Post on 09-Sep-2018
A comparison of CERES and GEOS clear-sky fluxes
A. E. DesslerDepartment of Atmospheric Sciences
Texas A&M University
The problem
• CERES calculates “clear sky” fluxes: what the flux would be if clouds were instantaneously removed
• important for calculating the effects of clouds
• “clear-sky” measurements from CERES are not pure measurements
• we would like to evaluate them
2
6
AIRS T and q profiles and Ts
CERES SSF data
Clear-sky only
Clear-sky Column Radiative Transfer
Model
7
AIRS T and q profiles and Ts
CERES SSF data
OLR
Clear-sky only
Clear-sky Column Radiative Transfer
Model
8
Clear-sky Column Radiative Transfer
Model
AIRS T and q profiles and Ts
CERES SSF data
OLR
Clear-sky only
9
Dessler et al., An analysis of the dependence of clear-sky top-of-atmosphere outgoing longwave radiation on atmospheric temperature and water vapor, J. Geophys. Res., 113, D17102, DOI: 10.1029/2008JD010137, 2008.
March 2005, over ocean, 70°N-70°S, nighttime
10
Dessler et al., An analysis of the dependence of clear-sky top-of-atmosphere outgoing longwave radiation on atmospheric temperature and water vapor, J. Geophys. Res., 113, D17102, DOI: 10.1029/2008JD010137, 2008.
March 2005, over ocean, 70°N-70°S
Chou minus CERES 3.7±2.1 W/m^2
Fu-Liou minus CERES -0.01±2.1 W/m^2
11
Dessler et al., An analysis of the dependence of clear-sky top-of-atmosphere outgoing longwave radiation on atmospheric temperature and water vapor, J. Geophys. Res., 113, D17102, DOI: 10.1029/2008JD010137, 2008.
March 2005, over ocean, 70°N-70°S
Chou minus CERES 3.7±2.1 W/m^2
Fu-Liou minus CERES -0.01±2.1 W/m^2
This method allows us to calculate the clear-sky flux
GEOS5 reanalysis
• Assimilates AIRS q and T fields• GEOS5 reanalysis includes a clear-sky
radiative flux calculation• this flux is another way to verify clear-sky
flux
12
14
GEOS5 reanalysis clear-sky fluxes
CERES SSF
OLR
8-times daily, interpolate nearest time to location of CERES
> 96% clear
16
January 2006
blue line is best fitCERES = 1.00039 GEOS + 7.1598.8% of variance explained
constant offset between GEOS and CERES for SW
17
January 2006400
350
300
250
200
150CE
RES
clear
-sky
lw fl
ux (W
/m2)400350300250200150
GEOS clear-sky lw flux (W/m2)
CERES 96% clear
18
January 2006
blue line is best fitCERES = 0.986 GEOS + 10.0598.7% of variance explained
400
350
300
250
200
150CE
RES
clear
-sky
lw fl
ux (W
/m2)400350300250200150
GEOS clear-sky lw flux (W/m2)
19
SSF vs. GEOS
For SW, the avg. difference is similar to the intercept because the slope = 1, for LW, the avg. difference is less because the slope is < 1.
MERRA vs. EBAF
• Energy Balance and Fill• CERES group’s best estimate of TOA flux• 3/2000-10/2005• Compare monthly global average clear-sky
fluxes from EBAF & MERRA
20
21
274
272
270
268
266LW
clea
r-sky
flux
2006200520042003200220012000Year
EBAF MERRA
56
54
52
50
48
SW cl
ear-s
ky flu
x
2006200520042003200220012000Year
EBAF MERRA
1.3 W/m2
0.2 W/m2
22
-1.0
-0.5
0.0
0.5
LW cl
ear-s
ky flu
x
2006200520042003200220012000Year
EBAF MERRA
0.80.60.40.20.0
-0.2-0.4-0.6
SW cl
ear-s
ky flu
x
2006200520042003200220012000Year
EBAF MERRA
23
-1.0
-0.5
0.0
0.5
LW cl
ear-s
ky flu
x
2006200520042003200220012000Year
EBAF MERRA
0.80.60.40.20.0
-0.2-0.4-0.6
SW cl
ear-s
ky flu
x
2006200520042003200220012000Year
EBAF MERRA
Slope = 0.41Variance = 84%
Slope = 0.79Variance = 90%
Conclusions• Comparisons to reanalysis provide useful tests of
measurements of TOA clear-sky fluxes• Comparisons of MERRA vs. EBAF show
– difference of clear-sky LW and SW shows distinct seasonal cycle
– after removing the seasonal cycle, EBAF shows greater variations than MERRA in both LW and SW
– EBAF and MERRA show similar interannual variations
• Next step: look at lat-lon differences24