CERES shortwave (SW), longwave (LW), and window (WN) channel radiative
fluxes are derived from empirical Angular Distribution Models (ADMs) that
convert a measured radiance in a given Sun-Earth-satellite viewing
configuration to a top-of-atmosphere (TOA) radiative flux. The first set of
CERES ADMs were developed from 9 months of CERES TRMM SSF measurements over the
Tropics (Loeb et al., 2003a,b). These ADMs were used to generate TOA fluxes on
the CERES TRMM Edition2B SSF product. The TRMM ADMs were also used to produce
TOA fluxes for the CERES Terra Edition1A SSF product while a new set of global
ADMs based on CERES Terra data were being developed. Once the new global CERES
Terra ADMs were completed (Loeb et al., 2004), these were used to generate TOA
fluxes on the CERES Terra Edition2B SSF product. Loeb et al. (2006) provide a
detailed assessment of the uncertainties in CERES Terra Edition2B SSF fluxes.
The CERES Aqua Edition1B SSF data product used the CERES Terra ADMs while a
new set of ADMs based on CERES Aqua data were being developed. The methodology
used to develop the new Aqua ADMs is based on the same approach as that used to
develop Terra ADMs (Loeb et al., 2004). The new Aqua ADMs are now complete and
are being used to generate TOA fluxes on the CERES Aqua Edition2A SSF product.
The new Aqua ADMs are based on 24 months (August 2002 - June 2004). The main
features of the new Aqua ADMs are given below. For a more detailed description
of the CERES Terra ADMs see Loeb et al. (2004).
Empirical ADMs over snow and sea-ice.
Increased angular resolution in nonpolar regions (2 degrees)
Use of "continuous" SW and LW ADM scene types over ocean, land, and desert.
Monthly 1 degree regional clear land+desert ADMs.
Neural network scheme to improve TOA flux estimates for footprints with
excessive "no retrievals". No retrievals can occur when imager data
is missing or when the cloud algorithm cannot provide a physical retrieval. If
the scene characteristics over 35% or more of a CERES FOV are unknown, TOA
fluxes are estimated using a neural network scheme. Approximately 3% of the
CERES data fall in this category.
While CERES Terra and CERES Aqua TOA flux errors are generally
quite similar (Loeb et al., 2006), differences are observed in polar regions.
Biases in daytime SW TOA fluxes from Aqua ADMs are significantly smaller
than Terra fluxes over sea-ice. In contrast, when the new Aqua ADMs were
first applied to infer LW TOA fluxes over the Antarctic Plateau, Aqua TOA flux
errors were found to be significantly larger than those from Terra ADMs. The
cause for these discrepancies is associated with differences between the
Terra and Aqua CERES polar cloud mask applied to MODIS. To avoid
the biases in Aqua LW TOA fluxes over the nighttime Antarctic Plateau, we've
replaced the original Aqua nighttime LW permanent snow ADMs with those
developed on Terra in Ed2A processing. A more detailed discussion of the
differences is provided in Loeb et al. (2006).
The tables below provide additional details on Aqua ADM scene classification
and validation results. Further details can be found in the following:
CERES Aqua Shortwave Channel ADMs for Different Scene Types
Scene Type
Description
Clear Ocean
Function of wind speed; Correction for aerosol optical depth included.
Cloud Ocean
Function of cloud phase; Continuous function of cloud fraction and cloud
optical depth (5-parameter sigmoid).
Land & Desert Clear
1°- regional monthly ADMs using Analytical Function of TOA BRDF (Ahmad
and Deering, 1992).
Land & Desert Cloud
Function of cloud phase; continuous function of cloud cover and cloud
optical depth; used 1°-regional clear-sky BRDFs to account for background
albedo.
Permanent Snow
Function of Cloud Fraction, Surface Brightness, cloud optical depth
Fresh Snow
Function of Cloud Fraction, Surface Brightness, Snow Fraction, cloud optical depth
Sea-Ice
Function of Cloud Fraction, Surface Brightness, Ice Fraction, cloud optical depth
CERES Aqua Longwave and Window Channel ADMs for Different Scene Types
Scene Type
Description
Clear Ocean, Land, Desert
Function of Ocean, Forest, Cropland/Grass, Savanna, Bright Desert, Dark
Desert, precip. water, lapse rate, skin temperature.
Clouds Over Ocean, Land Desert
Function of precip. water, skin temp, sfc-cloud temp. diff; continuous
function of parameterization involving cloud fraction, cloud and sfc
emissivity, sfc and cloud temp.
Permanent Snow, Fresh Snow, Sea-Ice
Each a function of cloud fraction, sfc temp, sfc-cld temp diff
Instantaneous TOA Flux Consistency (nadir vs θ=50-60 deg)
No-Glint Ocean Scene Type SW TOA Flux RMS (%)
Aqua Ocean: All-sky RMS = 4.23% (6.04%)
CLR
PCL
MCL
OVC
5.28
High
10.21
6.70
3.99
Mid
16.01
5.62
4.20
Low
8.33
10.06
10.63
6.13
25.50
3.78
3.61
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Ocean: All-sky RMS = 4.07% (5.21%)
CLR
PCL
MCL
OVC
6.27
High
11.64
6.84
3.72
Mid
10.73
5.65
4.15
Low
7.77
9.68
7.31
5.28
28.51
3.06
3.56
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Land SW TOA Flux RMS (%)
Aqua Land: All-sky RMS = 6.58% (5.03%)
CLR
PCL
MCL
OVC
3.05
High
3.01
9.79
8.11
4.65
Mid
4.63
7.85
4.77
3.06
Low
10.59
7.65
10.81
4.83
4.43
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Land: All-sky RMS = 5.77% (4.90%)
CLR
PCL
MCL
OVC
4.12
High
4.09
23.27
6.01
3.92
Mid
5.30
16.38
7.36
5.04
4.11
Low
8.65
8.10
16.53
5.97
6.43
4.36
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Snow/Ice SW TOA Flux RMS (%)
Aqua Snow/Ice: All-sky RMS = 8.23% (7.40%)
CLR
PCL
MCL
OVC
5.02
High
6.04
8.65
8.79
Mid
5.37
10.12
6.19
10.16
7.35
6.97
Low
9.79
17.50
11.54
14.18
8.85
5.17
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Snow/Ice: All-sky RMS = 8.50% (8.17%)
CLR
PCL
MCL
OVC
7.22
High
11.26
9.97
6.19
9.40
Mid
6.21
12.42
7.08
9.54
10.28
4.87
15.27
Low
10.05
17.01
9.04
10.77
8.50
5.98
6.47
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Ocean LW TOA Flux RMS (%)
Aqua Ocean: All-sky RMS = 2.39% (2.30%)
CLR
PCL
MCL
OVC
1.06
High
1.05
6.04
7.69
5.97
Mid
1.02
4.72
2.28
Low
0.73
0.71
1.05
0.99
1.52
1.96
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Ocean: All-sky RMS = 2.66% (2.63%)
CLR
PCL
MCL
OVC
0.97
High
1.95
6.35
7.79
7.31
Mid
5.13
4.47
4.84
Low
0.96
1.08
1.06
2.42
2.51
2.24
2.42
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Land LW TOA Flux RMS (%)
Aqua Land: All-sky RMS = 2.94% (1.97%)
CLR
PCL
MCL
OVC
1.60
High
2.42
2.41
6.75
5.90
Mid
2.15
3.42
3.77
3.35
Low
1.57
1.53
1.18
1.35
1.64
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Land: All-sky RMS = 53.41% (2.09%)
CLR
PCL
MCL
OVC
1.85
High
2.65
5.62
8.51
7.58
Mid
2.00
1.88
3.69
3.94
2.07
Low
2.00
1.96
3.26
1.43
1.27
2.25
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Snow/Ice LW TOA Flux RMS (%)
Aqua Snow/Ice: All-sky RMS = 2.13% (2.26%)
CLR
PCL
MCL
OVC
2.27
High
2.20
5.01
7.46
Mid
2.96
1.52
2.05
2.50
1.84
3.40
Low
2.19
1.37
1.96
1.61
2.43
1.61
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Terra Snow/Ice: All-sky RMS = 2.78% (2.87%)
CLR
PCL
MCL
OVC
2.85
High
2.58
2.39
5.15
5.16
4.10
Mid
2.35
5.59
2.36
2.23
4.00
4.85
3.84
Low
2.12
2.34
2.45
2.22
5.48
2.19
5.54
Thin
Mod
Thick
Thin
Mod
Thick
Thin
Mod
Thick
Regional TOA Flux Uncertainties
Regional mean all-sky SW TOA flux bias and RMS error
for Aqua and Terra by season for December 2002 - November
2003
Aqua
Terra
Season
Bias (W m-2)
RMS (W m-2)
Bias (W m-2)
RMS (W m-2)
January
0.13
0.75
0.11
0.98
April
-0.16
0.70
0.00
0.99
July
-0.02
1.08
0.02
1.37
October
-0.10
0.68
0.11
0.76
Regional mean all-sky LW TOA flux bias and RMS error for Aqua
and Terra by season for December 2002 - November 2003
Aqua
Terra
Season
Bias (W m-2)
RMS (W m-2)
Bias (W m-2)
RMS (W m-2)
January
0.26
0.51
0.17
0.57
April
0.31
0.58
0.35
0.66
July
0.25
0.56
0.33
0.63
October
0.22
0.55
0.27
0.58
Regional mean all-sky WN TOA flux bias and RMS error for Aqua
and Terra by season for December 2002 - November 2003.
Aqua
Terra
Season
Bias (W m-2)
RMS (W m-2)
Bias (W m-2)
RMS (W m-2)
January
0.16
0.24
0.17
0.28
April
0.19
0.27
0.23
0.34
July
0.17
0.27
0.22
0.32
October
0.16
0.25
0. 20
0.29
Comparison Between Edition1B and Edition2A Aqua TOA Fluxes
CERES Aqua Edition2A mean instantaneous all-sky TOA flux and CERES Aqua
Edition1B minus Edition2A mean instantaneous all-sky TOA flux difference
and RMS difference for January 2003 and July 2003
January 2003
Mean Flux (W m-2)
Mean Diff (W m-2)
RMS Diff (W m-2)
SW
248.6
-0.53
1.5
LW
240.3
0.55
0.61
WN
66.8
-0.043
0.092
July 2003
Mean Flux (W m-2)
Mean Diff (W m-2)
RMS Diff (W m-2)
SW
222.5
-0.82
1.8
LW
248.3
0.57
0.63
WN
71.1
-0.047
0.094
CERES Aqua Edition2A mean all-sky SW TOA flux (top), CERES Aqua
Edition1B minus Edition2A mean all-sky SW TOA flux difference (middle) and
RMS difference (bottom) for January 2003.
CERES Aqua Edition2A mean all-sky SW TOA flux (top), CERES Aqua
Edition1B minus Edition2A mean all-sky SW TOA flux difference (middle) and
RMS difference (bottom) for July 2003.
CERES Aqua Edition2A mean all-sky LW TOA flux (top), CERES Aqua
Edition1B minus Edition2A mean all-sky LW TOA flux difference (middle) and
RMS difference (bottom) for January 2003 (daytime only).
CERES Aqua Edition2A mean all-sky LW TOA flux (top), CERES Aqua
Edition1B minus Edition2A mean all-sky LW TOA flux difference (middle) and
RMS difference (bottom) for July 2003 (daytime only).