Biomass burning in Southeast Asia

from field studies to satellite data analysis

And the SPRINTARS model

Haruo Tsuruta, Eiji Oikawa, Syugo Watanabe, Toshiro Inoue,

Makiko Hashimoto,Teruyuki Nakajima (AORI, Univ. of Tokyo)

Jinchula Chotpitayasunon & Boossarasiri Thana

(Chulalongkorn University, Thailand) Daisuke Goto, Nobuo Sugimoto (NIES)

Toshihiko Takemura (RIAM, Kyusyu University)

Koichiro Sera (CRC, Iwate Medical University)

Shigeto Sudo, Seiichiro Yonemura (NIAES)

Yuichiro Shirasuna, Koichiro Hirano

(Yokohama Environment Research Institute)

Pradeep Khatri & Tamio Takamura (CeRES, Chiba University)

Masanori Yabuki (RISH, Kyoto University)

S. Katagiri, Tadahiro Hayasaka (CAOS, Tohoku University) Wataru Takeuchi (IIT, Univ. of Tokyo)

Impacts of biomass burning on local/regional/global environment

Collaborative study with Southeast and South Asia

Malaysian Meteorological Department (MMD): 1994- (NASA: SHADOS program: with MMD) Indonesia: BMG, BAU, 1995- Drs. Nul Hayati, Murdiyarso, Saharjo India: NPL, NRSA, 1995- Drs. Mitra, Gupta, Badarinath, and Vadrevu Singapore: National Environment Agency 1996- Thailand: (DOA, JGSEE), Chulalongkorn Univ. 2006-

An intensive field study through one year is needed to understand a comprehensive picture on the effect of human activity Including LULC and BB on atmospheric environment.

Enhancement of tropospheric ozone influenced by the trans-boundary air pollution over KL, Malaysia

(Yonemura et al, JGR, 2002)

Fig. 8 Methyl bromide and CO concentrations by aircraft and ground measurements over/in Sumatra in 1997-1998

(Sudo, Yonemura, and Tsuruta, 2002)

Objectives of our study in Phimai, a SKYNET station

1. Chemical characterization of atmospheric

aerosols at Phimai in rainy and dry seasons

2. AOT and single scattering albedo ( SSAo)

from optical measurements by skyradiometer

3. Single scattering albedo (SSAc) calculated from

chemical compositions

4. Comparison of chemical/optical properties

between the field study, and the SPRINTARS model

and the CALIOP data

Phimai

Hedo

Hefei

Minami

-TorishimaMiyako

Mandalgovi

DunhuangYinchuan/

Shapatou

Gosan

Anmyon

Nagqu

Osaka

Shirahama

Huanshang

Taishang

Sri-Samrong

Nepal

Hanimaadhoo

ABC

SKYNET

AERONET

ADEC, CEOP

ABC-EAREX

Ochi-Ishi

Gwangju

Hateruma

Wan-Li

Lu-LinHok Tsui

Xianghe

Kampur

Pune

Bac Giang

Mukdahan

Bangkok

Qira

Aksu

Dalanzadgad

Beijing

Taipei

Chao-Jou

Ussuriysk

Kao-Hsiung

Sapporo

Chichijima

Tsukuba

Tokyo

Chiba

Kanazawa

Toyama

Tateyama

Fukue

Om Koi

Amami

Phimai

Hedo

Hefei

Minami

-TorishimaMiyako

Mandalgovi

DunhuangYinchuan/

Shapatou

Gosan

Anmyon

Nagqu

Osaka

Shirahama

Huanshang

Taishang

Sri-Samrong

Nepal

Hanimaadhoo

ABC

SKYNET

AERONET

ADEC, CEOP

ABC-EAREX

Ochi-Ishi

Gwangju

Hateruma

Wan-Li

Lu-LinHok Tsui

Xianghe

Kampur

Pune

Bac Giang

Mukdahan

Bangkok

Qira

Aksu

Dalanzadgad

Beijing

Taipei

Chao-Jou

Ussuriysk

Kao-Hsiung

Sapporo

Chichijima

Tsukuba

Tokyo

Chiba

Kanazawa

Toyama

Tateyama

Fukue

Om Koi

Amami

P

A

Monitoring stations for atmospheric aerosols by ABC, AERONET, and SKYNET in Asia. The station of Sri-Samrong and Amami-Oshima has moved to Phimai (since 2004) and Cape Hedo (since June 2005), respectively.

II. Atmospheric Aerosols at the Observatory of Atmospheric Research

at Phimai (15.18°N, 102.57°E )

Sampling period

1. June 2006-Feb. 2007

2. July 2007-July 2008

A series of a half day sampling in the

daytime and nighttime for three days

was made twice in every month.

Phimai

Measurements and analysis

Surface aerosol measurements

Sampling sites: Phimai, Thailand (15.18N,102.57E)

Sampling period: Jul. 2007-Aug. 2008

Aerosol sampler: Multi-nozzle cascade impact

sampler with PM1.0, PM2.5 and PM10 impactor

Chemical analysis: Mass concentration

EC/OC (by optical/thermal analyser)

Water soluble ions (by IC)

Trace elements (by PIXE)

Optical measurements by Skyradiometer

Data analysis for: Aerosol optical thickness (AOT)

Single scattering albedo (SSA)

Angstrom exponent (AE)

Trace gases: CO and O3

Backward trajectory analysis: HYSPLIT MODEL by NOAA

Satellite data analysis: Hot spots of fires by MODIS (by Dr. Takeuchi) Vertical distribution of aerosols/clouds by CALIOP

Vertical profile of aerosols: LIDAR by NIES

Verical profiles of meteorological data (T,RH,WD,WS) Radio sonde at Ubon Ratchathani 200km east of Phimai

Surface data: Phimai Observatory:meteorological data(RH, T, WD, WS) Soil chemical properties

Backward trajectories by NOAA HYSPLIT MODEL

(W1) 00 UT 10 Oct. 2007 (D1) 00 UT 15 Oct. 2007 (D3) 00UT 20 Mar. 2008

Definition of wind pattern： Dry season: D1: northeasterly monsoon from East Asia or South China Sea D2: Transition stage of D1 and D3 D3: From South China Sea Rainy season: W1, W2: southwesterly monsoon from Indian Ocean

0

1

2

3

4

5

J-0

7

A-0

7

S-0

7

O-0

7

N-0

7

D-0

7

J-0

8

F-0

8

M-0

8

A-0

8

M-0

8

J-0

8

win

d di

rect

ion

date (month-year)

NNE-NE

ENE-E

S

SW

W1 D1 D2 D3 W2

Conclusion 1 : Chemical properties in Dry season,

Analysing the ratio of nss-SO4/EC, D1: Major source: Air pollutants from east Asia by NE monsoon, D3: Major source: Biomass burning from Indochina D2: Transition stage between D1 and D3 Aerosols from severe biomass burning was frequently trapped in the lower boundary layer

0

5

10

15

20

25

30

35

W1 D1 D2 D3 W2

me

an c

on

cen

trat

ion

(u

g m

-3)

wet and dry periods

K(BB)

soil dust

(NH4)2SO4

OC

EC

PM2.5measurement(obs.)

Time series of chemical composition in PM2.5

By the surface measurements

Fossil fuel mixed Biomass

Combustion (D1 & D3) burning

From East Asia in

Indochina

Dust from west Asia

0

2

4

6

8

10

0 5 10 15 20

EC

(ug

m-3)

nss-SO4 (ug m-3)

D1

D2

D3

W1

W2

AMAMI (W1)

MALAYSIA

Jambi (x 0.1)

D1

EC/nss-SO4 (2.5um>d)

D3

1/11

A

Scatter diagram between nss-SO4 and BC in PM2.5 from surface measurements (Phimai, Amami, Malaysia, Indonesia (Jambi))

0

2

4

6

8

10

0 5 10 15 20

nss-SO4 (ug m-3)

BC

(ug m

-3)

D1

D2

D3

W1

W2

D3

D1

SPRINTARS MODEL

(BC/nss-SO4)

Scatter diagram between

nss-SO4 and BC(EC)

in PM2.5

Surface measurements

SPRINTARS model

Large difference in

BC(EC)/SO4 between

D1 and D3 is well

correlated between

surface measurements

and SPRINTARS.

D1: fossil fuel combustion

from East Asia

D3: biomass burning

0

2

4

6

8

10

0 5 10 15 20

EC

(ug

m-3)

nss-SO4 (ug m-3)

D1

D2

D3

W1

W2

AMAMI (W1)

MALAYSIA

Jambi(x0.1)

D1

EC/nss-SO4 (2.5um>d)

D3

1/11

A

Biomass burning

Fossil fuel combustion

0

200

400

600

800

1000

1200

1/1

2/1

3/1

4/1

5/1

6/1

7/1

8/1

9/1

10

/1

11

/11

2/1

1/1

2/1

3/1

4/1

5/1

6/1

7/1

8/1

9/1

10

/1

11

/11

2/1

Nu

mb

er

of

wild

fire

Time of day (month/day, 2007-2008)

Thailand

D1 D2 D3

Dry season in 2007-2008

Number of wildfires from MODIS FireMAP During 2007-2008 (database from Prof. Takeuchi)

0

1

2

3

4

5

W1 D1 D2 D3 W2

wet and dry periods

mean c

oncentr

ati

on (

ug m

-3)

Obs.

S.model(O)

S.model(T)

EC (BC)(PM2.5)

1. Spatial distribution of mean BC(EC) concentration in W1, DI(D1+D2), DII(D3)

by SPRINTARS. Comparison of BC(EC) and OC between surface measurement

and SPRINTARS in wet and dry periods is in good agreement.

x

O(n)= 18 11 21 11 10

T(n)=105 48 103 50 60

W1

D1+D2

D3

0

5

10

15

20

W1 D1 D2 D3 W2

wet and dry periods

mean c

oncentr

ati

on (

ug m

-3)

Obs.

S.model(O)

S.model(T)

Dust partic les(total: 10>d)

4. Dusts: Large difference between the measurement and

the model in the dry periods(SPRINTARS << Measurements)

W1

D1+D2

D3

X

Conclusion 2 : Soil/mineral dusts D1: direct transport from east Asia, hardly from inner desert areas D2: re-suspension of soil dusts in Indochina D3: re-suspension of local soil dusts caused by thermal plume due to biomass burning in Indochina Wet season: Long range transport from west, possibly from the desert regions in west Asia

0

5

10

15

W1 D1 D2 D3 W2

mean

concentr

atio

n (ug

m-3)

wet and dry periods

Sea Salt

soil dust

NO3

OC

EC

Coarse mode (obs.)

(10um >d> 2.5um)

0

5

10

15

20

25

30

W1 D1 D2 D3 W2

mean

concentr

atio

n (ug

m-3)

wet and dry periods

K(BB)

soil dust

(NH4)2SO4

OC

EC

fine mode (obs.)

(2.5um >d)

Mean chemical composition for fine and coarse particles

in the wet and dry seasons

Coarse Particles PM2.5

y = 0.0495x - 0.0008

R2 = 0.9123

y = 0.0461x - 0.0002

R2 = 0.6815

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Ti

Zr

D1

D2

D3

W1

W2

線形 (D2)

線形 (W2)

10>d>2.5

D2

W2

Zr has not been detected in atmospheric aerosols at Amami in spring time,

And at Phimai in the D1 period when the surface wind was from east Asia.

In contrast, Zr was detected in D2 and D3 periods, and it strongly suggests that,

In the D1 period, soil was not dried well and biomass burning was not active.

In the D2 and D3 periods, however, soil became well dried and biomass burning

very active, and local soil dust and re-suspended dust by biomass burning

could strongly affect atmospheric aerosols at Phimai, in addition to that from east

Asia in D2.

y = 2.29x + 0.028

R2 = 0.68

y = 0.18x + 0.055

R2 = 0.52

0.0

0.5

1.0

1.5

0.0 0.5 1.0 1.5

K, T

i (u

g m

-3)

Fe (ug m-3)

■ K/Fe (1.0>d)

■ K/Fe(10>d>2.5)

+ Ti/Fe y = 0.10x - 0.001(1.0>d) R2 = 0.94

○Ti/Fe y = 0.093x + 0.001(10>d>2.5) R2 = 0.98

Phimai : daily mean data

Jul. 2007-Jun. 2008

Relationship for K/Fe and Ti/Fe in coarse particles and PM2.5

(July 2007- June 2008)

K: soil particles in coarse particles

biomass burning and bio-fuels in fine particles

20 June 2008

25.2 N 19.1 N 13.0 N 6.9 N 0.8 N

Phimai

Height (km) 10 5 0 Fig. 7 Vertical profile of Aerosol type along the orbits, day, 18 June 2008

37.3 N 31.2 N 25.1 N 19.0 N 12.9 N 6.8 N

Phimai

0 5 10 km Height

Fig. 8 Vertical profile of aerosol type along the orbit, night, 18 June 2008

Fig. 10 Backward trajectory analysis of air masses arrived at Phimai on 17-20 June 2008, for a week By NOAA HYSPLIT MODEL

17 June 18 June 19 June 20 June 210m, 1km, 2km

20 June 2,3,4km

00-00-05 18-07-53 23-04-28 21-25-38 19-46-43 Fig. 9 Global and vertical aerosol distribution by SPRINTARS MODEL

18 June 2008

Conclusion 3: Optical properties SSA: Clear seasonal variation with the minimum inD2 AOT: Highest in D3 possibly due to biomass burning AOT and PM2.5 : Positive correlation in D2 , and in D3, however, AOT was much higher than in D2. A possible reason might be multi layers of aerosols, as shown in CALIOP data, due to biomass burning in Indochina or transport of polluted air masses from the west.

0.70

0.75

0.80

0.85

0.90

0.95

1.00

J-07

A-07

S-0

7

O-07

N-07

D-07

J-08

F-08

M-08

A-08

M-08

J-08

SSA

day (1 Jul. 2007-31 Jun. 2008)

SSA(SKYNET500)

SSA(AERONET440)

SSA(AERONET676)

SSAex(chem)

SSAhalf(chem)

SSAin(chem)

9-11Jan. 5-7 Mar.

C D

SSAo by skyradiometer and SSAc by surface chemical compositions, assuming external (ex), half internal (half), and internal mixture (in). SSAc is only calculated for several cases, because the size distribution of atmospheric aerosols are used at Amami, not Phimai.

Cirrus effect

0

2

4

6

8

10

J J J A A A S S S O O O N N N D D D J J J F F F M M M A A A M M M J J J

sampling date (Jul. 2007-Jun. 2008)

Daily E

C c

oncentr

ati

on (

ug m

-3)

10>d>2.5

2.5>d>1.0

1.0>d

EC

11 Jan.

8

9

10

J J J A A A S S S O O O N N N D D D J J J F F F M M M A A A M M M J J J

W1 (SW) W2(SW)D1(NNE-E) D2 (NNE-E) D3 (S)

Wind direction

Time series of EC concentration and wind direction

× ×

(1) 06 h UT, 11 Jan. 2008 (2) 18 h UT, 11 Jan. 2008

Phimai

A thick aerosol layer between 0-2 km over Phimai In the daytime and nighttime

Phimai

Fig. 14 Total attenuation backscatter from CALIOP

0

2

4

6

8

10

12

14

16

18

20

0 5 10 15 20 25

Temp (℃)

Alt

itude (km

)

Phimai

Fig. 11 Time series of Aerosol Optical Thickness (AOT, 500nm)

and WD

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

J-07

A-07

S-07

O-07

N-07

D-07

J-08

F-08

M-08

A-08

M-08

J-08

Date (month-year)

daily m

ean A

OT

WD

NEESSW

W1 D1 D2 D3 W2

12 Feb.

11 Jan .

24 Mar .

AOT

B

A

AOT>0.80: 10-17 Feb. 15-24 Mar.

Fig. 1７

24 Mar. 2008

06h UT

No surface

measurement

H

(km)

15

10

5

0

Mie Lidar extinction coefficient over Phimai in March 2008

Soil dust

Spherical particles

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 10 20 30 40 50 60 70

PM2.5 (ug m-3)

AO

T (

0.5

um

)

19-21 Mar.

3-4 Apr.6-7 Mar.

8 Feb.

28-30 Nov.

9-11 Jan.

5 Mar.

W1, W2

20-22 Feb.

Scatter diagram between PM2.5 and AOT by skyradiometer

in the days when both data were simultaneously measured.

AOT could be positively correlated with PM2.5 near the surface.

More comprehensive study among field measurements,

satellite data analysis, and model simulation

should be performed in future.

Thank you very much for your attention!!