COL 12(11), 110101(2014) CHINESE OPTICS LETTERS November 10, 2014
1671-7694/2014/110101(5) 110101-1 © 2014 Chinese Optics Letters
Land surface albedo (LSA), characterizing the ener-
gy balance in the coupled surface–atmosphere system
(CSAS), is required by climate models at various spa-
tial scales from meters to kilometers
[1,2]
. Owing to its
strong feedback eects, LSA is also important for deter-
mining atmospheric conditions in the boundary layer.
To facilitate application of satellite-derived surface
albedos to climate studies, much research has been
done on products validation
[3]
, accuracy assessment
[4–6]
,
and method improvements
[7]
. Recently, eorts on the
albedo parameterization over ocean surface have been
made using the coupled ocean–atmosphere radiative
transfer (RT) code
[8,9]
. However, the parameterization
of LSA is still based on the assumption of isotropic
diuse skylight when calculating the surface blue-sky
albedo
[10]
. Recently, the concept of anisotropic sky ra-
diation has been proposed and preliminary results have
been obtained
[11,12]
.
With emphasis on the importance of considering aniso-
tropic distribution of sky radiation, impacts of diuse
skylight on the LSA estimation have been studied
[12]
.
Aiming to parameterize LSA for RT and climate mod-
eling, we rst simulated the angular distribution of sky
radiation using a benchmark Monte Carlo RT (MCRT)
model
[13,14]
, which accounts for three-dimensional (3D)
atmospheric eects. The simulated results gave us full
proof of coupling diuse skylight to LSA model. Then,
a rigorous anisotropic LSA remote sensing model was
constructed and LSA lookup table (LUT) was generat-
ed by coupling the DISORT code
[15]
and the bidirection-
al reectance distribution function (BRDF) assembly
library
[6]
. This object is to provide a fast and accurate
LSA parameterization for soil– vegetation– atmosphere
transfer and climate modeling.
To solve the 3D atmosphere eect on the RT pro-
cesses, the state of the art in modeling this eect is
Parameterization of land surface albedo
Shengcheng Cui (崔生成)
1
, Zhen Wang (王 震)
2
, and Shizhi Yang (杨世植)
1
1
Key Laboratory of Optical Calibration and Characterization, Anhui Institute of Optics
and Fine Mechanics, Chinese Academy of Sciences, Hefei 230031, China
2
College of Atmospheric Physics, Nanjing University of Information Science
and Technology, Nanjing 210044, China
Corresponding author: shchcui@126.com
Received March 6, 2014; accepted August 22, 2014; posted online October 27, 2014
Remote measurements of Earth’s surface from ground, airborne, and spaceborne instruments show that
its albedo is highly variable and is sensitive to solar zenith angle (SZA) and atmospheric opacity. Using a
validated radiative transfer calculating toolbox, DISORT and a bidirectional reectance distribution function
library, AMBRALS, a land surface albedo (LSA) lookup table (LUT) is produced with respect to SZA and
aerosol optical depth. With the LUT, spectral and broadband LSA can be obtained at any given illumination
geometries and atmospheric conditions. It provides a fast and accurate way to simulate surface reectance
over large temporal and spatial scales for climate study.
OCIS codes: 010.1310, 280.1310, 240.5698, 350.5610.
doi: 10.3788/COL201412.110101.
the Monte Carlo technique, which is often used as the
benchmark for testing and validating other RT mod-
els. Recently, a newly developed 3D MCRT model
has been validated and evaluated via the intercom-
parisons between the MCARaTS
[16]
, SHADOM
[17]
, and
SHARM-3D models
[18]
. It is found to be accurate in the
RT simulations for the CSAS, including the charac-
terization of the angular distribution of the irradiance
eld
[13,14]
.
Figure 1 shows diuse skylight distributions for four
dierent aerosol/cloud atmospheres at 550 nm with
an illumination angle at 60°, simulated using the 3D
MCRT model
[13,14]
. From the simulations, one can easily
nd that the angular distribution of diuse sky irradi-
ance exhibits dierent degrees of anisotropic properties.
This nding still holds true throughout the solar spec-
trum. From the RT theory, atmospheric particles, such
as aerosols, clouds, and air molecules, that all have
anisotropic scattering phase function, aect the radia-
tion eld in the atmosphere, resulting in the variation
of intensity in dierent propagation directions.
On the other hand, the underlying surface also
exhibits anisotropic light-scattering (reecting) behav-
iors, strongly depending on the geometries between the
illumination and the viewing direction. This anisotropic
surface reectivity may further modulate the outgoing
electromagnetic waves by the trapping mechanism, cor-
responding to the successive reections and scatterings
between the atmosphere and the underlying surfaces,
and have a considerable eect on the distribution of sky
radiation
[12,19]
. In this context, the atmospheric radiation
is in nature anisotropic.
Recent measurements from NASA airborne cloud
absorption radiometer instruments demonstrate that
BRDF of the CSAS is anisotropic
[20]
. The agreement
between the airborne observations and our simulation