Transcript PPT - Atmospheric Chemistry Modeling Group

Global Simulation of Secondary Organic
Carbon Aerosols
Hong Liao
California Institute of Technology
GEOS-CHEM meeting, April 2005
Why is Secondary Organic Aerosol (SOA)
Important?
 It is the least well-quantified and understood aerosol in
the atmosphere;
 It may contribute to a large fraction of organic carbon,
which in turn constitutes a large fraction of PM2.5 that
causes haze and adverse health effects;
 It influences climate directly through light scattering
or indirectly by serving as cloud condensation nuclei.
SOA Formation
Gas-phase reaction:
HCi  OX j   i, j ,1Pi, j ,1   i, j ,2 Pi, j ,2    
HCi  parent hydrocarbon
OX j  oxidant (O3, OH, or NO3)
 i, j , k  mass-based stoichiometric coefficient
Pi , j , k  semi-volatile product that partitions between
gas and aerosol phases
SOA Simulation in GEOS-CHEM
 5 reactive biogenic hydrocarbon groups:
HC1 = -pinene, b-pinene, sabinene and terpenoid ketones, 3D carene
HC2 = limonene
HC3 = -terpinene, -terpinene, terpinolene
HC4 = myrcene, terpenoid alcohols, ocimene
HC5 = sesquiterpenes
 28 organic products from O3, OH and NO3 oxidation :
6 (3 gases + 3 aerosols) from each of first four HC groups = 24
4 (2 gases + 2 aerosols) from oxidation of sesquiterpenes = 4
 9 tracers:
3 classes of biogenic VOCs: HC1, HC2, and HC4
SOG1 = lump of gas products from HC1, HC2, HC3 oxidation
SOG2 = gas product from HC4 oxidation
SOG3 = gas product from HC5 oxidation
SOA1 = lump of aerosol products from HC1,HC2, and HC3 oxidation
SOA2 = aerosol product from HC4 oxidation
SOA3 = aerosol product from HC5 oxidation
SOA Simulation in GEOS-CHEM (cont.)
 Concentrations of OH, O3, and NO3 from offline
fields or online simulation
 SOA Thermodynamic Equilibrium
[A]
k
[G ]i, j , k  K i , j ,M
om , j , j , k
o
[G ]i, j , k  gas-phase concentration (µg of gas per m3 of air)
[A ]i, j , k  aerosol-phase concentration (µg of aerosol per m3 of air)
K om,i , j , k  partition coefficient (µg-1 m3)
M o  [POA] 
[A]i, j,k
i, j , k
[POA]  POA concentration (µg of aerosol per m3 of air)
SOA Simulation in GEOS-CHEM (cont.)
 Dry deposition
Dry deposition of HCs, gas-phase products, and SOAs
follows the scheme in GEOS-CHEM.
 Wet deposition
80% of SOA dissolves into clouds [Limbeck and Puxbaum, 2000]
 Emission inventories
Emissions of monoterpenes are based on the work of Guenther
et al. [1995] and treated as a function of vegetation type, leaf
area index, and temperature;
Emissions of other reactive VOCs (ORVOCs) are also based
on Guenther et al. [1995], but are monthly fields from Global
Emissions Inventory Activities (GEIA).
Emissions of monoterpenes and ORVOCs are distributed into
each HC group following the study of Griffin et al. [1999].
Predicted SOA Concentrations (ng m-3)
GEOS-CHEM (offline)
GEOS-CHEM (online)
Unified Model (online)
Predicted Primary Organic Aerosol (POA) Concentrations (ng m-3)
GEOS-CHEM
Unified Model
Emission Inventories for POA
GEOS-CHEM
Unified Model
Emission (TgC/yr)
Fossil fuel
Biofuel
Biomass burning
Total
10.2
7.5
22.4
40.1
53.0
81.2
Burden (TgOM)
0.69
1.29
28.2
Parent Hydrocarbon Contributions to Global SOA
GEOS-CHEM (offline)
HC1
HC2
Unified Model (online)
52.0%
78.7%
HC3
20.0%
0.7%
HC4
10.3%
10.3%
HC5
11.0%
17.0%
0.16
0.23
SOA Burden (TgOM)
Conclusions and Suggestions
 Compared to the 33-tracer SOA scheme in the unified model,
the simplified 9-tracer scheme in GEOS-CHEM predicts well the
geographical distribution of SOA.
 SOA concentrations predicted in the GEOS-CHEM model are lower
than those predicted in the unified model, which can be explained
by the differences in predicted POA concentrations.
 Further investigation of SOA predicted in the upper troposphere
is needed.
 The specified emissions of ORVOCs should be replaced by a
scheme that is similar to what the GEOS-CHEM has for
monoterpenes. Coupling ORVOC emissions with meteorological
variables is necessary for studying the effect of climate change on
SOA formation.
Acknowledgements
U.S. EPA STAR grant
Bob Yantosca, Harvard University
Colette Heald, Harvard University
Rokjin Park, Harvard University