Indication of aerosol aging by Aethalometer optical absorption measurements

Luka Drinovec1, Griša Močnik1, Irena Ježek1, Jean-Eudes Petit2,3, Jean Sciare2, Olivier Favez3, Peter Zotter4, Robert Wolf4, André S.H. Prévôt4, and Anthony D.A. Hansen1,5

1. Aerosol d.o.o., Kamniška 41, SI-1000 Ljubljana, Slovenia2. LSCE (CEA-CNRS-UVSQ), Orme des Merisiers, Gif-sur-Yvette, France;

3. INERIS, Verneuil-en-Halatte, France4. Paul Scherrer Institut, CH-5232 Villigen PSI, Switzerland

5. Magee Scientific, 1916A M.L. King Jr. Way, Berkeley, CA 94704, USA

Keywords: Aethalometer, source apportionment, ACSM, AMS, PSCFContact author email: luka.drinovec@aerosol.siPresenting author email: irena.jezek@aerosol.si

ACCENT Symposium 2013, Urbino (Italy)

1. Introduction to BC measurements

Sources- Combustion

Effects of black carbon (BC):- Public health effects- Climate change

How to reduce harmfull effects:- Indentify sources: traffic vs household heating- Indentify sources: local vs. regional

dpp=20 nm

Note change in scale

dm=472 nm

3

Analytical Instrument : Aethalometer™

• Collect sample continuously.• Optical absorption ~ change in ATN. • Measure optical absorption continuously : λ = 370 to 950 nm.• Convert optical absorption to concentration of BC:

BC (t) = babs(t) / - mass absorption crossection

• Real-time data: 1 s/1 minute

ATN = ln (I0 / I)Reference I0

Sensing IBC

Light Source

Filter with Sample

Light Detectors

babs ~ ATN

Filter loading effect

0 100 200 300 400 5000

100

200

300

400

Tape a

dva

nce

b abs

(Mm

-1)

t (min)

babs

at 370 nm

raw data compensated

Tape a

dva

nce

Roxbury Feb - June 1999

R2 = 0.117

0

200

400

600

800

1000

1200

1400

0 5 10 15 20 25 30 35 40 45

ATN

Ave

rag

e B

C in

1-A

TN

bin

Average BC

Linear fit

k = 1.4

London Oct-Dec 2006

R2 = 0.90

0

2000

4000

6000

8000

10000

12000

14000

16000

0 5 10 15 20 25 30 35 40 45 50

ATN

Ave

rag

e B

C in

1-A

TN

bin

Average BC per 1 unit of ATN

Linear fit

k = 4.6

BC vs ATN analysis – ambient data

5

Linear reduction of the instrumental response due to loading of the filter fiber. Jump at the tape advance (similar to Virkkula (2007) model).

•ambient data – no dependence of BC on ATN

•slope k variable: site, source, aerosol age, composition

•need to determine it dynamically – do not assume, rather measure

BC (reported) = BC (zero loading) · { 1 - k · ATN }

Large loading effect Small loading

effect

k=0.005

k=0.001

6

Dual spot Aethalometer – AE33

ATN1 = ln (I0 / I1)Reference I0

Sensing I1BC1

Light Source

Filter with Sample

Light Detectors

Sensing I2

ATN2 = ln (I0 / I2)

Two parallel spots with different flow, therefore ->

From different loading and attenuation loading compensation parameter k(λ) is calculated.

Absorption data is compensated:babs=babs1/(1-k*ATN1)

12.1.2013 19.1.2013 26.1.2013 2.2.2013 9.2.20130,000

0,002

0,004

0,006

0,008

0,010k(

)

time

370 mn

470 nm

520 nm

590 nm

660 nm

880 nm

950 nm

Payerne Winter 2013

BC2

• measure attenuation with the Aethalometer

• absorption coefficient - babs

• for pure black carbon: babs ~1/λ

• generalize Angstrom exponent: babs

~1/λα

diesel: α ≈ 1

wood-smoke: α ≈ 2 and higher

BC source apportionment

7

J. Sandradewi et al., A study of wood burning and traffic aerosols in an Alpine valley using a multi-wavelength Aethalometer, Atmospheric Environment (2008) 101–112

b(λ) = bwb (λ,wood) + bff (λ,fossil) λ = 470 nm, 950 nm

BC source apportionment

8Sandradewi 2008

bi(470 nm) / bi (950 nm) = (470 nm / 950 nm) -

α = 1,0 ± 0,1 (fossil) Bond & Bergstrom 2004

α = 2,0 -0,5/+1,0 (wood) Kirchstetter 2004, Day 2006,

Lewis 2008

BCwb

BCff

Measurement campaignEMEP: summer 2012 & winter 2013

Payerne site- Payerne aerological station- Rural background site- NW Swiss

Paris site - SIRTA Atmospheric Research Observatory- located in a semi-urban environment- 25 km south of the Paris city center

Site Campaign BC (ng/m3) Biomass burning (%)

Payerne Winter 2013 789 29

Summer 2012 593 10

Paris Winter 2013 968 25

Summer 2012 671 4

Back trajectory analysis

Back trajectory analysis using Potential Source Contribution Function (PSCF)• Represents the probability that an air parcel may be responsible for high

concentrations observed at the receptor site• 72h back trajectories calculated with Hysplit v4.9• starting at 500m AGL

An example:- PSCF analysis of BC- Paris winter 2013

Indentification of source locations

- Angstrom exponent α obtained from AE33 spectral data - PSCF (Back trajectory analysis using Potential Source Contribution Function)

α < 1.3 (traffic emissions) α > 1.3 (biomass burning)

Paris – EMEP winter campaign 2013

16.6.2012 20.6.2012 24.6.2012 28.6.2012 2.7.2012 6.7.20120,000

0,002

0,004

0,006

0,008

0,010

k()

370 nm 470 nm 520 nm 660 nm 880 nm 950 nm

12.1.2013 19.1.2013 26.1.2013 2.2.2013 9.2.20130,000

0,002

0,004

0,006

0,008

0,010

k()

370 mn

470 nm

520 nm

590 nm

660 nm

880 nm

950 nm

Differentiation of fresh and aged aerosolsPayerne summer

300 400 500 600 700 800 900 10000,000

0,001

0,002

0,003

0,004

0,005

0,006

0,007

0,008

0,009

0,010

Ave

rage

k(

)

(nm)

winter summer

Payerne winter

Spectral fingerprint

Summer and winter aerosols have different optical properties - k(λ)

For background locations with aged aerosol loading effect at 880 nm (where BC is measured) is small!

Differentiation of fresh and aged aerosols

k880nm>0.002 (fresh aerosols) k880nm<0.002 (aged aerosols)

Paris – EMEP summer campaign 2012

- Compensation parameter k880nm obtained from AE33

- PSCF (Back trajectory analysis using Potential Source Contribution Function)

Particle coating hypotesis

Changes in k(λ) are caused by transparent coating

SMPS:Fresh soot particle size = 20-50 nmAged particle size > 100 nm

10 100 10000

200

400

600

800

1000

1200

Par

ticle

num

ber

--

Payerne Summer2013 Measurement fit1 fit2 fit1 + fit2

Particle diameter [nm]

Particle coating hypotesis

Aerosol mass spectrometers: ACSM & AMS (Aerodyne)-> Aerosol chemical composition is obtained

16.6.2012 23.6.2012 30.6.2012 7.7.20120

2

4

6

8

10

12

Co

nc.

(g

/m3 )

Org NH

4

SO4

NO3

Coating factor (CF) – ratio between the sum of nonrefractory aerosol mass to BC:

CF = (Org + NH4 + SO4 +NO3)/BC

13.6.2012 20.6.2012 27.6.2012 4.7.2012 11.7.20120

5

10

15

20

25

30

(SO 4 +

NH 4 +

OR

G)

/ B

C

0,005

0,004

0,003

0,002

0,001

0,000

-0,001

-0,002

-0,003

k 88

0 n

m

Particle coating hypotesis – summer data

AE33 compensation parameter

ACSM

Paris Summer2012 campaign

Compensation parameter k880nm and coating factor correlate well.

Summary

Spectral absorption data from Aethalometer AE33 was used for BC source apportionment during EMEP campaigns in Paris (France) and Payerne (Switzerland).

Back trajectory analysis using Potential Source Contribution Function (PSCF) was used to determine fossil fuel and biomass burning locations.

PSCF: Aged aerosols have small k880nm

Aethalometer and ACSM/AMS measurements were used for calculation of the coating factor (CF): Big CF = Small k880nm

Acknowledgements

The work described herein was co-financed by the EUROSTARS grant E!4825 and JR-KROP grant 3211-11-000519. Measurements performed at SIRTA (LSCE) were funded by CNRS, CEA, the EU-FP7(2007-2013) 'ATRIS' project under grant agreement n°262254, the Primequal Predit 'PREQUALIF' project (ADEME contract n°1132C0020), and the DIM R2DS (AAP 2010) 'PARTICUL'AIR' project. Measurements in Payerne were conducted by the Swiss Federal Office for the Environment (FOEN).

Thank you for your attention!