Piet Stammes,
KNMI, De Bilt, The Netherlands
7 November 2012
1

• Importance of aerosols
• Aerosol microphysics
• Spectral absorption by aerosols
• GOME-2
• Absorbing Aerosol Index
• First results on Aerosol Height
Acknowledgements to:
Martin de Graaf, Gijs Tilstra, Ping Wang, Olaf Tuinder (KNMI)
Eyk Boesche (FUB)
Marloes Penning de Vries (MPIC)
2

Absorbing Aerosol Index map from SCIAMACHY
Canadian and Alaskan forest fires June-July 2004
Californian forest fires
Smoke and Dust weak events strong events
Siberian forest fires in July 2006
Libian desert
Taklamakan desert
Thar desert
Desert dust
Sahara
Bodélé
Sahel biomass burning and desert dust storms
Amazonian rainforest biomass burning biomass burning smoke
Saudi Arabian lowlands biomass burning smoke
Indonesian forest fires
Rice straw burning
Smoke from forest fires
3 more data and information can be found at www.temis.nl

Why are aerosols important?
Air quality / Health Climate
Air traffic safety
Visibility
4

Many aerosol types: chemical compositions, sizes and shapes http://alg.umbc.edu/
5

Dust aerosols
Sahara dust event Size distribution
©nasa earthobservatory
Absorbing aerosols:
• Desert dust
• Smoke
• Volcanic ash
Fine mode aerosols: around 0.1 micron
- Coarse mode aerosols: around 1 micron
6

Observation by SCIAMACHY of absorption spectrum of smoke aerosols.
This absorption leads to heating of the troposphere up to 125 W/m2.
De Graaf et al., JGR, 2012
7

GOME-2 on Metop since 2006
• UV-visible-near-IR spectrometer
• 4 spectral channels, covering 240 - 790 nm
• 0.2-0.4 nm resolution
• Polarization Monitoring Devices (PMDs) at 15 bands
• Main products: ozone, NO2, SO2, minor gases
• Additional products: aerosols, clouds, surface albedo http://www.esa.int/esaLP/SEMTTEG23IE_LPmetop_0.html
8

• GOME(-1)
ERS-2
• GOME-2
Metop-A+B
• SCIAMACHY
Envisat
80 km
40 km
320 km
GOME-2 PMD
10 km
40 km
30 km
40 km
60 km
• OMI
EOS-Aura
24 km
13 km
Along track
9

Absorbing Aerosol Index (AAI)
Definition: residue r
 
100



10 log



R
340
R
380


 meas
 10 log



R
340
R
380


 Rayleigh

 where the surface albedo A for the Rayleigh atmosphere simulations is such that:
R meas

380
R
Rayleigh
380
( A )
A is assumed to be wavelength independent:
A
340
= A
380
The residue represents the observed 340/380 nm colour as compared to the pure Rayleigh colour (OMI: 354/388 nm)
AAI is the positive part of the residue 10

Reflectance at TOA with absorbing aerosols
Doubling-Adding KNMI
Radiative Transfer Model
Solar zenith angle = 30
°
Viewing zenith angle = 0
°
Surface albedo = 5%
Absorbing aerosols: altitude = 3-4 km optical thickness

= 2 single scattering albedo

0
= 0.75
11

Reflectance at TOA with absorbing aerosols and matched Rayleigh reflectance
As
Match at reference wavelength
To match the reflectance in the absorbing aerosol atmosphere at 380 nm , the surface albedo is decreased in the Rayleigh atmosphere:
Rayleigh atmosphere
Surface albedo = 0.6%
12

Reflectance at TOA with absorbing aerosols and matched Rayleigh reflectance
As
The curves don’t match at
340 nm:
Absorbing aerosols create a positive residue.
Residue
13

Generally:
• no clouds, no aerosols
• clouds, no absorbing aerosols
• absorbing aerosols
AAI: r > 0
: r = 0
: r < 0
: r > 0
Pros and Cons:
+ AAI can detect UV absorbing aerosols: volcanic ash, desert dust and smoke.
+ AAI works in cloudy scenes.
+ AAI works over ocean and land.
- AAI is an index: it depends on AOT (

), SSA (

) and altitude (

).
- AAI is very sensitive to absolute calibration.
14

Simulations of AAI for biomass burning aerosols
Clear-sky case
Nadir view
Aerosols at 4-5 km
Clouds at 1-2 km
DAK RTM simulations
Cloudy case
AAI increases with AOT
AAI decreases with SZA
Wang et al., ACP, 2012
15

Daily AAI map of GOME-2 spectral channels http://www.temis.nl/airpollution/absaai/ 16

Daily AAI map from GOME-2 PMDs
PMDs have 8x higher spatial resolution than the spectral channels http://www.temis.nl/o3msaf/vaac_pmd/
17

Information for the VAAC
(volcanic ash advisory centre)
Eyjafjolleruption of April-May
2010 http://www.temis.nl/o3msaf/vaac_pmd/
18

Smoke over Borneo from AAI, 1995 -2010
1997/1998 El Niño: drought caused many forest fires; 120.000 km 2 forest burned.
Satellite data sources: GOME, SCIAMACHY, GOME-2
Figure: L.G. Tilstra, KNMI
19

UV residue has two parts:
Absorbing Index & Scattering Index
Scattering aerosols and clouds Absorbing aerosols
GOME-2 Aerosol Indices for July, 2011, cloud fraction < 0.2.
Work of Marloes Penning de Vries (MPIC, Mainz).
Penning de Vries et al., ACP, 2012
Penning de Vries, Visiting Scientist report of O3MSAF, 2012
20

Effect of instrument degradation on the AAI
The global mean residue, the mean of all residues on a day between 60°N and
60°S, is about constant, showing only a very mild seasonal variation.
GOME-2
(for individual scan mirror positions)
Instrument degradation has a very large impact on the residue/AAI:
2.3 % reflectance change ~ 1 AAI point.
Tilstra et al. (JGR, 2012) developed an in-flight degradation correction method.
21

Approach: use cloud algorithm FRESCO for aerosol height
- FRESCO algorithm: fit of O
2
A-band at 760 nm using a
Lambertian reflector as cloud model.
- FRESCO v6 has two retrieval modes for 2 retrieved quantities:
Normal: Effective cloud fraction (cloud albedo

0.8) and Cloud height
Alternative: Scene albedo (cloud fraction

1) and Scene height
22
Wang et al., ACP, 2008

FRESCO retrievals using simulated O
2 for dust aerosols
A band spectra
Aerosol layer
Cloud layer
Clear-sky
Cloudy
23
Wang et al., ACP, 2012

Puyehue volcano (Chile), 20110606, Westerly Box
Wang et al., ACP, 2012
24

Puyehue volcano (Chile), 20110606, Easterly Box
Wang et al., ACP, 2012
25

• Absorbing aerosols, like desert dust, smoke, and volcanic ash can be detected by GOME-2
• GOME-2 provides near-real-time monitoring information on these aerosols, with the products:
- AAI for absorbing aerosols
- SCI for scattering aerosols (if cloud mask is used)
- FRESCO for aerosol height.
26

• O3MSAF GOME-2 data products: http://o3msaf.fmi.fi
• TEMIS GOME-2 data products: http://www.temis.nl
• GOME-2 and Metop: http://www.eumetsat.int
• GOME-2 L0 data quality information: http://gome.eumetsat.int
27

References on GOME(-2) aerosol retrievals
M. de Graaf, P. Stammes, O. Torres, and R.B.A. Koelemeijer, Absorbing Aerosol Index: Sensitivity analysis, application to GOME and comparison with TOMS, J. Geophys. Res. 110, D010201, doi:10.1029/2004JD005178 , 2005 .
M. de Graaf, L.G. Tilstra, P. Wang and P. Stammes, Retrieval of the aerosol direct radiative effect over clouds from space-borne spectrometry , J. Geophys. Res., 117, D07207, doi: 10.1029/2011JD017160 , 2012
M. de Graaf and P. Stammes and E.A.A. Aben, Analysis of reflectance spectra of UV-absorbing aerosol scenes measured by SCIAMACHY, J. Geophys. Res. 112, D02206, doi: 10.1029/2006JD007249 , 2007 .
M. Penning de Vries, Beirle, S., and Wagner, T.: UV Aerosol Indices from SCIAMACHY: introducing the SCattering
Index (SCI), Atmos. Chem. Phys., 9, 9555-9567, doi:10.5194/acp-9-9555-2009, 2009
M. Penning de Vries, and Wagner, T.: Modelled and measured effects of clouds on UV Aerosol Indices on a local, regional, and global scale, Atmos. Chem. Phys., 11, 12715-12735, doi:10.5194/acp-11-12715-2011, 2011 .
L.G. Tilstra, M. de Graaf, I. Aben and P. Stammes, In-flight degradation correction of SCIAMACHY UV reflectances and
Absorbing Aerosol Index , J. Geophys. Res., 117, D06209, doi: 10.1029/2011JD016957 , 2012 .
L.G. Tilstra, M. de Graaf, O.N.E. Tuinder, R.J. van der A, and P. Stammes, Studying trends in aerosol presence using the Absorbing Aerosol Index derived from GOME-1, SCIAMACHY, and GOME-2 , Proceedings of the 2011
EUMETSAT Meteorological Satellite Conference, EUMETSAT P.59, ISBN 978-92-9110-093-4, 2011.
L.G. Tilstra, O.N.E. Tuinder, and P. Stammes, A new method for in-flight degradation correction of GOME-2 Earth reflectance measurements, with application to the Absorbing Aerosol Index , Proceedings of the 2012 EUMETSAT
Meteorological Satellite Conference, EUMETSAT P.??, ISBN ??????????, 2012.
P. Wang, P. Stammes, R. van der A, G. Pinardi, M. van Roozendael, FRESCO+: an improved O2 A-band cloud retrieval algorithm for tropospheric trace gas retrievals, Atmospheric Chemistry and Physics, 8, 6565-6576, 2008
P. Wang, O.N.E. Tuinder, L.G. Tilstra, M. de Graaf, and P. Stammes, Interpretation of FRESCO cloud retrievals in case of absorbing aerosol events , Atm. Chem. Phys., 12, doi: 10.5194/acp-12-9057-2012 , 2012 .
28

29

AAI products from GOME, SCIAMACHY,
GOME-2, and OMI
GOME–1
SCIAMACHY
Wavelength pair (nm)
Equator crossing time
Pixel size
(km)
Days needed for global coverage
Platform / Operation period
340 / 380
340 / 380
10 : 30 LT
10 : 00 LT
320 × 40
60 × 30
3
6
ERS-2
(1995 – 2003*)
Envisat
(2002 – 2012)
GOME–2
OMI
340 / 380
354 / 388
09 : 30 LT
13 : 30 LT
80 × 40
13 × 24
1.5
1
MetOp-A
(2006 – present)
Aura
(2004 – present)
* GOME-1: loss of global coverage on 22 June 2003 ; instrument retired on 4 July 2011
30

FRESCO retrievals using simulated O
2
A band spectra for biomass burning aerosols
Aerosol layer
Cloud layer
Clear-sky
Cloudy
31
Wang et al., ACP, 2012

Australian Wildfires
Feb 7 th – Feb 12 th 2009
Figure: O. Tuinder, KNMI