No name

1
Evolution of PM2.5 Components in the Long-range
Transport Plume accompanied with the Southward
Asian Continental Outflow
Mr. Ming-Tung Chuang (GIEE, NCU, Taiwan) Dr.
Joshua S-Yuan Fu (CEE, UT, USA) Dr. Carey
Ji-Cheng Jang (EPA, USA) Ms. Pei-Cheng Ni
(CTCI Cooperation, Taiwan) Dr. Chang-Chuan
Chan (CPH, NTU, Taiwan) Dr. Chung-Te Lee
(GIEE, NCU, Taiwan)
CAMS conference, UNC at Chapel Hill, October, 2007
2
Introduction

• In the past, long-range transport of air masses
  from Asian continent to Taiwan were focused on
  yellow dust event (Lin, 2001 Lin et al., 2005
  Liu et al., 2006).
• During dust transports, aerosol sulfate and
  nitrate were found rich which indicates
  pollutants were transported with yellow dust
  (Carmichael et al., 1995Nishikawa et al., 2000
  Han et al., 2004) .

3
Introduction

• Asia had been noted to emit most sulfur in the
  world due to the increasing consumption of fossil
  fuel in China (Dignon and Hameed, 1989).
• There are many successive research that found 20
  - 50 acid deposition in Pacific countries were
  contributed by China (Kitada et al., 1992
  Ichikawa and Fujita, 1995 Chung et al., 1996
  Chang et al., 2000 Holloway et al., 2002) .

4
PM2.5 Episodical Weather Pattern from LRT
TAS
If the leading edge of Asian continental high
pressure system moves fast from Asian continent
to Taiwan, PM2.5 episodes will occur by the
accompany of a strong prevailing northeast wind
in the greater Taipei area.
5
Taipei Basin (viewing from the south)
Da-Twen Mountain
Aerosol Supersite?
6
(53 episodes in 3 yrs)
139
153
7
Models

• Meteorological data is from MM5 (version 3.7,
  Grell et al., 1994) simulation.
• Emission data for Domain 1 and 2 are from Streets
  et al. (2003). It is estimated that yearly growth
  factors of 1.13 and 1.19 are used for 2001 to
  2003 and 2003 to 2004 for anthropogenic sources,
  respectively, according to the growth of energy
  consumption and vehicles in China (Hao and Wang,
  2005 Heo and Feng, 2005). For Domain 3 and 4
  TEDS (Taiwan emission data version 6.1, Fu et
  al., 2007 TWEPA, 2006) are adopted. TEDS were
  processed through SMOKE (version 2.1, Houyoux and
  Vukovich, 1999 ).
• Chemical transport model used in this study is
  CMAQ (version 4.4, Byun and Ching, 1999 ). The
  cb4_ae3_aq option is chosen for chemical
  mechanism.

8
Simulation Domain
(dx36 km)
(dx1.33 km)
WL
KL
TAS
KT
9
Simulated PM2.5 Contour from Domain 2
1400 on day 19
1400 on day 20
0200 on day 20
Surface PM2.5 from model simulation
Meteorological surface map
10
Comparison of Simulated and Observed PM2.5 in
Taipei Basin
LRT plume dominated Taipei air quality in this
episode
about 1.5 days
11
Comparison of Simulated and Observed PM2.5
Components at Taipei Aerosol Supersite
PM2.5 nitrate
PM2.5 sulfate
PM2.5 nitrate was overestimated from vaporized
HNO3 reacting with local NH3
PM2.5 OC
PM2.5 EC
Low PM2.5 carbons were due to instrument
malfunction
12
Evolution of PM2.5 Components Along with the Path
of LRT Plume
160
100
70
13
Evolution of PM2.5 Components Along with the path
of LRT Plume
16-25
HNO3 was overestimated
1
NH3(g)HNO3(g)?NH4NO3(s) NH4NO3 ? due to
increasing temperature and low NH3. Kin and Park
(2001), Zhuang et al. (1999) found lifetime of
nitrate in fine mode is short.
Yao et al. (2003), Shimohara et al. (2001),
Jordan et al. (2003), Talbot et al. (2003), and
Dibb et al. (2003) all found nitrate is rich in
coarse mode. HNO3(g)NaCl (s) ? NaNO3HCl (sea
salt uptake) 2HNO3(g)CaCO3(s) ?Ca(NO3)2H2CO3
(dust uptake)
14
Evolution of PM2.5 Components Along with the Path
of LRT Plume
The decrease of N2O5 at night was about 100 to
150 pptv, which contributed HNO3 less than what
was evaporated from find mode nitrate during
daytime.
15
Evolution of PM2.5 Components Along with the Path
of LRT Plume
13(20 to 9µg m-3 )
35
16-19
HO2HO2?H2O2 (1)SO2HO2 ?HSO3-
HSO3-H2O2?H2SO4 (2)SO2 ?SO3 SO3H2O ?H2SO4
NH3(g)HNO3(g)?NH4NO3(s) NH3H2SO4
?NH4(H)SO4 2NH3H2SO4 ?(NH4)2SO4
Our modeling results verify the observations in
TRACE-P, ACE-Asia, Taiwan (Chuang et al., 2007),
and Hong Kong (Pathak et al., 2003 Ho et al.,
2003).
16
Evolution of PM2.5 Components Along with the Path
of LRT Plume
6
PM2.5OC/PM2.5 22-24
SOA formation
21
3.7
OC/EC 3.7-4.1
2.7
As a fraction of OC is water-soluble, this WSOC
is expected to decrease through cloud processing.
More studies are needed for the loss of OC during
transport. Process Analysis may help solve this
problem.
17
Conclusions

• We have successfully simulated a PM2.5 episode on
  20 December 2004 contributed from a long-range
  transport of pollutants from China to Taiwan.

• Model simulation shows that PM2.5 OC was only
  slightly decreased from 22-24 to 21. However,
  PM2.5 nitrate was decreased from 16-25 to a very
  lower value of 1. In contrast, PM2.5 sulfate was
  increased from 16-19 to 35.

• PM2.5 nitrate was overestimated from simulation
  which indicates that CMAQ should include the
  uptake of HNO3 on sea salt and soil especially
  for coastal area.

18
Acknowledgements

• We would like to thank Dr. Streets and Taiwan EPA
  for providing Asian emission database and Taiwan
  emission database, respectively.
• In addition, we acknowledge NCEP (National
  Centers for Environmental Prediction), DBAR (Data
  Bank of Atmospheric Research, managed by the
  Department of Atmospheric Sciences, National
  Taiwan University), and JMA (Japan Meteorology
  Agency) for accessing their atmospheric
  monitoring data.

19

Thanks for Your Attention
Comments Are Welcomed