Greg Carmichael

Intercontinental Transport of Pollutants Out &
Into Asia (emphasis on particles)
Image D. Anderson,
NASA, from Seinfeld et
al., BAMS, in press, 2003
Intercontinental Transport of Pollutants Out &
Into Asia
• What do we know about the transport
mechanisms?
• How good are the models?
• What are the sources of uncertainty?
• What do the observations tell us?
• What are some next steps?
NASA-Seawifs
The CFORS forecast (upper left) of the two dust systems are shown above. The dust plume (pink) represents the
region with dust concentrations greater than 200 mgrams/m3. White indicates clouds. The SeaWifs satellite image
(upper right) also clearly shows the accumulation of dust spiraling into the Low Pressure center. Also note the strong
outflow of dust in the warm sector “ahead” of the front over the Japan Sea. The two systems are clearly seen in the
satellite derived TOMS-AI (aerosol index) (lower right). The dust event is clearly seen in the China SEPA air pollution
Transport Mechanisms: As informed by field experiments &
models (e.g., Trace-P, Ace Asia, ITCT-2k2/Peace, ICARTT,
ABC)
• Convection
• Warm conveyor belt
lifting
• Post-frontal boundary
layer transport
• Low level pre-frontal
• Advection in the
westerlies
• Cold front subsidence
• Large-scale subsidence
• Mountain wave
subsidence
Liang et al., JGR, 2004
• Boundary layer
transport
Example: Results from Peace/ITCT2K2
Convection 8% of time
accounts for ~35% of
outflow flux
Oshima et al.,JGR,
2004
One Model’s View: One Spring
Dust
Sulfate
BC
Transport to Asia
100°E
Wild, et al., 2004
viz. Newell and Evans [2000]
How Good Are the Models?
Model inter-comparison studies focused
on Asia: e.g., MICS-Asia, DMIP, TraceP/Ace-Asia
Comparisons of predictions with
observations
MICS Phase II Results: Concentrations agree better than deposition fluxes
Model-1
Model-2
Model-3
Model-4
Model-5
Model-6
Model-7
Model-8
Sulfate concentration in March 2001 (mg m-3)
http://www.adorc.gr.jp/adorc/mics.html
contents
DMIPS: Dust inter-comparison study
Spread in mean
vertical profiles
Uno et al., 2005
Model Intercomparison
Study (MICS) Asia:
Source/Receptor
Predictions
Receptor 7 - Oki (Japan)
100%
NW ASIA
80%
SE ASIA
http://www.adorc.gr.jp/adorc/mics.html
TAIWAN
60%
S CHINA
C-E CHINA
40%
NE CHINA
N & S KOREA
20%
JAPAN
0%
1
2
3
4
5
6
7
8
Receptor 17 - Taichung (Taiw an)
100%
NW ASIA
80%
SE ASIA
TAIWAN
60%
S CHINA
C-E CHINA
40%
NE CHINA
N & S KOREA
20%
JAPAN
0%
1
2
3
4
5
6
7
8
What are the Major Sources of Uncertainty
in the Calculation of Aerosol Export?
900%
Emissions
800%
SO2
NOx
700%
CO2
600%
CO
500%
CH4
400%
VOC
300%
BC
200%
OC
NH3
?
)
100%
0%
China
Japan
Other East Southeast
Asia
Asia
India
Other
South Asia
Ships
Streets et al., JGR, 2003
All Asia
Predictions of wet deposition are markedly
different
Removal Processes Remain
Poorly Characterized in Models
Impact of Wet Removal on
Predicted BC
Progress limited by lack of
understanding and observations
Summary of Major Sources of
Uncertainty in the Calculations
Summary of estimated relative uncertainties* for integrated aerosol
quantities (column amounts, fluxes)
*(uncertainty divided by mean value).
Emissions
Wet
Chemical Vertical
Model
removal Formation Transport Resolution
Total
nss SO4
0.3
0.3
0.3
0.5
0.1
0.7
BC
OC
Dust
Sea Salt
3
3
5
5
1
1
1
0.3
-3
---
0.5
0.5
0.5
0.5
0.1
0.1
1
1
3.2
4.3
5.2
5.13
Note: for analysis of specific points some of these
terms are larger…
How do the models perform with
respect to observations?
Sulfate
Obs
M
BC
Comparison of Predictions vs Obs for INDOEX
and Ace-Asia (Ron Brown ship data)
Sub-micron
Super-micron
Total
M
a
s
s
C
o
m
p
o
s
i
t
i
o
n
su
M bN
-s O
u 3
su bN
M bns O3
-s
ub sS
ns O4
sS
O
su 4
M bN
-s a
ub
N
su a
M bC
-s a
ub
C
su a
M bO
-s C
ub
O
su C
M bE
-s C
ub
su EC
M bN
-s H4
ub
N
H
4
Comparison of Observed and
Predicted Chemical composition
(sub-micron mode)
100
10
1
0.1
0.01
0.001
0.0001
Export of Particles: One Model’s
View -- One Spring
Models predict a larger fraction
of BC & OC (wrt sulfate) is
transported out of Asia
What do the observations tell us?
Model-based
Observation-based
40%
~40%
exported
Both approaches have
large uncertainties !!
Koike et al., JGR, 2003
Ace Asia Aerosol Column Means: Summary of
Uncertainty, Model to Model Variability, and
Predictability (NOAA CCSP, ABC)
2 Models:
MOZART
STEM
Final thoughts
 Analysis is highly uncertain due to
understanding, current state of models, inputs
and available observations.
 Presently observations are used to compare with
predictions --- good for process development,
confidence building.
 Observations by themselves can not provide the
answer -- models necessary …. but also can’t do
it alone.
 Improved understanding needed to reduce
uncertainty:
 Processes (deposition)
 Role of clouds (transport & removal)
 Emissions
Final thoughts (cont)
Enhanced measurements (systems and experimental
designs) needed to constrain the problem
Expanded Monitoring Activities Will Provide Valuable new Information
EANET
ABC
Final thoughts (cont)
 Integration of measurements and models needed…ensemble
and data assimilation (get uncertainies, inversion for emissions
and removal parameters, etc.)
McKeen et al., in
prep.
Amir
et al.,
JGR
2005