Upload
emma-neal
View
217
Download
0
Embed Size (px)
Citation preview
MODIS/Meteosat/MISR Surface Albedo Comparison Exercise
MODIS/Meteosat/MISR Surface Albedo Comparison ExerciseB. Pinty (1), M. Taberner (1),
S. Liang (2), Y. Govaerts (3), J.V. Martonchik (4), Lattanzio (5), C. Barker Schaaf (6),
M. M. Verstraete (1), R. E. Dickinson (7), N. Gobron (1), and J-L. Widlowski (1)
(1) Institute for Environment and Sustainability of EC-JRC, Ispra (VA) Italy(2) University of Maryland, College Park, USA
(3) EUMETSAT, Darmstadt, Germany(4) Jet Propulsion Laboratory, Caltech, Pasadena, USA
(5) Makalumedia gmbh, Darmstadt, Germany(6) Boston University, Boston, USA
(7) SEAC, Georgia Institute of Technology, Atlanta, USA
2nd CEOS/WGCV/LPV Workshop on Albedo Products, Vienna, April 27-28, 2005
Various types of Surface albedo (1)Various types of Surface albedo (1)
BHR : Bi-Hemispherical Reflectance is the ratio between the upward and the downward radiant fluxes, that is, accounting for the downwelling diffuse intensities from the sky.
Depends on both surface and ambient atmospheric radiative properties and …the Sun angle.
All quantities can be defined monochromatic or broadband
Various types of Surface albedo (2) Various types of Surface albedo (2)
DHR: Directional Hemispherical Reflectance is the ratio
between the upward flux and the downward collimated flux coming thus from one single direction (Black sky).
Depends on surface radiative properties and …the Sun angle.
All quantities can be defined monochromatic or broadband
BHRiso: If the downwelling diffuse intensities from the sky is assumed fully isotropic then the BHR is equal to the integral of the DHR over all incoming directions (White sky).
Depends on surface radiative properties only.
Surface albedo products from space agencies
Surface albedo products from space agencies
• MODIS delivers DHRs (Black sky) and BHRiso (White sky)
• MISR delivers DHRs and BHRs as flux ratios but under ambient conditions and for the Sun illumination conditions at time of observations
• EUMETSAT delivers DHRs for a fixed Sun angle
and all information needed to reconstruct the DHRs and BHRiso
and all information needed to reconstruct the DHRs at any other Sun angle as well as the BHRiso
to reconstruct the BHRs may require some investments or some level of assumption
The albedo products may also differ wrt the spectral bands of integration they refer to.
Parameterization of the surface-atmosphere radiative coupling
Parameterization of the surface-atmosphere radiative coupling
),;,();,(
);,();,(),;,(
0000
000000
sfcdiff
sfciso
dirsfcsfc
zfzBHR
zfzDHRzBHR
Assuming that the field of downwelling diffuse intensity reaching the surface is PERFECTLY isotropic yields a convenient parameterization for the BLUE SKY ALBEDO
Surface level
Sun angle
Atmospheric optical depth (type of atmosphere)
Surface BRF (amplitude and shape)
ratio of direct to total downward flux
ratio of diffuse to total downward flux
1),;,();,( 0000 sfc
diffdir zfzf with
Pinty et al., JAS, 2005
Surface albedo comparisonSurface albedo comparison
• Perform a comparison between MODIS-Meteosat-MISR surface albedo products : a user perspective.
• Compare similar physical quantities, e,g., BHRs, BHRiso, DHRs.
• Based on year 2001 products: latest public version available.
• For two large geographical regions : Africa-Southern Europe and North-East Europe.
Comparison of Surface BHRiso products from MODIS/Meteosat/MODIS
Comparison of Surface BHRiso products from MODIS/Meteosat/MODIS
• Select the same period of time and identical geographical regions
• Achieve the needed transformations (e.g., BHRs, spectral conversions) to ensure comparison of physical quantities having same meaning
• Identify the product values showing appropriate QA
Albedo comparison for an ‘Ideal’ Band (0.4-1.1μm)
Albedo comparison for an ‘Ideal’ Band (0.4-1.1μm)
Govaerts, pers. Com
Spectral Correction for the Meteosat large band effects
Spectral Correction for the Meteosat large band effects
Spectral conversion to the ‘Ideal’ Band (0.4-1.1μm)
Spectral conversion to the ‘Ideal’ Band (0.4-1.1μm)
Meteosat MISR
January 2001
MISR low & MODIS high
MISR high & MODIS low
January 2001
Histogram of BHRiso differencesHistogram of BHRiso differences
(MISR-MODIS) Albedo
Jun
e 2
00
1
Jan
uary
20
01
Using Shunlin’s conversion factors
Using Yves’s conversion factors
January 2001
Mean BHR values over common area with valid values from one of the two
other sensors
Mean BHR values over common area with valid values from one of the two
other sensors
Ratio of the mean valuesRatio of the mean values
Primary EigenvectorsPrimary Eigenvectors
Correlation between pairs of samplesCorrelation between pairs of samples
Results for year 2001 Shortwave domain
(0.3-3.0 μm)
Results for year 2001 Shortwave domain
(0.3-3.0 μm)
Africa –Southern Europe
Primary EigenvectorsPrimary Eigenvectors
October 2001
Results for year 2001 Visible domain (0.3-0.7 μm)
Results for year 2001 Visible domain (0.3-0.7 μm)
Results for year 2001 Near-infrared domain
(0.7-3.0 μm)
Results for year 2001 Near-infrared domain
(0.7-3.0 μm)
Results for year 2001 Shortwave domain
(0.3-3.0 μm)
Results for year 2001 Shortwave domain
(0.3-3.0 μm)
Northern –Eastern Europe
Full inversion
Magnitude inversion
Backup solution
Mean BHR values over common area with valid values from one of the two
other sensors
Mean BHR values over common area with valid values from one of the two
other sensors
Full inversion
Primary EigenvectorsPrimary Eigenvectors
Full inversion
Northern –Eastern Europe
Hexadecad 6: End of March 2001
Issues and caveatsIssues and caveats
• An error was recently identified in the MISR processing code: BHRs tend to be biased high by about 2 to 3% on average (season & latitude dependent).
• The nominal spectral conversion formulae (from Liang and Govaerts) agree well.
• MODIS BHRs are off when estimated from the backup algorithm.
Africa –Southern Europe
Magnitude inversion
January 2001
June 2001
Northern –Eastern Europe
Magnitude inversion
January 2001
June 2001
Conclusions and PerspectivesConclusions and Perspectives
• Albedo (BHRiso) comparison reveals very good agreement between MODIS-MISR-Meteosat (high QA) products.
• Extend the comparison exercise globally.
• Repeat the exercise for DHRs.