AIRBORNE MEASUREMENTS OF TRACE GAS AND AEROSOL PARTICLE EMISSION RATIOS FROM BIOMASS BURNING IN AMAZONIA P Guyon, M Welling, G Frank, P Artaxo, G Nishioka,

AIRBORNE MEASUREMENTS OF TRACE GAS AND AEROSOL

PARTICLE EMISSION RATIOS FROM BIOMASS BURNING IN

AMAZONIAP Guyon, M Welling, G Frank, P Artaxo, G Nishioka, J Lloyd,

O Kolle, M A F Silva Dias, L V Gatti, A M Cordova, M O Andreae

Max Planck Institute for Chemistry, Mainz, Germany

2Max Planck Institute for Chemistry, Mainz, Germany

The SMOCC 2002 experiment(Sep–Nov 2002)

Some of the main objectives

- Characterize emissions from deforestation fires, especially aerosols

- Investigate the impact of smoke aerosols on cloud microphysics and rain formation

- Study seasonal differences in atmospheric properties


Instrumentation• CO (Aero-Laser) accuracy/precision ca. 5%

• CO2 (LiCor 6251) overall flight precision ca. +/- 0.4 ppm

• CN (CPC TSI 3022) precision ca. 5%

• Particle number-size distribution (SMPS TSI)

• Aerosol Scattering (Nephelometer TSI) precision ca. 5%

• Aerosol inlet: isokinetic, accumulation mode < cutoff

• Normal temperature and Pressure (25 °C, 1 atm.)


Sampling fresh smoke…


0

1000

2000

3000

4000

0 1000 2000 3000 4000 5000 6000 7000

CN, cm -3

Alt

itu

de,

m

Green Ocean over W. Amazon

Smoky Air over Rondonia

Vertical distribution of CN over clean and smoky regions of the Amazon

Elevated plumes

Polluted Boundary layer

Low CN in background pristine air


CO to CO2 and CN to CO emission ratios from forest fires over the Amazon

0

10

20

30

40

50

60

0 -

10

10

- 20

20

- 30

30

- 40

40

- 50

50

- 60

60

- 70

70

- 80

80

- 90

90

- 100

100

- 11

0

110

- 12

0

120

- 13

0

130

- mor

e

DCO/DCO2 (ppb/ppm) and DCN/DCO (cm-3/ppb) emission ratios

Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

)

CO to CO2

CN to CO

Flaming Smoldering


In Cloud:

- DCO/DCO2 = 52

- DCN/DCO = 19

In-fire variations of emission ratios

From 4 passes through fresh plume:

- DCO/DCO2 = 40–120

- DCN/DCO = 9–35

Outflow:

- here: DCN/DCO = 44

- other cases:

- DCO/DCO2 = 50–90

- DCN/DCO = 30–50


0

5

10

15

20

25

30

35

40

0 -

10

10

- 20

20

- 30

30

- 40

40

- 50

50

- 60

60

- 70

70

- 80

80

- 90

90

- 100

100

- 11

0

110

- 12

0

120

- 13

0

130

- mor

e


Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

)CO to CO2

CN to CO

0

5

10

15

20

25

30

35

0 -

10

10

- 20

20

- 30

30

- 40

40

- 50

50

- 60

60

- 70

70

- 80

80

- 90

90

- 100

100

- 11

0

110

- 12

0

120

- 13

0

130

- mor

e


Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

)

CO to CO2

CN to CO

Above boundary layer

Within boundary layer

Transport of

gases and

aerosols


Smoldering in the evening remains at the surface



0

5

10

15

20

25

30

35

40

0 -

10

10

- 20

20

- 30

30

- 40

40

- 50

50

- 60

60

- 70

70

- 80

80

- 90

90

- 100

100

- 11

0

110

- 12

0

120

- 13

0

130

- mor

e


Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

)CO to CO2

CN to CO

0

5

10

15

20

25

30

35

0 -

10

10

- 20

20

- 30

30

- 40

40

- 50

50

- 60

60

- 70

70

- 80

80

- 90

90

- 100

100

- 11

0

110

- 12

0

120

- 13

0

130

- mor

e


Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

)

CO to CO2

CN to CO

Above boundary layer

Within boundary layer

Transport of

gases and

aerosols

20

25


0

10

20

30

0 - 0

.2

0.2

- 0.4

0.4

- 0.6

0.6

- 0.8

0.8

- 1

1 -

1.2

1.2

- 1.4

1.4

- 1.6

1.6

- 1.8

1.8

- 2

DScattering(550 nm)/DCO (Mm-1/ppb) emission ratios

Fre

qu

ency

(nu

mb

er

of

ob

se

rve

d p

lum

es

pe

r b

ins

) Scattering/CO below BL

Scattering/CO above BL

Processing of aerosol particles

Moreover:b(700) > b(700)

A(450/700) > A(450/700)

Processed particles become larger


Conclusions

• DCN/DCO = 10–40 (savanna: 20–40) – remarkably small variation!

• DCO/DCO2: – large variability for flaming phase with values

centered around 3–7% (efficient combustion)– smoldering emissions remain close to the surface

• Insignificant particle number losses during vertical

transport and cloud processing in lower troposphere

efficient upward transport of pollutants long range transport



Diurnal variation of emission ratios?Or morning bias?

14-15 15-16 16-17 17-18 18-19 19-20 20-21 21-22 22-230

20

40

60

80

100

120

140

13

915

329 66 28

30 6

DCO

/DC

O2

(pp

b/p

pm

)

time bins GMT (hours, LT + 4 )

Average +/- 1 Stdev n = number of plumes per time bin


Emission of particles as a function of combustion efficiency

0.000

1.000

2.000

3.000

4.000

5.000

0.860 0.880 0.900 0.920 0.940 0.960 0.980 1.000

Combustion efficiency

Em

iss

ion

ra

tio

of

CN

/(C

O+

CO

2)


• Slide 1:• - Read title/institutions/SMOCC2002• - Picture in the background taken at our base airport. • - The 2 a/c used.• - Nice fire plume turning into a pyrocloud, allowing vertical transport of aerosols trace gases pollutants from BL to

FT.• Slide 2:• - Few words about SMOCC2002 (BB season + transition, JP, Rondônia,…)• - 1 of main objectives = Characterize emissions from deforestation fires, especially aerosols. Not well known in

contrast to e. g. savanna fires.• Slide 3 (Instrumentation):• To determine the ERs, we made measurements of CO, CO2, total particle number and particle scattering.

Additionally, we measured particle number size distribution and CCN.• Slide 4 (vertical CN gradient):• Coming to some data, this graph presents the “vertical… (title)”• - Over green ocean (background pristine air): little aerosols and virtually no gradient• - In areas under BB influence, typically:• - very polluted BL• - elevated polluted plumes, cloud processed• We have studied these plumes both within and above the BL.• Slide 5 (overall frequency plot):• In order to better understand the contribution of smoke to the atmosphere, we have calculated ERs of CO/CO2

and CN/CO for all the plumes measured with the a/c. ERs are defined as the slope of the regression of CO to CO2 and CN to CO, respectively.

• The CO/CO2 ratios describe the combustion efficiency/phase of the fire since smoldering CO, flaming CO2. • This frequency plot shows an overview of all ERs obtained during this campaign.• - For CN/CO, we found values centered around ca. 10–40 # per cm3 and ppb CO. This compares to savanna

burning (e.g. Hobbs et al.: ca. 20 – 36 over Africa).• - CO/CO2 ERs showed a large scatter but were overall centered around 4–8% for the most, i.e. quite efficient

burning, comparable to savanna. • However, the large variation in ERs is not only due to the differences in burning conditions from fire to fire.


• Slide 6 pictures of SLF survey fire:• - Taking this fire as an example, • - after 4 passes over this fire, we obtained DCO/DCO2 values ranging 40–120, almost the whole range of values obtained for all measured fires.• - Same for CN/CO• - hard to give a number for a specific fire• - in cloud: middle, but less relevant• - But looking at the outflow out of the clouds, CN/CO ratio was even slightly higher than that found in the fresh plumes. This was also confirmed in other cases. Shows that hot fires are an efficient mode of transport for particles, with, at least, little loss, and, in some cases, maybe production of particles out of the cloud.• Slide 7 (Frequency plot ERs below and above BL)• Overall, when separating the ER data into plumes measured below and above the BL, we found that:• - no stat. significant difference for the CO/CO2 data.• - (smoldering circled) Little smoldering was measured from the a/c because smoldering remains close to the surface. • Slide 8 (picture smoldering):• This is because: • - smoldering fires are less vigorous than flaming ones.• - smoldering usually more in the evening when convection has ceased.• Slide 9 (same as Slide 7):• - CN/CO ratio are slightly lower above BL (but the difference is stat. significant) confirming that little particle number loss due to coagulation or scavenging occurs during the upward transport and cloud

processing. efficient transport to FT. • Slide 10 (Neph):• However, when looking at the ERs of scattering/CO, the slightly larger amount of particles in the BL (i.e. fresher plumes) are less scattering that the processed aerosols. • Moreover, simultaneous measurements show that the backscatter ratio and the Angstrom coefficient of these particles within the BL are larger than those above BL, pointing at smaller particles than the

detrained ones (stat. signif.). • Slide 11 (size distributions):• The size distribution of the detrained smoke coming out of the cloud is typically shifted toward higher diameters.• Slide 12 (CCN):• This has in turn a dramatic effect on the CCN efficiency spectrum of these particles. • Whereas fresh smoke tends to be moderately efficient, with some ca. 50% of the particles activated at 1% ss, aerosols from detrained plumes tend to be much more efficient at making CCNs, close to what

we actually measured for background aerosols. • Therefore, these particles which are too numerous and too small to, in the pyrocloud, undergo efficient coalescence as droplets and rainout, are released in the FT as efficient CCNs. • Slide 13 (Conclusions):• - CN/CO ERs are spread over a relatively narrow range between ca. 10 and 40 (both CN and CO primarily emitted during smoldering/less efficient burning). These values compare to savanna ERs.• - The CO/CO2 ERs showed:• - large variability, with values centered around 40–80 (4–8%), which is a quit efficient burning (also comparable to savanna). • - smoldering emissions remain close to surface• - During vertical transport and cloud processing, there is little loss of particles by coagulation and/or coalescence of droplets, and there is a significant growth of the particles. Therefore, this is an efficient

mean for upward transport of pollutants from the BL to the mid-troposphere, thereby favoring long-range transport of, among others, efficient CCN and largely scattering particles. • Optional slides:• Slide 14 (Morning bias):• The apparent diurnal pattern of CO/CO2 ERs is actually a bias accruing in the morning hours due to CO2 from the night transpiration being uplifted with the plume, thereby dramatically decreasing the

CO/CO2 values. Even examples of purely smoldering large fires showed very small CO/CO2 values in the morning hours. • Probably all ERs relative to CO2 ever measured so far in the morning hours suffered from this bias. These data were removed from the rest of our dataset presented. • Slide 15 (ER CN/(CO+CO2) vs CE)• Shown is the emission of particles as a function of combustion efficiency for all plumes measured in the BL (morning biased data of course removed). • Here, the emission of particles is presented as CN/(CO+CO2), CE is CO2/Ctotal. • There is a significant (r^2 = 0.61) dependence of the emission of particles on CE, with more particles emitted at lower CE (toward smoldering).

AIRBORNE MEASUREMENTS OF TRACE GAS AND AEROSOL PARTICLE EMISSION RATIOS FROM BIOMASS BURNING IN AMAZONIA P Guyon, M Welling, G Frank, P Artaxo, G Nishioka,

Documents

particle scattering

total particle number

vertical cn gradient

emission of particles

little aerosols

deforestation fires

savanna fires

impact of smoke aerosols