Global Modeling of Organic PM: Global Modeling of Organic PM: State of the Science and Major GapsState of the Science and Major Gaps

Daniel J. Jacobwith Rokjin J. Park1, Colette L. Heald2 Tzung-May Fu3, Hong Liao4

and funding from EPRI, EPA, NSF

1 now asst. prof. at Seoul National University2 now asst. prof. at Colorado State University3 now asst. prof. at Hong Kong Polytechnic University4 Caltech, now at Chinese Academy of Sciences

ORGANIC PM IN STANDARD GEOS-Chem MODELORGANIC PM IN STANDARD GEOS-Chem MODEL

fuel/industry open fires

OH, O3,NO3SOG SOA

POA

K

vegetation fuel/industry open fires

700

isopreneterpenesoxygenates…

30

alkenesaromaticsoxygenates…

alkanesalkenesaromatics…

VOC EMISSION PRIMARY EMISSION

VOC

5020 100

20

Global sources in Tg C y-1

secondaryformation

TOP-DOWN CONSTRAINTS ON U.S. SOURCES

IMPROVE obs (1998)

Park et al. [JGR 2003]

annual U.S. source:3.1 Tg yr-1

0.520.89

0.60

1.10

Fit GEOS-Chem sources of organic PM to monthly 1998 data at IMPROVE sites

IMPROVED MODEL WITH 1ox1o RESOLUTION,SOA FORMATION BY PANKOW/SEINFELD MECHANISM

(2001,annual)

Park et al. [AE 2006]

IMPROVE

g m-3

SENSITIVITY OF NATURAL OC PM CONCENTRATIONSTO PRE-EXISTING POA AVAILABILITY

Is SOA formation limited by availability of primaryorganic aerosol (POA)?

YES NO

Park et al. [AE 2006]

Major implications for natural visibility objective of the Regional Haze Rule

QUANTIFYING THE BIOMASS BURNING SOURCE OF PM2.5 FROM CORRELATION OF TOTAL CARBON (TC) WITH NON-SOIL K+

AT IMPROVE SITES

Non-soil K+:West: summer max,low background

East: winter and summer maxes, high background

Background ns-K+

(traces industrial biofuel)

Park et al. [AE 2007]

use TC vs. ns-K+ linear regression together w/satellite data to quantify TC from open fires

d[TC]/d[ns-K+]

FIRE AND BIOFUEL CONTRIBUTIONS TO TC PM2.5Annual mean C concentrations, 2001-2004

Mean values for W and E at top of plot

Park et al. [AE 2007]

IMPLICATIONS FOR AQ STANDARDS AND EMISSION INVENTORIES

PM2.5, g m-3

(x1.8 for TC)

West East

Summer wildfires 0.47 0.25

Other fires 0.49 0.56

Residential biofuel 0.13 0.52

Industrial biofuel 0.13 0.32

Total fire+biofuel TC (this work)

1.2 1.6

Observed TC 2.2 3.4

Observed total PM2.5

3.8 8.2

Tg C y-1

Top-down

This work

Open fires

Biofuel

Park et al. [2003]

Open fires

Biofuel

0.90

0.51 (40% industrial)

0.68

0.96 (80% industrial)

Bottom-up

Bond et al. [2004]

Open fires

Biofuel

NEI99

Biofuel

0.89

0.42 (<10% industrial)

0.23 (<10% industrial)

Source attribution at IMPROVE sites(annual means)

Emissions for contiguous U.S. (climatological estimate for fires)

Park et al. [AE 2007]

FIRST MASS CONCENTRATION MEASUREMENTSOF OC AEROSOLS IN FREE TROPOSPHERE

ACE-Asia aircraft data over Japan (April-May 2001)

Observed (Huebert)GEOS-Chem (Chung & Seinfeld for SOA)

Observed (Russell)

OC/sulfate ratio

Heald et al. [GRL 2005]

Chung and Seinfeld scheme:,T OC

VOC SOG SOA • Observations show 1-3 g m-3 background;model too low by factor 10-100

LITTLE TRANSPACIFIC INFLUENCE OF ASIAN ORGANIC PMON U.S. SURFACE AIR

Spring 2001 IMPROVE sulfate data

Days of max Asian influence (diagnosed by GEOS-Chem)

March-May mean

Days of max

Asian influence

sulfate 0.69 1.04

nitrate 0.22 0.24

OC 1.08 1.19

Mean PM concentrationsobserved in NW U.S. (g m-3)

[Heald et al., JGR 2006a]

Other studies (Randall Martin, Rodney Weber) find no OC enhancement in transpacific Asian plumes sampled from aircraft

ITCT-2K4 AIRCRAFT CAMPAIGN OVER EASTERN U.S. IN JULY-AUGUST 2004

water-soluble organic carbon (WSOC) aerosol measurementsby Rodney J. Weber (Georgia Tech)

2-6 km altitudeAlaskafireplumes

Values ~2x lower than observed in ACE-Asia; excluding fire plumesgives mean of 1.0 gC m-3 (3x lower than ACE-Asia)

Heald et al. [JGR 2006b]

MODEL OC AEROSOL SOURCES DURING ITCT-2K4

~10% yield ~2% yield

Large fires in Alaskaand NW Canada:60% of fire emissionsreleased above 2 km(pyro-convection)

Heald et al., [JGR 2006b]

ITCT-2K4 OC AEROSOL: VERTICAL PROFILES

SOx = SO2 + SO42-:

efficient scavengingduring boundary layerventilation

ObservationsModel

hydro-phobic

Data filtered againstfire plumes (solid)and unfiltered (dotted)

Model source attribution

TotalBiomass burningAnthropogenicBiogenic SOA

Heald et al. [JGR 2006b]

IRREVERSIBLE DICARBONYL UPTAKE BY AQUEOUS AEROSOLIRREVERSIBLE DICARBONYL UPTAKE BY AQUEOUS AEROSOL

Chamber AMS experiments of glyoxal uptake by Liggio et al. [JGR 2005]Organic aerosol mass growth with time Inferred reactive uptake coefficient

• median = 2.9x10-3 observed for aqueous surfaces; evidence for oligomerization• similar observed for methylglyoxal on acidic surfaces [Zhao et al. ES&T 2006]

glyoxal methylglyoxal

POSSIBLE MECHANISMS FOR DICARBONYL SOA FORMATIONPOSSIBLE MECHANISMS FOR DICARBONYL SOA FORMATION

GAS AQUEOUS

Oligomers

OHOrganic acids

H* ~ 105 M atm-1

Ervens et al. [2004]Crahan et al. [2004]Lim et al. [2005]Carlton et al. [2006, 2007]Warneck et al. [2005]Sorooshian et al. [2006, 2007]

Altieri et al. [2006, 2008]

Schweitzer et al. [1998]Kalberer et al. [2004]Liggio et al. [2005a,b] Hastings et al. [2005] Zhao et al. [2006]Loeffler et al. [2006]

glyoxal

H* ~ 103 M atm-1

methylglyoxal

oxidation

oligomeriz

ation

oli

go

mer

izat

ion

GLYOXAL/METHYLGLYOXAL FORMATION FROM ISOPRENEGLYOXAL/METHYLGLYOXAL FORMATION FROM ISOPRENE

ON

O

O

O OH

OOH

O O O

OO

HO

OO

O

O

OH

Isoprene

C5 carbonylsHydroxy-

methylvinyl ketone Methylvinyl ketone Methacrolein

Glyoxal Glycolaldehyde Methylglyoxal Hydroxyacetone

Methyl-nitroxy butenal

25% 12% 37% 26%

18%33% 32%62%

29% 51% 24%39%

87%16%

45%52%

+ OH+ NO3

+ NO + NO

16%

84%

16%

84%

2%

andisomers

0.5%

Organicnitrates

Organicnitrates

GEOS-Chem mechanism based on MCM v3.1

Fu et al. [JGR, in press]

6% 25%

molar yields

GLOBAL GLYOXAL BUDGET IN GEOS-ChemGLOBAL GLYOXAL BUDGET IN GEOS-ChemIncluding reactive uptake by aq. aerosols + clouds with =2.9x10-3 [Liggio et al., 2005]

Global SOA formation of 6.4 Tg yr-1 (1.0 in clear sky + 5.4 in cloud);compare to 16 Tg yr-1 from terpenes/isoprene by semivolatile mechanism

Fu et al. [JGR, in press]

= 2.9 h

CHOCHO

Production Emission [Tg y-1] Molar yield [%] 45 [Tg y-1] Isoprene 410 6.2 21 Acetylene 6.3 64 8.9 Glyoxal* 7.7 100 7.7 Ethylene 21 5.7 2.5 Monoterpenes 160 2.8 1.8 Benzene 4.8 25 0.9 Toluene 6.7 16 0.7 Xylenes 4.7 16 0.4 Glycolaldehyde* 5.6 9.9 0.5 Methylbutenol 9.6 5.4 0.3 Loss 45 [Tg y-1] Photolysis 28 Oxidation by OH 6.5 SOA formation 6.4 Dry deposition 2.2 Wet deposition 1.9

(biomass burning)

GLOBAL METHYLGLYOXAL BUDGET IN GEOS-ChemGLOBAL METHYLGLYOXAL BUDGET IN GEOS-ChemIncluding reactive uptake by aerosols and clouds with =2.9x10-3

Global SOA formation of 16 Tg yr-1 (2 in clear sky + 14 in cloud);compare to 16 Tg yr-1 from terpenes/isoprene by semivolatile mechanism

= 1.6 h

Fu et al. [JGR, in press]

CH3COCHO

Production Emission [Tg y-1] Molar yield [%] 140 [Tg y-1] Isoprene 410 25 110 Acetone 57 14 10 Methylglyoxal* 5.0 100 5.0 >C2 alkenes 31 7.7 4.1 Hydroxyacetone* 4.9 75 3.6 Monoterpenes 160 4.2 3.5 Propane 16 11 2.7 >C3 alkanes 26 3.2 1.0 Toluene 6.7 12 0.7 Xylenes 4.7 23 0.7 Methylbutenol 9.6 6.2 0.5 Loss 140 [Tg y-1] Photolysis 100 SOA formation 16 Oxidation by OH 15 Wet deposition 1.8 Dry deposition 1.7

(biomass burning)

0 50 100 150 200 250 300 350 400

Measured methylglyoxal [ppt]

0

50

100

150

200

250

300

350

400

Mo

del

met

hyl

gly

oxa

l [p

pt]

(b)

0 50 100 150 200

Measured glyoxal [ppt]

0

50

100

150

200

Mo

del

gly

oxa

l [p

pt]

(a)

MODEL COMPARISON TO IN SITU OBSERVATIONSMODEL COMPARISON TO IN SITU OBSERVATIONS

Glyoxal Methylglyoxal

Continental boundary layer (all northern midlatitudes summer)Continental free troposphereMarine boundary layer

Indication of a missing marine source in the modelFu et al. [JGR, in press]

SCIAMACHY SATELLITE OBSERVATION OF GLYOXALSCIAMACHY SATELLITE OBSERVATION OF GLYOXAL

• General spatial pattern reproduced over land, SCIAMACHY is 50% higher than model• SCIAMACHY sees high values over oceans correlated with chlorophyll: unidentified marine source?

100 pptv glyoxal in marine boundary layer would yield ~1 g C m-3 SOA;could contribute to observed OC aerosol concentrations in marine air

Fu et al. [JGR, in press]

SIMULATION OF WSOC AEROSOL OVER EASTERN U.S.SIMULATION OF WSOC AEROSOL OVER EASTERN U.S.Water-soluble OC (WSOC) aerosol observations by Rodney Weber (GIT)from NOAA aircraft during ICARTT campaign out of Portsmouth, NH (Jul-Aug 04)

Observed

Model w/ dicarbonyl SOA addedModel w/ standard SOA

Model hydrophilic primary OA

biomass burning plumes excluded Boundary layer data (<2 km)

Fu et al., in prep.

IMPROVE (surface) ICARTT

model w/ dicarbonyls w/out dicarbonyls

Dicarbonyl source in summer is mainly Biogenic (isoprene)

CORRELATIONS OF FREE TROPOSPHERIC WSOC CORRELATIONS OF FREE TROPOSPHERIC WSOC WITH OTHER VARIABLES MEASURED ON NOAA AIRCRAFTWITH OTHER VARIABLES MEASURED ON NOAA AIRCRAFT

ObservedModel with dicarbonyl SOAModel without dicarbonyl SOA

• WSOC is observed to correlate with• toluene and methanol (anthro+bio?)• sulfate (aqueous-phase production?)• alkyl nitrates (photochemistry?)

• Model does not reproduce observed WSOC variability but does better with correlations, particularly when dicarbonyl SOA is included (sulfate, alkyl nitrates)

Fu et al., in prep.0 0.1 0.2 0.3 0.4 0.5

Correlation coefficient r

Observed WSOC

CO

Methanol

Acetone

PAN

Sulfate

Ammonium

Benzene

Toluene

Methyl nitrate

Ethyl nitrate

Isopropyl nitrate