View
43
Download
0
Category
Preview:
DESCRIPTION
Diurnal asymmetry in the GERB(-like) fluxes: an update. C édric Bertrand Royal Meteorological Institute of Belgium, Brussels, Belgium. Cedric.Bertrand@oma.be. Radiance-to-flux conversion. Instantaneous SW flux estimation: (all scenes) RMIB: (1) (2) CERES:. - PowerPoint PPT Presentation
Citation preview
Diurnal asymmetry in the GERB(-like) fluxes: an update
Cédric Bertrand
Royal Meteorological Institute of Belgium,Brussels, Belgium. Cedric.Bertrand@oma.be
ADM Normalization Factor(to account for the nonlinear variation of the radiance within an angular bin)
Radiance-to-flux conversion
Instantaneous SW flux estimation: (all scenes)
RMIB: (1)
(2)
CERES:
),,(),,(
)( .3.3
mvmsmadm
mvmsmSEVIRIsm
SEVIRI RLF
),,(),,(),,(
)(
mvmsmmvmsm
adm
mvmsmmsm NFF
RL
),,(*)()(.3.3/
mvmsmH
smSEVIRIsmSEVIRIGERB cFF
GERB SW correction factor (resolution enhancement factor)
Interpolated CERES-TRMMADM ACF
RMIB/CERES COMPARISON
Instantaneous flux estimation: (clear ocean)
CERES:
),,,(
),(),(),,,(
),,()(
mvmsmj
admth
mth
mvmsmadm
mvmsmmsm wsNF
LwsRLwsRwsR
LF
Aerosol correction term
where Radm(ws,sm,vm,m) = determined from the wind speed-dependent ADMs
Rth(ws, Lm) and Rth(ws, Ladm) = anisotropic factors inferred from the
measured and ADM interpolated radiances, respectively. determined by comparing Lm(sm,vm,m) and Ladm(sm,vm,m) with LUT of theoretical SW radiances stratified by aerosol optical depth
MS-7 DATA ALL AVAILABLE 2003/07/02 DAY TIME SLOTS 4 SELECTED ZONES outside sun glint occurrence
acquisition geometry
SBDART RTM
clear sky SW radiancetime series
RMIB SBDART CERES
COMPARISON
R2F CONV.
Monthly mean clear sky SW flux difference at l.n. ± 2 hours (April 04)
GERB-likeSEVIRI
GERB
CERES-TRMMADMs
mW 2.
mW 2.
mW 2.
mW 2.
mW 2.
mW 2.
mW 2.
l.n. – (l.n.-2) l.n. – (l.n.+2)
Max: < 1.33 W.m-2
Avg.: 0.11 W.m-2
West: Max FSW TOA after local noon
East: Max FSW TOA before l.n.
Local noon = Min(SZA)
Selected ocean footprints (3x3 SEVIRI pixel resolution)
[38.74oN, 41.08oW]
[37.00oN, 0.00o E]
[37.70oN, 25.41oE]
[11.97oS, 28.09oW]
[11.69oS, 0.00oE]
[12.43oS, 46.78oE]
[20.80oS, 26.62oW]
[20.36oS, 0.00oE]
[21.37oS, 41,29oE]
SEVIRI spectral conversion (NB-to-BB)
LufSW(s,v,)=co(s) + c1(s) L0.6 m (s,v,) + c2(s) L0.8 m (s,v,) + c3(s) L1.6 m (s,v,)
Data base of theoretical SBDART spectral radiances curves at TOA (150/600 clear ocean scene types)
Fictitious AOD diurnal cycle
Retrieved AOD from LufSEVIRI
distinct diurnal cycle pattern = artifact
AOD sinusoidal shape function of sin(RAA): AODmin = 180o ; AODmax = 90o
Azimuth dependence of the diffuse radiation scattering phase function of the atmos.Aerosol scattering phase function aerosol type. Wrong model scattering angle dependent error in the retrieved AOD
Fictitious AOD diurnal cycle
LUTs of theoretical SW radiances DISORT RTM (assuming the Hess et al. (1998) maritim tropical aerosols model)
SEVIRI spectral modeling SBDART RTM (assuming all the default SBDART aerosols models)
Estimated SEVIRI BB SW radiances theoretical SBDART SW radiances
IF SBDART ocean surface = very different from the real ocean
ADM ACF adjustment sensitive to more than just the presence of aerosols
Ex.: ocean too bright: Rth (LSBDART (No Aero))/Rth (LADM) > 1 for the wrong reason
Sensitivity to the DISORT ocean surface is a second-order effect since
Rth (LSBDART (No Aero))/Rth (LADM) = ratio of DISORT ACFs in a particular angle
Numerical noise Cloud contamination
Old response enhanced sensitivity in the blue
Adoption of the new response will GERB SW radiance ADM ACF partial adjustment
),,,(
),(
),,(),,(.3,),,,(
),,(),,(
*
*.3
vsws
admws
vsvsSEVIRIwsvsws
vsvsNF
LR
LRR
L
thH
th
adm
HSEVIRI
cc
),,(),,,(
),(
).3,(),,,(
),,( *.3
vsvsws
admws
SEVIRIwsvsws
vs cH
th
th
adm
SEVIRI NF
LRLRR
L
),,,(),,(
),(
).3,(),,,(
),,( *.3
vswsvs
admws
SEVIRIwsvsws
vs NF
LRLRR
L cH
th
th
adm
SEVIRI
F3.SEVIRI(s)
FGERB/3.SEVIRI(s) estimation:
Numerical noise Cloud contamination
Conclusions:1. asymmetry in the diurnal evolution of GERB(-like) SW
flux is an ocean surface related problem
2. there is a discrepancy between the theoretical NB-to-BB conversion from SEVIRI and the expected (CERES) BB radiances (used as input in the ADM aerosol adjustment procedure)
3. since the aerosol adjustment is a ratio of theoretical anisotropic factors that depends upon the departure of the observed BB radiance to the mean radiance used to develop the ADMs, introducing new “observed” BB radiances that are inconsistent with CERES radiances can lead to problems.
Conclusions:
4. correcting the SEVIRI based spectral modeling by the GERB measurement allows to reduce part of the disagreement between the estimated and the ADM radiances
5. The GERB TOA reflected SW flux over clear ocean surfaces should be estimated from the corrected SEVIRI BB radiance and the RGP should consider the new spectral response functions
IF SBDART ocean surface = very different from the real ocean
ADM ACF adjustment sensitive to more than just the presence of aerosols
Ex.: ocean too bright: Rth (LSBDART (No Aero))/Rth (LADM) > 1 for the wrong reason
Sensitivity to the DISORT ocean surface is a second-order effect since
Rth (LSBDART (No Aero))/Rth (LADM) = ratio of DISORT ACFs in a particular angle
Recommended