ATS770-presentation3-nanfeng

Report
Earth Radiation Budget
Satellite Remote Sensing II - ATS 770 Presentation3
October 26th, 2011
By
Nan Feng
Department of Atmospheric Sciences
The University of Alabama in Huntsville
Huntsville, AL
Outline
I.
II.
III.
IV.
V.
Introduction
Uncertainties
Angular Distribution Models (ADM)
Validations
Conclusion
Introduction to Earth’s Radiation Budget
• Absorption of Insolation and Emission of terrestrial radiation drive the
General Circulation of the Atmosphere
Introduction to Earth’s Radiation Budget
Largely responsible for Earth’s weather and climate
IPCC report 2007
Medium-Low Level
Understanding
Direct and indirect effects of tropospheric aerosols
Increased
planetary
albedo:
Scattering solar
COOLING!
Decreased
planetary
albedo:
Absorbing solar
WARMING!
Surface
Impact clouds
and
precipitation
processes
COMPLICATED!
Aerosol climate impact
Direct effects
•Scattering solar energy
•Absorbing solar/terrestrial energy
Indirect effects (Modify cloud properties)
•More droplets----clouds are brighter (Twomey, 1977 )
•More droplets----longer cloud life time (Albrecht,1989)
Semi-direct effect
•Absorbing aerosols heat airs and evaporate clouds
(Hansen et al., 1997)
ARF Estimation
Aerosol Radiative Forcing = FCLEAR-SKY – FAEROSOL
How can we study earth radiation budget ?
•Global Climate Models
•Global Observations
•Satellite measurements
of radiative quantities.
Spectral Categories Instruments for Radiation budget :
• Narrowband Sensors
• Broadband Sensors
Field of View Categories :
• Wide field-of-view (WFOV) Nonscanner
- Often called FLAT-PLATE sensors
- Measure radiation horizon to horizon
- 120o angular resolution
- Example : Nimbus7 ERB, ERBE, CERES
- Longer lifetime due to less wear
• Narrow field-of-view (NFOV)  Scanner
- AVHRR, Nimbus 6,7 ERB, ERBE, CERES
The Earth Radiation Budget Experiment (ERBE)
Mission
The Goddard Space Flight Center built the Earth Radiation Budget
Satellite (ERBS) on which the first ERBE instruments were launched by
the Space Shuttle Challenger in 1984.
ERBE instruments were also launched on two National Oceanic and
Atmospheric Administration weather monitoring satellites, NOAA 9 and
NOAA 10 in 1984 and 1986.
Both had two instruments : Scanner & Non-Scanner
http://eosweb.larc.nasa.gov/PRODOCS/erbe/table_erbe.html
ERBE Observed Global Longwave Radiation
The Clouds and Earth Radiant Energy System (CERES)
CERES SCAN MODES
Unique feature!
Rotating Azimuth Plane
Cross-Track Scan mode
CERES Spatial Coverage
TRMM-PFM
Terra-FM1/FM2
Aqua-FM3/FM4
CERES Scan Modes
Cross-Track
(FAPS)
CLAMS-Scan
(RAPS & PAPS)
Special-Scan
(PAPS)
Design Specifications
Orbits: 705 km altitude, 10:30 a.m. descending node (Terra) or 1:30 p.m. ascending
node (Aqua),
sun-synchronous, near-polar;
350 km altitude, 35° inclination (TRMM)
Spectral Channels:
Shortwave :
Window:
Total:
0.3 - 5.0 µm
8 - 12 µm
0.3 to 200 µm
Swath Dimensions:
Limb to limb
Angular Sampling:
Cross-track scan and 360° azimuth biaxial scan
Spatial Resolution:
20 km at nadir (10 km for TRMM)
CERES has four main objectives:
• Provide
a continuation of the ERBE record of radiative fluxes at
the top of the atmosphere (TOA) , analyzed using the same
algorithms that produced the ERBE data
• Double the accuracy of estimates of radiative fluxes at TOA and
the Earth's surface
• Provide the first long-term global estimates of the radiative
fluxes within the Earth's atmosphere
• Provide cloud property estimates that are consistent with the
radiative fluxes from surface to TOA
Limitations:
• Inadequate diurnal variation, only twice daily observations
(diurnal problem)
• Satellite sensors do not measure exactly the wavelength
integrated radiation budget quantities (spectral correction
problem or unfiltering problem)
• Radiance-to-flux conversion (angular dependence problem)
The uncertainties of ERB studies
•
Radiance calibration
•
Filtered to Unfiltered Radiances
•
Cloud contamination
•
Clear Sky Estimation
•
Radiances to flux Conversion - ADM
Radiance to Flux Conversion
•
Satellite measures radiance (I(o,,)) at a
given sun-satellite geometry during overpass
•
•
This radiance must be converted to flux
If surface is Lambertian, then for isotropic
scattering, flux F(o) = π * I(o,,)
•
2
 /2
However, for non-lambertian
surfaces, the
F (not
 0 )isotropic
  d  dI( 0 , , ) cos  sin 
scattering is
 0
 0
Anisotropic
Scattering
Isotropic
Scattering
Forward
Backward
Radiance to Flux Conversion
•
•
•
Angular measurements can be integrated to obtain nonLambertian flux (F(o))
Anisotropic factor or angular distribution model = The Ratio
of the Lambertian flux to non-Lambertian flux
ADM = Ratio of equivalent Lambertian flux to actual flux
• R(o,,) = π*I(o,,)/ F(o)
ADMs (Sorting-into-Angular-Bins, SABs)
Large ensemble of radiance measurements are first
sorted into discrete angular bins and parameters that
define an ADM scene type and ADM anisotropic
factors for a given scene type(j) are given by
 I j(oi , k ,l )
R j(oi , k ,l ) 
Fj(oi )
where : I j is the average radiance (corrected for Earth-sun
distance in the SW) in an angular bin (oi , k ,l ) , Fj(oi )
is the upwelling flux in a solar zenith angle bin, which is
determined by directly integrating I j over all angles (Loeb
et al., 2003). The set of angles oi, k, and .
Examples ADMs as the function of 0.55
0.55: 0.0-0.1
0.55: 0.2-0.4
> 0.6
0.55:0.1-0.2
Glint
0.55: 0.2-0.4
< 0.6
ADM Scene Identification
The main reason for defining ADMs by scene type is to
reduce the error in the albedo estimate.
Earth scenes have distinct anisotropic characteristics
which depend on their physical and optical
properties. (e.g. thin vs thick clouds; cloud-free,
broken, overcast, etc.)
Scene identification must be self-consistent. Biases in
cloud property retrievals (e.g. due to 3D cloud effects)
should not introduce biases in flux/albedo estimates.
CERES Single Scanner Footprint (SSF) Product
Coincident CERES radiances and imager-based cloud and
aerosol properties
Use VIRS (TRMM) or MODIS (Terra or Aqua) to determine
following in up to 2 cloud layers over every CERES FOV:
Macrophysical: Factional coverage, Height, Radiating Temperature, Pressure
Microphysical: Phase, Optical Depth, Particle Size, Water Path
Clear Area: Albedo, Skin Temperature, Aerosol optical depth
Scene Types for CERES/TRMM SW ADMs
ADM Category
Clear
Ocean
Land
Desert
Snow
Ocean
Land
Cloud
Desert
Snow
Total
Scene Type Stratification
- 4 Wind Speed Intervals
- 2 IGBP Type Groupings
- Bright and Dark
- Theoretical
- Liquid and Ice
- 12 Cloud Fraction Intervals
- 14 Optical Depth Intervals
- 2 IGBP Type Groupings
- Liquid and Ice
- 5 Cloud Fraction Intervals
- 6 Optical Depth Intervals
- Bright and Dark Deserts
- Liquid and Ice
- 5 Cloud Fraction Intervals
- 6 Optical Depth Intervals
- Theoretical
Actual
Total
4
2
2
1
62 (L)
53 (I)
45
33
1
203
Scene Types for CERES/TRMM LW and WN ADMs
ADM Category
Clear
Broken Cloud
Field
(4 intervals)
Overcast
Parameter Stratification
3 Precipitable Water
Ocean
5 Vertical Temperature
Change
Land
3 Precipitable Water
5 Vertical Temperature
Change
Desert
3 Precipitable Water
5 Vertical Temperature
Change
3 Precipitable Water
6 DT (Sfc-Cloud)
Ocean/Land/De
sert
4 IR Emissivity
Ocean+
Land+Desert
3 Precipitable Water
7 DT (Sfc-Cloud)
6 IR Emissivity
Total
15
15
15
288 (O)
288 (L)
288 (D)
126
TRMM ADMs
Better scene identification and Increased ADM sensitivity to
anisotropy
• using collocated VIRS and CERES data.
• VIRS is a narrowband imager – 2km spatial resolution
•
CERES has footprint of 10km (TRMM) at nadir
• 200 shortwave and 100 longwave scene types
http://asd-www.larc.nasa.gov/Inversion
Loeb et al., 2003; JAM, 42, 240-265
Loeb et al., 2003; JAM, 42, 1748-1769
Comparisons between TRMM and Terra CERES
TRMM
• Only 9 months of data (Jan-Aug, 1998 + March 2000)
• Spatial coverage limited to ±38o only
• 350 km precessing orbit with 35o inclination  46 days
for full range of SZA
• land cover types = only 4 categories based on IGBP
TERRA
• global coverage
• increased sampling
• Data available since 2000
• need for new ADMs because spatial resolution and
geographic coverage different
Terra CERES ADMs
CERES Terra SW ADMs – (a) Ocean – (1) Clear
Conditions: MODIS pixel-level cloud cover fraction less or equal than 0.1%
Instantaneous TOA fluxes are determined using combination of empirical
and theoretical ADMs as follows:
(0 , , )
=
ℎ ( , )
  , 0 , ,  [ ℎ
]
 ( , )
  , 0 , ,  is determined from wind speed-dependent empirical ADMs
that are derived from CERES data
ℎ  ,   ℎ ( , ) are theoretical radiative transfer model anisotropic
factors evaluated at the measured CERES radiance (0 , , ) and mean CERES
radiance ( , 0 , , ) in a given ADM angular bin, respectively.
SW ADMs – (a) Ocean – (2) Clouds
Continuous ADMs using analytical functions that relate CERES radiances
and imager parameters (e.g. cloud fraction and cloud optical depth.)
  = 
Where,  is the retrieved
Cloud optical depth of the ith
Pixel within the CERES FOV
 Try to combine f and 
into a single parameter
 Third order polynomial
 Five paras sigmoidal fit

 = 0 +

− − 0 
 ]
[1+
SW ADMs – (a) Ocean – (2) Clouds
Continuous ADMs using analytical functions that relate CERES radiances
and imager parameters (e.g. cloud fraction and cloud optical depth.)
The sigmoidal fit relative error
remains less than 1% in every cloud
fraction interval
The polynomial fit relative error
reaches -3% at intermediate cloud
fractions
 The
Similar
are obtained
closeresults
relationship
btw SW
when
other
bins are
radiance
andangular
()occurs
in spite
considered
when
separate
fits
of the ratherorlarge
range
of cloud
are
derivedassociated
for mixed-phased
and
properties
with a given
ice
clouds.
(
) range
 In general, the rms error in
predicting instaneous SW radiances
using the sigmoidal fit is btw 5%
and 10%.
SW ADMs – (a) Ocean – (2) Clouds
Continuous ADMs using analytical functions that relate CERES radiances
and imager parameters (e.g. cloud fraction and cloud optical depth.)
SW ADMs – (a) Ocean – (2) Clouds
 In each solar zenith angle interval, the liquid water clouds show well-defined
peaks in anisotropy for  = - 30 to -60 and close to nadir due to the cloud glory
and rainbow features, while peaks in anisotropy occur for ice clouds between  =
30 to 60 in the specular reflection direction, also observed by Chefer et al. (1999)
in POLDER measurements. Likely due to horizontally oriented ice crystals.
ADMs for Terra CERES:
1. Shortwave:
- Clear Land: Stratify by IGBP type + vegetation index + taer
1×1 latitude and longitude equal area regions with a
temporal resolution of 1 month
- Clouds over Land: Continuous scene type using sigmoidal
functional fits to data.
- Clear Snow/ice: Stratify by NDSI (permanent snow, fresh snow,
or sea ice. Further stratified into ‘bright’ and dark subclasses)
- Clouds over Snow: greater dependence on vza than cloud free
scence.
2. Longwave and Window:
- Cloud-free conditions: more surface types and high angular bins
resolutions (Stratified by precipitable water, imager-based surface
skin temperature and etc.)
- Cloudy conditions: a function of precipitable water, surface and
cloud top temperature, surface and cloud top emissivity and
cloud fraction.
Terra ADMs
Improvements :
•
using collocated MODIS and CERES data.
•
MODIS is a multispectral (36) imager with
250m, 500m, 1km spatial resolution
•
CERES has footprint of 20 km (Terra, Aqua) at nadir
•
scene type information from MODIS
•
angular bin resolution sharpened to 2o in shortwave
•
wind-speed resolution (over ocean) increase to 2 m/s
•
over land, ADMs built for 1ox1o lat-lon regions at 1 month temporal
resolution
• NDVI used to separate sub-regions within 1ox1o regions
Terra CERES ADMs: Validation
A series of consistency tests are performed to evaluate
uncertainties in TOA fluxes derived with the CERES SW
and LW ADMs:
•
•
•
•
Regional Mean TOA Flux Error Test (SW, LW and WN)
Instantaneous TOA Flux Uncertainties Test
Comparisons with ERBE-Like TOA Fluxes
Comparison with radiative transfer model
Regional Mean TOA Flux Error (Direct Integration)
•
•
•
Regionally averaged ADM-derived TOA fluxes are
compared with regional mean fluxes obtained by direct
integration of observed mean radiances (DI fluxes).
regions of 10×10 latitude and longitude, over several
months.
The regional all-sky ADM is constructed by sorting the
radiances in a region by viewing geometry (, 0 ,) and
evaluating the ratio of the mean radiance in an angular
bin to the DI flux, obtained by integrating radiances in all
angular bins.
Instantaneous TOA Flux Uncertainties Test
• Compare ADM-derived TOA fluxes over 1 regions from
different viewing geometries.
• Comparing CERES Terra ADMs and surface observations
(Programmable Azimuth Plane Scans Over ARM-SGP TEST)
• Terra-Aqua Instantaneous TOA Flux Comparison over
Greenland (69.5N)
• Multi-angle TOA Flux Consistency Tests
(Merged dataset of MISR-MODIS-CERES)
Instantaneous TOA Flux Uncertainties Test
Clear-sky multiangle SW TOA flux consistency: (a) Relative difference
[F(=50-60) –F(Nadir)]/F(Nadir); (b) Relative RMS difference
Validation results:
•
•
Based on all results and a theoretically derived
conversion btw TOA flux consistency and TOA flux
error, the best estimate of the error in CERES TOA flux
due to the radiance-to-flux conversion is 3% (10Wm-2)
in the SW and 1.8% (3 to 5 Wm-2) in the LW.
Monthly mean TOA fluxes based on ERBE ADMs are
larger than monthly mean TOA fluxes based on CERES
Terra ADMs by 1.8 Wm-2 and 1.3 Wm-2 in the SW and
LW, respectively.
To summary
The Angular Characteristics of TOA Radiance depends on
•
•
•
•
Viewing Geometry [Loeb et al., 2002; Suttles et al., 1988]
Surface characteristics (snow is brighter
vegetation) [Loeb et al., 2002; Suttles et al., 1988]
than
Atmospheric Characteristics (clouds, aerosols) [Loeb et
al., 2002; Li et al., 2000; Zhang et al., 2005; Falguni et al.,
2011]
Current CERES ADMs = f(geometry, surface, clouds)
References
• Leob, N.G., N.M. Smith, S. Kato, W.F. Miller, S.K.Gupta, P.Minnis,
and B.A. Wielicki, 2003: Angular distribution models for top-ofatmosphere radiative flux estimation from the Clouds and the
Earth’s Radiant Energy System instrument on the Tropical Rainfall
Measuring Satellite. Part I: Methodology. J. Appl. Meteor., 42, 240265.
• Leob, N.G., S. Kato, K. Loukachine, and N.M. Smith (2005),
Angular distribution models for top-of-atmosphere radiative flux
estimation from the Clouds and the Earth's Radiant Energy
System instrument on the Terra satellite. Part I: Methodology,
J.Atmos. Oceanic. Technol., 22, 338-351
• Loeb, N. G., Kato, S. et al., Angular Distribution Models for Topof-Atmosphere Radiative Flux Estimation from the Clouds and
the Earth’s Radiant Energy System Instrument on the Terra
Satellite. Part II: Validation, American Meteorological Society DOI:
10.1175/JTECH1983.1, 2007
Questions
Backup slides
Filtered to unfiltered radiance
Radiometric count conversion
algorithms convert the detector
digital count into filtered
radiances.
•
• For
use
in
science
applications, radiances from
earth
scenes
should
be
independent of the optical path
in the instrument.
Filtered To Unfiltered
Radiance
Unfiltered radiance
Filtered radiance
Conversion
Anisotropy in Satellite Observations
MISR
F1 = πL1 L1
L1 ≠ L2
L2
F2 = πL2
F1 ≠ F2
MISR L1B IMAGE
Therefore, Lambertian assumption will not work !
Sampling issues
CERES provides two overpasses over a given scene per day. How cloud
the limited observations represent the diurnal variation of solar reflected
and earth emitted radiation? (temporal sapling problem)
Solution: Using CERES observations from multiply satellites (EOS-AM,
EOS-PM, and TRMM), reduce time sampling error by 78%.
CERES has a larger footprint on the order of 10-20 km at nadir. In aerosol
forcing studies, part of samples are discard due to cloud contamination.
This, however, induce a spatial sampling issue.
ERBE ADMs
The Model
The parameters were calculated as a function of 12 scene types.
Scene type
Acronym
•Clear over ocean
clo
•Clear over land
cll
•Clear over snow
cls
•Clear over desert
cld
•Clear over land-ocean mix
clm
•Partly cloudy over ocean
pco
•Partly cloudy over land or desert
pcl
•Partly cloudy over land-ocean mix
pcm
•Mostly cloudy over ocean
mco
•Mostly cloudy over land or desert
mcl
•Mostly cloudy over land-ocean mix mcm
•Overcast
ovr
Cloud coverage (%)
0-5
0-5
0-5
0-5
0-5
5 - 50
5 - 50
5 - 50
50 - 95
50 - 95
50 - 95
95 - 100
Day-night LW flux difference divides overcast into overcast over
ocean (ovo) and overcast over land (ovl).
ERBE SW ADMs
Solar zenith angle
Viewing zenith angle
Relative azimuth angle
0 - 25.84 deg.
0 - 15
0-9
25.84 - 36.87
15 - 27
9 - 30
36.87 - 45.57
27 - 39
30 - 60
45.57 - 53.13
39 - 51
60 - 90
53.13 - 60.00
51 - 63
90 - 120
60.00 - 66.42
63 - 75
120 - 150
66.42 - 72.54
75 - 90
150 - 171
72.54 - 78.46
78.46 - 84.26
84.26 - 90.00
171 - 180
deg.
deg.
ERBE LW ADMs
For each of the twelve scene types, the LW anisotropic factor
and LW Standard deviation were derived as a function of:





four seasons
winter northern hemisphere (Dec., Jan., Feb.)
spring northern hemisphere (Mar., Apr., May.)
summer northern hemisphere (Jun., Jul., Aug.)
fall northern hemisphere (Sep., Oct., Nov.)
 10 colatitude regions
 7 viewing zenith angles
Scanner - A set of three co-planar detectors (longwave, shortwave and
total energy), all of which scan from one limb of the Earth to the other,
across the satellite track (in it's normal operational mode).
The ERBE Scanning Detectors :
1). One Total wavelength (0.2 – 50 μm)
2). One Long wavelength (5 – 50 μm)
3). One Short wavelength (0.2 – 5 μm)
Nonscanner - A set of five detectors
• one which measures the total energy from the Sun
(0.2 – 50 μm)
• two of which measure the shortwave and total energy
from the entire Earth disk (0.2 – 5 μm)
• two of which measure the shortwave and total energy
from a medium resolution area beneath the satellite

similar documents