Correction of Cloud Contamination
on AMSU Measurements
Xiaofan Li and Fuzhong Weng
NOAA/NESDIS/ORA
Camp Springs, MD 20746, USA
Presented at ITSC-XII 2/27,2002, Lorne Australia
AMSU Weighting Functions
Advanced Microwave Sounding Unit
Imaging and Temperature Sounding Channels
23.8 GHz, Ch1
emission from vapor,
clouds, surface
31.4 GHz, Ch2
emission from clouds, surface
52.8 GHz, Ch4
emission from surface
53.7 GHz, Ch5
cloud contam.
cloud contam
Advanced Microwave Sounding Unit
Imaging and Moisture Sounding Channels
89 GHz
surface emission
150 GHz
cloud scattering
rain scattering
183+-1 GHz
upper level vapor & cloud
ice scattering
183+-3 GHz
middle level vapor & cloud
ice scattering
Motivation
• Investigate the effects of precipitating
clouds on AMSU sounding channels using
cloud resolving model outputs,
• Correct the cloudy AMSU radiances to the
clear ones so that vertical temperature and
moisture profiles are retrieved.
Cloud Resolving Model
•
•
•
•
Non-hydrostatic and Anelastic (Tao and Simpson 1993)
Prognostic equations for T, qv, qc, qr, qi, qs, qg
Radiation (Chou and Suarez 1994, Chou et al. 1998)
Cloud Microphysics (Rutledge and Hobbs 1983 1984,
Lin et al. 1983, Tao et al. 1989, Krueger et al. 1995)
• Turbulence Closures
• Imposed spatial-uniform large-scale vertical velocity,
zonal wind, SST, horizontal temperature and moisture
advections
Cloud Resolving Model
•
•
•
•
Two dimension: x-z
Domain: 768 km
Horizontal Resolution: 1.5 km
Vertical resolution: 200 m near surface to 1 km
above 15 km
• Time step: 12 Seconds
Model Initial Fields
W (cm/s)
U (m/s)
SST
Simulated Surface Rain Rate
2pm
Cloud Model Outputs
Radiative Transfer Modeling
Fast Atmospheric and Surface POlarimetric Radiative
Transfer Model (FASMPORT) (Liu and Weng, 2002; JAS)
250
16-stream (I)
pol. two-stream (I)
16-stream (Q)
pol. two-stream (Q)
two-stream (I)
10
Q
240
230
9
8
220
7
210
6
I
200
190
4
3
170
2
160
1
0
150
30
60
90
Local zenith angle (degree)
•
•
Atmosphere:
– Gaseous constituents
– Hydrometeors
5
180
0
•
Polarimetric two-stream
Various surface & atmospheric
components
•
Surface:
– Coherent media: new sea
ice, smooth bare soil
– Clustering media: canopy
– Random media: snow, desert
– Rough surface: ocean
Microwave Emissivity Model
(Weng et al, 2001; JGR)
0
Surface Emissivity Spectra (θ=53 )
0.9
Snow
Canopy
0.8
0.7
Bare Soil
Wet Land
0.6
0.5
Desert
Ocean
0.4
0.3
0.2
0
20
40
60
80 100 120 140 160 180 200
Frequency (GHz)
0
Surface Emissivity Spectra (θ=53 )
1.0
Emissivity at H-Polarization
• Open water – two-scale
roughness theory
• Sea ice - Coherent
reflection
• Canopy – Four layer
clustering scattering
• Bare soil – Coherent
reflection and surface
roughness
• Snow/desert – Random
media
Emissivity at V-Polarization
1.0
0.9
0.8
Snow
0.7
Canopy
Bare Soil
0.6
Wet Land
0.5
Desert
0.4
Ocean
0.3
0.2
0
20
40
60
80
100 120 140 160 180 200
Frequency (GHz)
Scattering and Emission Processes
Phenomenology.
• Large gravity waves, whose
wavelengths are long
compared with the radiation
wavelength.
• Small capillary waves, which
are riding on top of the largescale waves, and whose RMS
height is small compared with
radiation wavelength.
• Sea foam, which arises as a
mixture of air and water at the
wind roughened ocean surface,
and which leads to a general
increase in the surface
emissivity.
Methodology: two-scale model
• Large-scale wave is simulated
as an ensemble of tilted facets
each acting as an infinitely large
specular surface
•Small scale-wave is approximated
by small perturbation theory
Ocean Roughness Model
Large-scale roughness is dependent on gravity waves whereas
small scale irregularities are affected by capillary waves. There
are coherent reflection and incoherent scattering associated with
the waves in both scales
coherent
Small scale
foam
incoherent
downwind
Large scale
upwind
crosswind
Oceanic Emissivity
NRL Measurements
Two-scale Simulations
Variation of U at 37 GHz with relative azimuth angle for wind
speeds of 5m/s, 10m/s, and 15m/s. SST = 300 K.
5
4
4
3
3
15m/s
2
2
6m/s
1
0
-1 0
4m/s
90
-2
180
270
360
U (unit: K)
Stokes Component U (K)
Variation of U at 37 GHz with relative azimuth angle for wind
speeds of 4m/s, 6m/s, 10m/s, and 14m/s. SST = 300 K.
10m/s
5m/s
1
0
-1
0
90
270
360
10m/s
-2
-3
-3
-4
14m/s
-5
-4
Relative Azimuth Angle (degree)
RElative Azimuth Angle (degree)
Variation of V at 37 GHz with relative azimuth angle for wind
speeds of 5m/s, 10m/s, and 15m/s. SST = 300 K.
Variation of V at 37 GHz with relative azimuth angle for wind
speeds of 4m/s, 6m/s, 10m/s, and 14m/s. SST = 300 K.
0.8
1.5
10m/s
0.6
1
5m/s
0.4
0.5
0
0
-0.5
90
4m/s
6m/s
180
10m/s
-1
14m/s
-1.5
270
360
V (unit: K)
Stokes Component U (K)
180
0.2
0
-0.2
-0.4
0
90
180
15m/s
-0.6
-0.8
Relative Azimuth Angle (degree)
Relative Azimuth Angle (degree)
270
360
Channel Correlation from
Simulation and Measurements
Observation
CH1 vs. CH2
CH4 vs. CH5
Observation
Simulation
Observation
Observation
CH16 vs. CH17
CH19 vs. CH20
Cloud Effects on AMSU-A
Sounding Channels
Channel 4
Channel 5
Cloud Effects on AMSU-B Sounding
Channels
Tb at CH 19 vs. IWP
Tb at CH 20 vs. IWP
Effects of Cloud Top
Correction to Cloudy TB
Summary and Conclusions
• AMSU sounding channels are sensitive to
precipitating clouds,
• The brightness temperature depression is
highly correlated with cloud ice water path
than cloud liquid water path,
• Thus, the correction is best made with cloud ice
and liquid water paths if both are required at
a sufficient dynamic range.