Jana Milford
Department of Mechanical Engineering
University of Colorado, Boulder
Atmospheric Science Progress Review
Meeting
Atmospheric Aerosols from Biogenic Hydrocarbon Oxidation
June 21-22, 2007
AcknowledgmentsAcknowledgmentsTanarit Sakulyanontvittaya,* CUTanarit Sakulyanontvittaya,* CU--Boulder, Mechanical EngineeringBoulder, Mechanical EngineeringDetlev Helmig,* Tiffany Duhl, Ryan Daly, John Ortega, David TannDetlev Helmig,* Tiffany Duhl, Ryan Daly, John Ortega, David Tanner, er, CUCU--Boulder, Institute for Arctic and Alpine ResearchBoulder, Institute for Arctic and Alpine ResearchChristine Wiedinmyer,* Alex Guenther,* Peter Harley,* John OrlanChristine Wiedinmyer,* Alex Guenther,* Peter Harley,* John Orlando, do, Louisa Emmons, Louisa Emmons, SouSou MatsunakaMatsunaka, National Center for Atmospheric , National Center for Atmospheric ResearchResearchBob Bob ArntsArnts, Chris , Chris GeronGeron, Jeff Herrick, J. Harvey, U.S. Environmental , Jeff Herrick, J. Harvey, U.S. Environmental Protection AgencyProtection AgencyMark Mark PotosnakPotosnak, Desert Research Institute, Desert Research InstituteBill Stockwell, Howard UniversityBill Stockwell, Howard UniversityGreg Greg YarwoodYarwood, Environ, EnvironRaoulRaoul ZavariZavari, Pacific Northwest National Laboratory, Pacific Northwest National LaboratoryTed Russell, Helena Park, Georgia TechTed Russell, Helena Park, Georgia TechGail Gail TonnesonTonneson, University of California at Riverside, University of California at RiversideNSF, Atmospheric Chemistry Program (ATM #0304704) and the U.S. NSF, Atmospheric Chemistry Program (ATM #0304704) and the U.S. EPA (#RDEPA (#RD--8310790183107901--0). 0).
* Coauthor* Coauthor
OutlineOutline
BackgroundBackgroundBiogenic hydrocarbon emissions and secondary Biogenic hydrocarbon emissions and secondary organic aerosol (SOA) formation organic aerosol (SOA) formation Sesquiterpenes (SQTs) and Sesquiterpenes (SQTs) and monoterpenesmonoterpenes (MTs)(MTs)
Project objectivesProject objectivesMethodsMethods
Chemical transport modeling Chemical transport modeling MeasurementsMeasurementsEmissions modelingEmissions modeling
ResultsResultsEmissions comparisonsEmissions comparisons
ConclusionsConclusions
Sesquiterpene (SQT) and Monoterpene (MT) Sesquiterpene (SQT) and Monoterpene (MT) Emissions from VegetationEmissions from Vegetation
Significant emissionsSignificant emissionsNorth America total Monoterpene North America total Monoterpene emissions 17.9 Tg C yremissions 17.9 Tg C yr--11
(Guenther et al., 2000)(Guenther et al., 2000)Sesquiterpene emissions? Sesquiterpene emissions?
Highly reactiveHighly reactiveOxidation products can partition to Oxidation products can partition to the aerosol phasethe aerosol phase
MonoterpenesMonoterpenes CC1010 HH1616Sesquiterpenes Sesquiterpenes CC1515 HH2424
α-pinene
Parent terpenoidParent terpenoid Aerosol Yield (%)Aerosol Yield (%)
ΔΔ33--CareneCarene 2.3 2.3 –– 10.910.9
ββ--CaryophylleneCaryophyllene 17.2 17.2 –– 62.562.5
αα--HumuleneHumulene 20.0 20.0 –– 66.766.7
LimoneneLimonene 6.1 6.1 –– 22.822.8
MyrceneMyrcene 7.6 7.6 –– 12.712.7
αα--PinenePinene 2.4 2.4 –– 7.87.8
ββ--PinenePinene 4.2 4.2 –– 13.013.0
SabineneSabinene 4.7 4.7 –– 10.610.6
Griffin et al., 1999
Aerosol yields for biogenic MTs and SQTsAerosol yields for biogenic MTs and SQTs
What are the regional landscape fluxes of MTs and SQTs?What are the regional landscape fluxes of MTs and SQTs?Environmental controlsEnvironmental controlsSpatial and temporal variationsSpatial and temporal variations
What is the contribution of BVOC oxidation to SOA What is the contribution of BVOC oxidation to SOA formation in the eastern U.S.?formation in the eastern U.S.?
Diurnal and seasonal trendsDiurnal and seasonal trendsDifferences in the contributions from MTs and SQTsDifferences in the contributions from MTs and SQTs
How sensitive is secondary aerosol formation from BVOC How sensitive is secondary aerosol formation from BVOC to anticipated changes in:to anticipated changes in:
Process model assumptions?Process model assumptions?Emissions of nitrogen oxides?Emissions of nitrogen oxides?Land cover?Land cover?
Key QuestionsKey Questions
Regional Chemical Transport ModelingRegional Chemical Transport Modeling
MM5/CMAQMM5/CMAQDomain ResolutionDomain Resolution
Horizontal: 36 km x 36 Horizontal: 36 km x 36 kmkmVertical layers: 9 Vertical layers: 9
Chemical MechanismChemical MechanismSAPRC99 with 3rd SAPRC99 with 3rd generation aerosol generation aerosol model and aqueous model and aqueous chemistrychemistry
EpisodesEpisodesJuly 2001July 2001January 2002January 2002
CMAQCMAQ ModificationsModifications
Photooxidation
Advection
Emissions
Deposition
Chemical Reactions
Aerosol Formation
•SAPRC99 mechanism with aqueous chemistry and aerosol module•BCARL, AHUMUL, SQSTwith O3 , NO3 , and OH
•Aerosol Yields•Partitioning Parameters•Aerosol Properties•BCARL, AHUMUL and SQST
•New SQT emissions•MEGAN model
IncorporateSQTs into
otherprocesses
Dry Deposition Wet Deposition
CMAQ InputsCMAQ Inputs
Initial Conditions
Meteorological Data – MM5
Anthropogenic Emissions DataSMOKE 2.0 (U.S.) - NEI 1999
•Last hour output
T. Russell and Sun-Kyoung Park (GA. Tech)
•January 2002•July 2001Area, Point, Mobile, Nonroad, and Point sources
•January 2002•July 2001
T. Russell and Sun-Kyoung Park (GA. Tech)
Boundary Conditions•MOZART2.2 output
SMOKE 2.1 - Mexico (1999), Canada (1996)
•January 2002•July 2001Area, Point, Mobile, Nonroad, and Point sources
Louisa Emmons
Model Evaluation: Focus on Eastern U.S. Model Evaluation: Focus on Eastern U.S. Supersites Supersites
Atlanta, Baltimore, NY, Pittsburgh, St. LouisAtlanta, Baltimore, NY, Pittsburgh, St. Louis
intensive periodsintensive periodsJuly 2001, January 2002 July 2001, January 2002
IMPROVE IMPROVE 24 h avg PM2.5, SO24 h avg PM2.5, SO44
==, NO, NO33--, OC, EC, OC, EC
SEARCH SEARCH urban/rural pairs in AL, FL, GA, MIurban/rural pairs in AL, FL, GA, MICC--14 data at three sites14 data at three sites
TVA CTVA C--14 data (Look Rock, TN)14 data (Look Rock, TN)
Biogenic Emissions Inventory DevelopmentBiogenic Emissions Inventory Development
MEGANMEGAN
0200400600800
100012001400
0 5 10 15 20 25 30 35 40
MM5MM5Meteorological
Data
Emission Data
Emission Model
Guenther, A.
Helmig et al., and Harley, P.
Land Use Land Use
& Cover& Cover
Measurement of SQT and MT emissionsMeasurement of SQT and MT emissions
Bag and cuvette Bag and cuvette enclosure systemsenclosure systemsCalibration systemCalibration system
Helmig et al, 2003Helmig et al, 2003
Cartridge and onCartridge and on--line line samplingsamplingGCGC--MS, GCMS, GC--PTRMS, PTRMS, GCGC--FIDFIDLaboratory and field Laboratory and field measurementsmeasurements
Branch enclosure measurements at Duke Branch enclosure measurements at Duke Forest (summer Forest (summer –– fall 2004)fall 2004)
Loblolly pineLoblolly pineFour FEB Teflon film branch enclosure systems Four FEB Teflon film branch enclosure systems operated simultaneouslyoperated simultaneously
Two towerTwo towerTwo groundTwo ground--levellevel
Double ozone scrubDouble ozone scrub--bingbing in all experimentsin all experiments
Aromatic doping used Aromatic doping used to test recoveriesto test recoveries
Possible 5Possible 5--10% wall 10% wall loss for SQTloss for SQT
isop
ropy
l-ben
zene
6.72
0
250
500
13.6
5
18.5
6
23.4
3
α-pi
nene
7.35
Retention Time (min)
FID
resp
onse
(mV
)
sabi
nene
8.56
9.30
β-m
yrce
ne
1,8-
cine
ole
limon
ene
10.2
410
.34
cis-
linal
oolo
xide
trans
-lina
lool
oxid
elin
alol
tri-is
opro
pyl-b
enze
ne
11.4
911
.92
12.3
6
tetra
-hyd
ro-n
apth
alen
e
nony
l-ben
zene
α-te
rpin
eol
14.6
5
tetra
deca
neβ-
cary
ophy
llene
β-fa
rnes
ene
γ-ca
dine
nege
rmac
rene
-D
20.2
520
.35
21.2
220
.35
22.2
922
.50
δ-ca
dine
ne
α-hu
mul
ene
21.0
5
*
*
* *
O
OH
O
OO
H
HH
OH
OH
Chromatogram (plotted as the flame ionization detector (FID) response) from a ponderosa pine emission sample. Monoterpene retention times 7.2–14.7 min; sesquiterpene retention times 18.9-22.5 min. Shaded peaks are the aromatic compounds from the reference standard.
Helmig et al., ES&T, 41:1545, 2007
1.
β-Caryophyllene
2. α/β- Bergamotene
3. β-Farnescene
4. α-Farnescene
5. α-Humulene
6. α-Muurolene
7. Germacrene
D
8. Δ-Cadinene
9. β-Selinene
10. γ-Cadinene
Helmig et al., in progress Atkinson and Arey, 2003, Griffin et al., 1999
Prominent Sesquiterpenes in Recent Measurements
Beta Caryophyllene Emissions Rate v. TemperatureLoblolly Pine, Duke Forest
ER = 1.16e0.17T
R2 = 0.87
0
200
400
600
800
1000
1200
15 20 25 30 35 40 45Temperature (oC)
Em
issi
on ra
te (n
gC g
-1 h
r-1)
Sesquiterpene (SQT) emission rate (ER) data from an enclosure experiment on a loblolly pine tree at Duke Forest showing total SQT emission rates plotted against the mean needle temperature inside the enclosure. Helmig et al., ES&T, 41:1545, 2007
Adsorbent Cartridge
Field GC
Field GC, Nighttimeexperiment
BVOC Emissions Modeling: MEGANBVOC Emissions Modeling: MEGANModel of Emissions of Gases and Aerosols from Nature: Model of Emissions of Gases and Aerosols from Nature: MEGANMEGAN
1 km resolution1 km resolutionImproved evaluation of LAI and Land Cover inputsImproved evaluation of LAI and Land Cover inputsAvailable through the NCAR Community Data PortalAvailable through the NCAR Community Data Portal
ργγγε ••••= SMageCEEM TPLAICE γγγγ ••=
EM: Emission ( μg m-2 hr-1)ε: Emission Factor ( μg m-2 hr-1)ρ: Loss and Production within plant canopyγCE : Canopy Factorγage : Leaf Age FactorγSM : Soil Moisture FactorγLAI : Leaf Area Index FactorγP : PPFD Emission Activity Factor (light-dependence)γT : Temperature Response Factor
Guenther, 2006
MEGAN v2.0MEGAN v2.0Meteorological Data
TemperatureSolar Radiation
YX
Grid Information
Land Cover Data
4 PFT categories-Needle leaf
-Broad leaf-shrub-Bush
-Grass-HerbFractions of plants in grid cells
Emissions Data
Emission FactorsLAI
Speciation &Mechanism Conversion
1) CBMZ
2) SAPRC993) RACM 4) RADM2
MEGANMEGAN•Linux REDHAT WS•FORTAN PGF90•IOAPI 3.0•NETCDF 3.5.1
γCE
γage
γSM
γLAI
γP
γT
ASCII format
I/O API -
NETCDF
I/O API -
NETCDF
MEGAN v. BEIS3MEGAN v. BEIS3
Additional emission activity algorithmsAdditional emission activity algorithmsSensible heat flux, leaf age, long term effects of Sensible heat flux, leaf age, long term effects of temperature and PARtemperature and PAR
Simplified canopy model to account for leaf Simplified canopy model to account for leaf temperature and canopy light extinctiontemperature and canopy light extinctionUpdated emissions factors Updated emissions factors
Includes speciated SQT and MT emissions from Includes speciated SQT and MT emissions from measurementsmeasurementsEF for individual chemical species vary spatiallyEF for individual chemical species vary spatially
Multiple options for Multiple options for landcoverlandcover inputs including high inputs including high resolution satellite data (MODIS, SPOT)resolution satellite data (MODIS, SPOT)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Gray "D
igger"
Pine
Scots
pine
Dougla
s Fir
Ponde
rosa P
ine
Ponde
rosa P
ine
Wes
tern R
ed C
edar
Redwoo
dDou
glas F
irSho
rtleaf
PineBea
ch P
ineRed
wood
Grand F
ir
Baby B
lue S
pruce
Whit
e Pine
Whit
e Pine
Whit
e Pine
Red P
ineRed
Pine
Balsam
FirRed
Pine
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
white p
ine @
tower(
shad
e)
Loblo
lly pi
ne #
1
Loblo
lly pi
ne #
2A
Loblo
lly pi
ne #
2B
Loblo
lly pi
ne - (
CU grou
nd)
Loblo
lly pi
ne - (
EPA grou
nd)
Loblo
lly pi
ne - (
Tower
1)
Loblo
lly pi
ne - (
Tower
2)Mon
del P
ineMon
del P
ineJu
niper
Junip
erPiny
on P
inePine
eJu
niper
Blue S
pruce
Blue S
pruce
Easter
n Whit
e Pine
Easter
n Whit
e Pine
Easter
n Whit
e Pine
Basal SQT and MT Emissions Rates for Needle Leaf Treesμg
C/g
dw-h
rμg
C/g
dw-h
r
MTP ER
SQT ER
Helmig et al., 2007, Matsunaka, Potosnak et al.
SQT and MT Emissions Rates Summer 2006
0.00
0.50
1.00
1.50
2.00
2.50
Pin Oak
#5Pin
Oak #6
Cottonle
ss C
otton
wood #1
Cottonle
ss C
otton
wood #2
Narrow Le
af Cott
onwood
#1
Narrow Le
af Cott
onwood
#2
Blue Spruc
e #1
Blue Spruc
e #2
Easter
n Whit
e Pine
#1
Easter
n Whit
e Pine
#2
Easter
n Whit
e Pine
#3
Octber
Glory R
ed M
aple
#1
Octber
Glory R
ed M
aple
#2Silve
r map
le
Sunflo
wer #1
Sunflo
wer #2
Sunflo
wer #3
Sunflo
wer #4
Corn #1
Corn #2
Corn#3
Corn #4
Bean P
lant #
1
Bean P
lant #
2Whea
t #1
Wheat #
2Tob
acco #
1Tob
acco #
2μ
g C
g(d
w)-1
h-1
MT
SQT
Helmig and Daly, 2007
Creekside NurseryJune – Aug. 2006
NCAR GreenhouseAug. 2006
Participants Affiliation Sample Collection Analysis
Detlev Helmig CU/INSTAAR On-line On-line Tenax
adsorbent
GC-MSGC-FIDGC-FID
Peter Harley NCAR Tenax
adsorbent GC-FID
Alex Guenther NCAR On-line GC-MS
Thomas Karl NCAR On-line PTR-MS
Jim Greenberg NCAR On-line O3 Reactivity
Sou
Matsunaga NCAR Super Q GC-FID
Tiffany Duhl NCAR Super Q GC-FID
Monica Madronich NCAR Super Q GC-FID
Nicole Bouvier-Brown UC Berkeley SPME Fibers GC-MS
Rei
Rasmussen Oregon Health & Science Univ.
Tenax
adsorbentCanisters
GC-MSGC-MS; GC-FID
Chris GeronBob Arnts
USEPA Tenax
adsorbent GC-MS
Hannele
Hakola Finnish Meteor.Institute
Tenax
adsorbent GC-FID
NCAR SQT Measurement Comparison, April 30 - May 4, 2007
Courtesy of P. Harley
Emission Factors for 4 PFTs
0.050.0
100.0150.0200.0250.0300.0350.0400.0450.0500.0
MEGAN2.0-06b SQT MEGAN2.0-L SQT
ug/m2-hr
BroadleafNeedle LeafShrub-BushGrass-Crop
Comparison of SQT Emission Factors from RecentMeasurements v. Prior Literature
Recent data Literature through 2004
Emission Factors for 4 PFTs
0.0100.0200.0300.0400.0500.0600.0700.0800.0900.0
1000.0
MEGAN2.0-06b SQT MEGAN2.0-L SQT MEGAN2.0 MTP
ug/m2-hr
BroadleafNeedle LeafShrub-BushGrass-Crop
SQT and MT Emission Factors by Plant Functional Type
Recent data Literature through 2004 Recent data
Max 465.8 g/km2-hr Max 154.4 g/km2-hr Max 154.4 g/km2-hr
Emissions Modeling Results -- JanuaryBEIS 3.0 MEGAN2.0-06b MEGAN2.0-L
Emissions Modeling Results – JulyBEIS 3.0 MEGAN2.0-06b MEGAN2.0-L
SOA Precursor Speciation for SAPRC99-S, July Average Emissions - CMAQ Domain
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
BEIS3.0 M EGAN2.0-06b M EGAN2.0-L
TRP1
SSQT
AHUM UL
BCARL
CRES
ARO2
ARO1
OLE2
ALK5
Emissions Modeling Results
ConclusionsConclusionsSQT emissions are highly variableSQT emissions are highly variable
Emissions likely dependent on leaf age and other environmental Emissions likely dependent on leaf age and other environmental variablesvariablesSeasonal dependence is uncertain but maybe importantSeasonal dependence is uncertain but maybe important
Measured SQTs appear to have stronger temperature dependency Measured SQTs appear to have stronger temperature dependency than MT emissionsthan MT emissionsLightLight--dependency observed in some dependency observed in some MTsMTs and and SQTsSQTsSome crops appear to be strong SQT emitters Some crops appear to be strong SQT emitters –– need more need more measurementsmeasurementsMEGAN provides an easily adaptable framework for BVOC emissions MEGAN provides an easily adaptable framework for BVOC emissions estimationestimationSpeciation schemes available for most popular chemical mechanismSpeciation schemes available for most popular chemical mechanismssSQT estimated to contribute 7 SQT estimated to contribute 7 –– 16% of SOA precursor emissions 16% of SOA precursor emissions (anthropogenic and biogenic, excluding isoprene) for continental(anthropogenic and biogenic, excluding isoprene) for continental U.S. U.S. in Julyin JulySQT estimated to contribute 1 SQT estimated to contribute 1 –– 2% of SOA precursor emissions in 2% of SOA precursor emissions in JanuaryJanuarySOA contributions to be determined!SOA contributions to be determined!
DisclaimerDisclaimer
Although the research described in this presentation has Although the research described in this presentation has been funded in part by the United States Environmental been funded in part by the United States Environmental Protection Agency, it has not been subjected to the Protection Agency, it has not been subjected to the AgencyAgency’’s required peer and policy review and therefore s required peer and policy review and therefore does not necessarily reflect the views of the Agency and does not necessarily reflect the views of the Agency and no official endorsement should be inferred.no official endorsement should be inferred.