedas to spring and summer and slightly stronger than in winter
a result of a larger contribution of IEPOX-OA at JST and (Table S2) and the diurnal profile appears similar to that in
91Fac at LRK (Table 2). The time series of LV-OOA was winter(Fig.9).ThepresenceoftheHOAfactorthroughout
weakly correlated with sulfate (r2∼0.2) at JST, but more theyearatJSTisexpectedduetotrafficemissionsinurban
strongly correlated with sulfate at LRK (r2=0.6–0.7; Ta- areas(Xuetal.,2015a).
www.atmos-chem-phys.net/16/5171/2016/ Atmos.Chem.Phys.,16,5171–5189,2016
5180 S.H.Budisulistiorinietal.:Seasonalcharacterizationofsubmicronaerosolchemicalcomposition
60
Winter Spring Summer (SOAS) Fall
40 BBOA
20
0
6
-3m) 42 91Fac
g
c. (µ 120
on 8 IEPOX-OA
c
ss 4
a
M 0
8
LV-OOA
6
4
2
0
1/3/2013 1/5/2013 1/7/2013 1/9/2013 1/11/2013
Date and time (local)
Figure7.AnnualtemporalvariationofPMFfactorsresolvedfromOAmeasuredatLRKin2013.OAmeasurementsinthesummerincluded
resultsfromtheSouthernOxidantAerosolStudy(SOAS)campaignthathavebeenpublishedinBudisulistiorinietal.(2015).
-3HOA (µg m) 543210 -3BBOA (µg m) 543210 -3SV-OOA (µg m) 543210
wtr spr smr fall wtr spr smr fall wtr spr smr fall
-3LV-OOA (µg m) 543210 -3EPOX-OA (µg m) 543210 -391Fac (µg m) 543210
I
wtr spr smr fall wtr spr smr fall wtr spr smr fall
Figure8.SeasonalaveragemassconcentrationsofPMFfactorsresolvedfromJST(solidsquares)andLRK(opentriangles).Errorbarsare
shownas±1standarddeviation.
AtJST,theBBOAfactorwasresolvedagainfromOAwith stant from summer to fall at both the urban and rural sites
averageconcentrationandfractionalcontributiontototalOA (Fig.8).However,thecontributionofLV-OOAtototalOAat
less than observed in winter. The diurnal profile of BBOA LRKincreasedduetodecreasingconcentrationsofotherOA
duringfallatJSTappearedsimilartothatinwinter,suggest- factors (i.e., IEPOX-OA and 91Fac; Table 2). JST LV-OOA
ing similar emission sources as well as possible PBL effect didnotshowdiurnalvariation,whereasXuetal.(2015a)ob-
duringthesetwocolderseasons.ThelackoftheBBOAfac- servedasmalldiurnalvariationbyHR-ToF-AMS.Themass
toratLRKcouldbeattributedtotheinabilityoftheACSMto resolutionoftheACSMinstrumentisnotashighastheHR-
captureafactorwithlowconcentration.Inwinter,theACSM ToF-AMS; thus, it might not be able to capture the diur-
couldcapturethestrongsignalofBBOAduetosomeperiods nalvariability.LRKLV-OOAincreasedinmid-morningand
ofintenseburningwhichwouldnotbeexpectedinfall. reached a maximum around mid-afternoon. Temporal vari-
SV-OOA was also resolved from JST OA with slightly ation of LV-OOA was weakly correlated (r2∼0.2) with in-
higheraverageconcentrationandfractionalcontributionthan organics at JST, but moderately correlated (r2=0.4–0.5) at
that observed in winter. The diurnal profile of fall SV-OOA LRK. Strong correlation of LV-OOA mass spectra (r2∼1,
wassimilartothatinwinter,suggestingsimilarsourcesand slope=0.8–1.1,Fig.S25)atJSTandLRKindicatesasimi-
the role of the PBL. The return of SV-OOA might be influ- laroridenticalsource.
encedbydecreasesintemperaturefrom∼26◦Cinsummer Theconcentrationof91FacatLRKdroppedsignificantly
to∼15◦Cinfall(TableS1),resultinginlessevaporationof infall.ThedropcoincidedwithadecreaseoftotalOAcon-
semi-volatilespecies. centration and ambient temperature – from around 20◦C
LV-OOA was resolved from OA at both JST and LRK. to around 10◦C (Fig. 2). Temperature has been shown to
AverageconcentrationsofLV-OOAremainedrelativelycon- haveanegativeeffectonSOAformationfrommonoterpenes
Atmos.Chem.Phys.,16,5171–5189,2016 www.atmos-chem-phys.net/16/5171/2016/