Application of advanced EARLINET
methodologies on space-borne lidars for the
retrieval of higher level aerosol products
V. Amiridis
with input from the EARLINET
NOA - National Observatory of Athens
Vassilis Amiridis, CAMS GA, 15 June 2016
EARLINET can advance space-lidar retrievals
EARLINET
How?
1. By providing optical and microphysical properties per aerosol type in the form of
advanced aerosol models, of paramount importance for aerosol type classification
from space (e.g. developments within the ESA-CALIPSO, LIVAS, NATALI, HETEAC)
2. By using these multi-wavelength aerosol models for homogenizing NASA and ESA lidar
missions in order to derive a homogeneous multi-decadal climatic dataset.
3. By applying the EARLINET models for satellite product optimization.
4. By providing advanced methodologies and algorithms for the derivation of higher level
products. Some examples include the retrieval of:
- pure dust product
- fine and coarse particle mode
- particle concentrations
- CCN/IN particle concentrations
Vassilis Amiridis, CAMS GA, 15 June 2016
CALIPSO:
L1 product and L2 aerosol type
Vassilis Amiridis, CAMS GA, 15 June 2016
Aerosol classification scores
Aerosol Type Lidar
Ratio (sr)
(Omar et
al., 2009)
Agreement with
EARLINET and
airborne HSRL
Dust 40 80%
Marine 20 55%
Polluted Continental 70 54%
Polluted Dust 65 35%
Smoke 70 13%
Clean Continental 35 -
Aerosol classification is based on:
- Particle depolarization ratio
- Layer-integrated backscatter (~aerosol load)
- Surface information (land/ocean)
Vassilis Amiridis, CAMS GA, 15 June 2016
Dust case: Depolarization corrections
-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
2000
4000
6000
8000
10000
12000
14000
16000
Polluted Dust
Nu
mb
er
of
laye
rs
Particle depolarization ratio
CALIPSO
Eq. (4)
-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
2000
4000
6000
8000
10000
12000
14000
16000
Nu
mb
er
of
laye
rs
Particle depolarization ratio
CALIPSO
Eq. (4)
Dust
Amiridis et al., 2013Tesche et al., 2013
Vassilis Amiridis, CAMS GA, 15 June 2016
Dust case
CALIPSO LR = 40 sr
EARLINET LR = 58 sr
Multiple Scattering
effect for CALIPSO
Tesche et al., 2013
Wandinger et al., 2012
Vassilis Amiridis, CAMS GA, 15 June 2016
CALIPSO vs AERONET
Schuster et al., 2012
Vassilis Amiridis, CAMS GA, 15 June 2016
Dust case
Amiridis et al., 2013
Consolidating all EARLINET
findings and applying
corrections for dust on
CALIPSO observations
Vassilis Amiridis, CAMS GA, 15 June 2016
CALIPSO vs AERONET for dust
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Pearson's r = 0.905, Slope = 0.660
CA
LIO
P A
OD
- L
R=
40
sr
(53
2 n
m)
AERONET AOD (532 nm)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Pearson's r = 0.901, Slope = 0.629
CA
LIO
P A
OD
- L
R=
40
sr
(35
5 n
m)
AERONET AOD (355 nm)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0Pearson's r = 0.905, Slope = 0.956
CA
LIO
P A
OD
- L
R=
58
sr
(53
2 n
m)
AERONET AOD (532 nm)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0Pearson's r = 0.901, Slope = 0.912
CA
LIO
P A
OD
- L
R=
58
sr
(35
5 n
m)
AERONET AOD (355 nm)
CALIPSO-AERONET
Collocation
In pure Dust cases
from CALIPSO typing
Amiridis et al., 2013
Vassilis Amiridis, CAMS GA, 15 June 2016
CALIPSO vs MODIS for dust over ocean
CALIPSO-MODIS
Collocation
Red overpasses rejected
Amiridis et al., 2013
Vassilis Amiridis, CAMS GA, 15 June 2016
Pure dust discrimination from dust mixtures
0
1
2
3
4
5
0 1 2 3 4 5 6
0
1
2
3
4
5
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
0
1
2
3
4
5
0.000 0.001 0.002 0.003
Polluted
Dust
Dust
Clear Air
Aerosol Subtype
He
igh
t (k
m)
Particulate Depolarization Ratio
CALIPSO L2
mean Layer CALIPSO
mean Layer corrected
BSC. COEF. 532 nm (km-1sr
-1)
Total
Pure Dust
Other Type
with depolarization 0.03
=( − )(1 + )
( − )(1 + )Tesche et al., 2012
Vassilis Amiridis, CAMS GA, 15 June 2016
BSC-DREAM8b vs LIVAS dust product
0
1000
2000
3000
4000
5000
6000
7000
8000
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Version I
Version II
Version III
Pearson's r for Extinction
Heig
ht (m
)
Sahel0.27
Atlantic0.09
NW Africa-0.14
W.IP W.Med C.Med
E.Med
Abs. bias, AERONET - BSC-DREAM8b
Mo
de
l o
ve
restim
atio
nM
od
el
un
de
restim
atio
n
|bias|<0.1
Amiridis et al., 2013
Vassilis Amiridis, CAMS GA, 15 June 2016
10-year pure-dust LIVAS climatology
Marinou et al., 2015Vassilis Amiridis, CAMS GA, 15 June 2016
⇒ appearance of localized
regions of increased
extinction coefficient values
over mountains (the Alps,
Pyrenees, Carpathian
mountains)
Dust LIVAS product using EARLINET methods
Marinou et al., 2015
Vassilis Amiridis, CAMS GA, 15 June 2016
Domain Jan-Feb-Mar Apr-May-Jun Jul-Aug-Sept Oct-Nov-Dec
Dust Layer Height
Marinou et al., 2015
Vassilis Amiridis, CAMS GA, 15 June 2016
MACC: Using the LIVAS dust product for evaluation
Georgoulias et al., 2016
Vassilis Amiridis, CAMS GA, 15 June 2016
MACC: Using the LIVAS dust product for evaluation
Georgoulias et al., 2016Vassilis Amiridis, CAMS GA, 15 June 2016
In total: MACC overestimates by 22 %
EU CE EE SWE CM EM ATL CWSah ESah ME
MACC 0.132 0.033 0.043 0.049 0.068 0.071 0.071 0.178 0.173 0.248
LIVAS 0.108 0.012 0.016 0.026 0.051 0.050 0.047 0.186 0.123 0.215
MB 0.024 0.021 0.027 0.023 0.017 0.021 0.024 -0.008 0.050 0.033
NMB 22 % 175 % 169 % 89 % 33 % 42 % 51 % -4 % 41 % 15 %
Evaluation of MACC DOD550
MACC: Using the LIVAS dust product for evaluation
Georgoulias et al., 2016Vassilis Amiridis, CAMS GA, 15 June 2016
MACC underestimates DOD over regions with very high dust loadings
Evaluation of MACC DOD550
0
1
2
3
4
5
6
7
8
9
0.000 0.025 0.050 0.075 0.100 0.125 0.150
Extinction coefficient (km-1)
He
igh
t (k
m)
MACC
LIVAS
EUROPE
0
1
2
3
4
5
6
7
8
9
0.0000 0.0025 0.0050 0.0075
Mean Bias (km-1)
He
igh
t (k
m)
MACC: Using the LIVAS dust product for evaluation
Georgoulias et al., 2016
Vassilis Amiridis, CAMS GA, 15 June 2016
MACC: Using the LIVAS dust product for evaluation
Georgoulias et
al., 2016Vassilis Amiridis, CAMS GA, 15 June 2016
Evaluation of MACC profiles (EU)
• MACC overestimates ext. coeff. during the cold period
• MACC underestimates ext. coeff. during the warm period (high DODs)
• Over ~ 5 km MACC steadily overestimates ext. coeff.
Climate modeling (RegCM):
Fine-tuning using LIVAS dust climatology
Tsikerdelis et al., 2016
Vassilis Amiridis, CAMS GA, 15 June 2016
CCN/IN-relevant aerosol parameters
Mamouri and Ansmann, ACP 2016
Pure Dust Separation
+
Ext. Coef. →Ae
roso
l Pa
rtic
le C
on
cen
tra
tio
n APC=const*AECd
Vassilis Amiridis, CAMS GA, 15 June 2016
Vassilis Amiridis, CAMS GA, 15 June 2016
Ongoing validation of CCN/IN retrievals from remote sensing:
CATS under-flight of FAAM during ICE-D
Vassilis Amiridis, CAMS GA, 15 June 2016
First IN retrievals using CALIPSO
Vassilis Amiridis, CAMS GA, 15 June 2016
Conclusions
EARLINET provides the means and the methods for:
• Advancing space-borne lidar retrievals including CALIPSO, CATS and future
EarthCARE
• Optimizing satellite products using EARLINET-measured intensive properties (e.g.
lidar ratios)
• Providing higher level products using the advanced methodologies developed
within EARLINET and in the framework of the ACTRIS RI (e.g. pure-dust product,
dust concentrations in the fine and coarse mode, CCN/IN particle
concentrations)
Vassilis Amiridis, CAMS GA, 15 June 2016
Top Related