Inorganic and black carbon aerosols in the Los Angeles Basin during CalNex

27 Pages·2013·7.16 MB·English

JOURNALOFGEOPHYSICALRESEARCH:ATMOSPHERES,VOL.118,1777–1803,doi:10.1029/2012JD018136,2013
Inorganic and black carbon aerosols in the Los Angeles
Basin during CalNex
J. J. Ensberg,1 J. S. Craven,1 A. R. Metcalf,2 J. D. Allan,3,4 W. M. Angevine,5,6
R.Bahreini,7J.Brioude,5,6C.Cai,8H.Coe,3J.A.deGouw,5,6R.A.Ellis,9J.H.Flynn,10
C. L. Haman,10 P. L. Hayes,5,11 J. L. Jimenez,5,11 B. L. Lefer,10 A. M. Middlebrook,6
J.G.Murphy,12J.A.Neuman,5,6J.B.Nowak,5,6J.M.Roberts,6J.Stutz,13J.W.Taylor,3
P. R. Veres,14 J. M. Walker,15 and J. H. Seinfeld1,15
Received22May2012;revised4December2012;accepted10December2012;published19February2013.
[1] WeevaluatepredictionsfromtheCommunityMultiscaleAirQuality(CMAQ version
4.7.1) model against a suite of airborne and ground-based meteorological measurements,
gas- and aerosol-phase inorganic measurements, and black carbon (BC) measurements
over Southern California during the CalNex field campaign in May/June 2010. Ground-
based measurements are from the CalNex Pasadena ground site, and airborne
measurements took place onboard the Center for Interdisciplinary Remotely-Piloted
Aircraft Studies (CIRPAS) Navy Twin Otter and the NOAA WP-3D aircraft. BC
predictions are in general agreement with observations at the Pasadena ground site and
onboard the WP-3D, but are consistently overpredicted when compared to Twin Otter
measurements. Adjustments to predicted inorganic mass concentrations, based on
predicted aerosol size distributions and the AMS transmission efficiency, are shown to be
significant.Owingtorecentshippingemissionreductions,thedominantsourceofsulfatein
the L.A. Basin may now be long-range transport. Sensitivity studies suggest that severely
underestimatedammoniaemissions,andnottheexclusionofcrustalspecies(Ca2+,K+,and
Mg2+),arethesinglelargestcontributortomeasurement/modeldisagreementintheeastern
part of the L.A. Basin. Despite overstated NO emissions, total nitrate concentrations are
x
underpredicted, which suggests a missing source of HNO and/or overprediction of
3
deposition rates. Adding gas-phase NH measurements and size-resolved measurements,
3
up to 10 mm, of nitrate and various cations (e.g. Na+, Ca2+, K+) to routine monitoring
stations in the L.A. Basin would greatly facilitate interpreting day-to-day fluctuations in
fine and coarse inorganic aerosol.
Citation: Ensberg,J.J.,etal.(2013),InorganicandblackcarbonaerosolsintheLosAngelesBasinduringCalNex,
J.Geophys.Res.Atmos.,118,1777–1803,doi:10.1029/2012JD018136.
Allsupportinginformationmaybefoundintheonlineversionofthisarticle. 6Chemical Sciences Division, Earth System Research Laboratory,
1DivisionofChemistryandChemicalEngineering,CaliforniaInstituteof NOAA,Boulder,Colorado,USA.
Technology,Pasadena,California,USA. 7Now at the Department of Environmental Sciences University of
2Now at the Combustion Research Facility at Sandia National California,Riverside,Riverside,California,USA.
Laboratories,Livermore,California,USA. 8Planning and Technical Support Division, Air Resources Board,
3National Centre for Atmospheric Science, University of Manchester, California Environmental Protection Agency, Sacramento, California,
Manchester,UK. USA.
4SchoolofEarth,AtmosphericandEnvironmentalSciences,University 9Now at Harvard University, Cambridge, Massachusetts, USA.
ofManchester,Manchester,UK. 10Department of Earth and Atmospheric Sciences, University of
5Cooperative Institute for Research in Environmental Sciences, Houston, Houston, Texas, USA.
UniversityofColoradoBoulder,Boulder,Colorado,USA. 11DepartmentofChemistryandBiochemistry,UniversityofColorado,
Boulder, Colorado, USA.
12Department of Chemistry, University of Toronto, Toronto, Ontario,
Correspondingauthor:J.H.Seinfeld,DivisionofChemistryand Canada.
ChemicalEngineering,CaliforniaInstituteofTechnology,Pasadena, 13Department of Atmospheric Sciences, University of California, Los
California,USA.([email protected]) Angeles, California, USA.
14NowattheMaxPlanckInstituteforChemistryinMainz,Germany.
©2012.AmericanGeophysicalUnion.AllRightsReserved. 15DivisionofEngineeringandAppliedScience,CaliforniaInstituteof
2169-897X/13/2012JD018136 Technology, Pasadena, California, USA.
1777
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
1. Introduction [5] Toevaluatetheextenttowhichpredictionsofastate-of-
the-scienceCTM,drivenbycurrentemissioninventoriesand
[2] The Los Angeles mega-city has consistently experi- resolved meteorological fields, agree with measured concen-
enced among the highest particulate matter levels in the trations, we employ the Community Multiscale Air Quality
United States and the highest ozone levels (http://www. (CMAQ) model version4.7.1 (http://www.cmaq-model.org/,
stateoftheair.org/2012/city-rankings/most-polluted-cities.html). Foleyetal.[2010])tosimulatethethree-dimensionaldistribu-
Severalmeasurementcampaignshavefocusedoncharacteriz- tionofaerosolsandgasesoverSouthernCaliforniaduringthe
ingparticulateairqualityintheLosAngelesBasin(e.g.Appel CalNexfieldcampaign.Predictionsarecomparedtoasuiteof
et al. [1982]; Turpin and Huntzicker [1991]; Chow et al. airborne and ground-based meteorological measurements,
[1994];Watsonetal.[1994];Elderingetal.[1994];Liuetal. gas- and aerosol-phase inorganic measurements, and black
[2000]; Hughes et al. [2002]; Pastor et al. [2003]; Croes carbon (BC) measurements over Southern California during
andFujita[2003];Neumanetal.[2003];Jacobetal.[2010]; the CalNex field campaign. Airborne measurements took
Docherty et al. [2011]), and in other urban areas, such as place onboard the Center for Interdisciplinary Remotely-
Pittsburgh, PA (Pittsburgh Air Quality Study, PAQS; e.g. Piloted Aircraft Studies (CIRPAS) Navy Twin Otter aircraft
Wittig et al. [2004]; Cabada et al. [2004]; Modey et al. (hereafter referred to as the Twin Otter) and the NOAA
[2004];Pekneyetal.[2006];Beinetal.[2006]),MexicoCity, WP-3D (hereafter referred to as the P3) aircraft during
Mexico (Mega-city Initiative: Local and Global Research May and June 2010. A Lagrangian particle dispersion
Observations,MILAGRO;e.g.Salcedoetal.[2006];DeCarlo model, FLEXPART (http://transport.nilu.no/flexpart, Stohl
et al. [2008]; Querol et al. [2008]; Molina et al. [2010]), et al. [2005]), is used to track the origins of measured and
Houston,Texas(2006TexasAirQualityStudy,TexAQS;e.g. predicted species in the Los Angeles Basin during CalNex
Parrish et al. [2009]; Nowak et al. [2010]; Washenfelder bycalculatingbacktrajectoriesbasedonadvectionandturbu-
et al. [2010]), and Beijing, China (Campaign of Air Quality lentmixingprocesses.Ourgoalistoidentifythemajorsources
Research in Beijing, CAREBEIJING; e.g. van Pinxteren of measurement/model disagreement for BC and various
et al. [2009]; Yue et al. [2009,2010]; Ianniello et al.[2011]; inorganicaerosol species andtosuggestadditional measures
Zheng et al. [2011]). Many studies have used data gathered that address these sources of error. The organic component
during these field campaigns to evaluate the fidelity of three- oftheparticulatematterwillbeaddressedinafuturestudy.
dimensional atmospheric chemical transport models (CTMs)
(e.g. Sarwar and Bhave [2007]; Fountoukis and Nenes 2. Model Description and Application
[2007]; Nolte et al. [2008]; Matsui et al. [2009]; Fountoukis
et al.[2009]; McKeenet al.[2009]; Fast et al. [2009]; Foley 2.1. CMAQ
et al. [2010]; Renner and Wolke [2010]; Zhang and Ying [6] Inthisstudy,weuseCMAQversion4.7.1[Foleyetal.,
[2010]; Karydis et al. [2010]; Kelly et al. [2010]; Lee et al. 2010] on a domain that includes a large portion of Southern
[2011];Huangetal.[2011];Pfisteretal.[2011]). California as well as parts of Mexico (Figure 1), covering
∘ ∘ ∘ ∘
[3] TheCaliforniaResearchattheNexusofAirQualityand theareafrom(31.83 N,121.43 W)to(35.69 N,114.43 W)
ClimateChange(CalNex)studywasconductedduringMay– with 4 km by 4 km horizontal grid cells (102 x 156 grid
June 2010 to address both air quality and climate change points). Simulations cover the time periodof May 2010.All
throughcoordinationandcollaborationbetweenseveralgov- simulations include a minimum spin-up period of four days
ernmentagencies,suchastheCaliforniaAirResourcesBoard to mitigate the influence of initial conditions, except for the
(CARB),theNationalOceanicandAtmosphericAdministra- P3flightduring4May2012whichincludedonlythreedays
tion (NOAA), and academic institutions (www.esrl.noaa. of spin up due to lack of MM5 meteorology for 30 April
gov/csd/calnex/). During CalNex, state-of-the-art airborne, 2010.Themeteorologicalfieldsusedtodrivethemodelwere
ship-based, and ground-based measurements of atmospheric generatedbythe5thgenerationPennState/NationalCenterfor
species,andoftheirtransportoverandoffthecoastofCalifor- Atmospheric Research Meso-scale Model (MM5 version
nia,wereconductedwiththegoalofunderstandingtheimpact 3.7.4;Grelletal.[1995]).Threenestedgrids,withhorizontal
ofairbornepollutantsonairqualityandclimate. resolutionsof36,12,and4km,wereusedtogeneratemeteo-
[4] Themotivationforthepresent workistoevaluate the rologicalfieldsatthedesiredresolution(FigureS1).Nestingis
extent to which we understand the observed behavior of theprocessbywhichamodelsimulationisrunoveracertain
inorganic aerosols in the Los Angeles airshed, one of the domain (the parent domain) at a given resolution. From the
world’s most important urban areas from the point of view predictedconcentrationswithintheparentdomain,boundary
of air quality. The suite of ground-level and airborne mea- conditions are extracted and used to drive a separate finer-
surements made during CalNex represent, by far, the most resolution model simulation that is run over a portion of the
complete characterization of Los Angeles air quality yet parent domain (the nested domain). The MM5 model was
carried out. Such a complex suite of measurements, gas initialized from NARR analysis data (http://nomads.ncdc.
and particle, surface and aloft, can only be placed in a noaa.gov/data/narr/)withanalysisnudgingoption,butobser-
unifiedcontextthroughtheintegrationofchemistryandphys- vational nudging was not used. The 36 km and 12 km grids
ics provided by a state-of-the-science atmospheric chemical were first run together, via two-way nesting, using the Grell
transport model, driven by appropriate meteorology for the cumulus, simple ice microphysics, NOAH soil scheme,
daysoftheexperimentandoperatingontheemissionsinven- MRFPBL,andRRTMradiationoptions.The4kmgridwas
tory of the region assembled by the relevant governmental then run, via one-way nesting, using boundary conditions
agencies. In many respects, the current work can be seen derivedfromthe 12 kmgridwithall options identicaltothe
as a parallel to the Mexico City MILAGRO study [Molina coarse domain simulations, except that the cumulus parame-
etal.,2010]. terizationwasturnedoffsincethesufficientlyfinehorizontal
1778
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
36
Coastlines
Freeways
2500
35.5
35 Tehachapi Mtns MDeosjaevrte
2000
34.5 San Gabriel Mtns El
Cajon
Pass San Bernardino
°de[] 34 Downtown L.A. Banning PMastnss 1500on (m)
Latitu33.5 Long Beach Elevati
Imperial
Valley 1000
33
32.5
500
32
31.5 0
−122 −121 −120 −119 −118 −117 −116 −115 −114
Longitude[°]
Figure1. CMAQmodelingdomain(coloredarea)usedforsimulationsduringtheCalNexFieldCampaign.
∘ ∘ ∘ ∘
Thedomaincoverstheareafrom(31.83 N,121.43 W)to(35.69 N,114.43 W)with4kmx4kmhorizontal
gridcells (102x156gridpoints).The starrepresentsthe Pasadena groundsiteandthe triangle represents
Bakersfield.
resolutionofthe4kmgridisexpectedtocapturethesmaller ThelimitationsofthecurrentCARBinventoryareexpected
cloudscaleatmosphericmotions.CMAQ-compatiblemeteo- toimpactagreementbetweenobservationsandpredictionsof
rological fields were then generated by processing MM5 anthropogenicBCandsulfurcontainingcompounds,butnot
output using the Meteorology-Chemistry Interface Program sea-salt emissions of SO2(cid:3) which are modeled interactively
4
(MCIP) version 3.6 [Otte and Pleim, 2010]. Vertically, in CMAQ [Gong, 2003; Sarwar and Bhave, 2007; Kelly
the meteorological fields extend from the surface to 100 mb etal.,2010].Allmeteorologicalfieldsandgriddedemission
((cid:1)18 km above sea level) using 30 layers. This vertical inventorieswerepreparedandprovidedbyCARB.
resolutionistypicalforregional-scaleandmeso-scalemodel- [8] Gas-phasepredictionsare basedona modifiedversion
ing studies such as this. For instance, Kelly et al. [2010] of the Statewide Air Pollution Research Center (SAPRC)
configured CMAQ to use 30 vertical layers for the coast of chemical mechanism (version SAPRC07TC, Carter [2010])
Florida. Similarly,Foley etal. [2010] testedCMAQ4.7 over implemented in CMAQ with the Rosenbrock numerical
theeasternUnitedStatesusing12,24,and34verticallayers. solver.Themainmodificationconsistsofusinganupdated
In the present study, there are 11 layers in the lowest 1000 isoprenephotooxidationmechanism basedon Paulotetal.
m,andthesurfacelayeris(cid:1)30mdeep. [2009a,2009b].Atmosphericmassdistributionsofparticu-
[7] Theemissioninventoryusedinthisstudyisamodified latematterbysizearerepresentedinCMAQasthesuperpo-
versionoftheday-specificARCTAS-CARBinventoryfrom sition of three log-normal distributions, referred to as
June 2008. The modification consists of averaging June modes. These are the Aitken mode (typical D range is
p
emissions to produce emissions for one representative 20nmto90nm),theaccumulationmode(typicalD range
p
week-day and one representative weekend-day for use dur- is90nmto1-2.5mm),andthecoarseaerosolmode(typical
ing May 2010. In so doing, it is implicitly assumed that D rangeis1-2.5to10mm)[BinkowskiandRoselle,2003].
p
emissions did not change appreciably from 2008 to 2010. Each mode is defined by its geometric standard deviation,
Thisassumptionmaybeinquestionduetotherecentemis- geometricmeandiameter,andthemagnitudeofmasswithin
sions control programs, such as the diesel truck rules and themode.Allparticlesareassumedtobesphericalandeach
ocean going vessel (OGV) fuel regulations (http://www. modeisassumedtobeinternallymixedchemically.Aerosol
arb.ca.gov/msprog/operators.htm). As a result, emissions of processes such as evaporation, condensation, coagulation,
BC and sulfur containing compounds may have decreased nucleation, advection, and wet and dry deposition affect
significantly from 2008 to 2010 [Dallmann et al., 2011; the total number of particles, total surface area, and total
Lack et al., 2011]. The CARB recognizes this limitation mass within each mode. The majority of primary PM
2.5
andisworkingtowardreleasinganewerversionoftheemis- emissions (99.9% by mass), including BC, are assigned to
sion inventory. Daily total emission rates for each species theaccumulationmode,andasmallfraction(0.1%bymass)
are given in Tables S4 and S5, and the emission inventory is assigned to the Aitken mode according to section 1.3 of
usedinthisstudyisavailablefromtheauthorsuponrequest. BinkowskiandRoselle[2003].
1779
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
[9] ThethermodynamicmodelISORROPIA-II[Fountoukis [11] Since the nested GEOS-Chem horizontal grid resolu-
andNenes,2007]isusedinCMAQtocomputethethermody- tion of 0.5∘ latitude x 0.667∘ longitude ((cid:1)55 km x (cid:1)60 km
namicequilibriumoftheNHþ-Na+-SO2(cid:3)-NO(cid:3)-Cl(cid:3)-H Oaero- at latitude 33∘N) is considerably coarser than the 4 km x 4
4 4 3 2
sol system. The assumption of thermodynamic equilibrium kmresolutionofCMAQ,andthedomainshowninFigure1
betweenfineinorganicparticulatenitrateandammoniumwith isrelativelysmallcomparedtotheNorthAmericancontinent,
gas-phase nitric acid and ammonia is commonly invoked in thepotentialdouble-countingofspeciesintheL.A.Basinwas
atmospheric CTMs. The validity of this assumption for fine taken into consideration. Coarse resolution acts to smooth
particles(D typically<2.5mm)wasconfirmedonthebasis concentration gradients via dilution, thereby reducing peak
p
ofdataobtainedduringthe1999AtlantaSupersiteExperiment valueswithintheregionandincreasingspeciesconcentrations
[Zhangetal.,2002].Karydisetal.[2010]usedthePMCAMx attheboundaries.Thesespeciescanthenpartiallyreenterthe
model with ISORROPIA-II to model inorganic aerosol L.A. Basin as boundary conditions, while simultaneously
measured during the Mexico City MILAGRO campaign. beingemittedwithintheL.A.BasinviatheARCTAS-CARB
They concluded that explicitly treating mass transfer to emissioninventory,whichisphysicallyunrealistic.However,
and from coarse aerosol as a dynamic process is essential itispossibleforspeciesemittedintheL.A.Basintobetrans-
forcapturingthecompetitionbetweensmallandlargeparti- ported outside the domain (Figure 1), and then reenter via
cles for condensible inorganic vapors. To account for this recirculation. To determine the impacts of both of these
competition, CMAQ partitions mass between the gas and potential issues, additional sensitivity simulations have been
aerosol phases according to the hybrid method [Capaldo conducted with CMAQ using two sets of boundary condi-
et al., 2000], inwhichinstantaneousequilibriumisassumed tions: (1) one set is derived from a nested GEOS-Chem
between the gas and aerosol phases in the two fine modes, simulation over North America that includes emissions in
anddynamicmasstransfergovernsthecoarseaerosolmode. the L.A. Basin, and (2) one set is derived from the same
SinceISORROPIA-IIiscapableofsimulatingaerosolsystems nested GEOS-Chem simulation over North America with
thatincludeK+-Ca2+-Mg2+,weconduct anadditionalsensi- emissionsinthedomainshowninFigure1settozero(the
tivitysimulation on the impact of dust emissions and crustal lattersetwasusedintheresultstobepresented).Theresults
speciesoninorganicaerosolconcentrationsintheL.A.Basin. (notshown)indicatethattheimpactofturningoffemissions
intheL.A.BasininthenestedGEOS-Chemsimulationhas
virtually no impact on black carbon concentrations, and
2.2. GEOS-Chem onlyaveryslightimpactontheboundaryinflowofsulfate.
[10] Dynamicchemicalboundaryconditions(1-htemporal Thereforerecirculationandthepotentialdouble-countingof
resolution) used in the CMAQ simulations were generated species are not issues in the current model configuration,
from the GEOS-Chem global chemical transport model although this may not be true for modeling configurations
(version9-01-01,http://acmg.seas.harvard.edu/geos/)viaone- withdifferentdomainsizesanddifferentgrid-cellsizes,orif
way nesting. GEOS-Chem was used to simulate global gas- therearesignificantemissionssourcesneartheboundariesof
∘ ∘
andaerosol-phaseconcentrationsat2 latitudex2.5 longitude thenesteddomain.
horizontalresolution,with47verticallevels.Boundarycondi-
2.3. FLEXPART
tions were extracted from the global simulation and used to
drive a nested GEOS-Chem simulation over North America [12] Totracetheoriginsofmeasuredandpredictedspecies
at 0.5∘ latitude x 0.667∘ longitude horizontal resolution, with during CalNex, a modified version of the FLEXPART
47verticallevels.Finally,CMAQ-consistentboundarycondi- Lagrangian particle dispersion model [Stohl et al., 2005] is
tionswerethenextractedfromthenestedGEOS-Chemsimula- usedtocalculatebacktrajectoriesofairmasses.FLEXPART
tionsforthedomainshowninFigure1.Allthreedomainsare hasbeenusedextensivelytoquantifytheimpactsofmeteoro-
shownonasinglemapinFigureS2.ThecoarseGEOS-Chem logical processes on pollution transport (e.g., de Foy et al.
parentgridsimulationspin-upperiodwas1year,andthespin- [2006]; Palau et al. [2006]; Ding et al. [2009]; Brioude
upperiodforthenestedGEOS-Chemgridsimulationwas4 etal.[2009]).AdetaileddescriptionoftheFLEXPARTmodel
months (Jan-Apr 2010). Goddard Earth Observing System usedinthisstudycanbefoundintheSupplementaryMaterial.
(GEOS-5)assimilatedmeteorologicaldatafromtheNASA
GlobalModelingandAssimilationOffice(GMAO)wereused 3. Observations
for all GEOS-Chem model simulations, which included
3.1. PasadenaGround-Site Data
ozone-NO -hydrocarbonchemistry[Beyetal.,2001]coupled
x
withsulfate-nitrate-ammoniumaerosolchemistry[Parketal., [13] Planetaryboundarylayer(PBL)heightsweremeasured
2004]. Outside North America, global emissions used in the atthePasadenagroundsitewithaVaisalaCeilometerCL31,
coarse simulations are from the Emissions Database for whichusestheminimum-gradientmethodtodetermineaero-
Global Atmospheric Research (EDGAR) inventory [Olivier sol backscatter profiles [Emeis and Schafer, 2006; Schafer
andBerdowski,2001].Anthropogenicemissionsdataforthe etal.,2004].ObservedPBLheightsarecomparedtopredicted
United States, used in GEOS-Chem, were from the EPA PBLheightsinFigure2.Previousstudieshaveshownoverall
National Emissions Inventory (NEI) 2005, scaled to the agreementbetweenthistechniqueascomparedtoradiosonde
simulation period according to trends in the EPA Acid Rain and sonic detection and ranging estimates of PBL height
Program (http://camddataandmaps.epa.gov/gdm/) and the [Haman, 2011; van der Kamp et al., 2010; Martucci et al.,
NEI Air Pollutant Emissions Trends Data (http://www.epa. 2007;Münkeletal.,2006].Forthisstudy,theaveragePBLheight
gov/ttn/chief/trends/). The nested GEOS-Chem simulation uncertainty and the minimum detection limit are (cid:4)5m and
also includes aircraft, biofuel and natural emissions of inor- (cid:1)80m,respectively.Adetaileddescriptionoftheinstruments
ganicaerosolprecursors,asdescribedbyParketal.[2004]. andsettingsusedinthisstudycanbefoundinHaman[2011].
1780
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
3500 by measuring the incandescence signal emitted from single
ve m)3000 BC-containing particles heated to their boiling point when
aboel (2500 passing through an intense Nd:YAG laser beam (l =1064
ht,ev2000 nm).BCvolume-equivalentdiameter(VED)iscalculatedfrom
BL Heigground l11050000 t1h.8edgetcemct(cid:3)ed3m[BaosnsdasasnudmBinegrgastsrpohmer,ic2a0l0p6a]r.tiTchleewSiPth2dinecnasnitdyeosf-
P 500
cence channels were calibrated in the same way as described
0
15 22 29 in Section 2 of the Supplemental Material to McMeeking
310 et al. [2010], with the main difference being that Alfa Aesar
K) 305 glassycarbonsphereswereusedinsteadofAquadag.Thedata
e ( 300 have been corrected for mass above and below the detection
atur 295 limit of the instrument by fitting a log-normal distribution to
per 290 the primary mode in the BC mass distribution and another
em log-normaldistributiontotheresidualofthisfit(i.e.asecond-
T 285
arymode).Thetwofitswerethenaddedtogether,andthefrac-
28015 22 29 tion of the fits above and below the SP2 detection limit were
100 calculated as 8.0% of BC mass below the detection limit and
1.9%above.ThemeasuredBCmassconcentrationshavebeen
80
divided by (1- 0.099). Based on Shiraiwa et al. [2008] and
%) 60 Schwarz et al. [2008b], the uncertainty in the determination
RH ( 40 of the mass of a single BC particle measured at the Pasadena
ground site is estimated to be (cid:1)30%. All BC measurements
20
fromthePasadenagroundsiteareshowninFigure3.
015 22 29 [15] Inorganic aerosol measurements were made by the
University of Colorado-Boulder Aerodyne high-resolution
Figure 2. Observed (black) and predicted (red) planetary time-of-flight aerosol mass spectrometer (Aerodyne HR-
boundary layer (PBL) heights, temperature, and relative ToF-AMS, Aerodyne Research, Inc., Billerica, MA USA,
humidity (RH) at the Pasadena ground site from 15 May– [DeCarloetal.,2008])attheCalNexPasadenagroundsite.
31May 2010. The inlet is designed to transmit particles with vacuum-
aerodynamic diameters (D ) of 60 to 600 nm with unit
va
[14] Refractory black carbon (BC) aerosol mass was mea- transmission efficiency, although particles with Dva above
suredatthePasadenaground-sitewithaDropletMeasurement 600 nm are also detected [Canagaratna et al., 2007;
Technologies(DMT,Boulder,CO,USA)SingleParticleSoot DeCarloetal.,2004].Acomposition-dependentcollection
Photometer(SP2).Briefly,theSP2detectsrefractoryBCmass efficiency (CE) was applied to the AMS data based on
2
Pasadena ground−site observed
Pasadena ground−site predicted
Mean Observed
1.8 Mean Predicted
1.6
−3m)1.4
g
µ
n (
o1.2
ati
ntr
ce 1
n
o
C
n
o0.8
b
ar
C
ck 0.6
a
Bl
0.4
0.2
0
05/19 05/20 05/21 05/22 05/23 05/24 05/25 05/26 05/27 05/28 05/29 05/30 05/31
Local Time (month/day)
Figure3. Measured(blackdots)andpredicted(reddots)BCconcentrationsatthePasadenagroundsite
from 19May –31 May 2010.
1781
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
recent work [Middlebrook et al., 2012]. The composition describedbyRobertsetal.[2010].Datawereacquiredevery
based method for collection efficiency addresses the issue 10seconds,andwereaveragedtooneminute.Thecalibrations
of particle bounce in the AMS and that particle bounce is wereperformedwithpermeationtubescalibratedasdescribed
a function of particle phase. The technique presented by by[Neumanetal.,2002].Thetimeconstantfortransmission
Middlebrook et al. [2012] encompasses the four main of HNO through the inlet was found to be several minutes.
3
factors influencing particle phase: relative humidity in the The overalluncertaintyofthe HNO measurementwas 34%
3
sampling line, acidity/neutralization of the sulfate content, +0.05ppbv.SO mixingratiosweremeasuredatthePasadena
2
ammonium nitrate content, and organic liquid content. For site with a commercial pulsed fluorescence detector (Model
thisdataset,therewereseveralinstanceswhereammonium 43i-TL,ThermoElectronCorp)operatedasdescribedinLuke
nitratedominatedtheaerosolmass,andasshowninFigure3 [1997]. All inorganic species measurements from the
in Middlebrook et al. [2012], when the ammonium nitrate PasadenagroundsiteareshowninFigure4.Hourlymeasure-
fractionexceeds 0.4,the CEfor the aerosol increases from mentsofNO andSO atthreegroundsitesintheL.A.Basin
x 2
0.45 to 1. Using a constant CE value during these periods arealsoreportedbytheCARBgroundnetwork(http://www.
would cause an overprediction of aerosol mass. The arb.ca.gov/aqmis2/aqmis2.php)andshowninFigure5.
ground-site AMS measurements are reported as 5-min
averages and have an uncertainty <30%. NH was 3.2. CIRPASTwin Otter
3
measuredat1Hzbyquantumcascadetunableinfraredlaser [16] BC aerosol mass was measured onboard the Twin
differential absorption spectroscopy (QC-TILDAS from Otter aircraft with a Droplet Measurement Technologies
Aerodyne Inc.) with an overall uncertainty during the (DMT,Boulder,CO,USA)SingleParticleSootPhotometer
CalNexcampaignof10%+0.42ppbv.[Ellisetal.,2010]. (SP2).ThemajorfindingsfromthisSP2duringCalNex,in-
HNO measurementsweremadewiththeacetateionCIMS cludingcalibrationoftheinstrument,aredetailedelsewhere
3
Boundaries Aq,O3 Aq,H2O2 Aq,O2(FE,MN) Gas,OH Primary SO42−
7 10
6
8
3) 5
−g m 4 ppb) 6
2−µO (43 (SO24
S 2
2
1
0 0
05/19 05/25 05/31 05/17 05/23 05/29
14 14
12 12
−3)10 b)10
µg m 8 (pp38
−NO (346 HNO 46
2 2
0 0
05/15 05/21 05/27 05/21 05/27
8 15
7
6
−3m) 5 pb)10
+µg H (434 NH (p3
N 5
2
1
0 0
05/21 05/27 05/23 05/29
Local Time (month/day)
Figure4. Observed(black)andpredicted(red)particulatesulfate,nitrate,ammonium,sulfurdioxide,nitric
acid,andammoniaconcentrationsfromtheCalNexPasadenagroundsite.Inthelegend,“Boundaries"refers
to sulfate attributable to boundary conditions, “(Aq,Gas),OX" refers to secondary sulfate produced by
aqueous-phase(Aq)orgas-phase(Gas)oxidationofSO byoxidantOX.“PrimarySO2(cid:3)"referstosulfate
2 4
emittedwithinthebasin.
1782
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
34.4
34.2
°e [] 34
d
atitu33.8
L
33.6
L.A. Westchester
L.A. North Main Street
33.4 North Long Beach
−120 −119.5 −119 −118.5 −118 −117.5 −117
Longitude [°]
North Long Beach North Long Beach
40
100 30
20
50
10
0 0
05/02 05/12 05/22 05/02 05/12 05/22
bv) L.A. North Main Street bv) L.A. North Main Street
p p
p p
(OX150 O (2 15
N 100 S 10
50 5
0 0
05/02 05/12 05/22 05/02 05/12 05/22
L.A. Westchester L.A. Westchester
30
100
20
50
10
0 0
05/02 05/12 05/22 05/02 05/12 05/22
Date (month/day)
Figure5. Measured(black)andpredicted(red)NO andSO mixingratiosforMay2010atthreeloca-
x 2
tionsintheLosAngelesBasin.GaseousmeasurementsweretakenfromtheAirQualityandMeteorolog-
ical Information System (AQMIS, http://www.arb.ca.gov/aqmis2/aqmis2.php).
[Metcalfetal.,2012].InMetcalfetal.[2012],1-minaver- sub-micron, non-refractory, size-resolved aerosol composi-
agedataarereported,butinthisstudy,wehavere-sampled tion.TheinletandcollectionefficiencyoftheAMSonboard
the dataset at 1 Hz to take advantage of the highest time- the Twin Otter were similar to those of the AMS at the
resolution available. To account for the BC mass outside Pasadena ground site. In an effort to measure aerosol mass
the SP2 detection range, a single log-normal function is fit distributions, the AMS onboard the Twin Otter was periodi-
to each 1-s histogram of single-particle BC mass between cally run in particle-time-of-flight (PToF) mode. Due to
0.48 and 290 fg (50-675 nm VED, assuming a spherical relatively low aerosol loadings, the signal-to-noise ratio was
particle density of 1.8 g cm(cid:3)3), and integrated to give bulk not of sufficient quality for a meaningful comparison to size
BCmassconcentrations.UnliketheSP2measurementsatthe distributionspredictedbyCMAQ.Instead,wefocusonbulk
Pasadena ground site, a single log-normal mode is sufficient particulateammonium(NHþ),particulatenitrate(NO(cid:3)),and
4 3
to adjust the measured BC size distributions for mass above particulate sulfate (SO2(cid:3)) mass concentrations reported as
4
andbelowtheTwinOtterSP2detectionlimits[Metcalfetal., 10-saverages.Wenotethatduetotheattemptedsize-resolved
2012]. As discussed in Metcalf et al. [2012], this adjustment measurements,narrowplumesmayhavebeenmissedbythe
increases bulk BC mass concentrations by 15–20%. Based AMSonboardtheTwinOtter.
on the calibration standards available, uncertainty in single- [18] All Twin Otter measurements reported here are from
particleBCmassdeterminationandbulkmassconcentrations instruments inside an unpressurized cabin. All instruments
isestimatedtobe(cid:1)40%. sampled downstream of a two-stage diffusion inlet with a
[17] Non-refractory particle mass and composition mea- transmission efficiency near unity for particle diameters up
surementsweremadebyanAerodynecompacttime-of-flight to about 3.5 mm [Hegg et al., 2005]. Sampling lines inside
aerosolmassspectrometer(C-ToF-AMS,AerodyneResearch, the cabin are kept reasonably uniform to all instruments, so
Inc., Billerica, MA USA) [Drewnick et al., 2005; Murphy further corrections for diffusional losses in these lines have
et al., 2009]. The AMS onboard the Twin Otter measures not been made. The Twin Otter conducted 18 research
1783
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
flights from Ontario, CA between 4 May and 28 May 2010 (cid:129) linear decrease in transmission vs ln(D ), from 100% at
va
duringCalNex.TheAMSwasonboardduring8oftheflights, D =550 nm to 0% at D =2mm.
va va
three of whichweretoSanJoaquinValley,whichisoutside (cid:129) 0% transmission above D = 2mm;
va
thedomainshowninFigure1.Therefore,thisanalysismakes
use of 5 Twin Otter flights during which inorganic aerosol [21] Thistransmissionfunctionisanaverageofthetrans-
concentrationsweremeasuredwithintheL.A.Basin. missioncurvesusedinseveralAMSstudiesasdescribedin
Knoteet al.[2011] andreferences therein, and isapplied to
allinorganicaerosolpredictions(CMAQ)thatarecompared
3.3. NOAA P3
to AMS measurements from the Pasadena site, the Twin
[19] The P3 aircraft conducted 18 research flights from Otteraircraft,andtheP3aircraft.Wenotethatduetovaria-
Ontario,CAbetween4Mayand20June2010duringCalNex tionbetweenspecificaerodynamiclensesindifferentinstru-
2010 (esrl.noaa.gov/csd/calnex/). This study uses daytime ments,theexacttransmissionefficienciesoftheAMSatthe
measurementsfrom 5P3flights thatfocusedonsamplingL.
Pasadena site, onboard the Twin Otter, and onboard the P3
A.Basinemissionsandtheresultingphotochemicalproducts.
will be slightly different than the one used in this study.
NH ,HNO ,NHþ,NO(cid:3),andSO2(cid:3),andvariousmeteorolog-
3 3 4 3 4 Appendix B presents the equations used to modify CMAQ
ical parameters were measured onboard the P3 aircraft. NH
3 predictions based on the AMS transmission window and
wasmeasuredat1Hz(equivalentto100mspatialresolution) efficiencies, and a derivation of these equations is given in
bychemicalionizationmassspectrometry(CIMS)withtypi-
theSupplementalMaterial.MeasuredBCconcentrationsfrom
cal uncertainties of (30% + 0.2 ppbv) and a 1s imprecision
all platforms, adjusted for mass above and below the SP2
of0.08ppbv[Nowaketal.,2010].HNO3wasmeasuredat1 detection limits, are compared directly to the unmodified
Hz by a separate CIMS instrument with an uncertainty of
predictedBCconcentrationsinallresultstobepresented.
(15% + 0.040 ppbv) and a 1s imprecision of 0.012 ppbv
[Neuman et al., 2012]. CO measurements were made by a
vacuum ultraviolet fluorescence instrument with (cid:4)5% 4. Results and Discussion
uncertainty and 1 ppbv imprecision [Holloway et al., 2000]. [22] TheAerosolModelingTestbedanalysis toolkit [Fast
SO2(cid:3), NHþ, and NO(cid:3) were measured from a pressure- et al., 2011] was used to map three-dimensional CMAQ
4 4 3
controlled region downstream of a low turbulence inlet meteorological parameters, predicted gas-phase concentra-
using a compact time-of-flight aerosol mass spectrometer tions, and predicted aerosol-phase concentrations onto each
(Aerodyne, Billerica,Massachusetts)[Bahreinietal.,2009]. flightpath(aswellasforvariousgroundsites).Thetemporal
The AMS data are reported as 10-s averages with 2s uncer- resolutionoftheCARBemissioninventoryandMM5meteo-
tainty (1-s imprecision) of 34% (0.06 mg m(cid:3)3), 34% (0.01 rologyis1h.Inthiswork,CMAQpredictedspeciesconcen-
mg m(cid:3)3), and 36% (0.01 mg m(cid:3)3) for ammonium, nitrate, trationfields,averagedovertheprevioushour,arecompared
andsulfate,respectively.BCmeasurementsinthesizerange to observations according to observational time-stamps (i.e.
of95-720nmVED(stillassumingaBCvoid-freedensityof observational points with time-stamps of 12:20 and 12:40
1.8gcm(cid:3)3)weremadebyanSP2similartothatusedonboard would both be compared to predictions averaged between
the TwinOtter [Schwarzet al.,2006].As withthe SP2mea- thehoursof12:00-13:00).Thediscrepancybetweenmeasured
surements onboard the Twin Otter, a single log-normal func- andsimulatedspeciesconcentrationfieldsandmeteorological
tionissufficienttoaccountformassaboveandbelowtheP3 parametersisquantitativelyassessedusingthefollowingfour
SP2 detection limits, and this adjustment increases bulk BC statisticalmetrics:
mass concentrations by 10–25%. As described in detail in
Metcalf et al. [2012], due to differences in calibration, mea- 1XN
ME¼ jP (cid:3)Mjðmean errorÞ (1)
surements from the SP2 onboard the Twin Otter are poten- N i i
tially biased low by (cid:1)12%, as compared to those from the i¼1
SP2 onboard the P3 (e.g. if the P3 measurement is 1 mg 1XN
m(cid:3)3,theTwinOttermeasurementwouldbe0.88mgm(cid:3)3). MB¼N ðPi(cid:3)MiÞðmean biasÞ (2)
i¼1
[20] TheAMSsusedinthisstudy,bothground-basedand X
airborne,measuredonlyparticleswithvacuumaerodynamic N jP (cid:3)Mj
diameters between 60 nm and 600 nm diameter with 100% NME¼ Xi¼1 i i ðnormalized mean errorÞ (3)
N
efficiency. Particles with aerodynamic diameters above 600 i¼1Mi
nm were also measured, but with reduced collection effi- X
N ðP (cid:3)MÞ
ciency. However, since the predicted Aitken,accumulation, NMB¼ Xi¼1 i i ðnormalized mean biasÞ (4)
and coarse aerosol modes are expressed in CMAQ in terms N M
of log-normal functions, each mode is defined for particle i¼1 i
diametersrangingfromzerotoinfinity.Therefore,allaerosol whereN,P,andMstandforthenumberofdatapoints,pre-
predictions are adjusted to match the transmission efficiency dicted quantity, and measured quantity, respectively.
oftheAMSbasedonthefollowingpiece-wisedefinedtrans-
missionfunction: 4.1. Meteorological Variables
4.1.1. GroundSite
(cid:129) 0% transmission below Dva =40nm; [23] Observed and predicted temperature, relative humid-
(cid:129) linearincreaseintransmissionvsln(D ),from0%atD ity (RH), and planetary boundary layer (PBL) height from
va va
=40 nm to 100% at D =100 nm; thePasadenagroundsiteareshowninFigure2andstatisti-
va
(cid:129) 100%transmissionfromD =100nmuptoD =550nm; cal metrics are given in Table 1. Overall, the agreement
va va
1784
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
betweenobservedandpredictedPBLheightinPasadenahas observed and predicted temperature has a 15-day average
∘
a 15-day average bias of -80 m (-9%). Agreement between bias of 1.06 C. The agreement between observed and pre-
dicted RH has a 15-day average bias of -10.3%. The discre-
panciesinRHaremostlikelyacombinationofunderpredicted
Table1. StatisticalMetricsBasedonMeasurementsandPredictions water vapor mixingratios andof theexponentialdependence
atthePasadenaGroundSiteDuringMay2010 of saturation-vapor pressure on errors in temperature (1∘C
temperatureerrorleadstoapproximately5%RHerror).Addi-
Parameter N ME MB NME NMB
tional ground-site comparisons of predicted and observed
PBLHeight(m) 1179 129 -80 0.14 -0.09 meteorology are given in the Supplemental Material. The
Temperature(k) 24697 1.88 1.06
resultsshowthattemperatureandRHareconsistentlyoverpre-
RH(%) 24697 11.6 -10.3
BC(mgm(cid:3)3) 3918 0.19 0.03 0.48 0.07 dicted and underpredicted, respectively, during the first week
SO2(cid:3)(mgm(cid:3)3) 1860 0.93 -0.36 0.44 -0.17 of May, with much better agreement during the last three
NNOH4(cid:3)3þ((mmggmm(cid:3)(cid:3)33)) 11886600 21..1057 --10..5706 00..7325 --00..5205 weeks. Predicted w∘ind speeds and wind directions agree to
4 within2m/sand60 ,respectively,atallsurfacesites.
SO (ppbv) 22491 0.72 0.64 2.50 2.01
2 4.1.2. TwinOtter and P3
HNO (ppbv) 22761 0.98 -0.38 0.85 -0.38
NH3(3ppbv) 366 1.51 0.45 0.74 0.22 [24] The Twin Otter and P3 flight paths and altitudes are
shown in Figures 6–11 and Appendix A Figures A1, A2,
ME=MeanError,MB=MeanBias,NME=NormalizedMeanError,
A3, A4. Temperature and relative humidity (RH) measure-
NMB=NormalizedMeanBias.Nisthenumberofdatapointscollected
duringMay2010. ments from the Twin Otter and P3 flights are compared to
May 24
35 4000 1 1
n
°Latitude[]343.54 11. 15 1 315.25 1311 24 1..554 Altitude a.s.l. (m)123000000000 mmonium Fractio0000....6248 Nitrate Fraction0000....2486
A
33.5
0 0 0
−119 −L1o1n8git−u1d1e7[°]−116 11:00 12:00 13:00 14:00 15:00 11:00 12:00 13:00 14:00 15:00 11:00 12:00 13:00 14:00 15:00
2 1.5 4 6
3µSulfate (g/m)10..155 3µBlack Carbon (g/m) 0.15 3µAmmonium (g/m)123 3µNitrate (g/m)12345
0 0 0 0
11:00 12:00 13:00 14:00 15:00 11:00 12:00 13:00 14:00 15:00 11:00 12:00 13:00 14:00 15:00 11:00 12:00 13:00 14:00 15:00
2
e3m) 21%
ulfatµg/1.5
Predicted Sontribution (0.15 58% 2%7%11%1%
C
0
11:00 12:00 13:00 14:00 15:00
Boundaries Aq,O3 Aq,H2O2 Aq,O2(FE,MN) Gas,OH Primary SO42−
Figure 6. From left to right and top to bottom: Twin Otter aircraft flight path for May 24, Twin Otter
altitudes(withrespecttosealevel)withtheflighttrackandaltitudetracearecoloredbythetime(Pacific
Standard Time) of day and time-stamps printed along each flight path in 30 min increments, Fraction of
predicted particulate ammonium within the AMS transmission window, Fraction of predicted particulate
nitratewithintheAMStransmissionwindow,predicted(red)andobserved(black)sulfateconcentrations,
predicted(red)andobserved(black)blackcarbonconcentrations,predicted(red)andobserved(black)ni-
trate concentrations, predicted (red) and observed (black) ammonium concentrations, predicted sulfate
source apportionment, Pie chart indicating the relative contribution from routes to sulfate averaged over
a given flight. In the bottom legend, “Boundaries" refers to sulfate attributable to boundary conditions,
“(Aq,Gas),OX"referstosecondarysulfateproducedbyaqueous-phase(Aq)orgas-phase(Gas)oxidation
of SO by oxidant OX.“Primary SO2(cid:3)"refers to sulfate emitted within the basin.
2 4
1785
ENSBERGETAL.:CALNEXINORGANICANDBCAEROSOL
May 25
35 2000 1 1
n
°Latitude[] 343.54 1111215 . 52.15.415 3 1. 5 31514.5Altitude a.s.l. (m)11505000000 mmonium Fractio0000....2846 Nitrate Fraction0000....4826
A
33.5
0 0 0
−119 L−1o1n8gitu−d1e17[°]−116 12:00 13:00 14:00 15:00 12:00 13:00 14:00 15:00 12:00 13:00 14:00 15:00
2 1.5 4 8
3m) 3)
3µSulfate (g/m)01..155 µck Carbon (g/0.15 µmmonium (g/m123 3µNitrate (g/m)264
Bla A
0 0 0 0
12:00 13:00 14:00 15:00 12:00 13:00 14:00 15:00 12:00 13:00 14:00 15:00 12:00 13:00 14:00 15:00
2
ulfate3µ)g/m1.5 24%
d Son ( 1 46%
Predicteontributi0.5 3%15%
C 4%8%
0
12:00 13:00 14:00 15:00
Boundaries Aq,O3 Aq,H2O2 Aq,O2(FE,MN) Gas,OH Primary SO24−
Figure 7. Same as Figure 6,but for the Twin Otter May 25 flight.
MM5predictionsinTable2.Observedandpredictedtempera- good,witha CalNex-average NMEandNMBof47.8%and
tures(averagedovereachflight)typicallyagreetowithin4∘C, 6.6%,respectively(Table1).ThisNMEof47.8%isslightly
whileobservedandpredictedRH(averagedovereachflight) higher than the measurement error inherent to the SP2
typicallyagreetowithin15%.Zhangetal.[2006]evaluated ((cid:1)30–40%)owingmostlytolargemissingpeaksinpredicted
MM5 meteorology coupled with CMAQ against measure- BConspecificdays(e.g.27and28May).Becausepredicted
ments taken during the Southern Oxidants Study, and found BCischemicallyinert,thePasadenaground-siteisfixed,and
similar levels of agreement between predicted and observed allweekdayemissionsareassumedtobeidentical,day-to-day
temperatureandRH.Inthepresentstudy,thelargestdisagree- variations in BC predictions at the Pasadena ground site can
mentoccurredduring the 21MayTwinOtter flightandthe only be caused byvariations in the predicted meteorological
14 May P3 flight, during which predicted RH was biased fields (e.g. wind fields). For instance, May 20 and May 27
low by 26.6% and 22.1%, respectively. Wind speed and were both Thursdays. However, the 1-h average BC predic-
wind direction measurements onboard the Twin Otter and tionson27Maydidnotgetabove0.4mg(cid:3)3,whileBCpredic-
the P3 are shown in Table 3. Predicted and observed wind tionson20Maywereupto1.1mg(cid:3)3.SincethepredictedPBL
speeds generally agree to within (cid:1)50% when averaged over heightsonthesedayswerecomparable(Figure2),variationin
eachflight.Althoughthemeandiscrepancybetweenpredicted thepredictedwindfieldsistheprimarycauseoftheday-to-day
∘
and observed wind direction ranges between 22-65 , the variation in predicted BC concentrations. Therefore, if the
average predicted and observed wind direction is that of the differences between the predicted and observed wind fields
daytime southwesterly sea breeze which advects emissions onanygivendayarecomparabletotheday-to-daydifferences
towards the north and northeast, exiting the Basin through inthepredictedwindfields,largeerrorsinBCpredictionsmay
passesintheSanGabrielandSanBernardinomountainranges occuratanygivenpoint(e.g.Pasadenagroundsite).However,
[Lu and Turco, 1995]. However, it is difficult to quantify theoverallagreementbetweenpredictedandobservedBCat
errors in species concentrations attributable to discrepancies the Pasadena ground site (NMB = 6.6%) suggests that on
betweenmeasuredandobservedwindvectors. average, both wind fields and upwind BC sources are repre-
sentedwellbyCMAQ.
4.2. Black Carbon
4.2.2. TwinOtter and P3
4.2.1. Pasadena Ground Site [26] Measured and predicted BC concentrations from the
[25] MeasuredandpredictedBCconcentrationsatthePasa- Twin Otter flights and the P3 flights during May 2010 are
denagroundsiteareshowninFigure3.Theoverallagreement shown in Figures 6–11 and Figures A1, A2, A3, A4. The
between predicted and observed BC concentrations is very illusory differences in noise levels of the Twin Otter BC
1786