SourceSource--Oriented Chemical Oriented Chemical Transport Model for Primary and Transport Model for Primary and

Secondary Organic AerosolSecondary Organic AerosolMike KleemanMike Kleeman

Civil and Environmental EngineeringCivil and Environmental EngineeringUC DavisUC Davis

Rob Griffin,Rob Griffin,EOS/Earth SciencesEOS/Earth Sciences

University of New HampshireUniversity of New Hampshire

Simon CleggSimon CleggUniversity of East AngliaUniversity of East Anglia

June 2007June 2007

RD831082

Unique Tracers For Source Unique Tracers For Source Apportionment of Particulate CarbonApportionment of Particulate Carbon

Unique Tracer Source Comment

LevoglucosanMethoxyphenolsBetulinJuvabione, DehydrojuvabionDehydroabietic acid

General Wood CombustionGeneral Wood CombustionPaper BirchBalsam FirConifers

Emissions rate depends on combustion conditions.

Hopanes and Steranes Lubricating oil in Gasoline-and Diesel powered engines

Does not distinguish between gasoline and diesel vehicles.

Isoprenoids and Tricyclic Terpanes Gasoline-powered Motor Vehicle Exhaust

Trace amounts present in diesel vehicle exhaust.

C31 HopanesDivanillyl1,2-divanillylethane

Coal

CholoesterolAcyl Monoglycerides

Grilling/Charring Meat Emission rate depends on the cooking method: charbroiling or frying.

1,6-anhydro-2-acetamido-2- deoxyglucose

Grilling/Charring Crustacean Seafood

High molecular weight, odd carbon number n-alkanes

Leaf Abrasion

Iso- and anteiso-alkanes Tobacco Smoke

Key Questions:Key Questions:Are tracers modified in the atmosphere?Are tracers modified in the atmosphere?

Effect of oxidantsEffect of oxidantsAqueous processingAqueous processing

How do we distinguish between sources that do How do we distinguish between sources that do not have unique tracers?not have unique tracers?

How do we quantify secondary organic aerosol How do we quantify secondary organic aerosol (SOA)?(SOA)?

How can we reduce the cost of source How can we reduce the cost of source apportionment?apportionment?

Project ObjectivesProject Objectives

Predict the formation of secondary organic Predict the formation of secondary organic aerosol using a stateaerosol using a state--ofof--thethe--science chemical science chemical transport modeltransport model

Identify improvements to secondary organic Identify improvements to secondary organic aerosol prediction algorithmsaerosol prediction algorithms

Determine regional source contributions to Determine regional source contributions to primary organic aerosol and secondary organic primary organic aerosol and secondary organic aerosol using a sourceaerosol using a source--oriented approachoriented approach

Summary of Project AccomplishmentsSummary of Project Accomplishments

1.1. R.J. Griffin, D. R.J. Griffin, D. DabdubDabdub, and J.H. Seinfeld, Development and initial evaluation of a dyn, and J.H. Seinfeld, Development and initial evaluation of a dynamic speciesamic species--resolved resolved model for gasmodel for gas--phase chemistry and sizephase chemistry and size--resolved gas/particle partitioning associated with secondary orgresolved gas/particle partitioning associated with secondary organicanicaerosol formation, J. aerosol formation, J. GeophysGeophys. Res., 110: D5, 2005.. Res., 110: D5, 2005.

2.2. J. Chen and R.J. Griffin*, Modeling secondary organic aerosol foJ. Chen and R.J. Griffin*, Modeling secondary organic aerosol formation from oxidation of rmation from oxidation of --pinenepinene,, --pinenepinene,,and dand d--limonene, Atmospheric Environment, 39: 7731limonene, Atmospheric Environment, 39: 7731--7744, 2005.7744, 2005.

3.3. T. Held, Q. Ying, M.J. T. Held, Q. Ying, M.J. KleemanKleeman, J.J. , J.J. SchauerSchauer, M.P. Fraser, A comparison of the UCD/CIT air quality model and, M.P. Fraser, A comparison of the UCD/CIT air quality model and the CMB sourcethe CMB source--receptor model for primary airborne particulate matter, Atmosphereceptor model for primary airborne particulate matter, Atmospheric Environment, 39: 2281ric Environment, 39: 2281--2297, 2297,2005. (supported jointly by EPA project #RD831082 and CARB proje2005. (supported jointly by EPA project #RD831082 and CARB project #2000 ct #2000 –– 05PM).05PM).

4.4. Q. Ying, T. Held, M.J. Q. Ying, T. Held, M.J. KleemanKleeman, Source contributions to the regional distribution of secondary, Source contributions to the regional distribution of secondary particulate matter in particulate matter in California, Atmospheric Environment, 40: 736California, Atmospheric Environment, 40: 736--752, 2006. (supported jointly by EPA project #RD831082 and CARB 752, 2006. (supported jointly by EPA project #RD831082 and CARB project #2000 project #2000 –– 05PM).05PM).

5.5. S.S. VutukuruVutukuru, R.J. Griffin, and D. , R.J. Griffin, and D. DabdubDabdub, Simulation and analysis of secondary organic aerosol dynamics , Simulation and analysis of secondary organic aerosol dynamics in the in the South Coast Air Basin of California, J. South Coast Air Basin of California, J. GeophysGeophys. Res., 111, D10S12, . Res., 111, D10S12, doidoi: 10.1029/2005JD006139, 2006.: 10.1029/2005JD006139, 2006.

6.6. J. Chen, H. Mao, R.W. Talbot, and R.J. Griffin, Application of tJ. Chen, H. Mao, R.W. Talbot, and R.J. Griffin, Application of the CACM and MPMPO modules using the CMAQ he CACM and MPMPO modules using the CMAQ model for the Eastern United States, J. model for the Eastern United States, J. GeophysGeophys. Res., 111, D23S25, . Res., 111, D23S25, doidoi: 10.1029/2006JD007603, 2006.: 10.1029/2006JD007603, 2006.

7.7. Q. Ying, M.P. Fraser, J. Chen, R.J. Griffin and M.J. Q. Ying, M.P. Fraser, J. Chen, R.J. Griffin and M.J. KleemanKleeman. Verification of a source. Verification of a source--oriented externally mixed air oriented externally mixed air quality model during a severe photochemical smog episode. Atmosquality model during a severe photochemical smog episode. Atmospheric Environment (7): 1521pheric Environment (7): 1521--1538 MAR 2007.1538 MAR 2007.

8.8. M.J.M.J. KleemanKleeman, Q. Ying, M.J. , Q. Ying, M.J. MysliwiecMysliwiec, R.J. Griffin and J. Chen. Source apportionment of secondary o, R.J. Griffin and J. Chen. Source apportionment of secondary organic rganic aerosol during a severe photochemical smog episode. Atmosphericaerosol during a severe photochemical smog episode. Atmospheric Environment 41, 576Environment 41, 576--591, 2007.591, 2007.

9.9. S.L. Clegg, M.J. S.L. Clegg, M.J. KleemanKleeman, R.J. Griffin, and J.H. Seinfeld. Effects of uncertainties in t, R.J. Griffin, and J.H. Seinfeld. Effects of uncertainties in the thermodynamic properties he thermodynamic properties of aerosol components in an air quality model. I. Treatment of iof aerosol components in an air quality model. I. Treatment of inorganic electrolytes and organic compounds in the norganic electrolytes and organic compounds in the condensed phase, and the results of an atmospheric simulation. Acondensed phase, and the results of an atmospheric simulation. Atmospheric Chemistry and Physics Discussions, tmospheric Chemistry and Physics Discussions, submitted for publication, 2007.submitted for publication, 2007.

10.10. S.L. Clegg, M.J. S.L. Clegg, M.J. KleemanKleeman, R.J. Griffin, and J.H. Seinfeld. Effects of uncertainties in , R.J. Griffin, and J.H. Seinfeld. Effects of uncertainties in the thermodynamic properties the thermodynamic properties of aerosol components in an air quality model. II. Predictions oof aerosol components in an air quality model. II. Predictions of pure component f pure component vapourvapour pressures of organic pressures of organic compounds. Atmospheric Chemistry and Physics Discussions, submitcompounds. Atmospheric Chemistry and Physics Discussions, submitted for publication, 2007.ted for publication, 2007.

SourceSource--Oriented Model for Primary Oriented Model for Primary and Secondary Organic Aerosoland Secondary Organic Aerosol

Track the primary and secondary organic Track the primary and secondary organic species through a mathematical simulation of species through a mathematical simulation of emissions, transport, chemical reaction, phaseemissions, transport, chemical reaction, phase--change, and deposition.change, and deposition.

Keep particles emitted from different sources Keep particles emitted from different sources separateseparate

Realistic simulation of atmospheric processing for Realistic simulation of atmospheric processing for organic speciesorganic species

Source apportionment toolSource apportionment tool

Air Quality ModelAir Quality Model

Transport

FogProcessing

ChemicalReactions

Photo-chemistry

DepositionGas-PhaseEmissions

AerosolEmissions

Condensation&

Evaporation

Particles of each size, source, and age are tracked separately

Crustal Material Other than Paved Road Dust Paved Road Dust Diesel Engines Meat Cooking

The Source Oriented External Mixture Model The Source Oriented External Mixture Model Tracks Particles Separately in the AtmosphereTracks Particles Separately in the Atmosphere

Non-cat Gas Engines

OtherSources

Acidic IndustrialCat Gas Engines

The Source Oriented External Mixture Model The Source Oriented External Mixture Model Tracks Particles Separately in the AtmosphereTracks Particles Separately in the Atmosphere

External MixtureExternal Mixture

Atmospheric Transformation ProcessesAtmospheric Transformation ProcessesFog Chemistry:-based on model of Jacob et al. (1989)-58 active chemical species-177 kinetic reactions-29 equilibrium relationships

Small Aerosol Water Content:-based on AIM model of Wexler et al. (1991)-kinetic exchange between gas and particle phase-particle phase species in equilibrium with one another

Secondary Organics:-based on model of Griffin et al. (2005)

-kinetic gas - particle exchange-temperature dependence for vapor pressure-activity coefficient calculations

Improvements for SOA ModelImprovements for SOA Model

Improvements to gasImprovements to gas--phase chemical phase chemical mechanism describing production of semimechanism describing production of semi--volatile speciesvolatile species

Partitioning into aqueous and organic Partitioning into aqueous and organic particle phasesparticle phases

Activity coefficient calculations for organicsActivity coefficient calculations for organics

SOA Species

HydrophobicHydrophilic

Biogenic Anthropogenic

Dissociative,low C# (1)

Dissociative,high C# (2)

Non-dissociative (3)

Dissociative (4)

Non-dissociative (5)

Biogenic (10) Anthropogenic

Polyaromatic (8)

Aromatic

High volatility (7)

Low volatility (6)

Aliphatic (9)

Original SOA Surrogate Species Classifications

SOA Model Adaptations

•Chemical changes (stoichiometry and kinetics) to better match observed product distribution, to provide a distinction between high- and low-SOA-yield compounds, and to better match temporal behavior of SOA formation in chambers

•Vapor pressures recalculated using structure activity relationships

•Certain very reactive species are no longer included in partitioning calculations

•NET: Expected decrease in SOA formation, more in-line with other methods based on chamber parameterizations

SOA Species Low C# (1)High C#

Polyaromatic (8)

Aromatic

Alkane-derived(9)

High volatility (7)

Low volatility (6)

BiogenicAnthropogenic

Non-dissociative (3)

Dissociative (2)

Aromaticfragments Non-

dissociative (5)

Dissociative (4)

Ring-retaining(new)

New SOA Surrogate Species Classifications

September 7September 7--9, 1993 9, 1993 SoCABSoCAB Study Study DomainDomain

Model Performance for OzoneModel Performance for Ozone

Ozo

ne

(pp

m)

Model Performance for Model Performance for NOxNOxN

Ox

(pp

m)

Model Performance for Individual Model Performance for Individual Organic Compound ClassesOrganic Compound Classes

PH

EN

(pp

m)

BA

LD

(pp

m)

AR

OH

(pp

m)

AR

OL

(pp

m)

Model Performance for Carbonaceous AerosolModel Performance for Carbonaceous Aerosol

LGBH

LGBH

CELA

CELA

AZUS

AZUS

CLAR

CLAR

Sensitivity of SOA to VOC EmissionsSensitivity of SOA to VOC Emissions

NOx Scaling Factor0.5 1.0 1.00.5

VO

C S

calin

g F

acto

r

0.5

0.5

1.0

1.0

Estimated Boiling Points For Surrogate Estimated Boiling Points For Surrogate Compounds Using Different MethodsCompounds Using Different Methods

Source: S. L. Clegg,1 M. J. Kleeman,2 R. J. Griffin,3 and J. H. Seinfeld, Effects of uncertainties in the thermodynamic properties of aerosol components in an air quality model. II. Predictions of pure component vapour pressures of organic compounds, Atmospheric Chemistry and Physics, 2007

METHODS:

1 – Nannoolal et al. (2004);

2 – Cordes and Rarey (2002);

3 – Kolovanov and Petrauskas(undated), and ACDLabs software v8.0 (Advanced Chemistry Development Inc., 2004);

4 - Stein and Brown (1994);

5 – Joback and Reid (1987);

6 – Wen and Qiang (2002a,b);

7 – Constantinou and Gani (1994);

8 - Marrero-Morejon and Pardillo-Fontdevila (1999).

Effect of Boiling Point Uncertainty Effect of Boiling Point Uncertainty on Predicted SOAon Predicted SOA

Source Apportionment of Secondary Particulate Matter

SOA SourceSOA Source--Oriented ClassesOriented Classes

BIOL

RO224

RO225

RO226 UR17 RPR3

AP7

RO256

PAN8UR6

UR5 UR3

BIOH

UR8

RO227

RO228 UR7

AP8

RO229 RO240 RPR8

RO257 PAN9

UR23

RO212 PN10

ISOP

RO210

MCR RO253

RO254 RP16

RO258 UR28

ALKH RO223 AP11 UR34

RO271 AP12

UR20

PAH

RAD7 RO238 RO247 RP15 UR27

UR15 RP14 RP19 UR31

UR11

RO231 AP10

UR19

AROHPHEN

AROLRO221

RAD2

RO217

AP1

RPR2 UR2

BALD ARAC

RAD5

RO222

RO223

RAD6

RPR5 UR14

AP6

RPR6 RPR7

ADAC

RAD4

RAD3

RO235

RO234

RP11

UR26

RO233 RPR9 RP17

UR29

AP5

RO236 RP12

RO237 RP13

RP18

UR30

Emissions of SOA PrecursorsEmissions of SOA Precursors

SOA Concentrations By Precursor FamilySOA Concentrations By Precursor Family

So

urc

e C

on

trib

uti

on

s to

SO

A

Diurnal Variation of SOA Source Diurnal Variation of SOA Source ContributionsContributions

ConclusionsConclusions

SOA calculations currently underSOA calculations currently under--predictpredictOC concentrations during typical air OC concentrations during typical air pollution episodes in Californiapollution episodes in California

Incorrect parameterization of surrogate Incorrect parameterization of surrogate compoundscompoundsMissing entire formation pathways (Missing entire formation pathways (oligomeroligomerformation)formation)

Known sources of SOA in central Los Known sources of SOA in central Los Angeles are dominated by transportationAngeles are dominated by transportation

Catalyst and nonCatalyst and non--catalyst equipped gasoline catalyst equipped gasoline enginesengines

ConclusionsConclusions

Biogenic sources make a surprisingly Biogenic sources make a surprisingly significant contribution to SOA in at the significant contribution to SOA in at the north and south ends of the north and south ends of the SoCABSoCAB

Water soluble materialWater soluble material

Highest concentrations at nightHighest concentrations at night

SourceSource--oriented methodology can be oriented methodology can be applied anywhere a chemical transport applied anywhere a chemical transport model is usedmodel is used

AcknowledgementsAcknowledgements

US Environmental Protection Agency Science US Environmental Protection Agency Science To Achieve Results Grant # RD831082To Achieve Results Grant # RD831082

San Joaquin San Joaquin ValleywideValleywide Air Pollution Study Air Pollution Study AgencyAgency

California Air Resources Board (Karen California Air Resources Board (Karen MaglianoMagliano,, AjithAjith KaduwelaKaduwela, Vernon Hughes), Vernon Hughes)

ChevronChevron

Tony Held, Tony Held, QiQi Ying, and Jin LuYing, and Jin Lu